Zia |e Males Ehrenreich__--- VAMOS 222553 14 | 159.9 | 93.0 Se) aS | 1222) | borse Siac |) 7O0Ld) |G |os9) | 70s Ehbrenreich____- Camaiura AD 64a COIS iets. Bul tae Gul ell Ones 78.9 | 64.3 | 4.8 | 3.8 | 79.1 Von den Stein- | Jurund_____- PAN elaVegtey eee ASF eLoT6n | to) | Seaeee 84.3 | 62.1 | 5.3 | 3.6 | 67.9 Ry MGODOS: =22--- eT esse ee TGOSO a Ee oe | at ce | PR Rese Le 7 (Mer fd (ese onl Ce eee 69.9 ruben <2 eee a US a I ae | a oe Soa (2) El ee ees (eee (ee 77.5 Females Ebrenreich__--_- ARLELO)- Base = 2 | 148.0 | 82.7] 18.5 | 14.6 | 11.6 | 51.9 | 78.9 | 62.7 | 4.1 | 3.4 | 82.9 Ehrenreich___~_- Camaiura_._| 4 | 153.7 | 79.8 | 18.0 | 14.0 | 11.6 |------ 77.7 | 64.4) 4.4 | 3.4 | 77.2 Vonden Steinen_} Jurund___.-_- 240i 0Aeeesn ASSO PS 30n 1d Ge see 83.3 | 64.4 | 4.8 | 3.3 | 68.7 From table 13, it may be seen that the Tupi are short, brachy- cephalic or subbrachycephalic, leptorrhine or mesorrhine, and have an accentuated sexual dimorphism in stature, the women being shorter than the men. Summary.—From the paleoanthropological and somatophysical data which we have cited, it is fair to infer that the first inhabitants of Brazil, and probably of South America, were probably the Lacida or the ‘““Homem Lagosantense’”’ (Mattos, 1941) or else the ‘““Homo lago- maritimus” (Eickstedt, 1934). The L&cida, a dolichocephalic and platyrrhine type discovered during the memorable explorations made by Lund in the District of Santa Luzia in Minas Gerais, may be 794711—_50——_7 84 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 distinguished from the ‘‘SSambaqui’’ type, found in the shell mounds of the southeastern coast of Brazil, also dolichocephalic and platyrrhine, by the height of the skull. The Lacida is definitely hypsicephalic; the ‘“Sambaqui”’ is chamecephalic. They were, it would seem, con- temporary. The craniometric analysis of the “Lagoa Santa” speci- mens shows that they average two hypsicephalic to one chamecephalic. Kickstedt classified both the Lacidas and the Sambaquis under the designation “Homo lago-maritimus Americanus,” although he be- lieves the ‘“SSambaqui” men may belong to a more remote period. Later, though long before historical times, new elements came to Brazilian territory, the Arawak, Carib, and Tupi-Guarani. We do not possess exact data to determine the exact period in which they ap- peared and came in contact with the first inhabitants, but along the Atlantic Coast in the southern part of the country as far as the Doce River, the skulls which have been found in more recent strata of kitch- en-middens are all pronouncedly brachycephalic, precluding any mis- taking them for skulls of the “Sambaqui’’ men. These later groups were driven to the Central-Eastern Plateau of Brazil, and together with the Lacidas they disappeared as distinct groups leaving only the Ge or ‘‘Tapuya’’ as a result of crossing. Other groups probably worked their way, in successive waves, into the Brazilian “sertdes,” an hypothesis which would explain the extraordinary number of languages spoken in South America. From the purely anthropometric point of view, the only incontesta- ble classification that can be made of the present-day aborigines is that which separates the dolichocephalic and platyrrhine or sub- platyrrhine Ge, living on the Central-Eastern Plateau, from the brachycephalic and leptorrhine or mesorrhine individuals of the coast, the basins of the great rivers, and the Guianas. Possibly, under the present regime of ‘‘Ajudancias,” so wisely cre- ated by the Servico de Proteccao aos Indios, from which the Indians receive medical and social help, interesting data may be collected that will serve as a basis for a better and more detailed classification of the remaining Indians of Brazil, before they are completely absorbed into the mass of the Brazilian population. BIBLIOGRAPHY Barbosa Rodrigues, 1882; Bastos d’Avila, 1938 a; Canestrini and Moschen, 1879; Carvalho, 1929, 1931; Ehrenreich, 1897 a; Eickstedt, 1934; Ihering, H. v., 1904; Lacerda, 1879 (1881), 1885; Lacerda and Rodrigues Peixoto, 1876; Mattos, 1941; Morton, 1839; Moschen (see Canestrini and Moschen, 1879); Pinto (see Roquette-Pinto); Péch, 1938; Rey, 1880; Rodrigues Peixoto, 1885 (also see Lacerda and Rodrigues Peixoto, 1876); Roquette-Pinto, 1917, 1935; Steinen, 1940, 1942; Virchow, 1872 a, 1874 a, 1874 b (?), 1875. See also the following: Bastos d’Avila, 1938 b; Rufz, Broca, and Martin de Moussy, 1860. THE PIGMENTATION AND HAIR OF SOUTH AMERICAN INDIANS By Morris STEGGERDA In describing the skin, eyes, and hair of South American Indians, one must, of necessity, quote from the writings of anthropologists who are acquainted with the various Indian tribes. This short paper, therefore, is chiefly a compilation of descriptions gathered from such sources. I shall consider first the Indians living in the western part of South America, beginning in the north and continuing south along the Andes to the tip of the continent, and then describe those in the north- central and central parts of South America. It is to be regretted that the descriptions are largely subjective. Very rarely have the anthro- pologists used any of the more objective color scales which might have proved useful in describing the skin, eye, and hair color. However, the terms used are at least descriptive, and as such are reviewed. SKIN COLOR Buschan (1922) makes the general statement that the skin color of South American Indians is lighter than that of the Indians of North America. He describes the skin color of South American Indians as being light yellowish gray, or the color of clay or of tanned leather, and says that this is especially true of the tribes living in the dense forest regions. The Indians of the Orinoco, for instance, have a very light skin color, almost white; while other Indians, whose habitat is mostly in the sun, are of a coppery or even purple-brown color. Asa result of crossings with Whites and Negroes, which have taken place over a period of time, all variations of skin color can be noted in South America, from the very dark tribes to those which are very light- colored. Describing the Indians of the present country of Panama, Hrdlitka (1926) speaks of the Cuna as having a medium-brown skin. He notes the frequent occurrence of albinism among these Indians, a condition which is present among other tribes of this area as well; for example, the San Blas Indians, who according to Harris (1926) range from ‘‘a normal red to white.” The Cayapé Indians, a Chibchan tribe living in the present country 85 86 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 ‘of Ecuador, are described by Barrett (1925) as being “‘reddish brown.”’ Their neighbors, the Tunebo, are said to have an olive-colored skin (Rochereau, 1924); whereas the Puinave, another group living in this area, are of dark skin color, according to Pericot y Garcia (1936), who describes the Tucano as having a brown skin. The Colorado, also Chibchan, have a reddish skin color. This is to some extent accentuated by their custom of painting their faces and bodies a bright red with a color obtained by grinding the seeds of the achiote plant. Hagen (1939) states that these Indians have generally a lighter skin color than other tribes of the upper Amazon region. The tribes living in the present country of Pert are dark brown, ac- cording to Eberhardt’s (1910) observations. The Aymara, living in the Andean Highlands of Pert, have been studied by McMillin (1927) who speaks of them as having a “swarthy brown to coppery”’ skin, or even a “dark olive” color, some being lighter with a yellowish hue. Ferris (1921) mentions that there are light as well as dark-brown individuals among this tribe, according to the degree of mixture with other races. Pericot y Garcia (1936), on the other hand, calls their coloring olive gray, and adds that they are lighter than the neighboring Quechua. The Araucanians, living farther to the south, are described as light to olive brown by Pericot y Garcia (1936), reddish brown by Latcham (1904 b), and light brown to yellowish by Ten Kate (1904). The Tehuelche of Argentina are said by Ten Kate (1904) to be light brown or yellowish brown. Of the Guato of the upper Paraguay River, Markham (1894) says the skin color is very light. On the extreme southern tip of South America are the Ona, Yahgan, and Alacaluf. Lehmann-Nitsche (1927) records the skin color of the Ona as very light yellow to brown, whereas Garson (1885) says that they are reddish brown, the shade of mahogany or bronze. Gusinde (1937), an authority on the Fuegian Indians, describes the Yahgan as yellowish brown; and the same term is used by Hyades and Deniker (1891), who add that they may also be reddish, and that the females are of lighter color than the males. The Alacaluf are classed by Skottsberg (1913) as being of the same color as the Yahgan, while Pericot y Garcia (1936) calls them merely “brown.” Turning to the Indians of the northern and central regions of South America, we shall first consider the Carib linguistic family, to which the Taulipdéng belong. According to Pericot y Garcia (1936), they have a delicate brown skin color; Koch-Griinberg (1923) describes the skin of the pure-blooded Taulipdng as “velvety, soft, and light brown.” The Makiritari, also a Carib tribe, are said to be ‘almost white.” In describing the Central Carib, Farabee (1924) discerns two main Vol. 6] PIGMENTATION AND HAIR—STEGGERDA 87 groups: one of the savannah and one of the forest. The savannah tribes are of darker skin color than the forest tribes, who live in the dense and shady jungles. Hawtayne (1886) found two types of Carib in the Island of St. Vincent: the yellow or red, and the so-called “‘ Black Caribs,’’ who are a hybrid race, a cross between yellow Carib and Negroes. Gillin (1936), in speaking of the Barama River Carib, mentions among their characteristics a light to red-brown skin color. The Yaruro in Venezuela are Mongoloid in appearance, and have a very dark skin color (Petrullo, 1939). Writing of the Arawak, Pericot y Garcia (1936) points out that they have a lighter skin color than the Carib. Ferris (1921), in describing the Machiguenga, an Arawakan tribe of eastern Pert, states that about 30 percent of them are of light skin color, 3 percent dark, and the remaining 67 percent medium brown. Another Arawakan tribe, the Paressi, have been studied by Roquette-Pinto (1938) who says that they are copper-yellow, of a darker shade in the Cozarimi and lighter in the Uaimare regions. The Passé, also Arawak, are described by Markham (1895) as nearly white, resembling Caucasians. The same author also speaks of the Puru-Puru (Arawak) as having among them a number of individuals who show white and brown patches of irregular size and shape on the skin. This, however, is a pathological condition. He also considers the Tacuna, and calls them darker than most Indians located on the Marajfién River. The Tariana have a glossy brown skin color; the Pawumwa or Chapacura, living along the Guaporé River, range from a dark, coppery color to a very light brown (Haseman, 1912). The Yamamadi, Ipurina, and Botocudo live in the dense forest regions of Brazil, and here again we find a yellowish-gray skin color (Ehrenreich, 1897); according to Manizer (1919), the Botocudo are light brown. The Anambé, a Tupi group living on the lower Tocan- tins River, show an almost white skin color. The Chiriguano, another Tupt tribe, are yellowish-brown (Ten Kate, 1904). The Sirioné of eastern Bolivia are often very light; in fact, Wegner (1934 a) makes the statement that there are some white Siriond. The skin of the Bororo and Carajd, as described in Ehrenreich’s (1897) observations, is reddish in color, the Bororo being darker than is usual among the Tupi. The Caingang were found to be both light and dark, depending upon the region in which they lived (Pericot y Garcia, 1936). The Nambicuara are described by Roquette-Pinto (1938) as having a skin color of burned-yellow, which is darker in the Cocuzu subtribe, and in the Tagnani tends to show shades of pink. Markham (1895) states that the Indians of the entire Amazon Valley have skins of ‘ 88 SOUTH AMERICAN INDIANS [B. A. B. Bull. 148 various coppery or brown shades, which often appear to have the color of smooth Honduras mahogany. EYE COLOR AND FORM The majority of the authors quoted in the section on skin color have also reported an eye color, and in many cases on the shape of the eye and form of the eyelid. Their remarks are, in the main, very uniform. The most typical descriptions of eye color are ‘dark brown,” “black,”’ and “dark brown to black’’; and the terms most frequently used to describe shape are ‘‘Mongoloid eyes,’’ “eye slits oblique,’”’ and ‘‘small eyes.” Therefore, in this section, mention will be made only of those descriptions that deviate from the general usages. One of the tribes of which something unusual has been said is the Chipaya, of highland Bolivia, among whom, according to Posnansky (1918), the Mongolian eyefold is unknown. Harris (1926), in describing the eyes of the San Blas Indians, says that in the ‘‘brown Indians’”’ the iris is medium to dark brown, while in the partial albinos it varies from hazel (blue with brown spots) to dark blue and dark violet. Describing the eyes of the Aymara, Ferris (1916) states that the areola of the iris is medium maroon, and the periphery greenish- yellow. Pericot y Garcia (1936) differs from this slightly, saying that the iris is dark brown and the cornea yellowish. Chervin (1913) has observed that the Aymara have lighter eyes than the Quechua, whose eye color has incorrectly been called ‘‘black,’’ whereas it is really of various shades of chestnut. Among the Tehuelche described by Ten Kate (1904), only one had a Mongolian eyefold, and all had eyes of a light-brown color. In speaking of the eyes of newborn Yahgan, Hyades and Deniker (1891) say that the iris is black, and becomes light brown after a few days, then turning to dark gray with a bluish shade after 2 weeks. After the tenth month the eyes are dark brown in all children. Gusinde (1931) states that the eyes of the adult Yahgan are either light hazel or various shades of brown to deep black. Gusinde, who has studied the Alacaluf, observed that the eyes of these Indians range from light hazel to deep black. Skottsberg (1913) remarks that their eyes are dark blue in children and be- come deep brown in adulthood. Both Gusinde (1931) and Pericot y Garcia (1936) point out that the Alacaluf’s eyes are Mongoloid in appearance. Concerning the Nahucua and Bororo Indians of Brazil, Ehrenreich (1897) says that the iris showed a blue color in a few of the individuals observed, while the majority had different shades of brown. The Botocudo have eyes of a medium light color, according to Manizer (1919). Vol. 6] PIGMENTATION AND HAIR—STEGGERDA 89 HAIR COLOR AND FORM In describing the hair of Indians in both North and South America, anthropologists very frequently use the general adjectives: “‘coarse,”’ “black,” and “straight.’”’ Many tribes, however, do differ from this description; and those comments of anthropologists that deviate significantly are recorded below. According to Harris (1926), who has made a thorough study of the brown and white Indians of San Blas, the hair is black in brown Indians, and from flaxen to straw-colored in the white Indians. Commenting on the hair of the Aymara and Quechua, Ferris (1921) states that it is generally black, straight, and abundant, but that the hair on the face is scant in both mixed and pure Quechua. There is very little grayness among these Indians, he says. Similar remarks have been made by Pericot y Garcia (1936) and Forbes (1870). The Araucanians, who also have dark and straight hair, have no body hair (Latcham, 1904 b); but Ten Kate (1904) has seen some individuals among them with a slight beard or moustache. The same author made identical observations on the hair of the Tehuelche: that it is abundant, straight, and black, and that here also some individuals have a slight moustache. The Yahgan have the same general characteristics, and Hyades and Deniker (1891) emphasize the Mongoloid appearance of these Indians. Gusinde (1937) also mentions the scant body hair, which he says is typical of all Fuegians. Skottsberg (1913) describes the hair of the Alacaluf as brown, in contrast to the uniformly black hair of the Fuegians. Ferris (1921), an authority of the Arawak, reports that the hair of the Machiguenga tribe is long, thick, dull, black, and fine. The scalp hair is straight in all males, but he found one female individual who had slightly wavy hair. He found that beards were entirely absent in 38.8 percent of the males, sparse in 18.8 percent, and present to a small extent in 42.4 percent. Similar observations have been made upon the Machiguenga by Farabee (1922), and he adds that face hair is pulled out by the men, whenever it appears. The hair of the Caingua is both wavy and straight, as reported by Pericot y Garcia (1936). That of the Cayua, of the Tupi linguistic family, as described by Koenigswald (1908 a), is always straight, black, and thick, and is worn long. Baldness and grayness of hair is unknown even in the oldest individuals of this tribe, according to the same author. In the Guayaki, the hair varies from brown to shiny black (Vellard, 1934). Serrano (1930) has made an interesting observation on the Mataco and Choroti; the hair of adults is black, while in the children itisreddish. Karsten (1932) describes the hair of these adults merely as long, black, and coarse. 90 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Roquette-Pinto (1938) comments on the hard, straight hair of the Nambicuara, but mentions that he has seen a few of these Indians with wavy hair. The Puinave, an independent group in Colombia, is worthy of special comment; Pericot y Garcia (1936) quotes a remark taken from Tastevin, in which the hair color of these Indians is described as chestnut brown to almost blond. Pericot y Garcia (1936), however, states that their hair is black and straight; and he also mentions that the men have scarcely any beard, except occasionally on the upper lip. The Tucano have short, black, and often frizzly hair, according to Pericot y Garcia (1936). The most usual description of the Indians of the Mato Grosso, the Purtis River region, and the Xingti district of Brazil, indicated that their hair is coarse, black, and straight; but this is apparently not true of all these tribes. Only among the Bororo and Carajd is this type of hair predominant. According to Ehrenreich (1897), other groups show individuals with thick, wavy, and fine hair; there are frequently people with curly hair, mostly among the Bacairi. The Bacairi hair color, though apparently black, shows a brown hue in strong sunlight, and the children always have this shade of hair. Old people have gray hair, but white hair has not been observed by this author. Wavy and frizzly hair has also been seen among certain individuals of the Arawak tribes; according to Pericot y Garcia (1936), the inci- dence of wavy and frizzly hair is rather great. Among the Indians whose hair sometimes diverges from the norm are also the Botocudo. There is a frequency of reddish-brown hair in this tribe (Manizer, 1919). Stegelmann (1903) also discovered a tribe of peculiar appearance living on the upper Envira River in Brazil. Their hair was light red, similar to that found in certain Jewish types. Their skin was red also. The other Indian tribes called them “Coto,” which means “howling monkey,” because of the similarity of their color to that of this particular monkey. BIBLIOGRAPHY Barrett, 1925; Buschan, 1922; Chervin, 1913; Eberhardt, 1910; Ehrenreich, 1897; Farabee, 1922, 1924; Ferris, 1916, 1921; Forbes, 1870; Garson, 1885; Gillin, 1936; Gusinde, 1931-37; Hagen, 1939; Harris, 1926; Haseman, 1912; Hawtayne, 1886; Hrdlitka, 1926; Hyades and Deniker, 1891; Karsten, 1932; Koch-Grinberg, 1923; Koenigswald, 1908 a; Latcham, 1904 b; Lehmann-Nitsche, 1927; Manizer, 1919; Markham, 1894-95; MeMillin, 1927; Pericot y Garcia, 1936; Petrullo, 1939; Posnansky, 1918; Rochereau, 1924; Roquette-Pinto, 1938; Serrano, 1930; Skottsberg, 1913; Stegelmann, 1903; Ten Kate, 1904; Vellard, 1934; Wegner, 1934 a. See also the following: Aichel, 1932 b; Henckel, 1934; Trotter, 1943. BLOOD GROUPS OF SOUTH AMERICAN INDIANS By WiuiaM C. Boyp The hereditary blood groups (O, A, B, AB) represent a set of physi- cal characteristics determined by three allelomorphic genes (O, A, B). The study of blood groups is of value to physical anthropology because they are genetically determined by a known mechanism, absolutely objective in character, completely unaffected by environment, not subject to mutation at any rapid rate (Wyman and Boyd, 1935; Hal- dane, 1940), and nonadaptive as far as extensive investigation indi- cates. These merits should render blood groups a most useful criterion in elucidating the classification of mankind and human evolution. The genes determining the four blood groups vary in frequency in different human populations, with B highest in Asia, and A highest in Spain, Australia, the Scandinavian countries, and North American Indians. Maps showing these distributions will be found in Boyd (1939 b) and Haldane (1940). The more recently discovered M, MN, and N types (Landsteiner and Levine, 1927) are determined by two allelomorphic genes without dominance. These types are independent of the O, A, B, AB blood groups. World distribution of the M, N types has not been as exten- sively studied as that of the O, A, B, AB groups, but the existing evi- dence suggests that their frequency is mainly more uniform. N type is highest in Australian aborigines; M type is highest in American Indians (Boyd, 1939 b; Birdsell and Boyd, 1940). The study of the blood groups as a physical anthropological charac- teristic goes back to the pioneer studies the Hirszfelds (1919) made during World War I. The inevitable extension of such studies to American aborigines was made in 1923 by Coca and Deibert (1923), who reported results on Indians of the Haskell Institute, Lawrence, Kansas, U.S.A. Although these and later authors found fair amounts of group A, the fact that A was much lower than in peoples of Euro- pean stock, and that in some Indian groups nothing but group O was found, led to the suggestion that originally American Indians possessed only group O, and that the A and B found were due to admixture of White blood. This idea received a serious set-back when the reports of Rahm (1931 c) and Matson and Schrader (1933), dealing with Yahgan in Tierra del Fuego, and Blackfoot and Blood in Montana, 91 92 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 respectively, were published. Mathematical analysis (Wyman and Boyd, 1935; Boyd, 1939 a) shows that in fact a majority of the results on Indians cannot be explained on this hypothesis. Boyd (1940) has suggested that the present distribution of blood groups in American aborigines is due to (a) several successive migra- tions of small numbers of persons, the various migrations representing somewhat different stocks; (6) the effects of isolation (Wright, S., 1931 a, 1931 b) of such immigrants in America, which led occasionally to the loss of one or more genes. Such a suggestion would account for the uneven character of the blood-group distribution in South America, where it seems evident that many tribes possessed only group O, others had O and A, one or two perhaps had some B, and some perhaps possessed both A and B. The available results on blood groups in South American Indians are presented in table 1, and those on the M, N frequencies in table 2. For discussion of statistical evaluation of the relative reliability of various results, see Boyd (1939 a, 1939 b), Haldane (1940). The results are arranged by countries, except that Tierra del Fuego is listed separately. The O, A, B, and AB percentages suggest the uneven distribution already mentioned. A number of tribes seem to have originally possessed only group O. Such were probably the Chunupi, Mataco, Toba, Guarant, and Macd, perhaps most of the Indians of Ecuador, and the Indians at Naranjalin Colombia. The Carajd: (Golden, 1930) may have possessed considerable B, although the exact percentage is hard to estimate from the tests reported, since such small numbers are involved. In answer to the doubt often raised as to the reliability of Golden’s results, it may be pointed out that he used the same sera to test peoples of European stock and achieved the expected results. Nevertheless, it is now the only tribe in which any significant amount of B seems at all credible, judging by the results available. A reexamination of the Yahgan indicates that the non-O groups are due to racial admixture. Prof. A. Lipschutz found that of 20 pure-blood Yahgans, all were group O. Some non-O, although not a high percentage, was found in 20 Yahgans known to be mixed. A similar check upon the Carajdé would be highly desirable, but for the meantime it seems best to provisionally accept the original report. It is not too easy to say which South American tribes certainly possessed group A before White contact, but we may speculate that the Colla and Pilagé are among them. (26d coed) 0S—TTL¥6L ‘qyS11 uo Bb yyoT uo d ‘suumnjoo prjos jo yysIoy Aq uMOYS so8ejueoied oyeutxoiddy ‘gq 10} 984} 5 ‘y 103 oues ay} jo Aouenbeay oy} syuesoidor d joquiAs yy, ‘“SUBIPUL UBOTIoUTY YPNOY Suowe uorznqiiysip dno1s-poo[q yuereddy—'g avy o=35 YFAO ¥0 OS'O= 7 o (26 "A cova) OS—ITL¥BL ‘4811 uo B "43, wo d ‘suummjoo prjos jo 3yZ1ey Aq uMOoYs soSeqjueoied oyeuntxoiddy ‘gq 103 4843 5 ‘y 103 oues ey} jo Aouenbesy oy} sjueseider d joquids oy, ‘“SUBIPU] UBOTJeUTY Y}NOg Zuowe uorjnqiystp dnoi3-poo,q yuersddy—'g avy Com red yD ———— ed aAlz eZAO 0 tl evFAO O=-d ¥0 OS O= oan szo=d sz'o=7 pte all ste A O-d Zo=d A , Pe iv 93 BOYD BLOOD GROUPS Vol. 6] Ss=sc==5 c16° = 9 ¢z0° 0 0 6'F 1'c6 Gia [pte ek Sie a Re ee ODT Leal te ac aa es OE SOE EOE ee OI AL OOLMS | marcetmamrnte 201009 oy Ors * 690° ZOL* OT £01 T'8T 9°02 FOG eas |e eo eae eee ODF SS arar eS arr er ee ca ee ee 0}seq “obnujuDy 8° 086° 8£0° ¥£0° 0 ZL +9 F 98 19% AGHHO)) [OMG CER Ge [Neds Lap | PSS Os Sees Sas RSS ae os osejueg pue Aopunqig | pus hopungig “(€61) Bons OOnu #6 816° €80° 970° 0'T z9 9°8 Z'F8 ¥8¢¥6 0 0 | 990°0 0 0 6 ‘0 T 68 hat THRO VV AINE momen ==" "990 L0UD qUaddad | qUaIsag | UWIILA | JUIILIT | 4AQUNNT DuijUuablpy (1) (D) (d) O d ee te Mf O | peysey 1o/d __|srenpra UOl}VsI}SOAUT avid equ, | -Ipul J —oues Jo Aouenbelg | —edA} poolq Sulaey s[enplarpuy sdnoib gy pun ‘g ‘py ‘QO :sdnoib poo7g fiq sunipuy updriuamp yynog fo woynoifissnjj— | TIAV I, [B. A. B. Bull. 143 SOUTH AMERICAN INDIANS 94 “qd 686 ‘PAO : ZE6T ‘S999”'T ¢ “poxlui=,,UI,, ‘oind=,,d,, y *po}B[NI[VIeyY ¢ “d 6861 ‘PAOg ‘Og6T ‘WOSITIOM PUB UeyZEYS z “SOUIT} 0Z UI eou0 Ub} 210M 00'Z Peedxe Jou P[NoYs enyjwA sz ‘cANTIqI[INbe o1oues Ul suolye[ndod uo JOL9 qnoyjIM 8UOp S}[Nsel UL ‘UOl}eIAep PivpuBys sj pue (I+b+d)—T Jo 01481 944 SI o/C 1 THEE EE EERE 000 'T 0 ye 0 0 oot | 02 oo’ | 002" 0 | 0 16 0 6 | ge 216" 0 | 820° 0 0 |o¢ |e | 8I 000° 0 0} o 0 0 oot | oz or: |o80° |esr: |e |rer |r1s2 jose | zet os: | to | 820° Joe |aze [te |r. | set 000° 0 0 | 0 0 0 oor | 002 L6L" 0 | 202" 0 0 |s9¢ |¢-e9 | sit 136° | 600° | ¥00° Cet. 6: e-26 | I11 000°T 0 0 | 0 0 0 oor | 201 066° | 100° | 600° ean oe st |0's6 | 6 e16° |¥00° | €zo° | T° 8° cy | 96 | OFT 296° | 600° | ¥z0° | T° cy Se es 916° | 200° | 210° o |Ft |re |zse | ont 086° | 900° | to’ |¢° zt |z2 | | zee 916° | 400° | 210° | 9° 8" ye |e | srt 000°T 0 0 | 0 0 0 oot | 182 186" 0 |zt0° |2° o |e |e96 | 6eF ses‘ |szo’ |szo’ |oz |2z% |ee | ozs | 699 quadsag \quadsag \quaosad |yuaosag |4aQunny (1) (Db) (d) O a vol ee ie ¥ O | poysay sjenpia —| -1pur —oue8 Jo Aduenbe1y —adf4} poolq ZUIAvY S[eNplAIpuy Bere aa a ee ZATIU OSGI tee ct a A ee OD es eae eee Ct ia Sano a aaa (DHSS) AUS ence me es Sh ee OU A IN OS le em AG a aT cana eT Fah figs 200910 lh ae ean i cae Rear gee gg agen || 7a eee EIYO) Oban 1ap D1LALT, his cata (9861) UWeM |-------"-~"S[[Bq BAVSSeD O2IH pUe 9010g UBIp |---~------vuvinyy UDbIng Per (6261) Apouusy pUv SSOP |-~-~~~--~--- ~~~" ~~ -OdNIOg pues equivqvovuENny |-~~-~ ~~ 5 , UL, WO ea hts BE eee (1) 9 pages lata a ecg eat eae Scheer ae ne eR me mene Re i ry NSE fae OS ae Belo LICL OULU: lcneus = secs en) Sk Eases ee oaks aes ae | ae ae, Fe Se ae ta al WlOOCZZGI hs a oe ee ee 2 ee eemeee oo | Nepeateeemmanmmn 7700) Nad Sigh ee (GEOL) OZ Ta ae SS ae ae a ee OCD) eee eRe UL te la ae perme 71810: OLOQTH A ec a ta ee cee aa ee meee COLO) fhonboi0q Bee een See eo ne prio 9 Rana sae song e a ae ena an OLU IS HOON eet Ss ve ae gd Opie ele a eis cad oe a ee CIOL YG), ang pie egee Nem gee OPsrastilhacts: oc) Le ea ee ea ee LU LO Beg Sige pe fee Bee aT Os soca pe ee ee er ee ae ae SOLO OD) Sera ae cg aie eae ODF res lan et eer oe ee ee ee UC EOL ees ee FEO L) PUR USS cia se nae recur reese ia porter en: OAL R OLY, Loponog A enc lee, Pe (0) 0 wie| ie ce ig aan wn Ae ee ee mma Nie Roar cee AOL Ene See Gere rey eRe OD ce seo ees ee re ee LO ROG) lee eG ores PrGL) BOneMNO RIA atee | oa eee ak OOD NG ee a 2 0E O01} BSI}SOAUT aed eqs eo Se ee anal ‘ponutu0g—sdnoib gy pun ‘g ‘vy ‘Q :sdnos6 poo7g fig sunipuy uvowamy ynog fo uorpdyfisspjn—] ATAV], Vol. 6] BLOOD GROUPS—BOYD 95 TABLE 2.—Classification of South American Indians by blood groups: M and N groups Individuals hay- | Frequency of Indi- | ing blood type— gene— Tribe Place Investigators Lae | tc: tested M M |MN| N (m) (n) Argentina Num- | Per- | Per- | Per- ber cent | cent | cent Bnet. -| Chaco... 22 s-nc _<) se es ee 1s Indians (not wild; possibly mixed—AvurHor)-_----_- 2 Actual body measurements on mixed breeds have been taken by few investigators. Bastos d’Avila (1937) lists several measurements taken on Indian tribal crosses. He discusses a cross between two important linguistic groups: the Arawak, represented by Wapishana individuals, and the Carib, represented by Macusht. The dimensions, as recorded by Braulino (1929), for 5 adult hybrid females and com- pared with pure Wapishana and pure Macushi females, are given in table 2. TABLE 2.—Comparative measurements of WapishanaX Macushi cross and pure Wapishana and pure Macushi (from Braulino, 1929) Adult Adult Adult female (5) Dimension Wapishana female female ry 7 = | j *> H : +) é 4 @ BRAZILIAN MESTIZO TYPES By Maria Jutia PourcHet Since the second half of the 16th century, intermarriage has been going on between representatives of the three basic races of Brazil: the Portuguese colonizers and later on Spanish, Italian, and German elements; the aborigines; and the Negroes, who were subsequently imported as slaves. An intense process of interbreeding, due to this peculiar ethnic set-up, is creating the most varied types of Mestizos, since the proportions in which the different ethnic strains were crossed are also varied. An extensive sociological bibliography was started by Euclides da Cunha and was continued by Gilberto Freyre and disciples of these two great masters. It contains accurate studies of temperament, affinities, tendencies, practices, customs, and traditions of those products of interbreeding. Studies of physical anthropology, how- ever, have received less attention despite the enormous variety of physical types which challenges investigators. What appears imme- diately obvious is that the Mestizo population shows no signs of degeneracy, and what was unjustly attributed to miscegenation has been proved beyond a shadow of a doubt to have been determined by causes linked with social factors which today have been fortunately recognized and duly evaluated. The crossing of Black and White is closely linked with the economic- social formation of Brazil and first received the attention of anthro- pologists. The crossing of White and Indian which preceded it began only recently to be the object of more accurate investigations and the contributions devoted to its study have been few indeed. S& e Oliveira in Bahia seems to have been the first to have made compara- tive scientific investigations of the physical peculiarities exhibited by the mulatto, cridulo, caboclo, and Whites of Bahia (S4 e Oliveira, 1895). Craniometric features of the different Mestizo groups in Brazil have been recorded, and although the number of crania studied was slight, we must recognize the precision of the anthropometric technique used. A series of general considerations on the problem of interbreeding in his time occupied the attention of the Director of the Faculdade Médica of Bahia, who recorded the existence of various 111 112 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 products of interbreeding, including the Bahia Whites, especially those of the lower social classes, as a type more or less mixed with African or indigenous blood (Roquette-Pinto, 1923). Believing in the destiny of the products of interbreeding, he foresaw, however, a great lapse of time before the firm and perfect union of the characteristic lines takes place; social prejudices and organic aptitudes, under the influence of our climate and other influences will establish the most favorable selection, parallel with the evolutionary develop- ment of the organisms. Such in general are the ideas of the Bahian scholar. The White-Indian-Negro crossing is responsible for the majority of the so-called sertanejo type, which has been clamoring for the earliest possible anthropological study. Although Negro influence was slight in Cearé where exploitation of Indian manpower was more favorable, in other regions of the northeast there was present, on the contrary, a large percentage of Negro blood, especially in the northeast char- acterized by latifundia, the land of the casa grande and the senzala (the big house and the slave quarters). Here, contact between White European colonizers and the colored races was the result of two social factors, the monoculture characteristic of the latifundia and the scarcity of White women among the arrivals (Freyre, 1934, p. 74). Products of interbreeding begin to arise then. They first resulted in equal proportions of White and Negro blood and afterward in the most varied proportions. These Mestizos constituted the favorite material for chroniclers and historians of the period, who regarded the novel reactions of those “‘sui-generis”’ peoples as a spectacle new to their eyes and to their point of view as Whites, Europeans, and civilized men. First of all, it seemed to them that miscegenation affected robustness and physical efficiency, and that it was the cause of a lack of dynamic energy and of disharmony in the physical type. They were probably neither the first nor the last to form snap judg- ments. Our sociologists, however, are proving that deeper causes of a social order linked to diet—causes and not cause—probably produced that first impression. These causes include insufficient and poorly balanced diet and a patriarchal regime based on slavery— deteriorating factors to which might be added the influence of the milieu, those geophysical conditions which modern studies are clarify- ing. A hypopituitary condition linked to a hypoadrenal condition seems to have been determined by the adaptation of glands to the new milieu. The brackish water which the Mestizos drank and the salty food they ate, in other words an excess of chlorides, which determined a low activity of the adrenal cortex, created a rather intense activity (Bastos d’Avila, 1940 a, p. 131). The physical type also was probably strongly marked—a stocky Vol. 6] BRAZILIAN MESTIZO TYPES—POURCHET {13 stature or, if we wished to go further, that frequent occurrence of brevilinear types which we can surmise from the description of the chroniclers of the times: dark complexion, short and thick neck, an almost waistless trunk, and a prominent abdomen. Assuming the responsibility of the milieu for the physical types to be found in the northeast (State of Pernambuco), two investigators made comparative studies of various ethnic groups (white skins [leucodermos], dusky skins [faiodermos], and black skins [melano- dermos]) functioning in three regions: the littoral, the intermediate country, and the hinterland. Unfortunately, the yellow skins, xantodermos (result of White and Indian crossing), were ignored (Ferraz and Lima Junior, 1939, pp. 277-317). According to the techniques of the constitutionalist school, Viola’s method, each one of the groups was studied in each one of the three regions, and the authors arrived at the following general conclusions: In each of the ethnic groups, those who dwelt along the littoral are predominantly longilinear; in the intermediate zone (mata e agreste) the brevilinear varieties predominate; in the hinterland zone the longilinear type again predominates, which may be attributed to greater or lesser thyroid activity and to food rich in or lacking in iodine. The excess of manioc flour, beans, rice, and tortillas is probably responsible for the incidence of brevilinear types in the intermediate zone. Miscegenation was just as great in the State of Bahia, with the White and Black crossing predominating at the beginning. Later on, however, in several parts of the state, there was an increase in Indian-Negro mixed breeds (the latter had fled from the littoral and taken refuge in quilombos), or the crossing of caboclos with Negroes, giving rise to cafusos, in whom the proportions of one blood and the other are variable. During our investigations with colored children in Bahia, we used the genealogical method for the first time. In a regular number of cases we discovered the mixture of three bloods: White, Black, and Indian. We did not have enough time to analyze the influence of Indian blood in the distribution of morphological traces. Therefore, we restricted ourselves exclusively to the White-Negro crossing. According to the personal testimony of several investigators (Artur Ramos, Edison Carneiro), cafusos (Negro-Indian) are frequently found in cities of the interior. In the central state of Minas Gerais, contact between Whites and Mestizos has been increased; the latter, in turn, resulting from previous crossings (Senna, 1922, p. 208). This has been going on since the penetration of the bandeirantes and the sertanistas paulistanos from south to north and before them, the entradas coming from the Bahian 114 SOUTH AMERICAN INDIANS [B. A. BE. Bull. 143 coast and going east and northeast. The crossing and recrossing was intense: the Paulista, product of Portuguese-Spanish-Indian crossing, the mameluco (crossing of White and Indian), the curiboca or cafuso (crossing of indigenous blood with African). The history of the settling (povoamento) of Minas Gerais, which is intimately linked with the history of the Paulista people, was built around this amalgamation (caldeamento). In the 17th and 18th centuries, the people of Minas Gerais was already composed of a ‘‘strong and helter-skelter ethic mixture”: Whites, Mestizos, Indians, and Negroes. Among the Mestizos there were mulattos, cabras, bastards, curibocas, caboclos, and pardos, resulting from Negro-White, Indian-White, and Indian-Negro crossings. But miscegenation con- tinues and the three bloods flow together forming the most varied series of products of crossing for which the above cited author demands anthropophysical and social study. In the State of Minas Gerais they are called by a great variety of names: capidus, biribas, tabaréus, peoes, jaguncos, etc. With the exception of several general observations, we do not know of any contribution to the physical anthropology of these peoples, even though that lacuna is making itself strongly felt. The principal fields deserving the study of race crossings are the types in wooded regions (matas) of the Rio Doce and in the bush (brenhas) of the Mucury and Itambacary Rivers, the gorutubano cowboys (vaqueiros) of the Nortista sertdes of the Jahyba River, the wiry and muscular sertanejos of the banks of the Sao Francisco and Jequitinhonha Rivers, and the ox drivers (boiadeiros) of the west and of the Triangle between the Paranahyba and Grande Rivers (Senna, 1922, pp. 222-226). Escaped Negroes in 1721 and 1725 who had rebelled against their overseer (dominador) sought refuge in quilombos in the neighborhood of native tribes. As far as we know, the latter were the Aimoré tribes of the Mucury River and the Botocudo of the Doce River, with whom they intermarried and created cafusos. The anthropological charac- teristics of those mixed types were, according to the description of chroniclers of the period, more or less the following: Almost black, dark coloring of the epidermis, copper color or coffee-brown; small though narrow feet, muscular, especially in the chest and the upper limbs; more Negroid than Indian in appearance; oval face with promi- nent cheek bones; broad nose; lips not thick; black eyes; characteristic high frizzled hair, as though it had been artificially raised, reminiscent of the Papuans of New Guinea (Roquette-Pinto, 1915, p. 51). The Euro-American crossing which has been going on in the Paulista Plateau since the middle of the 16th century, giving rise to the first generation of mamelucos, is closely bound with the history of Vol. 6] BRAZILIAN MESTIZO TYPES—POURCHET 115 the bandeirante movement. The history of bandeirantismo appears to to be closely linked with the mamelucos and their incomparable attributes of “great fecundity, magnificent longevity, and amazing virility.” Relations of cause and effect between bandeira and interbreeding are being brilliantly debated by our historians, and according to them interbreeding operated in the following manner (Ellis, 1936, p. 53): (1) Unions legalized by marriage between a White man and an Indian woman, or between the latter and a mameluco. (2) Illegal union between the White or the mameluco and the Indian woman, equivalent to American common-law marriage. (3) Fortuitous and accidental unions between the White or the mameluco and the Indian woman, who was sometimes secretly sold. These unions produced an immense number of bastards. Concerning the fecundity and longevity of the products of Ibero- American crossing, which took place in the 16th and 17th centuries on the Paulista Plateau, documents of the period which patient historians are collecting treat them expressively (Ellis, 1936, pp. 77-83) There is a terrific dispute raging among historians of bandeirantismo concerning the presence of Negroes in the bandeirante movement. Historical reasons, in the light of reliable documents, lead us to believe that ‘‘whites, Indians, and Negroes all participated in the bandeira” (Ricardo, 1942, 2: 61). The bandeira, the first collective product of Euro-Amerindian interbreeding, probably gave rise after- ward to other ethnic crossings: cafusos, mulattoes, and mamelucos were probably the results. Many a cafuso was probably the result of marriage between Negroes and Indians. Some of these unions were legally sanctioned by the action of the Jesuit missionaries; others, illicitly, were the consequence of the rape of Indian women by Negroes who fled from the bandeiras. The existence of cafusos, cabras, mamelucos, forros Indians, and Negroes is attributed to the amazing and rapid penetration of the sertées by the bandeirantes in the south and the criadores in the north (Oliveira Vianna, 1938, p. 93). The bandeirante movement and the conquest of Amazonas were the accomplishments of Mestizos. Among the contributions to the physical anthropological study of our present Mestizo population the most important are those carried on by the Servigo de Satide of the Army and by the Labora- torio de Antropologia of the Museu Nacional. Among the former must be reckoned the studies of Murillo Campos, Romeiro da Roza, and Arthur Lobo da Silva. The latter are linked with the name of Roquette-Pinto, who greatly stimulated anthropological studies in Brazil. In all of these contributions we shall analyze the data referring 794711509 116 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 to the different ethnic groups, point out the differences, and sum up the conclusions of each investigator. In studying different ethnic groups—Whites, caboclos, Mestizos, and Blacks—Murillo de Campos expresses the opinion that distinc- tions between curiboca or cafuso Mestizos, mamelucos, and mulattoes are becoming more and more difficult to establish because a careful study frequently reveals stigmata of the three races in the same individual (Campos, 1919, vols. 9-12, pp. 1 ff.). His investigation of cephalic, facial, and nasal indices led to the following results: Ethnic group: Cephalic index Nasal index Facial index FG Fe Alpe a a nll af Pu 78. 9 50. 3 92. 8 Mestizon.\2.. Sassen eee 79. 0 66. 7 95. 9 Caboclogs2 22 eee 83. 7 61.1 90. 9 Blagg = 2h er ieee ee eae 79. 7 70. 7 88. 8 For height, chest measurement, weight, and Pignet index the results encountered in the four groups are as follows: Height Weight Chest measurement Pignet index Ethnic group: (em.) (grs.) (mm.) Withers oo ocr ees ee te 166 56, 300 816 28. 5 IMICSTIZOB es =e eres eee are 165 57, 100 824 25. 1 Cahocloss 22 os. OS 162 56, 200 821 23. 5 IBlnckse uit s Paee L oe. ee 167 63, 000 817 27. 1 Data were also collected referring to the average shoulder and pelvic girdle measurements and to the dynamometric force. In conclusion Campos (1919) states that the Brazilian Mestizo is in general the result of the fusion of, three races and that his supposed inferiority is a problem for preventive and socialmedicine rather than of racial genetics proper. A regional investigation conducted with a group selected from the wooded zone (Minas Gerais) based upon 290 individuals, most of whom were 21 years old, yields data concerning height,! chest measurement, and weight (Romeiro da Roza, 1920, pp. 53-61). Under the heading ‘Mestizos’” were included all the products of crossing in whom it was difficult to ascertain racial origins, although the author does not have the slightest doubt in affirming that there was a high percentage of individuals who had aboriginal ancestors. The results for the different groups were as follows: Chest Height Weight measurement Pignet (mm.) i Ethnic group: (mm.) (grs.) index Wie Ses oe oe ee 1, 670 55, 100 863 25. 6 MiestizO See fete Sk eee be 1, 656 55, 300 861 24, 2 Bac kt oth iets cae eae Oe ae Ss 1, 681 59, 000 866 22. 5 Proportionate to their height, the Mestizos had the finest chest measurements. Vol. 6] BRAZILIAN MESTIZO TYPES—POURCHET 117 In regard to the Pignet index an important discovery was made according to which the Mestizos had a higher average than the Whites, a fact which is pointed out by the author as disproving the bruited physical inferiority of the Mestizo. In an excerpt from the Report of the Secgao de Antropologia of the Museu Nacional, Roquette-Pinto presents data concerning 600 females and 1,127 males, from every State in Brazil and from 21 to 25 years of age. Each ethnic group was studied from the point of view of regional distribution, including three zones: the northern states, the central states, and the southern states. For the caboclos the regional variation was: Cc 3 Height Cephalic Nasal aboclos: (mm.) index index Southern@e 222 ene ee eee 1, 690 79. 36 70. 00 Centrale cee et oar oak ee 1, 656 80. 10 66. 03 {ig [lovig FLO) 20 Ua Ree ie RW Se a 1, 633 83. 15 66. 66 In our opinion, the differences revealed are quite important and a physical anthropological study of the indigenous groups of the region would clarify the matter (Roquette-Pinto, 1923, p. 30). To sum it up, we will analyze the two most important investigations made to date, which, by a happy coincidence, constitute all of volume 30 of the Arquivos do Museu Nacional (Lobo da Silva, 1928; Roquette- Pinto, 1928). The first and main work contains original abundant data—38,675 files on young men from 20 to 22 years of age, gathered in all the States of the Brazilian Federation. In addition, the files were divided according to region: the States were separated into three groups, taking into account not only the geographical position but certain analogies between them as well. The first group includes the States of the interior without a coast line and localized on the central plateau of Brazil; the second is the northern group from Amazonas to Bahia inclusive; the third extends from Espirito Santo to Rio Grande do Sul. The caboclos presented the greatest variation in height, greater in the south than in the north. The caboclos had the greatest chest measurements, the Mestizos being next. In regional distribution, the northern caboclos were more favored than the others. Despite its complete lack of biological significance, the Pignet index was the only one used. The caboclos had the best indices, as Romeiro da Roza has already discovered in his investigation with individuals selected from the forest region. Finally, we shall comment upon the work of Roquette-Pinto which has been incorporated in the now classic work, ‘‘Nota sobre os ytpos antropologicos do Brasil.’”? This Brazilian scholar, who has been con- cerned for a long time with the problems of our peoples, uses data 118 SOUTH AMERICAN INDIANS [B. A. E. Bull. 143 collected for about 20 years. Because of accusations launched against our Mestizos from Aryan sources, he always takes pains to verify “whether their anthropological characteristics show signs of anatomical or physiological decadence.”” The data were gathered from young men hailing from every State, sons and grandsons of Brazilians, all healthy and subject to the same living conditions, and for this collec- tion of data Martin’s technique was used. For this study we will select for analysis the two groups resulting from crossing, dusky skins (White and Negro) and yellow skins (White and Indian). Unfortunately, the other products of crossing, cafusos and caborés, were numerically unimportant and were not analyzed; according to calculations made by the Museu Nacional in 1922, the percentage of mulattoes and caboclos was, respectively, 22 percent and 11 percent of the Brazilian population. The anthropometric data considered were: Weight; height; facial height; cephalic and nasal indices; color of the skin, of the hair, and of the eyes; and type of hair. For Brazilian dusky skins (Negro- White crossing), Roquette-Pinto found that they were predomi- nantly around 1.64 m. tall and less frequently around 1.73 m. The cephalic index clusters around 78, which is mesocephalic. The nasal index is predominantly leptorrhine. To sum it up: The mulattos of Brazil form a group which is not homogeneous. Among them there is a marked tendency toward the White race which many of them approach by different characteristics. None of the characteristics studied (height, cephalic and nasal indices, chest measurement, length of the face, bizygomatic breadth, life span) lead us to consider them as involuted types. [Roquette-Pinto, 1928.] In conclusion, dusky skinned Brazilians may be characterized as follows: Individuals of brown skin, more or less dark (Nos. 20 to 30 on the Von Luschan scale), dark eyes (black or brown) and ulotrichous hair; medium height; meso- cephalic, mesorrhine, narrow face. [Roquette-Pinto, 1928.] In the xanthodermic group (resulting from White-Indian crossing), height clusters around 2 points, 1.63 m. and 1.69 m., visibly veering toward the latter. The Tupi, Arawak, and Carib of the far north, some Ge from the south of Bahia and from Minas Gerais made the greatest contributions to those crossed groups whose height varies from 158 cms. to 164 cms. The Bororo (173 cms.), Carajé (168 cms.), and Nahukwa (168 cms.) had little influence in the crossing. White blood seems to be respon- sible for the relatively tall height of some of our caboclos. A leptorrhine nasal index indicates a strong White influence. Strong brachyfacial characteristics predominated. Vol. 6] BRAZILIAN MESTIZO TYPES—POURCHET 119 In this, as in other investigations, the relatively high chest measure- mept is noted. This characteristic gives our caboclos an appearance of physical robustness which so many chroniclers have remarked upon. The brachycephalic index, which is quite homogeneous, is higher than that of the White skins. Roquette-Pinto made a résumé of the following occurrence of characteristics in those products of crossing: Skin 20 to 30 on the Von Luschan scale; black hair, lissotrichous; dark eyes, at times with the palpebral fold slightly oblique; short, broad face; medium or low stature; brachycephalic; leptorrhine or mesorrhine. After analyzing the physical anthropological data of the different groups, the Brazilian investigator Freyre arrives at the following general conclusions: From the physiological point of view the investigations proved that inter- marriage between White and Negro, and White and Indian always result in normal types, unless the progenitors are carriers of morbid heredity. The habit of considering Mestizos who are only ill or disgenicos as degenerate has been com- mon among those who are violently opposed to miscegenation. The fact is that this confusion in attributing responsibility must be removed. All we need is to recall the case of Ceard, where the greatest number of intermarriages took place between white skins and yellow skins. Th2 vitality of the races was in no way affected by the crossings. The physical resistance and moral vigor of the conquerors of Amazonia (northeastern Mestizos) who overcame every obstacle is absolute proof. We already referred to this when comparing it with the tenacity of the bandeirantes. Everything leads us to believe that miscegenation was a valuable contributing factor in the formation of the Brazilian, creating that ideal type of the modern man for the Tropics, the European with Negro or Indian blood to revive his energy. [Freyre, 1934, p. 74.] Despite the small number of anthropological investigations con- cerning the present Mestizo Brazilian peoples (in the present work we considered by preference the results of White and Indian cross- ing), neither here nor anywhere else did intermarriage cause degen- eracy save in those cases where unfavorable individual conditions entered into the picture. The evaluation of a human group should be made in the light of their achievements, and in Brazil, if nothing else were to speak in its favor, the two great sertanista movements, the conquest of Ama- zonia and the bandeirante penetration would prove that miscegena- tion has been advantageous rather than prejudicial. More accurate studies should be made in the meanwhile falling into the three following categories: (1) The evaluation of variability of the morphological traces of groups resulting from crossing. (2) A constant observation of the reactions of the present Mestizo peoples of Brazil. (3) The adoption of a genealogical method in certain regions wherever it is possible to do so, particularly in order to throw 120 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 light on the little-known subject of crossings of White with Indian, resulting in caboclos, and of Negro with Indian, resulting in cafusos. The study of the present Brazilian Mestizo peoples is therefore deserving of accurate investigation using the most modern scientific methods. BIBLIOGRAPHY Bastos d’Avila, 1940 a, 1940 b, ms.; Campos, 1919; Ellis, Jr., 1936; Ferraz and Lima Junior, 1939; Freyre, 1934; Lima Junior (see Ferraz and Lima Junior, 1939); Lobo da Silva, 1928; Oliveira Vianna, 1938; Ricardo, 1942; Romeiro da Roza, 1920; Roquette-Pinto, 1915, 1923, 1928; SA e Oliveira, 1895; Senna, 1922. See also the following: Lowrie, 1938; Pourchet, 1939; Roquette-Pinto, 1940. References on growth and constitution: Bastos d’Avila, 1940 b; Bastos d’ Avila et al., 1937; Bezerra, 1940; Castro, 1939; Juliani, 1939; Lacaz de Moraes, 1939; Lima Junior and Ignacio, 1939; Pourchet, 1942; Roquette-Pinto, 1942; Sussini et al, 1937; Thomaz, 1942. THE PHYSICAL ANTHROPOLOGY OF CHILE THE ANTHROPOMETRY OF THE INDIANS OF CHILE By Cartos HENCKEL The present study presents a brief summary of the physical an- thropology of the Indians of Chile who survive as ethnic groups at the present day. It does not take into account the Chilean Indians who have disappeared, either through becoming mixed or absorbed in the White population. The extreme south of the continent of South America was occupied by three ethnic groups, the Ona (Shelknam), the Yahgan (Yamana), and the Alacaluf (Halakwulup).! THE ONA (SHELKNAM) In pre-Columbian times the Ona occupied the large island of Tierra del Fuego but today are reduced to a very few individuals who live around Lake Fagnani, some of them in the Salesian Mission of Rio Grande. They numbered 276 persons in 1919 (Gusinde, 1939), 110 in 1931 (Rahm, 1931 a),*and?scarcely three dozen in 1938 (Gusinde, 1939). SOMATOLOGY Numerous observations have been made of the physical appearance of the Ona (Sarmiento de Gamboa, 1768; Banks, 1896; Darwin, 1875; Fitz-Roy, 1839; Serrano, R., 1880; Lista, 1887; Segers, 1891; Nor- denskidld, 1896; Wieghardt, 1896; Cook, F. A., 1900; Outes, 1909; Gallardo, 1910; Canas, 1911 b; Dabbene, 1911; Furlong, C. W., 1917 a; Barclay, 1926; Lothrop, 1928; Serrano, A., 1930; but complete studies using modern methods have been made only by Lahille (1926), Lehmann-Nitsche (1916 b, 1927), and Gusinde (1922 a, 1922 b, 1922 ¢c, 1924, 1926 a, 1932, 1939), who with Lebzelter followed Martin’s an- thropometric techniques (Martin, 1928). Early writers called attention to the tallness of the Ona. Lehmann- Nitsche found that 20 men had a mean stature of 174.1 cm. (168.0- 183.7) and 30 women of 159.6 cm. (149.0—-168.3). Gusinde’s results were similar: 24 men, 172.9 cm. (163.1-180.9), and 22 women, 160.3 em. (153.7-166.9). 1 See Cooper, 1917 a, for a complete bibliography of the physical anthropology and ethnography of these Fuegian tribes. 121 122 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 The Ona head is large and rugged (pl. 24, top, left, and bottom, right). The horizontal circumference is considerable: men, 590 cm.; women, 574 cm. (Gusinde); or men, 583.5 cm.; women 563.7 cm. (Lehmann-Nitsche). The maximum longitudinal diameter of the head is great: men, 199 mm.; women, 190 mm. (Gusinde); so is the maximum transverse diameter: men, 158 mm.; women, 152 mm. The average horizontal cephalic index is 78.8 for men, 79.8 for women (Gusinde); or 79.6 for men, 80.5 for women (Lehmann-Nitsche). This is mesocephaly approaching the lower limit of brachycephaly. The face is oval, the cheek bones wide and prominent. The average morphological height of the face is: men, 124 mm.; women, 117 mm. The bizygomatic breadth has a mean of 150 mm. for men, 142 for women. The facial index shows euryprosopy: men, 82.7; women, 82.2 (Gusinde). The nose is large, straight, sometimes slightly aquiline, and leptor- rhinic. The mean nasal index is: men, 67.5; women, 68.8 (Gusinde). The forehead is low owing to the down growth of the hair toward the orbits. The eye opening is small, somewhat oblique, and almond- shaped. All the Ona have a transverse fold of the upper eyelid (plica palpebralis media) and most of them a marginal fold (plica naso- marginalis). The eye color almost always corresponds to number 2 or 3 of R. Martin’s color chart. The mouth opening is large and the lips are thin. The naso-labial furrow is well marked, especially in persons of some age. The chin is massive and there is always a furrow between it and the lower lip. According to Gusinde’s detailed description of the ear (1926), in the lower half of the external ear the helix is outstanding for its con- siderable thickness. The tragus and antitragus are especially massive. The neck is regular, and the trunk is broad and strong (pl. 24, bottom, left.) The biacromial width has a mean of 438 mm. in men, 403 mm. in women (Gusinde). The length of the trunk has a mean value of 533 mm. in men, 508 mm. in women (Gusinde). Due to the great development of the thorax, the circumference of the chest is great; Gusinde found an average of 100.8 cm. in 15 men. The arms are strong, well formed, and muscular. The average length of the arm is 783 mm. for men, 692 mm. for women (Lehmann- Nitsche). The hands are small and well formed (Lehmann-Nitsche, 1904 d). The legs are thin in the calf, thick in the thigh. The total average length of the leg is 921 mm. or 53.2 percent of the total stature in men, 849 mm. or 52 percent in women. The feet are small and well formed. Vol. 6] ANTHROPOMETRY—HENCKEL 123 The skin is a brownish yellow, relatively clear, and corresponds to Nos. 10 and 11 of Von Luschan’s color chart. The hair is dark and blackish. The hairs are straight, strong and thick. (For details, see Saller, 1939.) The beard has very few hairs. Hairs are extremely sparse in the armpits and in the pubic region. CRANIOLOGY AND OSTEOLOGY For the craniology of the Ona see Hultkrantz (1907), Lebzelter (1925), Hilden (1930), and Gusinde (1939). According to Gusinde, who has studied the greatest amount of material to date, the average cranial capacity is 1,480 cc. in males and 1,356 cc. in females. These figures correspond to the aristencephaly of Sarasin. The greatest cranial length (pl. 25) is 192 mm. in males, 184 mm. in females; the width, 143 mm. and 137 mm., respectively. The hori- zontal cranial index has a mean of 74.5 for males, 74.9 for females (Gusinde), indicating dolichocrany approaching mesocrany. The mean basio-bregmatic height is 136 mm. for males, 135 mm. for fe- males (Gusinde). The auricular height is 116 mm. and 114 mm., respectively. The vertico-longitudinal index is 70.9 and 73.4, re- spectively, showing medium orthocrany. To judge by Gusinde’s median values, the Ona may be classified as follows: Metrio- to acrocranic, according to the vertico-transverse index (men, 95.1; women, 98.7); orthocranic, according to the index of auricular height in relation to cranial length (males, 60.3; females, 62.0); steno- to metriometopic, according to the fronto-parietal index (males, 65.5; females, 67.4). The average angle of frontal inclination of the Ona is only 48°. The forehead is low and retreating. There is an almost complete absence of lateral frontal protuberances. A supraorbital torus is common; also a sagittal crest (lophocephaly). The muscular relief of the occiput is well marked, and in most cases amounts to a torus. Mean facial dimensions of the skull (Gusinde) are: Morphological height of the face, males, 126 mm., females, 127 mm.; bizygomatic width, males, 144 mm., females, 138 mm.; facial index, 87.2 and 91.8, indicating meso- to leptoprosopy; superior facial index, males, 52.7, females, 55.3, indicating meseny or a slight lepteny. The angle of total profile in skulls of both sexes is 84°, which is near the upper limit of mesognathy: and close to orthognathy. The orbital index, 92.4 for males and 82.3 for females, shows that most skulls are mesoconchic. The nasal index shows that males are leptorrhinic (45.9) and females slightly mesorrhinic (47.1). The mean palatal index, 76.6 for males, 80.1 for females, is leptostaphylinic or slightly mesostaphylinic (Gusinde). All skulls show a palatal torus. 124 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 The mandible is very massive, the mean bicondylar width being 127 mm. for men, 122 mm. for women. Ona teeth are considerably worn, but caries occurs only rarely. For other Ona skeletal characteristics, see Hultkrantz (1907). THE YAHGAN (YAMANA) The Yahgan lived in the region included between Slogget Bay on the point southeast of Isla Grande, the Peninsula of Brecknock, and Cape Hornos. Today, only some 24 persons survive, living in the village of Mejillones on the Island of Navarino (Gusinde, 1939). SOMATOLOGY There are many descriptions and observations concerning Yahgan somatology: Abel (1934), Bove (1882), Bridges (1893), Colvocoresses (1852), Darwin (1875), Forster (1843), Fitz-Roy (1839), Gusinde (1939), King (1839), Hyades and Deniker (1891), Lehmann-Nitsche (1916 c), Martial (1888), Saller (1939), Spegazzini (1882), Snow (1857), Webster (1834), Weddell (1827), and Wilkes (1844). The stature of the Yahgan is relatively short. Various mean heights recorded are: Men Women Author US ieee ee LE ae Sd Be Bove, 1883. 1h iN ai ag ee fe 1484s Der Me ee Bove, 1882-83. WSs oon see 40ers oe ce Martial, 1888. 1S Bs ae ame ees ena ne, fe Ue Sed a ae ene Hahn, 1883. 17:7 jib ORO a Se ae iE: 17) See Ee fae ee Hyades and Deniker, 1891. TGOs Se esate LC Ly (a San A ee Gusinde, 1939. CB re ee cs ee TAO Mere AeA ise Dabbene, 1911. The head is large and of great capacity (pl. 24, top, right). The mean maximum head length is 190 mm. for men, 180 mm. for women (Hyades and Deniker, 1891), or 197 mm. and 186 mm., respectively (Gusinde, 1939). The greatest mean width of the head is 151 mm. for men, 142 mm. for women (Hyades and Deniker); or 155 mm. and 150 mm., respectively (Gusinde). The horizontal cephalic index is 79.6 for men, 79.1 for women (Hyades and Deniker), or 78.6 and 80.8'(Gusinde). The Yahgan are thus mainly mesocephalic, with a slight tendency to brachycephaly. The face (pl. 27, top, left) is generally ovaloid, somestimes round, but always angular with prominent cheekbones. The morphological height of the face has a mean value of 118 mm. in men, 116 mm. in women; the bizygomatic width has a mean of 150 mm. for men, 141 mm.forwomen. The facial index of 78.7 for men is hypereuryprosopic and 82.4 for women is euryprosopic. The superior facial index of 48.2 for men and 50.1 for women indicates meseny for both sexes (Gusinde, 1939). Vol. 6] ANTHROPOMETRY—HENCKEL 125 The Yahgan forehead is low, narrow, and retreating. The super- ciliary ridges are well developed, the frontal torus being conspicuous. The front limit of the head hair is very near the orbits. The nose is relatively narrow between the eyes, while farther down it is wide. The nasal profile is generally concave, and in rare cases straight. The nasal index, 77.1 for men, 74.2 for women, is mesorrhinic. The eyes are rather small. The opening between the lids is gen- erally spindle-shaped. The upper lid has a transverse fold, and in some cases there is also a marginal fold. The color of the iris corre- sponds to number 2 or 3 on Martin’s color chart. The mouth is large and the lips generally medium thick, but there are individuals both with thin and very thick lips. The upper lip almost always protrudes. The chin is quite massive. The neck is short and thick. The body is massive and cylindrical, and in adults almost without a waistline. The mean trunk length is 490 mm. for men, 472 mm. for women, and the biacromial width is 392 mm. and 371 mm., respectively (Gusinde). The circumference of the thorax is considerable: a mean of 92.3 cm. for men, 85.8 cm. for women (Hyades and Deniker). The mean ratio of the circum- ference of the thorax to the stature is 58.7 percent for the two sexes. The arms are well formed and muscular. The mean total length is 72.5 em. for men, 65.0 cm. for women (Gusinde). The hands are small. The legs are very weakly developed. The total leg length has a mean value of 85.2 cm. for men, 74.9 cm. for women, which is 53.2 percent and 51.2 percent, respectively, of the total stature (Gusinde). The ‘‘atrophy”’ of the legs, of which Hyades and Deniker speak, is caused by the custom of spending a great deal of time in canoes, paddling or fishing, and is not an hereditary characteristic (Gusinde). The skin color is somewhat dark. In most individuals, it is clearer than Nos. 10, 11, and 12 of Von Luschan’s color chart, but these tints are also observed. The hair is abundant, stiff, and generally straight, although wavy hair is seen on some persons. According to Spegazzini (1882), the Yahgan have no body hair, and only the old men have a few hairs on the upper lip and chin. In the armpits and pubic region there are very few hairs. The hair color is very black. (For details, see Saller, 1939.) CRANIOLOGY AND OSTEOLOGY Yahgan craniology has been dealt with by Garson (1885), Gusinde (1939), Hilden (1930, 1931), Hultkrantz (1907), Hyades and Deniker (1891), Ten Kate (1904), Mantegazza and Regalia (1886), Owen (1853), G. Sergi (1887), and Vignati (1927 b). The following measure- 126 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 ments are mean values from Gusinde, who has examined the greatest amount of Fuegian cranial material to date: The cranial capacity of males is 1,432 cc. and of females is 1,290 cc., indicating euencephaly (Sarasin). The greatest skull length is 186 mm. for males, 177 mm. for females (pl. 26, top), which is considerably less than for the Ona, whereas the greatest width is scarcely different, being 142 mm. for males, 136 mm.forfemales. The horizontal cranial index is 76.6 for male skulls, 77.0 for female, which is mesocranic and approximately two points higher than for the Ona. The basio-bregmatic height reaches 136 mm. in men, 131 mm. in women, and the auricular height 115 mm. and 112 mm., respectively. The vertico-longitudinal index is 73.6 and 74.1, corresponding to mesocrany. According to the vertico-transverse index of 96.1 for males, 96.3 for females, the Yahgan are orthocranic. The auricular height- length index, 62.1 for males, 63.3 for females, is between the limits of ortho- and hypsicrany. The fronto-parietal index, 67.4 for males, 67.5 for females, is metriometopic. The angle of frontal inclination in the skulls is 49° in males, 50° in females, and the forehead is very retreating. The lateral frontal protuberances are hardly noticeable, but frequently there is a frontal torus. Parietal eminences are present only in a very few cases, but a sagittal crest (lophocephaly) is frequently observed. The occipital bone is flat in its cerebral portion, which forms almost a right angle with its cerebellar portion (Hyades and Deniker). There is often an occipital torus. The morphological height of the face is 121 mm. for males, 111 mm. for females. The bizygomatic width is 142 mm. for males, 131 mm. for females. The facial index is 84.7 and 83.9, respectively, indicating euryprosopy. The superior facial index is 51.7 and 50.8, respectively, showing meseny. The angle of total profile is mesognathic, being 82° in males, 80° in females. The orbits of the Yahgan are mesoconchic, the index being 79.5 for males, 80.2 for females. The nose is mesorrhinic, the index being 47.9 in males, 47.3 in females. The palate is leptosta- phylinic, with an index of 74.8 for males, 71.7 for females. A palatal torus is always present, though sometimes weakly developed. The mandible is large and massive. The bicondylar width, 123 mm. in males, 116 mm. in females, is somewhat less than that of the Ona. Yahgan dentition is similar to that of the Ona. Hyades and Deniker noted that the permanent teeth appear earlier than in Whites. Osteometric observations on the other bones of the skeleton are contained in Garson (1885), G. Sergi (1887, 1888), Hyades and Vol. 6] ANTHROPOMETRY—HENCKEL 127 Deniker (1891), Hultkrantz (1907), S. Sergi (1928), Delle Seta (1938), Genna (1928, 1930-32), Jazzeta (1928), Pastore (1935-37), Scolni de Kliman (1938), and Sabatini (1933-34). THE ALACALUF (HALAKWULUP) The Alacaluf inhabit the islands, beaches, and channels from the Gulf of Penas south to the northwestern portions of Isla Grande in Tierra del Fuego. They number less than 90 persons (Gusinde, 1939). SOMATOLOGY Somatological studies and observations are contained in Bischoff (1882 a, 1882 b), Boehr (1881), Essendorfer (1880), Fitz-Roy (1839), Gusinde (1939), King (1839), Hyades and Deniker (1891), Lehmann- Nitsche (1916 d), Martin (1893-94), Manouvrier (1881), Outes (1909), Seitz (1833), Skottsberg (1910), and Virchow (1881). The Alacaluf stature is small, the mean values according to five authors varying between 151.0 cm. and 161.4 cm. for men and between 143.2 and 152.2 for women. Gusinde, who as usual made the greatest number of observations (15 men, 16 women), gives an average of 154.7 cm. for men and 143.2 cm. for women. The greatest head length has a mean of 192 mm. for men, 183 mm. for women, and the greatest head width is 148 and 143, respectively. The horizontal cephalic index has a mean of 77.4 for men, 78.2 for women, which makes the Alacaluf mesocephalic (Gusinde). The face is oval or round, and, especially in women, it is flat. (See plate 27, top, right, and center.) The cheek bones are not very promi- nent. The average morphological height of the face is 120 mm. in men, 108 mm. in women. The bizygomatic width is 140 and 131 mm., respectively. The face is mesoprosopic in men (average index, 85.6) and euryprosopic in women (average index, 82.3) (Gusinde). The nose is triangular as seen from the front; and either straight or somewhat concave from the side, though sometimes slightly convex. The nose is mesorrhinic, but close to the limit of leptorrhiny, the average index being 72.2 for men, 70.6 for women (Gusinde). The forehead is low because the hair grows down toward the supra- orbital ridges. The eye opening is generally fusiform. In almost all persons of this group there is a transverse fold on the upper eyelid (plica palpebralis media). A true Mongolian fold is not found among the Alacaluf. The color of the eye is dark brown, corresponding to No. 2 on R. Martin’s color scale. The mouth is large, and the lips generally thick. A slight protru- sion of the upper lip is frequently noted. The naso-labial furrow is well marked. The chin is rounded. 128 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 The body or trunk is 467 mm. long in men, 459 mm. in women. The biacromial width has an average of 387 mm. in men, 349 mm. in women (Gusinde). The thorax is generally almost flat. The arms are well developed, the mean total Jength being 688 mm. in men, 622 mm. in women (Gusinde). The legs are less developed than the arms, and the calves are thin. The authors constantly mention the poorly developed musculature of the legs, and they explain it in the same way as among the Yahgan, that is, the Alacaluf’s habit of spending a great deal of time in their canoes, paddling or fishing. The mean total length of the leg is 814 mm., or 53.0 percent of the stature in men, and 730 mm., or 51.2 per- cent of the stature in women (Gusinde). The feet are large. The Alacaluf skin color is light brown, corresponding to Nos. 13 to 15 on Von Luschan’s color scale (Outes, 1909). The fatty tissue is well developed. Bischoff (1882 a, 1882 b) gives data on the micro- scopic structure of the skin. For fingerprints, see Abel (1934). The hair is abundant and grows low on the forehead. Frequently it grows in on the temples, sometimes being continuous with the eyebrows. The hairs are thick, stiff, and smooth. For the micro- scopic structure of the hairs, see Martin (1893-94) and Saller (1939). The hair color corresponds to Nos. 27 and 4-27 of Fischer’s table. The beard is little developed, and in the armpits and the pubic region the hair is scarce. CRANIOLOGY AND OSTEOLOGY For the craniology of the Alacaluf, see Medina (1882), Martin (1893-94), Mehnert (1893), Garson (1885), Ten Kate (1904), Latcham (1911), and Gusinde (1939). Hoyos (1913) describes a Fuegian skull, but does not give its tribe. The mean cranial capacity of the Alacaluf is 1,530 cc. in males, 1,295 cc. in females (Gusinde). Unfortunately, these mean values were determined on the basis of only 13 skulls. According to Gusinde, the greatest skull length (see pl. 26, bottom) is 191 mm. in males, 180 mm. in females; the greatest width, 141 mm. and 138 mm., respectively. The mean horizontal cranial index is 74.6 for males, 76.7 for females, that is, dolichocranic for males, mesocranic for females. The basio-bregmatic height has a mean of 140 mm. for males, 130 mm. for females; the auricular height, 121 mm. for males, 117 mm. for females. The mean vertico-longitudinal index is 73.2 for males, 71.9 for females, showing orthocrany (Gusinde), To judge by Gusinde’s data, the mean indices of the Alacaluf may be classified as follows: Metrio- to tapeinocranic, according to the vertico-transverse index (97.6 for males, 91.5 for females); hypsicranic, according to the auricular height-length index (63.6 for males, 64.7 Vol. 6] ANTHROPOMETRY—HENCKEL 129 for females); metrio— to stenometopic, according to the fronto- parietal index (67.9 for males, 65.5 for females). The medium angle of inclination of the forehead is, for both sexes, 52°. The forehead is low and sloping. The lateral frontal protuber- ances are little developed. The glabella is only slightly rounded; usually it corresponds to No. 2 on Broca’s scheme (Martin). The Alacaluf skulls frequently have strongly developed superciliary ridges and a frontal torus. Parietal eminences are little developed and appear only in isolated cases. Lophocephaly is very frequent. The occipital bone often has a torus, and the muscular relief is always well marked. The mean morphological height of the face is 125 mm. in males, 113 mm. in females; the bizygomatic width is 143 mm. and 131 mm. respectively. The facial index, 86.5 for men, 85.9 for women, in- dicates mesoprosopy. The mean superior facial index, 51.1 for males, 52.0 for females, indicates meseny (Gusinde). The angle of the total profile for both sexes has a mean of 84°, corresponding to mesognathy (Gusinde). According to the mean orbital index (83.3 for males, 87.1 for females), masculine skulls are mesoconchic, feminine are hipsiconchic. The mean nasal index (46.2 for males, 48.6 for females) shows that males are leptorrhinic, females mesorrhinic. According to the mean palatal index (78.2 for males, 80.9 for females), males are leptosta- phylinic, females mesostaphylinic. The upper dental arch usually has the form of a ‘“U”’. A palatal torus is general and is characteristic in these skulls. The mandible is very massive, the bicondylar width being 125 mm. in males, 117 mm. in females. The teeth show considerable wear, especially in older persons. Caries occurred only rarely among the Alacaluf while they still re- tained their native culture. Today it is more frequent, owing to the influence of modern civilization. The weight of the Alacaluf skull is considerable, as among all the Fuegians, owing to the thickness of the walls of the cranial vault. For the remainder of the Alacaluf skeleton, see Martin (1892). Unfortunately a work by Vallois (1932) on the Fuegian humerus is not available to me. NATIVES OF THE SOUTHERN PROVINCES In the southern provinces of Chile, from Arauco to Llanquihue, live a considerable number of Araucanians or Mapuche, who are quite distinct from the foregoing tribes. This ethnic group is not homo- geneous physically, as we shall see. 130 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 The Araucanians were estimated by Latcham (1928) to number 120,000 and were increasing rapidly. Brand (1941 a) places their total at 300,000, which is probably somewhat high. The distribution of the Mapuche, according to the 1920 census, is as follows: Province*of-Araucoe. ve ee ee ee ee eae 4, 980 ‘Province:of Bio=Bios. le Yee. 2522 ee eee 1, 372 Proving: of Malleto.. =: 302 25.3532 5 5 11, 815 Province;of Cautine= 352-5. ee fae eee ene eee 58, 305 Province: of: Valdivia 2250 oF 2 te) cp see ee es 19, 723 Province of llanquiltie®) 2252.00 52.5 seer ee ee ee 8, 697 The older chroniclers, including de Olivares, Gémez de Vidaurre, and Molina, as well as the historians, Barros Arafia and others, referred to the physical appearance of the Mapuche. Special works on the physical anthropology of the Mapuche are, however, still very scarce. SOMATOLOGY The stature of the Mapuche, as the following mean figures show (table 1), has considerable variation according to locality. TaBLE 1.—Daia on the stature of the Mapuche (mean values) Males Females Locality Source Stature Number Stature Number (cm.) studied (cm.) studied Cordilleraee sete ese ee oe oN 168. 4 DBs ss 5 eee ae Bary ds Latcham (1911). Sub-Andean region. 220222222252 164.3 il 147.5 6 | Latcham (1911). Central Valleys 82. eee 160. 6 31 143. 2 19 | Latcham (1911). Contral' Valleys oii oo ee ee 162.2 41 143.7 11 | Guevara (1898). Oentralv siioy 22 ese ee ae 159. 2 LOOM) ee ees |e es Matus (1912). CO mS Cee a eee ee ey 168.0 CZ) ea Tae ee Lea ee Latcham (1911). According to these data, stature increases to a maximum in the regions of the cordillera, the habitat of the Pehuenche, and decreases toward the south, between the Toltén and the Gulf of Reloncavi, where the Huilliche live (Guevara, 1898). Schiuble (1939) noted that the inhabitants of the Coast are shorter than those of the Central Valley. On the other hand, Latcham (1911) considered that a Coastal subtype is taller. The head is generally large, its mean maximum length reaching 180.8 mm. in men (Matus), and its mean maximum width 150.9 mm. The mean horizontal cephalic index in men is, according to Matus’ data, 83.2, indicating brachycephaly. Matus (1912) found the follow- ing distribution of head form among 100 Mapuche: Dolichocephaly, 4 percent; mesocephaly, 14 percent; brachycephaly, 62 percent; hyperbrachycephaly, 20 percent. PLATE 24.—Physical types of Tierra del Fuego. Top, left, Ona woman; top, right, Yahgan type; bottom, left and right, Ona. (Bottom, left, after Pericot, 1936, p. 681; others, after Gusinde, 1939, pls. 1, 2, 3.) PLATE 25. Ona skulls, Tierra del Fuego. (After Gusinde, 1939, pls. 7, 8.) eng Bagh & % Sao? Vinh, PLatE 26.—Skulls from southern Chile. Top, Yahgan; bottom, Alacaluf. (After Gusinde, 1939, pls. 14, 13.) PLaTE 27.—Physical types from southern Chile. Top, left, Yahgan woman. Top, right, Alacaluf type. Center, Alacaluf woman. Bottom, left, Mapuche chief. Bottom, right, Mapuche woman. (After Gusinde, 1939, pls. 4, 5, 6; bottom pictures courtesy Carlos Henckel.) Vol. 6] ANTHROPOMETRY—HENCKEL 131 The face (pl. 27, bottom, left and right) is generally round, though in some it is angular. Seen in profile, the face is notably flat. Atten- tion is called to the prominent and well-developed cheek bones. The forehead is narrow and low. The eye opening is narrow; frequently there is a transverse fold of the upper lid (plica palpebralis media) and a marginal fold. The iris is dark in color, most commonly corre- sponding to the shades Nos. 13, 14, and 15 of Martin’s and Schultz’ chromatic chart. The conjunctiva is somewhat yellowish. For histological details of the eye, see Henckel (1942 a). The nose is generally broad; in profile, it is straight, never aquiline. The mouth is large, the lips generally thick. There is marked progna- thism. The chin is square and somewhat prominent. The neck is short and thick. The thorax is well developed, its mean lateral diameter in men being 31.8 cm. (Matus), and its antero-posterior diameter 22.0 cm. The circumference of the chest has a mean value of 93.0 cm., which is very high if we take into account the short stature of this group (Matus). The lung capacity of the men varies between 4,500 and 5,300 cc. The back is broad in both sexes, and the abdomen is somewhat enlarged, showing a certain propensity to obesity. The arms are short and thick, and the hands are relatively broad and short. The dynamometric hand pressure has been determined in some men by Matus; it averages 49 in the right hand and 47 in the left. The legs are short and, especially in women, the thigh is thick and rounded. The ankle is thick; the calf has not been described. The feet are short and thick. The Mapuche skin color has been characterized by the abbot Molina (1776), in the following words: ‘Although they are the lightest of all American natives of the south, their complexion is some- what olive-color.”” Many individuals have skins no darker than those of southern Europeans. Microscopic observations on the quantity and distribution of cutaneous pigment have been made by Henckel (1941). The Mongoloid spot is found very frequently in the newborn and nursing children; according to Mardones (1937), in 86.7 percent of the cases. The arrangement of the dermal ridges on the hand was studied by the present author (Henckel, 1933 b, 1942 b). In 246 Mapuche, the Galtonian types occur with the following frequency: Arches, 7 percent; loops, 56 percent; and whorls, 37 percent. The hair color is dark, usually corresponding to the shades V, W, and X on the chromatic table of Fischer and Saller. The head hair is very coarse: 94.3 uw average in women, 87.1 » in men. It is also 79471150 10 132 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 smooth and stiff. For its histological characteristics, see Henckel (1941). We must mention certain blond Mapuche, with white skin and light eyes, in Boroa, Cautin Province, mentioned by the old chroniclers, Goémez de Vidaurre and Molina. Guevara and Latcham (1911) regarded them as the result of accidental crosses of Indians and Whites. It is more likely, however, that this was a case of partial albinism, as has been observed in other native American groups. The head hair of the Mapuche is very abundant. It grows low on the forehead and temples. Kretschmer’s “‘fur cap” is quite common (Barrientos, 1942). Sometimes, especially in children, the head hair continues to meet the eyebrows (Schauble, 1939). In rare cases, there is baldness. Gray hair comes late in life (Poeppig, 1942). There is little beard, and body hair is scarce. There are few hairs in the armpits and pubic region, even in individuals well developed sexually (Pi-Sufier and Reyes, 1935). Barrientos (1942) has diagnosed Kretschmer’s constitutional types among the Mapuche as follows: ‘“‘Pyknic,” 46.8 percent; athletic, 18.9 percent; and leptosomic, 34.0 percent. For some other aspects of Mapuche somatology, see Houzé (1884), Latcham (1904), and Manouvrier (1883). CRANIOLOGY The present status of Mapuche craniology has been well character- ized by Brand (1941 a): . . & few small series of Araucanian skeletal material are present in Chile, Argentina, France, Germany, Great Britain, Spain, Scandinavia, Italy, and the United States. However, no museum has an adequate collection and not even one anthropometric index, ratio, or simple measurement has been made on an adequate number of specimens. [Brand 1941 a, p. 33.] Thus, unfortunately, we can give but few data on Mapuche crani- ology. Medina (1882) gives some figures and measurements on seven Araucanian skulls, but without the main details. The horizontal cranial indices of 18 male and female Mapuche from the Central Valley range from 72.2 to 84.4 and have a mean of 80.3 (Guevara, 1898, 1912). These include 3 which are dolichocranic, 3 mesocranic, and 12 brachycranic. In a series of 12 male and female crania from the cordillera, to the east, the same index varies from 70.5 to 85.0, and includes 6 which are dolichocranic, 3 mesocranic, and 3 brachycranic. Guevara concludes that brachycrany predominated on the Coast and in the northern part of the Central Valley and that dolichocrany was increasingly present in the south and in the cor- dillera, to the east. As descriptive craniological characteristics, Guevara mentions the Vol. 6] ANTHROPOMETRY—HENCKEL 133 weight, roughness, narrowness of the forehead, the flatness of the posterior curvature, the prominence of the inion, the well-marked superciliary ridges, the great development of the malar bone, and the width of the lower jaw. Vergara Flores (1902) made a comparative study of the crania of the Araucanians and Aymara. Verneau (1903) examined 6 Mapuche skulls (one male, five female), which came from Mochita graves in Concepcién and are now in the Museum of Natural History in Paris. Their mean horizontal cranial index is 81.5 and 81.4, respectively, indicating brachycrany. Their mean vertico-horizontal index is 79.1 and 77.3, respectively, indicating hypsicrany. Their mean vertico-transverse index is 92.0 and 94.9, respectively, indicating metriocrany. The average facial index for the female skulls is 70.7 (hypereuryprosopy). The mean orbital index is 87.2 and 89.4, respectively, which, according to Broca, is meso- conchic. The mean nasal index is 56.1 and 54.6, respectively, in- dicating chamaerrhiny. The average cranial capacity of the Araucanians is 1,425 ce. for 6 male skulls and 1,355 cc. for two female skulls, according to Quatre- fages and Hamy (cited by Verneau, 1903). Latcham (1911) gives 81.7 as the mean cranial index of 92 male and female Mapuche skulls from the Central Valley. The average cranial capacity for the male skulls is 1,350 cc. and for the female skulls, 1,230 ec. This, in Sarasin’s sense, is euencephaly. The mean height index is 86.6, the Mapuche thus being high-headed (‘‘hypsiacrocephalic’’). The facial skeleton, according to Latcham (1911), has the following characteristics: Bizygomatic diameter, 136 mm. in males, 132 mm. in females; orbital index, 85.2; nasal index, 48.5 (mesorrhinic) ; subnasal profile angle, 76.2° (prognathic). Deniker (cited by Martin, 1928) gives the average Araucanian cephalic index as 83.9 for males and females. Araucanian dentition has been studied morphologically by Mufioz (1936), who gives information on wear, caries, malocclusion, trema, diastema, Carabelli’s tubercle, etc. Although they have not been utilized in the present article, mention should be made of the work done by Ten Kate (1892) among the Argentine Araucanians and the studies made by Puccioni (1912) and Hoyos (1913). THE INDIANS OF NORTH CHILE North of the Rio Loa to the Chilean-Bolivian border are various Aymara Indians whose number Brand estimates (1941 c) at 40,000. The ancient Atacamefio of the Cordillera de Antofagasta and Tarapacdé and the Puna de Atacama (Latcham, 1911; Oliver, 1932) 134 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 have disappeared completely, despite the claim to the contrary by Brand, who confuses them with the present Indians living in the Puna de Atacama. For the last, see Philippi (1860). The Chilean Diaguita, who lived in the southern part of the Province of Atacama and the Province of Coquimbo, have disappeared as an ethnic group, but their physical type has been preserved in the rural population of this region (Brand, 1941 b). In the littoral north of the Rio Loa lived the Uro, who are com- pletely extinct today. According to Latcham (1912), they were brachycephalic. The coast between the Rio Loa and the Rio Choapa was occupied by the Chango, of whom a few descendants survive in the coves between Tocopilla and Taltal, where they are civilized and mixed with Whites (Oyarztin, 1927 b; Oliver, 1932; Brand, 1941 c; see also Handbook, vol. 2, pp. 595-597). Some of their somatological characteristics have been described by D’Orbigny (1839), Philippi (1860), Latcham (1910, 1911, 1926), and Gigoux (1927). Latcham (1911) gives the following description: The Changos were of low stature, 160 cm. for men, 145 cm. for women. The body is large in proportion to the arms and legs; the shoulders wide. But the chest is not so well developed as among the mountain people. The face is wide and the features arerugged. The forehead is not very narrow but it is low and receding. The eyes are small and dark, the nose is narrow at the base but broad at the end, and always straight, never aquiline. The mouth is large, the lips thick and everted. The skin is dark, being brown as burned by the sun and wind and not reddish. The hair is black, stiff, straight, and lusterless, and it grows so low over the temples that the forehead appears narrower than it really is. The mien is somber and sad. [Latcham, 1911.] For the craniology of the Chango, there are works by Latcham (1903, 1904 c, 1912, 1939), Vergara Flores (1905), Fonk (1906, (1912)), Barras de Aragén (1909), and Ibar (1933). Morphological characteristics which have received a great deal of attention are the!thickness of the walls of the skull (an average of 11.65 mm.), especially in the occipital and malar regions (‘skulls with thick walls,” says Vergara), and the great weight. According to Latcham (1912), the Chango skulls are long, ‘‘dolicho- or subdolicho- cephalic,”’ with a tendency to lophocephaly. The greatest width is between the parietal protuberances. The forehead is narrow but not depressed. The face is long, narrow, and somewhat flat. The orbits are rectangular. The palate is very wide, and prognathism is pro- nounced. For a series of 19 Chango skulls from the Rio Loa, Vergara (1905) gives the following mean values: Cranial index of males, 80.2, of Vol. 6] ANTHROPOMETRY—HENCKEL 135 females, 88.4; orbital index for males, 87.4, of females, 91; nasal index of males, 50.6, of females, 52.6. The average cranial capacity of 9 Chango skulls from Paquica is 1,302 cc. (Ibar, 1933). Ibar gives the average cranial index as 74.9, the orbital index as 87.7, and the nasal index as 48.4. According to Ibar, the long bones of the Chango are notable for their thickness and evidence of strong musculature, especially of the arms. BIBLIOGRAPHY Abel, 1934; Banks, 1896; Barclay, 1926; Barras de Aragén, 1909; Barrientos, 1942; Barros Arafia, 1884-1902, vol. 1; Beclard, 1863; Bischoff, 1882 a, 1882 b; Boehr, 1881; Bove, 1882, 1882-83, 1883; Brand, 1941 a, 1941 b, 1942; Bridges, 1893; Cafias, 1911 b; Colvocoresses, 1852; Cook, F. A., 1900; Cooper, 1917 a; Dabbene, 1911; Darwin, 1875; Delle Seta, 1938; Deniker (see Hyades and Deniker, 1891); Essendorfer, 1880; Fitz-Roy, 1839; Fonk, 1906 (1912); Forster, 1843; Fur- long, C. W., 1917 a; Gallardo, 1910; Garson, 1885; Genna, 1928, 1930-32; Gigoux 1927; Gédmez de Vidaurre, 1889; Guevara, 1898, 1912; Gusinde, 1922 a, 1922 b, 1922 c, 1924, 1926 a, 1926 b, 1939; Gusinde and Lebzelter, 1932; Hahn, 1883; Henckel, 1933 b, 1939, 1941, 1942 a, 1942 b; Hilden, 1930, 1931; Hyades and Deniker, 1891; Houzé, 1884; Hoyos, 1913; Hultkrantz, 1907; Ibar, 1933; Jazzetta, 1928; King, 1839; Lahille, 1926; Latcham, 1903, 1904 b, 1904 c, 1910, 1911, 1912, 1926, 1928, 1938, 1939 a; Lebzelter, 1925 (also see Gusinde and Lebzelter, 1932) ; Lehmann-Nitsche, 1904 d, 1916 b, 1916 c, 1916 d, 1927; Lista, 1887; Lothrop, 1928; Manouvrier, 1881, 1883; Mantegazza and Regalia, 1886; Mardones, 1937; Marguin, 1875; Martial, 1888; Martin, 1892, 1893-94, 1928; Matus, 1912; Medina, 1882; Mehnert, 1893; Molina, 1776, 1787; Mufioz, 1936; Nordenskiéld, 1896; Outes, 1909; Owen, 1853; Oliver, 1932; Orbigny, 1839; Olivares, 1864; Oyarziin, 1927 b; Pastore, 1935-37; Philippi, 1860, 1872; Pi-Sufier and Reyes, 1935; Poeppig, 1942; Puccioni, 1912; Rahm, 1931 a, 1931 b; Regalia (see Mantegazza and Regalia, 1886); Reyes (see Pi-Sufier and Reyes, 1935); Sabatini, 1933-34; Saller, 1939; Sarmiento de Gamboa, 1768; Schéuble, 1939; Scolni de Kliman, 1938; Segers, 1891; Seitz, 1883; Sergi, G., 1886-87, 1888; Sergi, S., 1928; Serrano, A., 1930; Serrano, R., 1880; Skottsberg, 1910; Snow, 1857; Spegazzini, 1882; Ten Kate, 1892, 1904; Topinard, 1881; Vallois, 1932; Vergara Flores, 1902, 1904, 1905; Verneau, 1903; Vignati, 1927 b; Virchow, 1881; Webster, 1834; Weddell, 1827; Wieghardt, 1896; Wilkes, 1844. i i ; 7 7 ay ov , * oy : (eC vy <—™ eh Cas hy 2 ARH Bg mt » a ie ; a ee Lio) Sai eee hae ; es ae 2 = a a “ 7 s - a a - *s a at . é hy MOTT RE? oe rs 7 : 7 We - 6 i. 7 j - 7 2 o ol _ ’ THE GEOGRAPHICAL PATHOLOGY OF CHILE By Ernesto Herzoa One of the principal tasks of geographical pathology is to study in each country the diseases that are related to climate, to conditions of life, to racial and other factors, and to investigate the differences of their course and form as compared with those known in other coun- tries. From this can be deduced the enormous importance of the comparative study of diseases in different parts of the world, a fact which can furnish many leads to the etiology, mode of course, and manifestation of many pathological patterns, which up to the present are not clear or are unknown. These conditions can be more easily studied in countries which are sparsely populated but have a great climatic variety due to their great extent of territory. In this sense, Chile is undoubtedly one of the most interesting from the point of view of its geographical pathology because of its great territorial expanse from lat. 18° to 56° S. In the northern part of the country there are tropical but very dry regions, besides desert zones, and great differences in altitude from a few meters above sea level up to 3,000 to 4,000 m. (about 9,000 to 13,000 ft.); in most of the southern part, in the wilderness and lakes, a humid climate prevails, and in the region of Magallanes, in the far south, there is less humidity but more winds, and the weather is cold. Fur- ther differences exist between the maritime zones on the littoral of the long coast, the continental zones of the central part, and the cordillera region. Unfortunately, the low density of the population and in some places its entire lack is as yet an obstacle to the extension of geographical pathological studies through the whole country. Accurate data can be obtained only on the basis of a great number of autopsies which up to the present time have been possible only in the large cities where there are institutes of pathological anatomy. The confidential observations made by some doctors in places where no anatomico-pathological services exist may also be of importance. Due to these difficulties, however, it has been impossible to obtain exact anatomical details about the constitutional and pathological anatomy of the natives of Tierra del Fuego and about the Mapuche (Araucanian) Indians scattered through the southern part of the country. With the establishment of new anatomico-pathological insti- 137 138 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 tutes in the future, we may be able to add further contributions to this subject. The racial components of the population of Chile are mainly the White race, particularly from Spaniards and other Europeans, and the native Indian. The Negro is virtually nonexistent in Chile. Thus, we deal in Chile mostly with a population of mixed White and native stocks. It is of the utmost interest to ascertain whether this mixture pro- duces special reactions to disease as well as characteristic morpho- logical types, and whether it has greater or lesser resistance with regard to definite pathological patterns. As yet anatomical investi- gations based on autopsies are insufficient to permit accurate general deductions. The determination of exact racial and constitutional factors is also complicated by a series of other factors, such as climatic or environmental ones, which cannot always be well defined or singled out. And so, our studies on geographical pathology refer to some thousands of autopsies in the course of 12 years in the central zone of Chile, practiced for the most part at the Universidad de Concep- cién and partly in the corresponding Institutes of Santiago and of Valparaiso, besides medical observations, without autopic control, in the different parts of the country. Hence, this is but a modest trial and only a beginning for the geographical pathology of Chile. First, we must direct our attention to certain infectious and parasitic endemic diseases. Typhoid fever.—This is an endemic disease throughout the country and is conditioned, in the first place, by certain hygienic deficiencies, for example, lack of potable water in many small towns and especially in rural zones. Also the old system of irrigating vegetable gardens and flower gardens by means of trenches, i. e., open canals full of dirty water that runs through entire communities, have contributed a great deal to the spreading of the typhoid bacillus, thus causing infec- tion of human beings directly as well as through contaminated fruits. Of course, this danger has diminished in the last few years in cities with good systems of potable water, but there are still other sources of infection by unkempt humans, purveyors of germs, and contami- nated food. Observations, verified also by autopsies, show that foreign- ers living in Chile are much less resistant to typhoid fever than the autochthonous Chileans; also the death-rate among foreigners is much higher. Although the course and anatomical pattern of this disease does not show any major deviation from that observed in other coun- tries, it is notable that mild cases are more frequent here than in Europe. The explanation of this phenomenon seems to be that people living in more primitive or at least less hygienic conditions and who, therefore, are always apt to have more contact with the typhoid Vol. 6] GEOGRAPHICAL PATHOLOGY—HERZOG 139 bacillus should be in a state of relative immunization, a fact that results in milder patterns than in the case of persons who have not been subjected to infection, i. e., who have not become allergic to them. Unfortunately, we have no available data of this disease among the natives so that we cannot substantiate this thesis. We do not think it probable that there is any racial influence in these cases, although this possibility cannot be entirely disregarded until it has been studied. Exanthematic typhus.—This infectious disease is also endemic in Chile, for isolated cases are always found, and its appearance depends primarily upon hygienic and social factors. Epidemics which appear now and then, such as that of 1932 to 1935 with a 20 percent mortality, are caused by unknown biological factors besides the hygienic and social ones (Herzog, E., 1937, pp. 574-600). We do not know as yet the extent of this disease among the natives, but epidemics among them have not been mentioned. There do not seem to be great differences in the appearance of exanthematic typhus in the various regions of the country, although during the last epidemics the northern and central zones were the more affected while no cases occurred in Magallanes. As yet, however, it is impossible to determine definite climatic influences. Epidemic meningitis—Meningococcic meningitis appears, from time to time, in the form of limited endemias, especially in the northern and central zones of the country, but it presents no particular charac- teristics differing from those known in other countries. In 1942, during a certain period, the cases of meningitis were more frequent and almost assumed an epidemic character. Tuberculosis.—The problem of tuberculosis in Chile is one of the gravest, for Chile ranks second among all the countries of the world in the frequency of tuberculosis and the mortality due to it (Holtheuer, 1934-35). Until now, the patterns obtained from autopsies show practically the same anatomical and clinical aspect observed in other countries, but whether tubercular preinfection in adults is not more frequent than had been thought is a matter of discussion. Observa- tions made at autopsies both in Concepcién and Santiago suggest prime infection in adults, particularly among adolescents, but the total number is not as yet sufficient to draw any general conclusions. The characteristic trait in these cases is an extensive and intense caseous lymphadenitis of the hilus nodes and _ tracheo-bronchial ganglia, with progredient tuberculosis of the lungs, well-known forms in the precocious generalization of primitive tuberculosis in children. Should this fact be further verified, it would point to an environment as yet not wholly tuberculized. It would be also interesting to make a comparative study of the course of this disease among the natives, 140 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 something that has not yet been done. As to the different parts of the country, there do not seem to exist any major differences in the tuberculosis pattern, although we still lack anatomico-pathological data about the extreme north and southern regions of the country to decide definitely on any possible climatic influences. Syphilis.—Anatomical manifestations of syphilis observed at present in Chile are few, which agrees with the observations of the last decades in many countries. Thus syphilitic gummas on the autopsy table are rare, and so are serious destructive osseous alterations. The only evidence rather frequently noted in adults is luetic mesaortitis with or without aortic aneurisms. The pattern of luetic mesaortitis has been observed in the Instituto de Anatomia Patoldégica in the Universidad de Concepcién, with 2.9 percent out of 1,152 autopsies and 1 percent of aortic aneurisms (Jara, 1936). This figure is more or less stable. The reason for a lesser frequency of organic syphilis in adults must be sought, first of all, in effective treatment. There are also biological factors which are not known in detail. There are no further data as to syphilis among the natives. Congenital syphilis plays an important role in the death rate of children. Anatomical statistics made at our Institute, based on autopsies with thorough histological examinations, have revealed 37.4 percent of congenital syphilis among 143 premature and newly born infants (Rojas, 1936). Undoubtedly, the role of congenital syphilis as a decisive factor upon infant mortality has been greatly exaggerated. Jt is known from numerous observations and especially autopsies that there is a series of other factors that account for such high mortality among children in Chile, such as malnutrition, lack of hygiene, and social conditions. Actinomycosis.—This infectious disease, caused by the actinomy- ces, is rarely apparent either in the autopsies or in the central wheat zone, but, unfortunately, there are no exact statistics to compare its frequency with that of other countries. Thus far this is only our impression based on autopsic data. Malignant pustula (anthrax).— on - oe eses nooo a ee gs 1310. 00 91.92 25.30 1. 67 1 Relative weight is equal to the weight of the organ in grams per kilogram of body weight. 2M indicates the average or arithmetic mean; m (M) the mean error from the average. Mahn (1933) studied the weight of this organ in newborn Chilean babies; he arrived at an average of 142 gm. for the boys and 146 gm. for the girls. RESPIRATORY SYSTEM The lungs.—Certain facts concerning the weight of the lungs in Chileans have been produced by N. Mufioz (1934) and Henckel and Skewes (1943). N. Mufioz (1934), after having studied 90 lungs entirely free from disease, concluded that 960 gm. was the average absolute weight of both lungs; that of the right lung being 500 gm. and of the left 450 gm. However, the author does not mention the sex of the bodies on which he made his investigations. Henckel and Skewes (1943) considered the weight of the lungs in 129 Chilean men, arriving at the conclusions shown in table 3. TABLE 3.—Weight of the lungs in Chilean men (in grams) (from Henckel and Skewes, 1948) Left lung Right lung Both lungs Age (years) Absolute weight| Absolute weight] Absolute weight} Relative weight M |m(M)| M |m(M)| M |m(M)| M |m(M) 20-800 = 2-5 oe Sean os oae eae ale 390.48 | 12.23 | 462.70 | 13.08 | 850.00 | 23.64 | 14.69 0. 40 5) te eee ee ee 423.33 | 17.74 | 490.00 | 21.39 | 900.00 | 38.28 | 15.13 0. 57 1 ee ee ee 458.33 | 26.46 | 530.56 | 29.71 | 977.78 | 55.87 | 17.44 1.17 Se ee ee eee eee ae 495.00 | 39.03 | 545.00 | 47.71 |1025.00 | 84.69 | 17.20 1,61 Ch Ite oie, bin US a 437.50 | 30.28 | 575.00 | 43.28 |1006.25| 68.11 | 20.00 2. 35 N. Mujfioz (1934) and Jirén (1935 b) made observations on the varia- tions in peripheral segmentation of lungs in Chileans. The horizontal fissure (fissure interlobularis horizontal) of the right lung is incomplete (fig. 3, left) in 58.82 percent of cases (N. Muiioz, 79471150 —11 148 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 1934) while in only 41.17 percent of cases it reaches the anterior edge of the lung (M. Tapia, 1932). According to Jirén (1935 b), a secondary fissure is seen rather fre- quently on the upper half of the lower lobe of the right lung. The so-called azygous lobe, an additional lobe at the base of the right lung, is found completely developed (as proved by N. Mufioz, 1934) in 5.29 percent of cases and incompletely developed in 3.52 percent. Figure 3.—Lung and aorta. Left: Incomplete horizontal fissure of right lung. Right: Origin of truncus brachiocephalicus and first left carotid artery from a common branch of the aorta. CIRCULATORY SYSTEM Heart.—Henckel and Skewes (1943) give the following information concerning the weight of the heart: Taste 4.—Weight of the heart in Chilean males (in grams) (from Henckel and Skewes, 1948) Absolute weight Relative weight Age (years) eee eae M m (M) M m (M) [ue BF Lea Ie ES Ae tir shee See 318. 73 5.00 5.55 0.09 SSSA Eee EE ee Oy Dia et eT Se eee ae 324.19 7. 46 5. 46 Ba! TST). See e OES See ee Ee ee a ee ee cee eee eee 338. 27 13.30 5.70 one ee ee ea eS a a eee cea 367. 50 12.74 6.14 pb!) LTE CO) ce ae > ES eee. oe ee ee 336. 36 11.91 6. 26 .32 Arch of aorta (arcus aortae).—According to N. Mufioz (1933), among the different types of aortic arch, one is particularly frequent in Chileans: the brachiocephalic trunk (truncus brachiocephalicus) and the first left carotid (arteria carotis communis sinistra) originate from a common branch of the arch of aorta (fig. 3, right). This for- mation has been verified in 25 percent of 300 female corpses. In 8 percent of the same group it was observed that the left vertebral Vol. 6] PHYSICAL ANTHROPOLOGY—HENCKEL 149 artery (arteria vertebralis sinistra) started from the arch of aorta. There are no percentages for other types of the aortic arch. Concerning variations of the coeliac artery (arteria coeliaca), the gastric coronary artery (arteria gastrica sinistra) and the hepatic artery (arteria hepatic communis), I might mention the excellent data furnished by Flores (1939). Spleen.—There are various works concerning the peculiarities of this organ in Chileans. Recently the newborn’s spleen was studied by Mahn (1933). Per- haps the most important conclusion reached by this author is that at birth the weight of this organ in Chileans (males, 11.2 gm.; females, 10.2 gm.—average) is no less than that in other nationalities (Hell- mann, 7.9 gm; Herrmann, 11.2 gm; Lucas, 10.6 gm.; Miller, 10.8 gm.). The weight of the adult spleen in Chileans is, however, less than that in other ethnic groups. Jirén says (1939): ‘Among us the weight of the spleen is from 90 to 100 gm.” Henckel and Skewes (1943) estab- lished an average of 106.17+3.0 gm. as the absolute weight of the spleen in Chilean males of all ages. Naturally, the weight of this organ depends on the age, a sharp decline in absolute and relative weight being noticed in older people. The authors mentioned have arrived at the averages shown in table 5. TaBLE 5.—Weight of the spleen in Chilean males (in grams) (from Henckel and Skewes, 1948) Absolute weight Relative weight Age (years) M m(M) M m(M) Rp ere eee a eee ieee See eh ee aes 108. 05 3. 68 05 0. 06 ie ok ) ee ee a er es eee ee ee ree ee 101. 91 6. 06 1,92 11 Ce 3 Se ES eS ae eee Pa ee ieee 102. 87 1.23 1.73 12 ee eee in ee eee oe Fe 79. 55 9. 29 1.34 14 WESTON OF Seo os oe 2 a Sle Ee a ee eee eaae 88. 33 20. 96 58 14 A comparison of the figures for weight of spleen in Chileans with the corresponding figures for other ethnic groups shows that they are lower than those for Europeans or inhabitants of the United States. (French, 148 gm., average; English, 150 gm.; Germans, 150 gm., etc. See Loth, 1931.) Thus, it is very clear that in adult Chileans the weight of the spleen generally is lower than that of the groups mentioned, although in the newborn no difference whatever is noticeable (Mahn 1933). Nevertheless, there are other peoples in whom the average weight of the spleen is likewise lower than that of the Europeans and inhabi- tants of the United States. These include the Negroes and the Japanese, for whom Loth gives the following average weights: Negroes, 150 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 males, 115 gm., females, 80 gm.; Japanese, males, 93 gm., females 86 gm. According to Jirén (1939), average dimensions of the adult male spleen are: length 9% cm., width 6% cm., and thickness 3 cm. There is less interest from the point of view of anthropology in the clefts, indentations, etc., on the edges and surfaces of the spleen. (See Loth.) Mahn (1933) and Jirén (1939) furnished some details on this subject. A complete absence of the spleen was observed by Jirén in only a single case (1939). Superfluous spleens are rather frequent (14.5 percent) in the new- born (Mahn, 1933). ENDOCRINE SYSTEM Unfortunately, we have at present very few data concerning the glands of internal secretion. Mahn (1933) has ascertained the weight of the adrenal glands (corpora suprarenalia) in newborn babies. The average for males is 5.55 gm. on the right side and 5.67 gm. on the left; corresponding weights for females are: 5.21 gm. and 5.21 gm. Superfluous adrenal glands (corpora interenalia accesoria) were observed in 4.5 percent of cases. UROGENITAL SYSTEM Henckel and Skewes (1943) have ascertained the weight of the kidneys. Their conclusions are shown in table 6. TABLE 6.—Weight of the kidneys in Chilean males (in grams) (Henckel and Skewes 1943.) Absolute weight Relative weight Age (years) M m(M) M m(M) 7 | as | Aa RR eA ee ee Sm kee ee ee 290. 51 36 5.05 0.11 2 Eee Se ae ta ee eee es Pe ee ens ee 310. 23 9.12 5.34 15 Ci Raise sue) 8 ee ee ee Pa eas ae ee ee 322. 50 10. 86 5.45 .19 {SSE SS SAE eee eS et ee eS eee ee 302. 73 11. 28 5.01 .23 Gland Over = — 250s eee ee ee ae ae Ee 265. 83 5.11 -29 Mahn (1933) gives the weight of the kidneys in newborn Chileans as follows: Males, right 13.2 gm., left 13.6 gm.; females, right 12.1 gm., left 12.5 gm. According to the same author, the horseshoe kidney is found in 0.25 percent of 800 newborn Chileans and the ring-shaped kidney in 0.125 percent of the cases. A bicornuate uterus was found by Mahn (1933) in two cases (0.58 percent) among 347 newborn Chilean girls. Vol. 6] PHYSICAL ANTHROPOLOGY—HENCKEL 151 NERVOUS SYSTEM W. Mufioz (1935) made a systematic study of the terminal branches of the lower maxillary nerve (nervus alveolaris mandibularis). He discovered several variations: for example, he was able to prove the anastomosis of the lower dental nerve with the lingual nerve in only 22 percent of cases; that of the mylohyoid nerve with the lingual nerve in only 30 percent of cases. Jirén (1935 a) calls attention to the fact that the classic description of the great sciatic nerve, i. e., that it arises from the top of the sacral plexus, passes beneath the pyramidal muscle, and divides in the poplit- eal hollow, applies in only 38 percent of cases among Chileans while in the remaining cases (62 percent) there are many variations, all characterized by numerous divisions of the great sciatic. BLOOD GROUPS There are still very few data on blood groups among the Chileans. Three statements refer to the frequency of the classic groups A, B, AB, and O in the people of Santiago (Meza, Sanhueza, and Dus- sert, 1930; Bunster, Sandoval, 1941). Analogous serological studies have not yet been made in other parts of this large country; only a small number of Mapuche (Araucanian) from the vicinity of Temuco have been examined by Onetto and Castillo (1930). The results of these studies appear in table 7. The most notable fact in this table is the high percentage of the O-group or the r gene in Chileans. TABLE 7.—Frequency of the classic blood groups and their genes People of Santiago Oo A B AB p q r n Source Per- | Per- | Per- | Per- cent cent cent | cent Baniigeore acess 52.90 | 32.64 | 12.80 | 1.66 | 0.198 | 0.084 | 0.727 | 242 reed ae and Dus- ser NantIAlOs es--- oes 54.50 | 33.30 | 10.10 | 2.10} .199 | .066) .738 | 5500 | Bunster (1941). Bantigvoe see 59.12 | 29.09 9.09 | 2.70 -170 . 057 . 769 | 4200 | Sandoval (1941). A Mapuche_---._---- 75.60 | 17.20 | 6.20] .60] .095]| .035 | .869]| 382] Onetto and Castillo (1930). Data on the serological factors M and N are even less abundant than those on classic blood groups. Sandoval (1941) after examining more than 500 people of Santiago obtained the following percentages: M, 41 percent; MN, 51.5 percent; N, 7.5 percent. Henckel, Castelli, and Dal Borgo (1941), who examined 100 Mapuche from the vicinity of Temuco, obtained the following percentages: M, 57 percent; N, 9 percent; MN, 34 percent. The respective genes are: for Chileans from Santiago, m=6.40, n=2.74; for Mapuche, m=7.55, n=3.00. With due reserve it can now be said that the frequency of the gene n is quite low among Chileans. 152 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 By combining the classic blood groups with the groups M and N, Sandoval (1941) was able to present the following percentages in the people of Santiago: AM, 14.25 percent; AN, 4.75 percent; AMN, 13.00 percent; BM, 2.75 percent; BN, 0.75 percent; BMN, 4.5 per- cent; ABM, 1.25 percent; ABN, 0.5 percent; ABMN, 0.75 percent; OM, 22.75 percent; ON, 1.50 percent; and OMN 33.25 percent. BIBLIOGRAPHY Castelli (see Henckel, Castelli, and Dal Borgo, 1941); Castillo (see Onetto and Castillo, 1930); Dal Borgo (see Henckel, Castelli, and Dal Borgo, 1941); Dussert (see Meza, Sanhueza, and Dussert, 1930); Figallo, 1940; Flores, 1939; Henckel, 1941, 1942 a; Henckel, Castelli, and Dal Borgo, 1941; Henckel and Skewes, 1943; Jirén, 1933, 1935 a, 1935 b, 1938, 1939; Loth, 1931; Mahn, 1933; Meza, Sanhueza, and Dussert, 1930; Mardones, 1937; Martin, 1893-94; Mufioz, N., 1933, 1934; Mufioz, W., 1935; Onetto and Castillo, 1930; Sandoval, 1941;¥Sanhueza (see Meza, Sanhueza, and Dussert, 1930); Seitz, 1883, 1886; Skewes (see Henckel and Skewes, 1943) ; Soenksen, 1936; Tapia, M., 1932, 1939; Weldt, 1934, 1935; Westen- hoeffer, 1911. See also the following.—Fallot and Alezais, 1890; Paula Pontes, 1939; Pontes, 1937; Vinelli-Baptista, 1937-38. GLOSSARY PHYSICAL ANTHROPOLOGY Acrocrany. Breadth-height indices of the skull (basion-bregma height X 100/ greatest breadth) of 98.0 and higher. For the living (acrocephaly), sub- stituting ear height, it refers to indices of 85.0 and above. Actinomycosis. A chronic infectious disease of cattle, sometimes transmitted to man, caused by a parasitic fungus. Allelomorph. One of a pair of Genes (q. v.) determining the development of con- trasted characters, and believed to occupy equivalent loci in homologous chromosomes. Allergy. The natural hypersensitiveness of an individual to a foreign substance (antigen), as contrasted with anaphylaxis, which is artificially induced. Amazénidos. A term used by Imbelloni to describe the physical type of the Indians of the Amazon and Orinoco Basins. Same as Von Eickstedt’s ‘‘Brasilide Rasse.” Amebiasis. The state of being infected with minute one-celled animal organisms called amebae. Aneurism. A sac formed by the dilation of the walls of an artery and filled with blood. Angina. Any disease or symptom characterized by spasmodic suffocative attacks. Angiology. The sum of knowledge regarding the blood- and lymph-vessels. Aristencephaly. Skull capacities of 1,451 ec. and above in males, or 1,301 cc. and above in females. Asthenic. See Leptosomic. Athletic. A term used by Kretschmer to describe the intermediate form between the extreme stout and thin constitutional types. Auricular. Pertaining to the auricle or ear. Australoid. Having the physical characters common to native Australians. Vol. 6] PHYSICAL ANTHROPOLOGY—HENCKEL 153 Biacromial. Relating to the tips of the acromion processes of the shoulder blades, in the sense of the line or distance between. Bizygomatic. Relating to the two cheek bones, in the sense of the line or distance between. Brachycephaly. Length-breadth indices of the head (greatest breadth X< 100/ greatest length) between 81.0 and 85.4. When applied to the skull (brachy- crany) this term refers to indices between 80.0 and 84.9. Brachyfacial. Having a short face. Brevilinear. A term used to indicate the short, stocky constitutional type according to Viola’s method. Calvarium. The skull without lower jaw. Carabelli tubercle. A small elevation of enamel occasionally seen on the lingual surface of a molar tooth. Caries, dental. Dissolution and disintegration of the enamel and dentin by the action of acid-producing bacteria and their products. Cephalic. Pertaining to the head. Chamaecephaly. Length-height indices of the head (ear height100/ greatest length) up to and including 57.6. When applied to the skull (chamaecrany) basion-bregma, height is used and the indices include 69.9. Chamaerrhiny. Nasal indices between 85.0 and 99.9 in the living and between 51.0 and 57.9 on the skull. Cirrhosis. A disease of the liver, ending in atrophy and degeneration of the organ, which give to it a granular, yellow appearance. Cholecystitis. Inflammation of the gall bladder. Cholelithiasis. The presence or formation of gallstones. Colimbidos. A term used by Imbelloni to describe the physical type of the Indians of the northwest coast of North America. Same as Von Eickstedt’s “‘Pazifide Rasse.”’ Cranial. Pertaining to the skull. Cretinism. A chronic condition associated with defects of the thyroid gland, marked by arrested physical and mental development. Cysticercosis. The condition of being infected with the larval form of the tape- worm enclosed in cysts. Dacryon. The metric landmark at the common junction of the frontal, maxillary, and lacrimal bones. Dermatoglyphics. The name for all the skin patternings of fingers, toes, palms, and soles, as well as for the study thereof. Diastema. A space. In dentistry, a space between the teeth. Diastole. The stage of dilatation of the ventricles of the heart. Diverticulum, Meckel’s. An occasional sacculation of the ileum derived from an unobliterated vitelline duct. Dolichocephaly. Length-breadth indices of the head (greatest breadth X 100/ greatest length) up to and including 75.9. When applied to the skull (dolichocrany) this term includes indices between 70.0 and 74.9. Dolichoid. Elongated. Dynamometry. The measurement of the force of muscular contraction. Echinococcus. A small tapeworm, the larval form of which is occasionally found in humans encased in cysts (cysticercosis). Embolism. The plugging of an artery or vein by a clot or obstruction which has been brought to its place by the blood-current. Endocarditis. Inflammation of the lining membrane of the heart. Etiology. The study or theory of the causation of any disease. 154 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Euencephaly. Skull capacities between 1301 and 1450 cc. in males, or between 1151 and 1300 cc. in females. Euryprosopy. Morphological face indices in the living (nasion-gnathion X 100/ bizygomatic breadth) between 79.0 and 83.9. Used in connecton with the skull it refers to indices between 80.0 and 84.9. Exanthematic. Pertaining to an eruptive disease. Fuéguidos. Term used by Imbelloni to describe an Indian physical type found especially among the Botocudos and the tribes of Tierra del Fuego. Same as Von Eickstedt’s ‘‘Lagide Rasse (Kiistentypus).” Gene. A unit of inheritance situated at some particular locus of a particular chromosome and transmitted according to the laws of Mendel. Glabella. A metric landmark at the most prominent point in the midline between the eyebrows. Also used to designate this region. Gumma. A soft gummy tumor occurring in the late stages of syphilis. Hilus. The depression at the entrance and exit of the vessels, nerves, and duct of a gland. Hominid. Pertaining to the group of mammals to which Homo, or mankind, belongs. Hydatinous cyst. A cystlike tumor, sometimes found in man, which is the encapsulated larval stage of the small tapeworm, Taenia echinococcus. Hyperbrachycephaly. Length-breadth indices of the head (greatest breadth X 100/greatest length) from 85.5 and up. When applied to the skull (hyper- brachycrany) this term includes indices between 85.0 and 89.9. Hypereuryprosopy. Morphological face indices in the living (nasion-gnathion X 100/bizygomatic breadth) up to and including 78.9. Used in connection with the skull it refers to indices up to and including 79.9. Hypsiacrocephaly (or -crany). A term coined by combining Hypsicephaly (or -crany, q. v.) and Acrocephaly (or -crany, q. v.) denoting the upper range of the mean height index (for the skull from about 83.5 up). Hypsicephaly. Length-height indices of the head (ear height 100/greatest length) of 62.6 and higher. As applied to the skull (hypsicrany), substitu- ting basion-bregma height, it refers to indices of 75.0 and up. Index. Usually the ratio between two measurements expressed as a percentage of the larger. Thus, the cephalic or cranial (horizontal) index is: Maximum breadth X 100/maximum length; the morphological face index is: Chin-nasion height X 100/bizygomatic breadth; the nasal index is: Breadth X 100/length. (See, however, the Pignet index). Istmidos. A term used by Imbelloni to describe the physical type of the Indians of Middle America. Corresponds to a part of Von Eickstedt’s ‘‘Zentralide Rasse.”’ Lacrimale. The metric landmark at the point of intersection of the posterior lacrimal crest with the fronto-lacrimal suture. Laguidos. Term used by Imbelloni to describe an Indian physical type found among the oldest skeletal remains, especially at Lagoa Santo, Brazil. Same as Von Hickstedt’s ‘‘Lagide Rasse (Berghéhlentypus).” Lepteny. Upper face indices in the living (nasion-prosthion X 100/bizygomatic breadth) between 53.0 and 56.9. When applied to the skull it refers to in- dices between 55.0 and 59.9. Leptoprosopy. Morphological face indices in the living (nasion-gnathion X 100/ bizygomatic breadth) between 88.0 and 92.9. Used in connection with the skull it refers to indices between 90.0 and 94.9. Leptorrhiny. Nasal indices between 55.0 and 69.9 in the living, and up to and including 46.9 on the skull. Vol. 6] PHYSICAL ANTHROPOLOGY—HENCKEL 155 Leptosomic. Pertaining to the thin constitutional type. Same as Asthenic. Leptostaphyliny. Palatal indices as determined on the skull, up to and including 79.9. Lissotrichous. Having straight hair. Longilinear. A term used to indicate the long, thin constitutional type according to Viola’s method. Lophocephaly. A condition of the skull in which a pronounced sagittal crest is present. Luetic. Pertaining to or affected with syphilis. Lymphadenitis. Inflammation of lymph glands. Maxillo-frontale. A metric landmark at the intersection of the anterior lacrimal crest, or this crest prolonged, with the fronto-maxillary suture. Melanesoid. Having the physical characters common to the natives of Melanesia. Mesaortitis. Inflammation of the middle coat of the aorta. Meseny. Upper face indices in the living (nasion-prosthion < 100/bizygomatic breadth) between 48.0 and 52.9. When applied to the skull it refers to indices between 50.0 and 54.9. Mesocephaly. Length-breadth indices of the head (greatest breadth & 100/great- est length) between 76.0 and 80.9. When applied to the skull (mesocrany) this term refers to indices between 75.0 and 79.9. Mesoconchy. Orbital indices between 76.0 and 84.9 (measuring breadth from Maxillo-frontale, q. v.), between 80.1 and 85.0 (measuring breadth from Dacryon, q. v.), or between 83.0 and 88.9 (measuring breadth from Lacrimale, Ghva)e Mesognathy. Gnathic indices (basi-alveolar length x 100/basi-nasion length) between 98.0 and 102.9. Mesoprosopy. Morphological face indices in the living (nasion-gnathion X 100/bizygomatic breadth) between 84.0 and 87.9. Used in connection with the skull it refers to indices between 85.0 and 89.9. Mesorrhiny. Nasal indices between 70.0 and 84.9 in the living, and between 47.0 and 50.9 on the skull. Mesostaphyliny. Palatal indices, as determined on the skull, between 80.0 and 84.9. Metabolism, basal. The minimal heat produced by an individual in the fasting and resting stage. Metriocrany. Breadth-height indices of the skull (basion-bregma height X 100/greatest breadth) between 92.0 and 97.9. For the living (metriocephaly) substituting ear height, it refers to indices between 79.0 and 84.9. Metriometopy. Transverse fronto-parietal indices on the skull (minimum frontal diameter X 100/greatest breadth) between 66.0 and 68.9. Mongoloid. Having the physical characters common to the Mongolian peoples. Morphologic. Pertaining to form and structure. Mutation. The germinal change resulting in the production of an individual having a different genotypic constitution from its parents. Myology. The study of muscles. Orthocrany. Length-height indices of the skull (basion-bregma height X 100/greatest length) between 70.0 and 74.9. As applied to the living (ortho- cephaly), substituting ear height, it refers to indices between 57.7 and 62.5. Orthognathy. Gnathic indices (basi-alveolar length X 100/basi-nasion length) up to and including 97.9. Paleoanthropology. The study of the ancient remains of man. 156 SOUTH AMERICAN INDIANS [B. A. B. Bull. 148 Pampidos. A term taken over from Von Eickstedt by Imbelloni to describe the physical type of the Indians of southeastern South America, especially Argentina. Peritonitis. Inflammation of the lining of the abdominal cavity. Pignet index. A measure of body build obtained by the formula: Stature (cm.) —[chest girth (em.) +body weight (kg.)]. Planidos. A term used by Imbelloni to describe the physical type of east- central North America. Same as Von Eickstedt’s ‘‘Silvide Rasse.”’ Platyrrhine. A broad nose, indicially corresponding to Chamaerrhiny (q. v.). Polymorph. Having many forms. Prognathous. Having projecting jaws. Pueblos-andidos. A term used by Imbelloni to describe the physical type of the Indians of the Pueblo area of North America and of the Andean area of South America. Combines a portion of Von Eickstedt’s “Zentralide’”’ with his “‘Andide Rasse.”’ Pyknic. A term used by Kretschmer to describe the stout constitutional type. Raphé. A ridge or furrow that marks the line of union of the halves of various symmetric parts. Sepsis, puerperal. Poisoning by the products of a putrefactive process occurring after childbirth. Somatology. The sum of what is known regarding the body. Splanchnology. The sum of knowledge in regard to the viscera. Stenometopy. Transverse fronto-parietal indices on the skull (minimum frontal diameter X 100/greatest breadth) up to and including 65.9. Subdolichocephaly (better, -crany). A term used to classify the length-breadth indices of the skull between 75.0 and 77.76. Systole. The stage of contraction of the ventricles of the heart. Tapeinocrany. Breadth-height indices of the skull (basion-bregma height * 100/ greatest breadth) up to and including 91.9. For the living (tapeinocephaly), substituting ear height, it refers to indices up to and including 78.9. Thrombosis. The formation or development of a plug or clot in a vessel. Torus. A swelling; a bulging projection. Ulotrichous. Having wooly hair. Uvula. A small, fleshy mass hanging from the soft palate above the root of the tongue. Part 3. THe LANGUAGES OF SoUTH AMERICAN INDIANS By J. Atpen Mason Imiroduchione = 22524. bane. SOURCES Bare a ee The Meso-American languages -__ Hokan-siouan. =. Utartecane sa. 22.325 Macro-Otomanguean- ---_-_- Lencan, Jicaquean, and Paves a nee eee MIDIDON GR 2 kecg sos sh Chibchan languages of Cen- tral ‘Amieries.. = = = Chibehba proper..= ===. -=- =-= Colombian subgroup--_----- Inter-Andine group_-___-_--- BAT DACOAeTOUD ss Andaki (Andaquf)___-_-__- SEtOl CrOUps= == see ees Languages probably of Chibchan TITIES re eer eh so, iyo (Q uije) so eee Wisumalnany 8 ask oe Cotman (Kotane) =): 2 o.3 Languages of doubtful Chibchan relationships. 2 == = 2a Se Coeche (Mocoa) 2. ...-2 22 Bsmerald@ 22 e235 seuss Uyaibons | oe ee ae Aconipa (Akonipa) ---_-_-- CONTENTS PAGE 159!) Yunea=Purohan 22626225. 282 169 WUMCAMH Sateen sue oo 173 Pamir ea he 173 Cafiari (Canyari)__________ 173 Atalan Po ee pn Ye a, Ue ne eee 174 Sec, Sechura, or Tallan____- 1774: Wethumraran-.2 2222907207) Ouechise ~ = oe ee 174 aAuprinagray 5 mea a Ta @hiquigoame = o-oo os ks 175 | Macro-Guaicuruan_-__________- Mataco-Mac4____________- 176 TARA Gr. os 5 t 178 Macé (Enimagdé, Coch- 179 aboth)i iis oe we 179 Guaicurd (Waicurt) ___-_--_- £30) ule=Vilelamas see 2 yop egy a 181 Tonocoté, Matard, and Gua- 181 CAngetimee Soked? Keak s age & Awa walian 24 Sr 2... <5. SA a Chané and Chan4_________ 184 Languages of probable Arawakan 184 CUTE E0012 sa la lel al ie RCH 184 ATAUS POU. — 2 > oo oe 186 Apolista or Lapachu--__-__-_- Ye 71.51(2°5) 0 NRA aa a eR ACNE 186 Dueuna: (Tikuna) - =. 186 fA Te Re 187 SUG UNESE os eae eS ee 197 | Languages of possible Arawakan isqi}) relationships. —_ 22 = 188 AUN CRI ee oe ee Dingyarde hoo See we, & 191 pL AT Gia cae a wen 192 Uru-Chipaya-Pukina______- 192 Ochosumae: 5-52 se. 192 Chango and Coast Uru- £Oo)) Caribana. anes. Jeeses) 7 158 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 PAGE PAGE Languages of probable Cariban Northern tropical, etc—Con. Sihlighionsse 2.22 sesso eee 231 Panoan—Continued Chocé, and Cariban of Co- Pane oe Poe. eS 271 [LG LEi| 01: Wath oh Selig lh Sag aL 232 Chanticuro.2...7 22222 271 Peba=) apuaooe. fs 2ab=22 233 | Southern tropical Lowland fami- PCa Aes eee A Soe 234] lies of presumedindependence. 271 trees 2S Clee Aa ae 235 Unclassified languages of Parner ee es BAF es Ie 235 astern Per@s 2/222 271 are GM) oo ee 235 Small ‘families’? of Bolivia. 272 Pamenteira.\- <2 22 eae ssk 236 itenauian. 222522202 272 Macro-Tupf-Guaranian-_-~______- 236 Caniehanan 2.222222 272 Tupf-Guaranian___________ 236 Cayuvayan..2..222222 273 Yurimagua (Zurima- Movyiman....; 5.) =. 222 273 Gs) See. 2.5 Jee 240 Mosetenan_-__..--- - 274 Arikeme Jo 3. Gte le F aes 240 ibecan =: =e Je 25o see 274 Miranyan or Boran_______-_ 243 Wuracaredn soe 275 Waoloan ish ic ae Ghm 8 244 Small languages of the Nomuye. 8 sc. 52 eee 245 Brazil-Bolivia border Muenane.2. 225252425 246 (Havari, Masdca, Capi- Pati aoe eS Bo ee 246 shané, Purubord, Ma- OrejGn 22: 30 wie 246 shubi; ete.) 2.2. =. 275 Coerunas oe ee 246 Catukinan-22 = 2222 22428 276 Mudgkes Si. SSA 246 Chapaeurane 22 oe 277 Resigeros = 222 ee 247 Wanyam (Huafiam) and eapeniade soos She eh 247 Cabishi (Kabi¢i)_.-.. 279 Omurano (Roamaina?)_ 250 Mascoin2 2c. 0s aca 279 panelae seas seme. 251 Zemnuconn =o) eee ee - 280 Waneloce oo eee 251 Guatosne 52222 se 281 Awishiras. 222222202222 252 Bororoan and Otuke_-_-__--- 282 Northern tropical Lowland fami- Coraveca and Covareca; lies of presumed independence. 252 Curucaneca and Cu- Wisirnunn a2 ote oa 252 TUMINACH. os 1-2 Se 283 AvmicGaN So hose 2" eis Dele Sor 253 Nambic¢uaran= 5-52.55 2es22 283 Calinnan fey nS 253 Cabishi+:<=>t2.-2eee— 285 LISTE ee Meet ae ee 253 Miuran: = i245 22 eo See 285 rina The 2222. ks Ae 254 Matanawi_____----- . 285 Sdlivan, Macu, and Piaréa_. 254 ‘Pramaian 2 eee 286 Pamigua and Tinigua_-__-___ 255 Carajane => se seun eee 286 Otomacan, Guamo or Guama Carizian: 2202s eee 286 ang: VAaruran 2 2. o2s5- 250 bh Macro-Ges25—- 42. 2 a eee 287 Guanine so. So oe ee 256 GOL cizeds ua Sees ee 288 Puinavean or Macti_____-_. 257 Caingang.2..-4=2 2 ae 291 Tucanoan (Betoyan)______- 258 Camacdin, Mashacali, and Coto sees thse Tes 261 Puri (Coroado) J22s22 3222 293 Cahuapanan=: - 5225202055. 261 Camaesn (3.2 eee 294 Muniches >i. 2820 2: 262 Mashaealii 3-2 26.43 295 Panos. | 222 =~ 2 Sameer ee 262 Puri (Coroado) £...=222 295 The Chama languages... 263 Patashiis 222 220.2 gees 296 Cashibe: 14s eee 264 Malai 2 Se soo 2 Ue 297 Mayoruna? 2 28522. 269 COPsPGs con maccwes. ceeeee 298 Itucale, Simacu, and Botoeude. 2424255 =- Sse 298 raring ee eee 270 | Shavanté (Chavanté, Savante).._ 299 Vol. 6] LANGUAGES—MASON 159 PAGE PAGE Shavanté—Continued Southernmost languages—Con. 1 Saal aes setae = _.299 Charrua, Kerandi, Chan4, CIDE Vestn ees Go sea 300 etc—Continued ONUG UIE asta emeieits Te Seba gabe 300 COLT se ee A a RD 305 Guattavane els es shoes ele 300 Allentiac or Huarpean_-_-_-_-- 306 Small languages of the Pernam- Sanavirén and Come- buco region (Fulnié, Natt, ehmgGnan Sos. cL secs 307 Pancarurti, Shocé, Shucuri, Sanavirone oo coe tk os 307 Tush4, Carapaté, Payact, Te- Comechingén___-.._--- 307 remembé, and Tarairiu or ATAVCAnIANe sooo ofesee sca 307 Ochuesg yang). 22 Sees e2 301 Chonoet Aste Poe 309 Southernmost languages----_--- 302 Puelighea nie) ssi seo 309 PBS U SR So Seley es 302 Het (Chechehet) ------ 309 ATACaING = 2=— Sa82 soe 302 Chonan or Tewelche (Tehuel- Omawaca (Omahuaca). 303 che) and Ona....-...... of Diaguita or Calchaqui. 303 Vahl vans yt 2526 ove 311 Charrua, Kerand{f, Chand, Aenean. so 528). ne 311 BUC Sres see ese ss 304) BibKosraphy. 2263-5. 2 eases 312 INTRODUCTION ? Even a relatively short sketch of the linguistic conditions of a large area should cover such points as: general features—phonetic, morphological, and lexical—that characterize the languages, and the main points in which they differ from languages of other regions; brief digests of the grammar and phonetics of each independent family or at least of the more important ones; a classification of these families in groups according to phonetic and morphological type; a classification of the component languages of each family in their proper subdivisions as dialects, languages, groups, and stocks, according to degree of linguistic relationship; and a reconstruction of linguistic history and migrations. As regards the aboriginal languages of South America it must be understood at the outset that, as comparatively little reliable data are available upon them, none of the above points can be treated with any approach to thoroughness, and on most of them little can be said at present. South American Indian languages have no uniform or even usual characteristics that differentiate them from North American lan- guages. The same may be said of American languages fundamentally, as opposed to Old World languages. Languages were formerly grouped into categories according to morphological pattern: isolating, ageglutinating, polysynthetic, and inflective, with an implication of evolution and betterment toward the inflecting ideal—of course, of 1 Under the title of ‘‘Status and Problems of Research in the Native Languages of South America’’ this Introduction, with slight revision, was read at the annual meeting of the American Association for the Advancement of Science at Cleveland, Ohio, September 15, 1944, as the author’s vice-presidential address as incoming chairman of Section H, the section on anthropology, It was later published in Science, vol, 101, No. 2620, pp. 259-264, March 16, 1945, 160 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 our own Indo-European languages. However, research has shown that, so far as there has been any evolution, the isolating is the last, not the first stage. American languages were once classed with the polysynthetic, with agglutinative tendencies. No such hard-and-fast distinctions can be made; few languages belong definitely to one or another class, and most of them show traits of several classes. This applies equally well to American and to Old World languages; some show tendencies toward inflection, more toward polysynthesis., It is impossible to give any description that would characterize the majority of American Indian languages or contrast them with Old World languages, either from a morphological or a phonetic point of view. Incorporation (of the nominal or pronominal object) was formerly considered one of the characteristics of American languages; this also is missing in many of them. A classification of languages according to patterns and types being impossible, the only possible one is genetic, based on relationship, common origin, and linguistic history.” The classification of human groups according to their languages is now accepted as the best system for reconstructing historical connec- tions. Cultural elements are too easily adopted to have much histori- cal value; somatological characteristics, though more permanent than linguistic ones, are less readily identifiable in mixture. On the other hand, a proved relationship of two languages at present widely sepa- rated indicates a former close connection or identity of the ancestors of their speakers and thus affords important data on human migration. But proof of linguistic relationship is fraught with innumerable difficulties. It is seldom absolute, but depends on acceptance by scholars; on the other hand, it is impossible to prove that two lan- guages are not related. Merely to ascertain the connection between two languages is far from sufficient to establish a good historical picture. If we knew no more than that Spanish, Italian, German, and Russian are related it would mean little. All the languages of South America may be related; all those of all America may be; conceivably all languages in the world may eventually be proved to have a common descent. In the same sense, all mammals are related, all animals are related, all life had a common origin. Relationship means little unless we know degree and nearness of relationship. A direct comparison of two distantly related languages seldom yields convincing proof of their connection. A comparison of Polish and English would probably result in a negative decision; it is only because we know the historical linguistics of the Indo-European lan- 2 On the classification of languages, and of American Indian languages in particular, see Boas, 1911; Hoijer, 1941; Mason, 1940; Voegelin, 1941; and references and bibliographies therein, Vol. 6] LANGUAGES—MASON 161 guages well, with reconstructed roots of words, that the relationship can be proved. On the other hand, no proof would be needed of the relationship of French, Spanish, and Italian; even if we did not know their descent from Latin, the resemblance is obvious. ‘The relation- ship of dialects such as Catalan, Provengal, and Gallego is even closer and more evident. Related languages are grouped in ‘‘families” or ‘“‘stocks,” presumed, on present evidence, to be unrelated. These families are then sub- divided into divisions, groups, branches, languages, types, dialects, varieties, etc. The terminology is indefinite and there are no estab- lished criteria. When families heretofore considered independent are determined to be related, a more inclusive term is required; phylum has been accepted. For instance, if Indo-European, Hamito-Semitic, and Finno-Ugrian are ‘‘proved” to be related, as has been posited with considerable ground, they would compose a phylum. Most of the 85-odd ‘families’? of North America, formerly considered in- dependent, are now grouped in relatively few phyla. Good scientific grammars of South American languages are prac- tically nonexistent, and grammars of any kind, even of the older type based on analogy with Latin grammar, are very few. Comparisons of morphology, one of the important criteria for linguistic connections, are, therefore, in most cases impossible. Most of the classifications are based on lexical grounds, on vocabularies, often short, usually taken by travelers or missionary priests, and generally with the help of interpreters. The recorders were almost always untrained in phonetics and each used the phonetic system of his native language— Spanish, Portuguese, French, German, or English; sometimes Dutch or Swedish. Scientific deductions made on the basis of such material have little claim to acceptance. Yet on many languages, extinct or living, nothing else is available. An independent family should not be posited on the basis of one such vocabulary, no matter how ap- parently different from any other language. (Cf. Mashubz.) Of many extinct languages, and even of some living ones, nothing is known; of others there are statements that the natives spoke a language of their own, different from that of their neighbors, but without any suggestion as to how different, or that the language was intelligible or unintelligible or related to that of other groups. Of some, only place and personal names remain; of others, recorded lexical data ranging from a few words to large vocabularies and grammatical sketches. Owing to the magnitude of the field it has been possible for me to make very few independent studies and comparisons of lexical and morphological data with a view to establishing linguistic connections, and even most of the articles published by others in support of such 162 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 relationships have not been critically studied and appraised. The greater number, and by far the most cogent, of these studies have been written by the dean of South American linguists, Dr. Paul Rivet. Similar studies in Macro-Ge languages have been published by Loukotka. In almost all of them the authors were, unfortunately, limited to comparing vocabularies collected by others and pregnant with the faults already herein set forth. Words from lists in one group of languages are compared with words from languages of another group. Rarely are the roots or stems isolated or known, and morpho- logical elements may often be mistaken for parts of stems. Rarely has it been possible to deduce any rules of sound-shift, the best proof of linguistic relationship, or the examples given are too few in pro- portion to the number of comparisons to carry conviction. Few of these proposed linguistic relationships can be said to be incon- trovertibly proved; good cases have been made for many, and many or most of them have been accepted by later authorities, and are accepted herein. Others are of doubtful validity, and all require reappraisal, and reworking, especially those in which new data may later become, or may already have become, available. It is a truism of linguistic research that, given large enough vocabu- laries to compare, and making allowances for all possible changes in the form of a word or stem, as well as in its meaning, a number of apparent similarities, convincing to the uncritical, can be found between any two languages. Especially is this true if the comparison is made between two large groups, each consisting of languages of admitted relationship. To carry conviction, laws of sound-shift must be deduced, obeyed by a large proportion of the cases in question, and a basic similarity in morphological and phonetic pattern must be shown. Few of the comparative works on South American languages attempt such obligations, and almost all suffer from the faults above listed. There is not a really thorough comparative grammar of any South American, or for that matter of any American, native linguistic family, except possibly Algonkian. One of the pitfalls to be avoided in linguistic comparison is that of borrowing. Languages easily adopt words from neighboring languages; these must be discounted in seeking evidence on genetic relationship. Words for new concepts or new objects are likely to be similar in many languages; ® generally their categories and very similar forms betray their recent origin. Phonetic pattern and morphological traits are also borrowed, but to a lesser degree. Grammatical pattern is the most stable element in a language, phonology next; vocabulary is most subject to change. There are several areas in America where a number § See Nordenskidld, 1922; Herzog, G. 1941. Such words as those for banana, cow, telegraph, are pertinent. Vol. 6] LANGUAGES—MASON 163 of languages with little or no lexical resemblance have a relatively uniform phonology, and/or similar morphological peculiarities. Many American languages, North as well as South, show resem- blance in the pronominal system, often n for the first person, m or p for second person. Whether this is the result of common origin, chance, or borrowing has never been proved, but the resemblance should not be used as evidence of genetic connection between any two languages. Many of the languages of central and eastern Brazil are characterized by words ending in vowels, with the stress accent on the ultimate syllable. In some cases, the amount of borrowed words and elements may be so great as practically to constitute a mixed language. Linguistic students are in disagreement as to whether a true mixed language with multiple origins is possible. Loukotka, in his 1935 classification, considers a language mixed if the foreign elements exceed one-fifth of the 45-word standard vocabulary used by him for comparison. Lesser borrowings he terms ‘‘intrusions’’ and “‘vestiges.’’ (See also Loukotka, 1939 a.) The situation is further complicated by the fact that, in a large number of instances, the same or a very similar name was applied by colonists to several groups of very different linguistic affinities. This may be a descriptive name of European derivation, such as Orején, ‘Big Ears”; Patagon, “Big Feet’’; Coroado, “Crowned’”’ or ‘‘Tonsured”’; Barbados, “Bearded”; Lengua, “Tongue.”’ Or it may be an Indian word applied to several different groups in the same way that the Mayan Lacandén of Chiapas are locally called ‘“‘Caribs,”’ and the rustic natives of Puerto Rico and Cuba ‘‘Gibaros’”’ and ‘‘Goajiros,’”’ respectively. Thus, ‘“Tapuya,” the Tupt word for ‘enemy,’ was applied by them to almost all non-Tupt groups, “Botocudo’” to wearers of large lip- plugs, etc. Among other names applied to groups of different languages, sometimes with slight variations, are Apiacd, Arard, Caripuna, Cha- vanté, Guand, Guayand, Canamart, Carayd, Catawishi, Catukina, Cuniba, Jivaro, Macui, Tapieté, not to mention such easily confused names as Tucano, Tacana and Ticuna. Many mistakes have been made due to confusion of such names. (Cf. especially, Arda.) America, and especially South America, is probably the region of greatest linguistic diversity in the world, and of greatest ignorance concerning the native languages. On the very probable presumption that each homogeneous group, tribe, band, or village spoke a recog- nizable variant dialect or variety, there may have been 5,000 such in South America. The index of Rivet (1924 a) lists some 1,240 such groups (including a few synonyms), and this is far from the total. For instance, in the above index, Rivet lists 13 component members of the small and unimportant Timote family of Venezuela; in his 7947115012 164 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 monograph on the Timote (Rivet, 1927 a) he mentions 128 names for local groups, apart from the names of the villages occupied by them. The multitude of languages in America has often been given as an argument for a comparatively great length of time of human occupa- tion of this hemisphere. This concept presupposes that the first immigrants to America had a common speech. This is unlikely; it is more probable that each migrating group had its specific language, and that the number of presumably independent linguistic families may originally have been even greater than at present. Such a reduction has been the linguistic history of the rest of the world. These ‘“‘families” may either have had a remote common ancestry or multiple unrelated origins; of the origin and early forms of speech we know nothing. All known “primitive” languages are highly complex and evidently have had a long period of development. Of course, the minor dialects and obviously related languages were differentiated in America. Since the main migration to America is believed to have been via Alaska, we would expect to find in South America languages of older migrations than in North America, the speech of the earliest migrants forced to the peripheries and to cul-de-sacs by later and more aggres- sive groups, and also small enclaves of moribund independent lin- guistic families. This applies especially to southernmost and eastern- most South America, and to the speech of natives of paleo-American physical type, such as the Ge and the Fuegians. Regarding extracontinental relationships, many ill-conceived at- temps have been made to show connections between South American native languages and Indo-European or Semitic ones; all these are so amateurish that they have been accorded no scientific attention. Dr. Paul Rivet is firmly convinced of the connection between Australian languages and Chon, and between Malayo-Polynesian and Hokan. Instead of by direct trans-Pacific voyages, he believes that the Australian influence came via the Antarctic during a favorable post- glacial period not less than 6,000 years ago.* This radical thesis has met with no acceptance among North American anthropologists. The data offered in its support fall short of conviction, but probably have not received sufficient careful consideration. It is possible that some of the South American languages belong to the great Hokan or Hokan-Siouan family or phylum of North America. (Cf. Yurumangut, Quechua.) Since isolated Hokan enclaves are found as far south as Nicaragua, evidence of migrations across Panamé would not be entirely unexpected. A number of languages from Colombia to the Gran Chaco have Hokan-like morphological patterns. Dr. J. P. Harrington is convinced of the Hokan affiliations of Quechua, 4 Rivet, 1925 b, and many other articles. (See bibliography in Pericot y Garcfa, 1936, p. 432.) Vol. 6] LANGUAGES—MASON 165 but his published article (1943) fails to carry conviction, and no other argument for Hokan in South America has been presented. Such Hokan migrations, if proved, were probably at a relatively early period. On the other hand, several of the great South American families have penetrated the southern peripheries of North America. Chib- chan languages occupied a solid area, with possibly a few small enclaves of other families or isolated languages, as far as the Nicaraguan border, and the probably affiliated “Misumalpan” (Miskito-Sumo-Matagalpa) would extend this area to cover Nicaragua. Arawak and Carib extended over the Lesser and Greater Antilles, and the former may have had a colony on the Florida coast. In 1797 the native Carib Indians remaining in the Lesser Antilles, mainly on St. Vincent Island, were transported to Roatan Island off the coast of Honduras. Mixing with the Negro population there, they have spread over much of the coast of Honduras and parts of British Honduras. They now number some 15,000, most of them speaking a Carib jargon. The trend in the classification of American languages has been quite opposite in North and in South America. In the former, radical scholars believe that all the many languages formerly considered in- dependent may fall into six great phyla: Eskimo, Na-Dene, Algonkian- Mosan, Hokan-Siouan, Macro-Penutian, and Macro-Otomanguean, plus the South American phylum Macro-Chibchan. In South America, on the contrary, the more recent classifications have increased rather than reduced the number of families or groups given independent status. Most of these new ones, it must be admitted, are one- language families, many of them extinct, and generally based on one or a few short vocabularies that show little or no resemblance to any other language with which they have been compared. These should be considered as unclassified rather than as independent families. It is certain that the number will be greatly reduced as the languages become more intensively studied, but doubtful if it will ever reach such relative simplicity as in North America. Almost certainly the lin- guistic picture will be found to be far more complex than in Europe and Asia. One of the main reasons for the great difference in the proposed number of linguistic families in North and South America is that the study of South American linguistics is‘now about in the same stage as that of North American languages thirty years ago. Since that time many trained students, both in the United States and in México, have studied the native languages intensively, largely under the direction 166 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 or example of the late Drs. Franz Boas * and Edward Sapir. Except for the indefatigable Dr. Paul Rivet and the late Curt Nimuendaji, South America has had few linguistic scholars of wide interests and scientific viewpoint, and until recently very few trained younger men. The North American languages have been grouped into six phyla, mainly on grounds of morphological resemblance and intuition, and in this the students have been aided by the fact that the languages are fewer, and fewer of them extinct, so that such morphological studies could be made. South America suffers not only from lack of students, paucity of grammatical studies, multitudes of languages, extinction of many of them, but also from the practical problems of linguistic research: immense distances, poor transportation, difficulties and expense of expeditions, lack of capable interpreters, and similar handicaps. The history of attempts to classify the languages of South America was reviewed by Chamberlain in 1907. The earlier classifications, such as those of Adelung and Vater, Balbi, Castelnau, Gilij, Herv4s, Ludewig, Von Martius and D’Orbigny, were not considered therein, and need not be here. Modern classification began with Brinton in 1891 (1891 a). With his usual far-seeing good sense, not ‘curiously enough” as Chamberlain remarks, Brinton refused to enumerate or list his “stocks,” but apparently recognized nearly sixty. In many later short articles Brinton continued to alter his groupings. Other lists published in the next few years were McGee, 1903 (56) ; Chamberlain, [1903] (57); Ehrenreich, 1905 (52). All these differ more than the slight variation in total would suggest. Chamberlain then gave his own list, totalling 83. Later (1913 a) he published a revision of this, which became the standard classification in English for a decade or more. Though the total of 83 stocks is exactly the same as in his earlier list (plus 77a), the number of alterations, deletions, and additions is great. Since 1922 a number of classifications have appeared. Krickeberg (1922) stressed only the 15 most important families; based on this Jiménez Moreno (1938) published a large distribution map in color. P. W. Schmidt (1926) also wisely did not attempt to enumerate and list every family, but discussed them under 36 families or groups. The late Curt Nimuendajii never attempted a complete linguistic classification of South America, and his unpublished map and index do not include the far north, west and south, but his first-hand knowl- edge of the rest of the continent is unexcelled. In this restricted region he recognizes 42 stocks, 34 isolated languages, and hundreds of 5 See especially ‘‘Handbook of American Indian Languages,” edited by Franz Boas, Parts 1 and 2, Bulletin 40, Bureau of American Ethnology, Washington, 1911, 1922; Part 3, New York, 1933. ee Vol. 6] LANGUAGES—MASON 167 unclassified languages, the latter generally without any known linguistic data. Two comprehensive classifications of all South American languages have been made in the last 20 years. Paul Rivet (1924 a), com- bining some of Chamberlain’s families, separating others, reached a total of 77. Pericot y Garcia (1936) follows Rivet very closely, but not in numerical or alphabetical order. The most recent classifica- tion and the most radical—or most conservative, according to the point of view—is that of Loukotka (1935). Dividing more of Rivet’s families than he combined, he enumerates 94 families with a total of some 558 languages. Later he revised the details somewhat, but only regarding the languages of Brazil. In this latter article he notes the linguistic sources for each language (Loukotka, 1939 a). In view of the great uncertainty regarding the relationships and classification of the South American native languages, and the great differences of opinion, the example of Brinton, Schmidt, and Kricke- berg is herein followed, in not attempting to enumerate and rigidly to separate the genetic families. The classification of the languages of South America herein given is, therefore, presented without any pretense of finality or even of accuracy; the data are too insufficient. Future research will indicate many errors and change the picture decidedly. It is hoped that the present article incorporates all the accepted revisions since the appear- ance of other classifications, and improves on the latter. As regards exactitude and finality I can but cite the opinion of a great linguist: To attempt to make an exact and complete classification of all languages in rigorously defined families is to prove that one has not understood the principles of the genetic classification of languages. [Meillet and Cohen, 1924, p. 10.] South American linguistic history or philology does not extend before the beginnings of the 16th century with the first words and observations made by European voyagers. No native alphabets had been developed; there were no hieroglyphs, and even pictographs, petroglyphs, and picture-writing seem to be less than in North America. The Peruvian quipus were arithmetical, astrological, divinatory, and mnemonic. There was a tradition among the Quechua at the time of the Conquest that they had once had a system of writing on tree leaves that was later forbidden and forgotten (Montesinos, 1920, chs. 7, 14, 15; Bingham, 1922, ch. 16; 1930, ch. 9.), but this is given little credence by modern scholars, and no trace of it remains. However, it has recently been suggested that painted symbols were employed by some natives of the North Peruvian Coast (Larco Hoyle, 1944). A system of writing has been claimed for the Chibcha also, based, not on tradition, but on the peculiar, and appar- ently nonpictorial character of many pictographs in Colombia; this 168 SOUTH AMERICAN INDIANS [B. A. E. Bull. 143 also has received no credence among archeologists. On the other hand, the modern Cuna of Panamé have developed an interesting existent system of mnemonic picture-writing.® Two of the native languages merit special mention as having be- come, after the Spanish Conquest, lenguas francas of wider extent and use than formerly. The Tupi of the Brazilian coast became the basis of the lingua geral, the medium of communication of priests and traders throughout the Amazon drainage; it is now generally re- placed by Portuguese. The Cuzco dialect of Quechua became the culture language of the ‘‘Jnca’”’ region and extended its area even before the Conquest; after the latter it continued its spread and was adopted as a second language by the Spanish in Pert. Neither language has today, however, the cultural position of the Maya of Yucatan, for instance, though both have added many native terms in the Spanish and Portuguese of their regions, and even throughout the world, such as tapioca, jaguar, llama, and quinine. It has been estimated that 15 percent of the vocabulary of Brazilian Portuguese is of Tupi origin. In Paraguay, Guarani is considered a culture language, and some newspapers are published in it. A description of the geographical area occupied by each language would take too much space. The approximate region may be seen by reference to the large linguistic map and to the tribal sections in the other volumes of this Handbook. In this connection, the point of temporal relativity must be taken into consideration. The habitat given is that of earliest record. Great changes in population and migrations took place during the 16th to 18th centuries and even later, and migrations on a lesser or equal scale must have occurred in preceding centuries. These cannot be recorded on the map, but some are noted in the regional articles in Volumes 1 to 4. These changes took place especially in eastern Brazil and in the Andean region. In the preparation of this article I have received help, great or little, generally information through correspondence, from Messrs. J. Eugenio Garro, Irving Goldman, John P. Harrington, Frederick Johnson, Jacob Bridgens Johnson, Claude Lévi-Strauss, Erwin H. Lauriault, William Lipkind, Alfred S. Métraux, Bernard Mishkin, Eugene A. Nida, Louis Rankin, A. F. Reifsnyder, Paul Rivet, John Howland Rowe, David B. Stout, Harry Tschopik, Jr., and Charles . Wagley. My sincere thanks to these and to all others who may have assisted in any way. ‘Two persons deserve especial mention, credit, and thanks: Betty J. Meggers (Mrs. Clifford Evans, Jr.), who helped greatly with the bibliography; and particularly Mrs. Maria Alice Moura Pessoa, who is largely responsible for the linguistic map as well as for much of the bibliography. A large part, possibly the major 6 Nordenskiéld, “Comparative Ethnological Studies,” 7, Géteborg, 1928-1930. Vol. 6] LANGUAGES—-MASON 169 part, of the latter was prepared by Works Progress Administration Project No. 18369 in 1939 under the direction of the late Dr. Vladimir J. Fewkes. Mr. Ronald J. Mason also assisted in checking the map. SOURCES In addition to earlier and outmoded classifications such as those of Gilij (1780-84), Hervds y Panduro (1800), Adelung and Vater (1806- 17), Balbi (1826), D’Orbigny (1839), Ludewig (1858), and Martius (1867), about a dozen authors have offered classifications of the South American languages, or of those of large parts of South America. Although their opinions are often mentioned in text, they are generally omitted in the language bibliographies herein because of their con- stant recurrence, except in those cases, particularly Adelung and Vater (1806-17), Martius (1867), Lehmann, W. (1920), Tessmann (1930), and Jijon y Caamano (1941-43), where they present source material. The more recent classifications, with a brief note on their natures, are as follows: Alexander Chamberlain, “Linguistic Stocks of South American Indians”’ (1913 a). This 12-page article is the last of several such by Chamberlain. It enumerates his 84 families with several bibliographi- cal references to each, all of which may be found in the references herein. The accompanying map is small. Paul Rivet, “Langues de l’Amérique du Sud et des Antilles” (1924 a)—a 69-page part of Meillet and Cohen’s “Les Langues du Monde” (1924). Under each of his 77 families Rivet briefly notes the - component languages in their groupings and with their locations, in text—not in tabular—form. Over a thousand languages (or dialects) and synonyms are mentioned, and the very full index, containing about 1,250 names of South American languages, is most useful. The bibliography consists of only 82 items, all of which are included herein. It is followed by a 4-page article on “‘L’ecriture en Amérique.” Several large folding maps are included. Cestmir Loukotka, “Clasificacién de las Lenguas Sudamericanas” (1935). This is a small and rare pamphlet of 35 pages. In tabular form he lists his 94 families with the component languages (Arawak has 89) in groups or divisions. Extinct languages are so marked. Loukotka notes languages that, in his opinion, are mixed, or that have “intrusions” or “vestiges” of other languages. This is in accord with his comparisons of a 45-word standard vocabulary; the language is “mixed’’ if it contains more than one-fifth of foreign words, has ‘“ntrusions” or ‘‘vestiges’” if foreign words are few or very few. There is no bibliography and no map. Wilhelm Schmidt, ‘‘Die Sprachfamilien und Sprachenkreise der Erde” (1926). The South American section comprises 59 pages. 170 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Schmidt wisely does not enumerate his families but mentions most of the languages with their locations, and has classificatory charts for the larger families. Many references are given in text, some of which may be missing in the bibliography herein. Maps are provided in a separate atlas. The index is large. The second half of the book is devoted to ‘‘Die Sprachenkreise und Ihr Verh4ltnis zu den Kultur- kreisen,”’ where the phonologies, grammars, and syntaxes of the languages of the world are compared. To my knowledge, this is the only place where the little that is known about the morphology of South American languages may be found in one work. A digest and critique of Schmidt’s kulturkreis as it applies to South American languages should have formed a section of the present monograph. Daniel G. Brinton, ‘“The American Race’’ (1891 a). Although Brinton covers briefly all phases of American anthropology, his groupings are on a linguistic basis and his linguistic interests very great. He wisely does not enumerate his families but gives tables of the component languages of the principal families, with their locations. To prove relationships he gives comparative vocabularies and considerable linguistic data, comments, and arguments. He was the first to suggest some relationships but naturally much of his work is out of date. The bibliographical references are rather numerous, and probably some are missing in the bibliographies herein. No map is provided. L. Pericot y Garcia, “América Indigena” (1936). Like Brinton, Pericot covers all phases of the American Indian. He follows Rivet in mentioning very many small groups with their locations, also not in tabular form. He has a section (pp. 94-106) on “CarActeres lingiifsticos.” Most valuable are his voluminous bibliographical references with digests which, for South America, fill 36 pages (pp. 692-727) of concise data. Probably not all the bibliographical references are included herein. There are many detailed maps of parts of South America. Walter Krickeberg, ‘‘Die Vélker Stidamerikas” (1922) in Georg Buschan’s “‘Illustrierte Volkerkunde,” vol. 1, pp. 217-423 (1922). Krickeberg devotes some pages, especially 219-227, to linguistic features, and other remarks, passim, but gives no classificatory tables or charts. A small map, which formed the basis for the map of Jiménez Moreno (México, 1936), shows most of the families, and the principal component members of each, according to his opinions, which are generally those of consensus. The bibliography is relatively small, Although not including all of South America, the following four works deserve especial mention for their large and full coverage: Cestmir Loukotka, ‘“Linguas Indigenas do Brasil” (1939 a). Like Vol. 6] LANGUAGES——MASON 171 Loukotka’s pamphlet issued in 1935 (see above) this is a concise table of families and component languages, rigidly restricted to Brazilian territory. The name, locale, and principal references for source material are given for each group. ‘‘Intrusions,”’ “vestiges,” and mixed languages are noted. Symbols denote whether a language is extinct, and if the data on it are poor. A map is included, and 10 of the 28 pages are devoted to a large and excellent bibliography of source material, all of which items are included herein. Jacinto Jijén y Caamaijio, ‘‘El Ecuador Interandino y Occidental’ (1943). This is volume 3 of Jijén’s monumental work of this title (1941-43). Half of the volume, chapter 30 (pp. 390-654), is de- voted to ‘‘Las Lenguas del Sur de Centro America y el Norte y Centro del Oeste de Sud-América.”” It covers east to longitude 60° (Wapishana-Nambicuara-Ashluslay), and to latitude 30° S. Thus he largely supplements Loukotka (1939 a), though both omit the Araucanian-Patagonian region. He is especially strong in the Colom- bia-Ecuador-Pert area. Territory and many source references are given for each language, together with arguments regarding their classification. Eight folding maps accompany the volume. Most if not all of the bibliographical references are included herein. Giinter Tessmann, ‘“‘Die Indianer Nordost—Perus” (1930). Tess- mann covers much of eastern Perti and Ecuador most thoroughly. Fifty tribes are considered. His section 76 under each of these gives the known linguistic data, together with vocabularies, known data on morphology, and the most important source references, most of which, naturally, are included herein. A special section (pp. 617-627) is on ‘‘Sprachliche Verwandschaft” and includes (pp. 624-626) a table giving his radical ideas regarding linguistic classification. The accompanying maps are small. Walter Lehmann, ‘‘Zentral Amerika; Die Sprachen Zentral- Amerikas” (1920). These two large volumes afford a mine of infor- mation on the languages from southern México to western Ecuador. All the source material, books, and documents have been studied, mentioned, digested, and much of it reproduced. The bibliography is probably nearly complete to that date. The several very large maps contain much printed information, and cover a wider area than the text, including parts of western Venezuela and Brazil and northern Pert. Many of the bibliographical references are not included herein. Curt Nimuendaji’s unexcelled first-hand knowledge of the peoples and languages of Brazil was apparently hardly superior to his knowl- edge of the literature. He had definite ideas on the classification of languages but, unfortunately, never published them. They often disagreed with those of others but, since his opinions were often based on actual acquaintance, they merit careful consideration. He sub- 172 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 mitted a very large and very detailed map, an alphabetical list of tribes with references to location on map, and a very complete bibli- ography. On the map the tribal names were underlined with color in accord with a linguistic family color chart. As a great number of colors were employed, it is possible that occasional errors were made in transferring the familial linguistic affiliation, according to his opinions, from the map to the tribal index. His map did not include the Andean region, or the far south. See also the following references, which are very incomplete, and mainly relatively recent: ADDITIONAL REFERENCES GENERAL Relationships with Old World.—Anonymous, 1930 d (Basque); Christian, 1932 (Pert-Polynesia); Dangel, 1930 (Quechua-Maori); Ferrario, 1933, 1938 (Altaic); Gancedo, 1922 (Japanese); Imbelloni, 1928 b (Quechwa-Oceania); Koppelmann, 1929 (East Asia); Rivet, 1925 a, 1925 b, 1925 c, 1926 a, 1926 b, 1927 b, 1927 «¢, 1928 (Melaneso-Polynesian, Australian); Tavera—Acosta, 1930 (Asia); Trombetti, 1928 (Asia); Zeballos, 1922 a (Japanese). America General.—Anonymcus, 1928 a, 1929; Aza, 1927, 1930 a, 1930 b, 1931; Brinton, 1885 a, 1885 c, 1886 a, 1886 b, 1887, 1889, 1894 a, 1894 b, 1898 ¢; Castro, A., 1935; Clarke, 1937; Ferrario, 1937; Gorrochotegui, 1918; Mitre, 1909-10; Rivet, 1921 b; Rochereau, 1932; Salas, 1918; Schuller, 1936; Vifiaza, 1892. South America General.—Bastian, 1878-89; Brinton, 1884, 1892 a; Cham- berlain, [1903], 1907, 1910 a; Ferrario, 1927; Goeje, 1935; Hestermann, 1927 a, 1938; Lafone—Quevedo, 1912 a; Loukotka, 1939 b; Mason, J. A., 1945; Norden- skidéld, 1922; Oiticica, 1933, 1934; Penard, T. E., 1926-27; Romero, 1931; Schmidt, W., 1925; Schuller, 1925; Talbet, 1926. REGIONAL Antilles.—Bachiller y Morales, 1883; Goeje, 1939; Penard, T. E., 1927-28. Argentina.—Boman, 1908; Campanella, 1938-39; Constancio, 1939; Dfaz and Diaz, 1939; Imbelloni, 1936; Lehmann-Nitsche, 1924; Martinez Orozco, 1938; Portnoy, 1936; Selva, 1922; Serrano, 1941. Bolivia.—Ter4n, 1917. Brazil.—Borba, 1904; Botelho de Magalhdes, 1946; Carvalho, 1929, 1931; Gillin, 1940; Koch-Griinberg, 1922, 1928; Krug, 1925; Loukotka, 1939 a; Martius, 1867; Nimuendaji, 1925, 1931-32, 1932 a; Nimuendaji and Valle Bentes, 1923; Pompeu Sobrinho, 1919, 1933; Santos, N. C. dos, 1935 a, 1935 b; Schuller, 1911 b; Senna, 1932; Snethlage, E. H., 1931; Tastevin, 1924. Chile.—Brand, 1941 c; Cineo-Vidal, 1916; Latcham, 1939 b; Lenz, 1904-10; Valenzuela, 1918-19. Colombia.—Anonymous, 1934; Beuchat and Rivet, 1910; Castellvf, 1934 a, 1934 b, 1934 c; Fabo, 1911; Igualada and Castellvf, 1940; Jij6n y Caamafio, 1941-43; Lehmann, 1920; Medina, M., 1919-20, 1920-21; Ortiz, 1937, 1938, 1938-39 a; Otero, 1938-39; Pinell, 1928; Rivet, 1912 a; Schuller, 1930 ¢; Triana, 1907. Ecuador.—Buchwald, 1921, 1924; Grijalva, 1921; Jij6n y Caamafio, 1919; Leén, A. M., 1930 a, 1930 b; Orejuela, 1934; Paz y Nifio, 1936-37; Rivet, 1934; Santa Cruz, 1921, 1923 a; Verneau and Rivet, 1912, Vol. 6] LANGUAGES—-MASON 173 Fuegia.—Brinton, 1892 c; Chamberlain, 1911 a; Cooper, 1917 a, 1917 b; Gusinde, 1926 c; Hestermann, 1914; Lothrop, 1928; Skottsberg, 1915; Steffen, 1923. Guiana.—Brett, 1868; Farabee, 1918 b, 1924; Im Thurn, 1883; Martius, 1867, 2:312-13 (Comparative vocabulary of 17 British Guiana languages) ; Schomburgk, 1847-48, 1849. Paraguay.—Brinton, 1898 a; Hanke, 1938; Kersten, 1905. Peru.—Bollaert, 1860; Farabee, 1922; Santa Cruz, 1922. Uruguay.—Lothrop, 1932; Perea y Alonso, 1937; Serrano, 1936 a. Venezuela.—Alvarado, 1919 b, 1921; Arcaya, 1918; Armellada and Matallana, 1942; Carrocera, 1935; Gillin, 1940; Jahn, 1927; Koch-Griinberg, 1922, 1928; Lares, 1918; Raimundo, 1934; Salas, 1919; Tavera-Acosta, 1921-22. THE MESO-AMERICAN LANGUAGES In the Meso-American area considered within the scope of this Handbook are found representatives of all four of the great linguistic phyla of México and Central America, the Hokan-Siovan, Macro- Penutian, Macro-Otomanguean, and Macro-Chibchan. 'The first two are also widespread in the United States. Only a very few of the languages of the first three phyla are here included; none of them extends south of Costa Rica. The Macro-Chibchan phylum is pri- marily a South American entity and is mainly treated of later herein; it did not extend north of Honduras. These languages, with their appropriate bibliographies, are discussed more fully in J. A. Mason (1940) and Johnson (1940). (See also Johnson, Handbook, vol. 4, pp. 63-67.) HOKAN-SIOUAN The two small Meso-American languages belonging to the great Hokan-Siouan phylum are of the Hokaltecan (Hokan-Coahwiltecan) subphylum. There are only afew small and widely separated enclaves of this phylum south of the large groups in northern México. Never- theless, there are indications of related languages in South America (see especially Yurumangut, and J. P. Harrington, 1943), and many “families” in a long belt from Colombia to the Gran Chaco seem to have a Hokan type of morphology. The two languages under consideration are Subtiaba and the tiny enclave Maribichicoa. The true and earlier name for the language is Maribio. They are grouped with the Tlapanec of Guerrero, México, under the name Supanec. MACRO-PENUTIAN The Macro-Penutian phylum is a rather hypothetical one, the rela- tionship of the putative components not yet proved to general satis- faction. One probable member, Utaztecan, has languages in the Meso-American area; another, the Mayan, and a less certain member, 174 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 the Xincan, abut on this area. Another doubtful component, the Lencan, isincluded in the region. Two other stocks in this area, Jica- quean and Payan, are also possibly Macro-Penutian but more likely Macro-Chibchan. However, all four, Xinca, Lenca, Jicaque, and Paya, are best considered unclassified for the present. UTAZTECAN This stock, of great importance in México, has several enclaves in the Meso-American area. They probably belong to two different migration periods, an older one of Nahuatl languages, including Nicarao, Nahuatlato, Bagace, and Pipil, and a later one of Nahuatl, consisting of a few small isolated enclaves, probably of Aztec traders or colonists, known as Desaguadero and Sigua. MACRO-OTOMANGUEAN The Meso-American Macro-Otomanguean languages all belong to the Manguean family. All are on the west coast and all extinct. Three languages are distinguished: Choluteca or Chorotega, Mangue (with the divisions or dialects Diria and Nagrandan), and Orotifa (with the divisions of Orosi and Nicoya). LENCAN, JICAQUEAN, AND PAYAN Authorities disagree greatly as to the affinities of these three ‘fami- lies,’”’ which consist of one language each, the dialects being negligible. Some see Chibchan elements in all, some Macro-Penutian (Mizo- cuavean) elements in all. The former are naturally stronger in Paya, the latter stronger in Lenca. All three may be related, but the dif- ferences between them, and between each and other languages, are so great that they had best be considered isolated or unclassified for the present. MACRO-CHIBCHAN Most of the languages of Panamé and Costa Rica are admittedly Chibchan, and most of those of Nicaragua and southeastern Honduras are Misumalpan, probably of the Macro-Chibchan phylum. These are treated later under Chibchan. The possible relationship of Paya, Jicaque, and Lenca, in descending order of probability, is considered above. Cacaopera is a Matagalpan enclave in Lenca territory. BIBLIOGRAPHY This field is completely covered in Walter Lehmann’s “Zentral-Amerika; Die Sprachen Zentral-Amerikas, II’ (1920). Lehmann not only gives practically every source until that date but republishes all the lexical material. Only the more important sources, all given in Lehmann, are noted here: Hokan-Siouan.—Subtiaba: Lehmann, W., 1915; Sapir, 1925; Squier, 1853. Vol. 6] LANGUAGES—MASON 175 Utaztecan.—(The bibliography of Utaztecan, especially of Nahuatl or Aztec, is enormous, but that of the Central American groups is small.) Pipil: Scherzer, 1855. Nikira: Squier, 1853. Macro-Otomanguean.—Chorotega: Squier, 1852,1853. Mangue: Brinton, 1886 c. Diria: Squier, 1852. Nagrandan: Squier, 1852, 1853. Lenca.—Hernandez, E., and Pinart, 1897; Membrefio, 1897; Peccorini, 1910; Sapper, 1901; Squier, 1858; Stoll, 1884. Jicaque.—Conzemius, 1923; Membrefio, 1897. Paya.—Conzemius, 1927-28; Membreiio, 1897; Stoll, 1884. For Central America in general, especially Costa Rica, see: Ferndndez Guardia, 1892; Gabb, 1875; Gatschet, 1900; Grasserie, 1904; Herzog, W., 1886; Lehmann, W., 1910 a; Sapper, 1901; Scherzer, 1855; Schuller, 1928; Squier, 1852, 1858; Stoll, 1884; Thiel, 1882. CHIBCHAN Chibchan is one of the stocks of major importance in South America. Its area is extensive, its members many and some of them large, and in former days it probably covered a wider area, especially to the south. Some of the languages have become extinct, a number of them without linguistic record, so that their Chibchan relationships are assumed from indications of geographical position, place names, statements of early sources, etc. The language of highly cultured peoples, among others the Chibcha or Muisca of the Bogota region, it failed to become a standard language, like Aztec or Quechua, or a lingua franca like Tupi. The Chibchan languages occupy a promi- nent position in the question of intercontinental relationships, since the family is the only one that extends into North America. The Chibchan languages extended over all Panamé, most of Costa Rica and Nicaragua, and may have included the Jicaque and Paya of Hon- duras. (See preceding section; also Mason, 1940; Johnson, 1940.) They may have come into contact with the Maya. This is important in view of Schuller’s belief in a great phylum that includes Maya, Chibcha, Carib, and Arawak (Schuller, 1919-20 a, 1928). The Chibchan “family” seems to be one of those (see Quechua) with a morphology somewhat resembling Hokan, though lexical proof of genetic connection still remains to be advanced. Jijén y Caamafio (1941-43), therefore, proposes a great “‘super-phylum”’ Hokan-Siouan- Macro-Chibcha. Rivet has been studying a new vocabulary of Yuru- mangut (q. v.) and comparing it with Hokan with some favorable results. Formerly almost all the languages of highland and coastal Colombia were considered to belong to the Chibchan group, but recent opinion assigns the Chocé’ and most of the other groups of northern Colombia, except for the Bogoté Chibcha and the Arhuaco region, to the Carib ™ The Cuna and Chocé are linked culturally, and apparently linguistically, in other sections of this Hand- book (vol. 4, pp. 49-51). 176 SOUTH AMERICAN INDIANS [B. A. E. Bull. 143] (q. v.). This is presumed to be the result of a relatively recent but pre-Columbian migration that supplanted former Chibchan-speaking peoples. The subdivisions of Chibchan differ very greatly in the former classifications of W. Schmidt (1926), Loukotka (1935), W. Lehmann (1920), Rivet (1924 a), and others, and the latter has changed his opinion greatly. As a tentative basis, therefore, the latest classifica- tion, that of the Ecuadorean Jijé6n y Caamajio (1941-43), who has made a special study of this region, is herewith presented, without implication of definite acceptance as proved. Jijén y Caamaijio places in his Macro-Chibchan phylum a number of languages heretofore considered as independent ‘families,’ and divides it into eight primary groups: A. Paleo-Chibcha (Esmeralda-Yarur6) . Chibcha . Timote . Cofan Murato . Mosquito-Xinca . Puruhé-Mochica . Cholona Of these, only group B was formerly considered Chibchan, and only that is considered immediately below. Jij6n y Caamafio divides his Chibchan languages into four groups: Archaic or Western, Pacific Intermediate, Inter-Andine Intermediate, and Evolved or Eastern. Each of these is divided into subgroups with numerous languages. Rivet in his latest Chibcha classification (1943 a) divides the Chibchan languages into 10 groups: . Barbacoa . Coconuco Péez . Chibcha Proper . Changina Cuna . Guaymi . Talamanca . Andaqut 10. Guatuso Many of these represent one of Jij6n y Caamaiio’s subgroups, but there is considerable disagreement. Harts yow CONANT WNHH CHIBCHAN LANGUAGES OF CENTRAL AMERICA ® Most of the languages of Panam4 and Costa Rica are of recognized Chibchan affinities, and most of those of Nicaragua belong to the 8 See alternative classification in Handbook, vol. 4, pp. 64-66. Vol. 6] LANGUAGES—MASON 177 “‘Misumalpan (q. v.) Stock,” a hybrid name proposed by Mason (1940) and Johnson (1940) for the Miskito (Mosquito), Sumo, and Matagalpa families. The Paya and Jicaque families of Honduras may also be related to Chibcha, and members of the “ Macro-Chibchan Phylum.” The true Chibchan languages of Central America are divided into a number of groups. No authors agree upon this point. Mason (1940) and Johnson (1940) propose four groups, Rivet (1943) six. Rama (vide infra) Rivet places in his fourth or Chibcha Proper Group. The other groups he terms ‘‘Changina,” ‘“‘Cuna,” “Guaymi,” and ““ Talamanca.” Jij6n y Caamajfio (1941-43) divides the Central American Chibchan languages into five groups. The languages of the Talamanca, Guatuso, and Cuna groups he places with the Barbacoan languages to form his Western or Archaic Group. Jijén y Caamajio does not differentiate Rivet’s Guaymi and Changina groups but puts them together with some western Colombian languages into his second, or Pacific, Group. He agrees with Rivet in separating Rama and Melchora from the others and places them, together with Chibcha Proper, in the Eastern Group of evolved languages. He and Rivet are in relative agreement as regards the component languages of each subgroup. The Cuna group is often termed “‘Cueva-Cuna.’”’ The subgroups seem to be: I. Island A. San Blas (Tule or Yule) B. Caimanes II. Mainland A. Cuna (Chucunake and Bayano) B. Cueva (Coiba) Cueva and Cuna were very closely related, yet separate. Chucunake and Bayano are local names, not dialects. Mandinga is a hybrid negroid group; Secativa is not a dialect. Cuna is isolating in general character; word order is fundamental in sentences. Reduplication is frequent. Suffixing clearly predomi- nates over prefixing. Mason’s (1940) Guaymt-Dorasque subfamily is accepted by Jijén y Caamafio, but Rivet (1943) divides it into two, Changina and Guaymi. In the former group, together with Chumulu, Gualaca, and Changina, probably go the extinct Dorasque (Torresque), and probably Burica and Duy. Bukueta is a synonym or dialect of Sabanero; Muite is a dialect in the Guaymt subgroup. W. Lehmann (1920) gives the following divisions of Dorasque: Dolega, Chumulue, Iribolo, Chiriluo, Suasimi, and Zuri. With Changina apparently belongs Chaliva (Saliba, Soriba, Sariba, Shelaba). Valiente, Talamanca, Viceita, Urinama, Tariaca, and Pocosi are probably dialects of Bribri. Tojar, Teshbi, Depso, Lari, and Uren 178 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 seem to be dialects of Térraba. Boruca is a synonym of Brunca; apparent dialects of this subgroup are the extinct Kepo, Coto, Burucaca, Turucaca, and Osa. Important languages not mentioned by Rivet that seem to fall in the Talamanca group are Guetar, Voto, and Suerre (Turricia). With Cabecar, according to W. Lehmann (1920), goes Corrhue; and with Tucurrike go Orosi, Cachi, Sakawhuak, and Seche- whuak. Guatuso, with its variety Corobici or Corbesi, and Rama with its dialect Melchora, are obviously very different from each other and from other Central American Chibchan languages, and Mason (1940) was evidently in error in making a Rama-Corobici subfamily. Both Rivet (1943) and Jijén y Caamafio (1941-43) place Rama with the languages of Chibcha proper. Rivet puts Guatuso in a class by itself, and Jijé6n y Caamafio makes it a subgroup of his Western Group. BIBLIOGRAPHY The standard work, reprinting most of the known vocabularies and citing the published work to that date, together with hitherto unpublished material, is W. Lehmann, 1920. Franco, 1882, and Pinart, 1887, also cover most of the languages, as do Thomas, C., and Swanton, 1911. See also Lehmann, W., 1910 a; 1910 b; Sapper, 1905. For Cuna also see: Berengueras, 1934; Cullen, 1851 b, 1866, 1868; Gasso, 1908, 1910-1914; J. P. Harrington, 1925; Holmer, 1946, 1947; Nordenskiéld, 1928, 1928-30 b, 1929 b, 1932, 1938; Pinart, 1890 a; Pinart and Carranza, 1890, 1900; Prince, J. D., 1912, 1913 a, 1913 b; Puydt, 1868, pp. 100-105; Rivet, 1912 a; Stout, 1947; Uhle, 1890, p. 485; Wassén, 1934 a, 1934 b, 1937, 1938. For the other languages see Brinton, 1897 (Guetar); Céspedes Marin, 1923 (Guatuso); Fernandez, L., 1884 (Guatuso); Gabb, 1886 (Bribri, Brunca, Terraba, Tiribi); Gagini, 1917 (Bribri, Brunca, Guatuso, Guetar, Terraba); Pinart, 1890 b (Dorasque, Changuina, Chamula, Gualaca), 1892 b (Guaymi, Muot, Move, Peno- nomé); Pittier de Fabrega, 1898 (Bribri), 1903 (Terraba); Pittier de Fabrega and Gagini, 1892 (Terraba); Skinner, 1920 (Bribri); Thiel, 1882 (Guatuso), 1886 (Terraba, Brunca, Guatuso); Zeledén, 1918 (Guetar). CHIBCHA PROPER The Chibchan languages that have been grouped in the Chibcha Proper group are widely scattered, containing not only some in central Colombia but those of the Sierra Nevada de Santa Marta and some of Nicaragua. Rivet (1943) lists the following five main languages: Muisca or Chibcha (Muysca, Mosca) Tunebo or Tame Guamaca Cégaba (Kéggaba) or Arhuaco (Aruaco, Aruak) Rama Jij6n y Caamafio (1941-43) also places all of these languages in his Eastern or ‘Evolved” group, which he divides into three subgroups, Cundinamarca (Muisca-Tunebo), Arhuaco (Cdégaba-Guamaca), and Central American (Rama-Melchora). Vol. 6] LANGUAGES—MASON 179 Other important languages or dialects of the Muisca-Tunebo sub- group are Dui, Sinsiga, Pedrazd, Guasico, Chita, Fusagasucd, and Morcote. Duit seems to be closely related to Muisca. Pedrazéd is claimed to be a Tunebo dialect. Morcote seems to be rather variant. Though Sinsiga is generally considered closely related to Tunebo, W. Lehmann (1920) believes it closer to the Cagaba-Arhuaco group, and to form a connecting link between the latter and the central Chibchan languages. Languages or dialects of the Cdgaba-Arhuaco subgroup are Guamaca, Atanke (Atanque), Bintucua, and Ica (Busintana). Chimila (q. v.) has been placed by some in the Dorasque-Guaymi group. W. Schmidt (1926) places Tunebo, Andagui, and Betoi in this central Chibchan subgroup. If the Rama on the border of Nicaragua and Costa Rica, apparently the northernmost of the true Chibchan languages, really belongs in the central subgroup, this has important historical implications. Melchora is apparently a dialect. BIBLIOGRAPHY Chibcha (Muisca).—Acosta Ortegén, 1938; Adam, 1878; Adelung and Vater, 1806-17; Bernal, 1919; Beuchat and Rivet, 1910; Forero, 1934, 1939; Grasserie, 1904; Lugo, 1619; Restrepo Canal, 1936; Rivet, 1920 c; Rozo, 1938; Uricoechea, 1854, 1871 (Duit, Sinsiga). Tunebo.—Rivet, 1924 b; Rivet and Oppenheim, 1948; Rochereau, 1926-27; Anonymous, 1926-27. Cagaba.— Anonymous, 1919 e; Bolinder, 1925 (Ica); Celedén, 1886 (Kéggaba, Guamaca, Bintukua, Atankez), 1892 a (Atanquez), 1892 b (Bintucua); Isaacs, 1884 (Binticua, Guamaca); Preuss, 1919-27, 1925. Rama.—Conzemius, 1929 a, 1930 a; Lehmann, W., 1914. COLOMBIAN SUBGROUP Jij6n y Caamano (1941-43) divides his Pacific Group into an Isthmian subgroup of Central American languages, and a Colombian eroup. In the latter he places Chimila, Yurumangut, and possibly a number of unimportant languages: Timba, Lile, Yolo, Jamundi, Yameci, and Aburrd. None of these is classified by Rivet in 1948. (See separate sections on Yurumangut, Chimila, and Tairona.) INTER-ANDINE GROUP Jij6n y Caamafio’s Inter-Andine Group consists of a number of languages that Rivet (1943) divides into two groups, the Coconuco and the Pdez. Like Barbacoa, independent Coconucan and Paniquitan (Pdez) families were formerly accepted. Some authorities placed all this group in their Barbacoan family. Totord, Coconuco, Moguex, and Guanaco are the important and generally accepted members of the 7947115018 180 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Coconuco subgroup. Other probable members are Guambiano, Polindara, and Puben or Pubenaro. Jijén y Caamaiio (1941-48) places in this or a closely related group Popaydn (Popayanense), Malvasa, Timbia, and possibly Panzaleo (q. v.) and Quijo (Kujo) (q. v.). Pdez and Paniquité are apparently closely related, as both Rivet and Jij6n y Caamajfio agree. Otero (1938-39) calls Paniquitd a subdialect of Pdez. W. Lehmann (1920) wrote that the relationship between Pdez and Moguex (Coconucan) is quite evident. Formerly Piao, Panche, and Patdngoro (Palenque) were also placed in this group, but both Rivet and Jijén y Caamajio agree that these belong, together with Chocé, to the Carib family. BIBLIOGRAPHY Paez and Paniquité.—Anonymous, 1879; Beuchat and Rivet, 1910; Castillo y Orozco, 1877; Lehman, H. 1945; Narvdez, 1944; Ortiz, 1938-39 b; Pittier de Fabrega, 1907; Rivet, 1912 a; Uricoechea, 1877, Mogeux.—Beauchat and Rivet, 1910; Donay, 1890; Rivet, 1912 a. Coconuco.—Beauchat and Rivet, 1910; Mosquera, 1866; Rivet, 1912 a, 1941. Totoré, Guambiano.—Anonymous, 1879; Beauchat and Rivet, 1910; Eraso Guerrero, 1944; Ortiz, 1938-39 b; Rivet, 1912 a. BARBACOA GROUP Barbacoa was considered a separate stock by Brinton (1891 a) and Chamberlain (1913 a), but is now generally accepted as related to Chibcha. Jijé6n y Caamafio (1941-48) places the Barbacoa languages with Talamanca, Guatuso, and Cuna to form his Western Group, and divides them into two divisions, Pasto and Caranki-Cayapa-Colorado. Rivet (1943) mentions only Coaiquer (Cuaiker), Cayapa, and Colorado. These seem to be the most important languages, but Jijén y Caamajio (1941-43) mentions Nigua in the Cayapa-Colorado branch, and Pasto, Colima, and Muellamuese in the Pasto branch. Rivet (1924 a) thinks that Pasto is Tucano. Jijé6n y Caamafio (1941-48) states that Telembi is the same as Coaiquer and that W. Lehmann (1920) was wrong in distinguishing them, but that Cayapa and Iscuande are not the same as Coaiquer, as Barrett (1925) believed. Pichilimbi probably belongs in this group. Other languages placed in this group by some authors but not accepted by either Rivet or Jijén y Caamajio are Manabita and Latacunga. In his 1948 classification, Jijén y Caamaiio places Quillacinga (Killacinga) and Sebondoy with his Eastern Group, but in a map (map II) he groups them with the Barbacoa languages. (See Coche.) The relationship of the Barbacoa languages to the doubtful Esmer- alda family and the Yunga-Mochica has been largely discussed; Vol. 6] LANGUAGES—MASON 181 Esmeralda may well be Chibchan. W. Lehmann (1920) compared Colorado and Mochica and found only three words that hint at affinity. BIBLIOGRAPHY Barbacoa.—Beuchat and Rivet, 1910; Rivet, 1912 a; Schuller, 1930 b. Pasto.—Jij6n y Caamafio, 1941-438, 1:144-234; 2:104-107; Ortiz, 1938. Colima.—Su4rez de Cepeda, 1923. Cayapa.—Barrett, 1925; Beuchat and Rivet, 1907; Jij6n y Caamafio, 1941-43, 2:289-384; Seler, 1902 b; Verneau and Rivet, 1912; Wilczynski, 1888. Colorado.—Beuchat and Rivet, 1907; Buchwald, 1908 a; Jij6n y Caamajfio, 1941-48, 2:119-288; Rivet, 1905; Seler, 1885, 19C2 b. ANDAK{ (ANDAQU{) The extinct Andakt of the southern Colombian Highlands must not be confused with the living Andoke of the southeastern Colombian forests; the latter are either Witotoan or independent linguistically. The Andaki were also formerly considered independent; following Brinton (1891 a), Chamberlain (1913 a) put them in the Andaquian family. All modern authorities agree that their language was Chib- chan, probably of the Chibcha-Arhuaco subgroup. Jijén y Caamafio (1941-43) and Schmidt (1926) place the language with the Chibcha Proper Group, Rivet (1924 a) in a class by itself. Loukotka (1935) considers it a mixed idiom and sees vestiges of Mashacali and Cain- gang in it, a rather unlikely possibility. Igualada (1940) says that no Andaki-speaking Indians were found up to 1940 in the Colombian Caqueté area; the modern Andaki and Agiienunga descendants speak ““Inga’”’ (Quechua) and Spanish. (See also Hernfndez de Alba, Hand- book, vol. 2, p. 922.) BIBLIOGRAPHY Albis, 1860-61; Igualada and Castellvf, 1940; Jij6n y Caamafio, 1941-43; Rivet, 1912 a, 1924 c. BETOI GROUP ® The extinct Betot adjoimed the Tucanoan Betoya, from whom the Tucano (q. v.) family was formerly named (Betoyan). The Betoi language is now generally believed to have been Chibchan in affinities. Jij6n y Caamajfio (1941-43) and Schmidt (1926) place Betoi with the Chibcha Proper languages. With them were probably associated Girara and Lache. W. Lehmann (1920) believes that Caquetio, generally classed as Arawak, was also related. Nimuendaji (index) leaves Lache unclassified. Loukotka (1935) adds Situfa. * See Hern4ndez de Alba, Handbook, vol. 4, pp."393-394. 182 SOUTH AMERICAN INDIANS Bibliography.—Gumilla, 1745. Cuiscua ! I. Western A. Talamanca 1. Guetar 2. Quepo 3. Cabecar 4. Estrella 5. Chiripé 6. Tucurrike 7. Suerre 8. Bribri a. Pocost, b. Tariaca 9. Terraba 10. Brunca (Boruca) 11. TVirribi 12. Voto 13. Coto B. Barbacoa 1. Pasto a. Pasto(?) b. Coatker(?) c. Muellamuese d. Colima e. Patia f. Sindagua (Malba) 2. Cayapa-Colorado a. Colorado b. Nigua c. Cayapa d. Caranki C. Guatuso 1. Guatuso-Corobici D. Cuna 1. Cuna (Coiba, Cueva, San Blas) II. Pacific A. Isthmian (Guaymi) PON AMA wD PE . Murire . Muoi Move Valiente Penonomefio Changuena Dorasco . Chumula . Gualaca 1 Based on Jijon y Caamafio, 1941-43. [B. A. B. Bull. 143 Vol. 6] LANGUAGES—MASON 183 CuipcHa—Continued II. Pacific—Continued B. Colombian 2 1. Timba 2. Lile 3. Yolo 4. Jamundi 5. Yameci 6. Aburrad III. Inter-Andine 3 A. Péez 1. Pdez 2. Pantkita 3. Killa B. Coconuco 1. Totoré 2. Polindara 3. Moguer (Guambia) 4. Coconuco 5. Guanaco 6. Pubenaro (?) C. Popayanense 1. Popaydn 2. Puracé IV. Eastern 4 A. Cundinamarca 1. Chibcha-Muisca 2. Duit 3. Sinsigé 4. Tunebo B. Arhuaco 1. Caégaba 2. Bintucua 3. Guamaca 4. Atankez 5. Sanha 6. Ica C. Central America 1. Rama 2. Melchora 2 All of the below are of very questionable affinities. Jijon y Caamafio (1941-43) also places in this group Chimila and Yurumangui, on which see separate articles herein. 3’ Hernéndez de Alba (Handbook, vol. 2, p. 922) places the Paéez and Coconuco subgroups, together with the Pijao subgroup (see ‘‘Chocé and Other Possibly Cariban Languages of Colombia’”’ herewith) in the Tala- manca-Barbacoa group of Chibcha. Jijon y Caamafio (1941-43) places in his Inter-Andine group also Panzaleo and Quijo, on which see separate articles herein. 4 Jijon y Caamafio (1941-48) also places in this group Andaki and Betoya, on which see separate articles herein, and Quilla, Quillacinga, and Sebondoy-Mocoa, for which see Coche herein. 184 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 LANGUAGES PROBABLY OF CHIBCHAN AFFINITIES Several other extinct languages of western Colombia and Ecuador are generally believed to have been of Chibchan affinities. Among these are: PANZALEO Jij6n y Caamafio (1941-43) believes that Panzaleo was most likely related to Chibcha, though it may have been affiliated with Puruhd- Mochica (Yunga). He places it questionably in his Inter-Andine group, probably most closely related to the Coconuco subgroup. Uhle suggested a relationship with Subtiaba (Hokan). (See Murra, Hand- book, vol. 2, p. 795.) Bibliography.—Jij6n y Caamajio, 1941-43, vol. 1, ch. 10; vol. 3, ch. 29. CARA AND CARANKI The cultured Cara (Scyri) had apparently given up their original language in favor of Quechua even before the Spanish Conquest. There are some reasons for the opinion that it was of the Barbacoa Group, where it is placed by Rivet (1924 a), but its affiliation will prob- ably never be certainly known. Murra (Handbook, vol. 2, p. 792) states that it was similar to Pasto and Cayapa. Bibliography.—Buchwald, 1908 b; Jij6n y Caamajio, 1941-48, 1:234-285. KIJO (QUIJO) The Kyo abandoned their native tongue in favor of Quechua very early, possibly before the Spanish Conquest; its nature is, therefore, very controversial. It is generally placed with Cofdn (q. v.), but may have been more closely related to Chibcha. Jijén y Caamafio (1941-43) places it questionably with Panzaleo in his Inter-Andine Group of Chibcha. (See Steward and Métraux, Handbook, vol. 3, p. 652.) Bibliography.—Jij6n y Caamafio, 1941-43; 1:290-295; Tessmann, 1930, p. 237 ff. MISUMALPAN “Misumalpan”’ was the new hybrid term proposed by Mason (1940) for the group consisting of the former linguistic families Misquitoan (Miskito), Suman, and Matagalpan in Nicaragua and southern Honduras. They were there considered to compose a stock of the Macro-Chibchan phylum. Paya and Jicaque of southern Honduras may be related more distantly. Jijé6n y Caamafio (1941-43) lists them as group F of the phylum. This group he terms “‘Group Misquito-Xinca,”’ evidently including in it the Xinca of San Salvador, and by inference the Zenca of Honduras, two groups considered by 10 The earlier term is Mosquito; Mason preferred the more modern form Miskito, but the editors of this volume, the former. Both forms of the word are used therein. Vol. 6] LANGUAGES—MASON 185 Mason (1940) to be more likely affiliated with the Macro-Penutian phylum of North America. MISUMALPAN I. Miskito A. Miskito 1. Tawira a. Tawira b. Mam c. Wanki d. Baldam e. Cabo II. Sumo A. Ulva 1. Ulva a. Ulva (Ulua) b. Prinsu ce. Cucra B. Yosco 1. Yosco C. Sumo 1. Tawahca a. Twahca b. Lacu c. Coco d. Wasabane e. Pispi 2. Panamaca a. Panamaca b. Carawala ce. Tunki 3. Boa 4, Bawahca III. Matagalpa A. Matagalpa 1. Matagalpa . Matagalpa . Cacaopera . Chato (?) . Dule (?) . Pantasma (?) oMaann»oe® BIBLIOGRAPHY W. Lehmann (1920) republishes most of the known vocabularies and other original material, and cites most of the published works. See also Thomas and Swanton (1911). Especially important, or of recent date, are the following: Miskito.— Adam, 1891, 1892; Bell, 1862; Berckenhagen, 1894, 1905, 1906; Brinton, 1891 b; Conzemius, 1929 b, 1932; Cotheal, 1848; Fellechner, Miller, and Hesse, 1845; Heath, G. R., 1913, 1927; Henderson, A., 1846; Henderson, G., 1811, pp. 227-229; Young, T., 1842, pp. 170-172; Zidek, 1894. », Sumo.—Conzemius, 1929 b, 1932; Membrefio, 1897. Matagalpa.—Brinton, 1895 b; Sapper, 1901. Ulua.—Squier, 1853. 186 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 COFAN (KOFANE) The extinct Cofdn has heretofore been considered by all authorities an independent family, though this is unlikely in view of their small area. The language has probably long been extinct, though there are a few hundred Cofdn still living. Both of the principal authorities on this region, Rivet and Jijén y Caamafio, are now convinced that Cofdn is related to Chibchan. Rivet has not yet presented his proof or intimated his opinion as to the closeness of the connection. Jijén y Caamafio (1941-43) makes Cofan one of the eight members of his Macro-Chibchan phylum. He gives no subsidiary languages. Con- nections with the Barbacoa Group of Chibcha have also been suggested. Two adjacent groups that have often been considered as Cofdn languages are Kijo and Latacunga. These have sometimes been identified with the historical Cara or Syri. Both may be more purely Chibchan. The Kijo (Quijo) (q. v.) were Quechuaized long ago. Jij6n y Caamafio (1941-43) places them in his Inter-Andine group of Chibcha. Bibliography.—Castellvi, 1938; Jij6n y Caamaiio, 1941-43. LANGUAGES OF DOUBTFUL CHIBCHAN RELATIONSHIPS COCHE (MOCOA) Synonyms: Koche, Kot&e, Kote, Mocoa, Mokoa, Sebondoy, Sitbundoy, Kamsé, Quillacinga, Kilasinga. The more important historical name Coche seems to have sup- planted Mocoa(n), which Brinton (1891 a) and Chamberlain (1913 a) gave to this supposedly independent family. Mocoa is retained by a few modern authorities, such as Krickeberg (1922) and Loukotka (1935). Most of them have accepted its independent position, but in his recent thorough study Jij6n y Caamafio (1941-43) reached the conclusion, occasionally previously suggested, that it is Chibchan, in- fluenced by Carib, closest to Chibcha Proper, to Cdgaba, and to Tala- manca, in this order. However, Ortiz (1941), the most recent writer, refuses to accept Jij6n y Caamafio’s conclusions and insists on the independence of Coche. He believes that the rather extensive Chib- chan resemblances are due to borrowing. Rivet has also not yet, to my knowledge, accepted the Chibchan affinities or decided to remove Coche from his list of independent families. The language is some- times called Camsd; the principal tribe is the Sebondoy; the related Quillacinga and Mocoa are extinct, but there are said to be some 1,700 Sebondoy, though probably not all speaking their native tongue. The Mocoa have adopted Quechua and are now known as Ingano, Vol. 6] LANGUAGES—-MASON 187 Jijén y Caamajio (1941-43) does not recognize the Coche as a separate group or mention the name in his classification, but lists the languages Quilla," Quillacinga, and Sebondoy-Mocoa together with the Chibcha Proper languages in his Cundinamarca subgroup of the Eastern Group. In his map II, however, he places them with the Barbacoa subgroup (Western Group). The lexical data are considerable (see Ortiz, 1941; Jijon y Caamaiio, 1941-43), but grammatical material is badly needed. Cocue (Mocoa) 1. Sebondoy. 2. Quillacinga. 8. Patoco. BIBLIOGRAPHY Buchwald, 1919; Castellvf, 1934 a, 1934 b; Chamberlain, 1910 a, pp. 191-192; Ernst, 1891; Igualada and Castellvi, 1940; Jij6n y Caamaiio, 1938, 1939, 1941- 43, 1:97-144; Ortiz, 1938, 1941 a; Rivet, 1912 a; Safiudo, 1923. ESMERALDA A tiny extinct group of the coast of Ecuador that has been considered as forming an independent family since the classification of Chamber- lain (1913 a). The data upon it are very few. Pericot y Garcia (1936) gives Atacame as a synonym; this can have no relation to the Atacama of the Chilean desert region. W. Schmidt (1926) believes that it may belong with the Barbacoa Group of Chibcha, a very probable connection, but it is unlikely that its exact affiliations will ever be proved. Jijén y Caamajfio (1941-43) believes it to be distantly related to Chibcha, forming, with';Yaruro, the :Paleo-Chibcha, division of his Macro-Chibcha phylum. (See Murra, Handbook, vol. 2, p. 802.) Bibliography.— Buchwald, 1920, 1922; Jij6n y Caamajio, 1941-43; Rivet, 1912 a; Seler, 1902 a, 1902 b. TAIRONA AND CHIMILA The long-extinct Tairona have generally been classified as Chibchan, doubtless because of their close geographical proximity to the Chib- chan-speaking Cdégaba. The same is true of the living Chimila, some- times regarded as the modern descendants of the Tairona. Thus Park (Handbook, vol. 2, p. 868) says that the Tairona and Chimila “although linguistically related, are not included in this [i. e., Cégaba- Arhuaco] designation.”” Both of these are low-altitude peoples, coast, foothill, or lowland-dwellers, the culture of the Chimila being mainly that of a forest people. The language of the Tairona is utterly un- 11 Hern4ndez de Alba (Handbook, vol. 2, p. 922) places Quilla in the Pdez subgroup of the Talamanca- Barbacoa group, but Quillacinga as a member of the Cochean family. 188 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 known; they may well have been Cariban or Arawakan." Reichel- Dolmatoff has recently done work among the Chimila. His linguistic material has not yet been published, but he informs me (personal communication) that Chimila is Arawakan. Arawakan affinities of Tairona would not be unexpected, since they were coterminous with the Arawakan-speaking Goajiro. Bibliography.—Bolinder, 1924, 1925; Celedén, 1886. YURUMANGUI This hitherto neglected and almost unknown group and language of the Colombian west coast has recently assumed considerable im- portance. A manuscript vocabulary was recently discovered in the Archivo Nacional in Bogot4 and published. Dr. Paul Rivet has been studying it for some years, finds no resemblances with any nearby language, and believes it to be Hokan and therefore related to Melaneso- Polynesian (Rivet, 1943). Ortiz (1946) does not consider the point as proved, and prefers to consider Yurumangui as an independent tongue. BIBLIOGRAPHY Archivo nacional . , ., n. d.; Arcila Robledo, 1940; Jij6n y Caamafio, 1941-43, vol. 3, appendix 2; Ortiz, 1946; Rivet, 1948. TIMOTE This small group of the Venezuelan highlands has been much more thoroughly investigated than most, but unfortunately the linguistic data are still limited to a few small vocabularies. No running text or grammatical study is known, andoneis urgently needed. Rivet (1927 a) has assembled all the information available. Regarding the opinion of Ernst (1885) that Timoie is related to Chibcha, Rivet reaches the same conclusion that Brinton (1891 a) did earlier, that there is some lexical resemblance, but not enough for proof, and that Timote had best be considered independent. In this all other authorities agree with him, except Jij6n y Caamafio (1941-43), who believes in the re- lationship and makes Timote Group C of his Macro-Chibcha phylum. Muku is a synonym for the family. There seem to be two main languages, Timote and Cuica (Kuika). With Timote are probably related Mirript (Maripi), Mukucht (Mococht), Migurt, Tiguiné, and Escaguey; with Cuica, Tosté, Escuque, and Jajé. However, Brinton (1891 a) lists 29 groups, the names taken from Ernst (1885), and 12 J wish here to make public confession and express regret that I identified the Santa Marta archeological culture with the Tairona. In this Handbook (see vol. 2, p. xxIx) the Santa Marta archeological culture is considered as one of the few that have been tied up with an historic people. This identification has not yet been proved, but seems probable. A careful study of the original historical sources, a thing I have not yet found time to do, will be the major factor in determining the question.—J. A. M. Vol. 6] LANGUAGES—MASON 189 Rivet (1927 a), making a more thorough study, compiles a list of names, synonyms, and variations of 99 dialects and 29 varieties, each probably linguistically distinguishable. This is one example of the tremendous complexity of language in South America. TimoTse Famity (VENEZUELA) ! I. Cuica (Kuika) A. Cuica Proper B. Tosté 1. Tosté Proper 2. Trranja 3. Tomoni C. Eskuke (Eskukey) 1. Eskuke Proper 2. Bomba 3. Moka 4. Tirandé a. Cobu b. Catike c. Catu d. Tirandé Proper e. Estiguate (Estiguati) D. Jajzé6 (Jakén, Jajén) 1. Jajé Proper 2. Esnijaque 3. Kikoke (Kikokz) 4, Mapen (La Vega) 5. Durt 6. Mikimboy II. Timote (Timott) A. Timote Proper 1. Mukurujiin 2. Mukusé 3. Mokoyupu 4. Mukuarsé 5. Ciribuy 6. Miyoy 7. Mukumbda 8. Kindoraé 9. Tafallé 10. Mukumbajt 11. Cino 1 From Rivet, 1927 a, 4:137-167. In this article, which includes a large map and bibliography, the Timote Family is divided into two groups: Timote Proper and Cuika. The Cuika he divides into the four groups above noted. The Timote group is divided into numerous subgroups, of which the only one he names is the Timote Proper. The five groups above: Timote Proper, Chama, Mocochi, Mucutu, and Tapano are distinguished on basis of Rivet’s grouping in text into five paragraphs of very different lengths. Names are mine, choosing a name in this group shown on his map, except Chama, which is accepted generic. Loukotka (1935) makes a fourfold division: (1) Timote; (2) Mokoéi; (3) Migurt; (4) Cuika. His (2) and (3) are included in Rivet’s Timote group. Miguri is probably equivalent to Chama. 190 SOUTH AMERICAN INDIANS Timorese FamILy (VENEZUELA)!—Continued II. Timote (Timoti)—Continued B. Cama (Miguri ?) 25. ONAaAhWNe . Mokunée (Mukunée, Mukuneée) . Mukurubé (Mokurugud) Tabay (Mukunuténe, Tabayon ?) Mukurumagua . Guake (Guaki) . Mukumba 5 Créuy . Mukutoke (Mukufio, Migurt ?) . Mukurufuén . Muka . Mukumpi . Mukutiri . Mukusnanda . Mukaikuy Mukus6, ete. moad op g. . Mukuranda . Mukuhiiun (Mukupine, Mokoion) . Ciguara . Insnumbi (Insumbt) . Estankes . Mukuéi (Makuti, Mokoéiz) a. Misanté b. Mokao c. Mosnaééd d. Misikea, ete. . Eskagtiey . Mukuziin . Tatuy (Tatey ?) . Mukaria . Mukaketa . Mukusiri . Kapart . Jazi (Mukundi) . Mukubaée (Mirript, Mirripuy, Maripii ?) . Mukiun (Mukumpi, Lagunillas) . Kasés . Mukuinamo . Aritkagua Tibikuay Makulare Mukusumpui . Barbudos . Jamuén, ete. . Kinaré . Tiguind Guaruni (Guarurt) wa. w. 09 mh © as ome.) iFor footnote 1, see page 189, [B. A. BD. Bull. 143 Vol. 6] LANGUAGES—MASON 191 TimoTEe FamMity (VENEZUELA)!—Continued II. Timote (Timott)—Continued C. Mocochi (Mokoéi) 1. Miyuse 2. Tukani 3. Mokot& (Torondoy) D. Mukutu (Escaguey) 1. Eskaguey 2. Kanagué 3. Kind 4. Mokoino (Mokino) 5. Mombun 6. Yarikagua 7. Arikagua 8. Mukutuy 9. Mukupati 10. Mukuéaéz 11. Trikagua 12. Mokoto (Mukutu, Mukuti) a. Guarake b. Bailadores E. Tapano 1. Aviamo 2. Mokombé (Mokobo) 3. Tapano III. Unclassified tribes A. Kiroré B. Mijure C. Montun D. Iguino 1 For footnote 1, see page 189. BIBLIOGRAPHY Bricefio-Iragorry, 1929; Brinton, 1892 a; Chamberlain, 1910 a; Ernst, 1885; Fonseca, 1914 (1920); Jahn, 1927; Marcano, 1891; Oramas, 1920; Rivet, 1927 a; Tavera-Acosta, 1907. CANDOSHI, CHIRINO, AND MURATO Each of these extinct or little-known languages of western Ecuador has been linked by some recent authority with some other, or others. Rivet (1924 a) considered Chirino as forming an independent family. Loukotka (1935) calls the family Candoshi (Kandosi), and composes it of two groups, one consisting of the Candoshi and Shapra, the other of the extinct Chirino (Cumbaraja), Sacata, and Rabona. He con- siders Murato a synonym of Candoshi. Tessmann (1930) makes Shapra and Murato divisions of Candoshi, which language, synonymous with Maina in his opinion, he considers a mixture of Ge, Arawak, and Pano. Rivet thinks that Chinchipe is a synonym of Murato, and Steward and Métraux (Handbook, vol. 3, p. 615) believe that 192 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Chinchipe and Bagua are probably related to Patagén. Brinton (1891 a) long ago placed Murato with Zéparo, and Steward and Métraux (Handbook, vol. 3, pp. 629, 633) call it a subtribe of Zaparoan Andoa. Jij6n y Caamafio (1941-43) regards this group, ‘“Lenguas Muratas,”’ as related to Chibcha, composing Group E of his Macro-Chibcha phylum. Bibliography.—Anonymous, 1897; Leén, A. M., 1928-29; Rivet, 1930 b; Tess- mann, 1930. CHOLON Synonyms: Cholona, TSolona, Colén, Tscholén. Cholén (an) is one of the small families early distinguished (Chamber- lain, 1913 a) and universally accepted. According to the majority of authorities, it consists of two languages, the Cholona Proper or Tinganes and the Hibito (Xibito, Chibito, etc.). Brinton (1891 a) quotes early sources to the effect that the Cholén spoke a different language from the Hihito. Tessmann (1930) calls it a language mixed with Quechua; he gives a vocabulary of 30 words. Jijén y Caamafio (1941-43) makes Cholona the last (H) component member of his Macro-Chibchan phylum. | A grammar has been recently published by Fr. Pedro de la Mata (1923); an earlier work on Cholén by Fr. Francisco Gutierrez is mentioned. J. P. Harrington has recently compared Cholén with Quechua and believes them related. His evidence has not been published. Bibliography.—Beuchat and Rivet, 1909; Brinton, 1892 a; Chamberlain, 1910 a; Mata, 1923; Tessmann, 1930, H{BITO The extinct Hibito (Chibito, Xibito, Jibito, Zibito, Ibito, etc.) is classed with Cholén(a) by most authorities. Brinton (1891 a) quotes the old sources to the effect that the Cholén spoke a different idiom from the Hibito. Tessmann (1930) calls it a mixed language (Pano-Ge), while Cholén he considers mixed with Quechua. He gives a 33-word vocabulary. Loukotka (1935) also believes it mixed with Panoan. It became extinct about 1825. A grammar was written by Fr. José de Araujo. Bibliography.—Izaguirre, 1927-29; Tessmann, 1930. COPALLEN Apparently only four words are known of the extinct Copallén, of Copallén, Llanque, Ecuador. Jijén y Caamajfio (1941-43), who has made a most thorough study of the languages of western Ecuador, 13 Pages 458-459. This was unfortunately omitted from his Table of Contents. Vol. 6] LANGUAGES—MASON 193 dismisses it with a word, but accords it independent position in his final classification (1943). The data on which Loukotka (1935) assigns it to an independent family must, therefore, be very slight; it had better be left unclassified. It seems to be ignored by all other authorities. Bibliography.—Jij6n y Caamafio, 1941-43. ACONIPA (AKONIPA) Aconipa is one of the almost unknown languages considered as an independent family by Loukotka (1935) and apparently mentioned by no other compiler. In his recent exhaustive study of pre-Columbian western Ecuador, Jij6n y Caamafio (1941-43) merely mentions it as one of the languages of Ecuador; he leaves it independent in his final (1943) classification. Extinct, the data on it are very few, and insufficient to warrant its classification, at any rate as a distinct family. Bibliography.—Jiménez de la Espada, 1897, p. 32. YUNCA-PURUHAN If the validity of the group of languages under consideration were established, ‘“‘Yuncahd’’ would be proposed as a cogent hybrid term. The classification of the extinct coastal languages of Ecuador and northern Perti has always been—and may always be—uncertain and controversial. The “family” consists of the five groups that were given independent status by Rivet (1924 a) under the names Ataldn, Cararr, Puruhd, Sek, and Yunka. Yunca and Cafiari are families of long standing, at least since the classification of Chamberlain (1913 a); Sek is proposed by Rivet alone (1924 a). Jijén y Caamafio (1941-43) comes to the conclusion, as a result of his exhaustive studies of pre-Columbian western Ecuador, that Puruhd, Canyari, and Manteita (Manabita) are closely related and go with Yunga to form an independent family. He claims that all these differ hardly more than dialectically. As all these “families” and their component languages are extinct with practically no lexical data, except for Yunca, and as Jij6n y Caamafio reaches these conclusions mainly on the basis of proper names, the degree of relationship will probably never be proved. The family also includes, in his opinion, Huancavilca, by which he apparently implies Rivet’s Ataldn family. He proposes the name Puruhd-Mochica for this family, which he considers a major division (G) of his Macro-Ohibcha phylum. Jij6n y Caamafio is by no means the first or only one to propose such a consolidation. W. Schmidt’s (1926) Yunca-Huancavilca Group consists of Huancavilca (Ataldn), Tallan and Sechura (Sec), and Yunca, Mochica-Chanco, Chimu, and Eten (Yunea); he does not 194. SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 mention Puruhdé or Canari. Loukotka (1935) establishes a Chimu family with a Yuncan southern division, and a Puruhd-Caiart northern division. YUNCA Synonyms: Yunga, Mochica, Chimu. The Yunca, Mochica, or Chimu language of the Northern Coast of Perti is fairly well known through De la Carrera’s grammar (1644). It is practically extinct, but a few words are said still to be used by some of the Coast fishermen. A relationship to Chibcha (Barbacoa, Colorado) has been suspected, but W. Lehmann (1920) compared Mochica with Colorado without any result. Uhle has suggested a relationship with Uro (q. v.). Chamberlain (1913 a) and Brinton (1891 a) both posited a Yunca(n) family. The former extent of the Ywnca languages to the south and inland is much disputed. Some authorities believe it extended south to Ica, including practically the entire Peri Coast. According to Jijén y Caamafio (1941-4), it reached to south of Lima. He also believes that it included the North Pert Highlands, including the provinces of Ca- jamarca and Ancachs, a region ordinarily ascribed to Quechua, and impinged on the Hibito and Cholona of the Montafia to the east. These deductions are drawn from study of place names and traditions, since these regions were Quechwaized in very early, probably pre- Conquest, days. The following regions or ethnic groups are thus of uncertain original language and are left unclassified on the linguistic map: Ayavaca, Huancapampa, Huambo,~:Chachapoya, \Cajamarca, Huamachuco, Conchuco, Huacrachuco, Huayla, Pinco, Ocro, Huamali, Huanuco, Cajatampo, Atavillo, Chinchaycocha, Tarma, and Yauyo. Dr. J. P. Harrington, after a study of De la Carrera’s grammar (personal communication), reports that the phonetics are almost identical with Quechua, and that there are many vocables and other features like Quechua. Most nouns, and also most verbs, are mono- syllabic, generally ending in a vowel. The morphological mechanism is generally by suffixes. Less acceptable is the opinion of Zeballos Quifiones that the place names of the region show Maya and Zapotec resemblances, and present proof of Central American influences in the Chimu region. BIBLIOGRAPHY Adelung and Vater, 1806-17; Altieri, 1939 a, 1989 b; Bdrcena, MS.; Bastian, 1878, pp. 169-173; Briining, 1913; Buchwald, 1915; Carrera, 1939 (1644); Christian, 1932; Harrington, J. P., 1945; Jij6n y Caamaifio, 1941-43; Kimmich, 1917-18; Larco Hoyle, 1938-39; Middendorf, 1892; Ore, 1607; Tschudi, 1884; Villareal, 1921. Vol. 6] LANGUAGES—-MASON 195 PURUHA A small group, established as an independent family by Rivet (1924 a) and accepted as such by a few others. It became extinct about the close of the 17th century. There is general agreement that it is related to Canari. (See Murra, Handbook, vol. 2, p. 797.) Bibliography.—Jijén y Caamafio, 1923-24, 1941-43, 1: 410-455, 3: 88-136. CANARI (CANYARI) Also a small group, but one of longer standing as an independent family, since at least the time of Chamberlain (1913 a). No linguistic subdivisions have been suggested. (See Murra, Handbook, vol. 2, p. 799.) Bibliography.— Chamberlain, 1910 a; Cordero Palacios, 1924; Jij6n y Caamajio, 1921, 1941-43, 2: 3-78, 3:5-140; Moreno-Mora, 1922; Rivet, 1912 a. ATALAN Apparently Atalén and Tallén must be distinguished, although the languages are adjacent in coastal Ecuador. Confusion and disagree- ment are great. The linguistic data on both are so slight that their true affiliations will probably never be certainly known. Atalén was first proposed as an independent family by Rivet (1924 a), consisting of the languages Manta, Huancavilca, Puna, and Tumbez. It is one of four language groups that Loukotka (1935), with unusual reticence, left unclassified. Jijén y Caamafio (1941-48) after thorough study placed the group with his Puruhd-Mochica group of Macro-Chibchan, a classification provisionally accepted herein. It is uncertain whether the Caraca group goes with the Atalén Manta or with the Barbacoan Cara. Dialects of Atalén seem to be Apichiqui, Cancebt, Charapoto, Pichote, Pichoasac, Pichunsi, Manabi, Jarahusa, and Jipyapa. Bibliography.—Jij6n y Caamafio, 1941-48. Yunca-PurvuHA I. Yunca-Puruhdé (Yunea-Wancavilca, Puruhéd-Mochica) A. Yuncan 1. North Group (Puruhé-Cafiart) a. Puruhé b. Canyari (Caftart) c. Manabita (Mantenya) 2. South Group (Yunca) . Yunga . Morropé . Eten (2) . Chimu . Mochica (Chincha) Chanco >moaoeoe 794711—50——14 196 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Yunca-PurvHi{—Continued I. Yunca-Puruhé (Yunca-Wancavilca, Puruhé-Mochica)—Continued B. Atalén 1. Wancavilca(Huancavilca) 1 a. Manta? b. Tumbez * ce. Puna! d. Carake: Apichiki, Cancebi t See Murra, Handbook, vol. 2, p. 806. 3 See Murra, Handbook, vol. 2, p. 803. 8 See Murra, Handbook, vol. 2, p. 807. SEC, SECHURA, OR TALLAN The small Sec “family” of restricted area in westernmost Ecuador was first proposed by Rivet (1924 a). Loukotka (1935) accepts the proposed family and calls it Sechura (Secura). W. Schmidt (1926) puts it in his Ywnca-Huancavilca group, together with several others of Rivet’s “independent” families in this region. Brinton (1891 a) also grouped it with Yunca. Jijén y Caamafio (1941-43) who, in his very complete study of pre-Columbian western Ecuador, accepts the Yunca-Huancavilca group under his proposed alternative name Puruhd-Mochica, passes Sec off with the brief note that it was a language of Tallana, Ecuador, extinct in the last century, implying that not enough is known of it to classify it with any degree of finality; this is probably true at present. In his final classification (1941-43), Jij6n y Caamaifio makes Talldn a separate phylum. The extinct languages Talldén, Chira (Lachira), Colan, Piura, and Sechura are generally classed with Sec. The Catacao, a little further inland, are said still to speak their presumably related language; their investigation is a great desideratum. There is an ipso facto presump- tion of connection between TYalldén and the extinct Ataldén “family” just to the north, but the differentiation must be kept in mind. The linguistic data seem to be limited to 40 words collected by Spruce and published in Markham (1864 a). Bibliography.— Markham, 1864 a. KECHUMARAN “Kechumaran” is a hybrid term here proposed for the first time to designate the yet unproved but highly probable subphylum consisting of Quechua and Aymara. It has long been believed that Aymara and Quechua have linguistic as well as cultural relations. The extent of this relationship still awaits study. Phonetics and morphology show a relatively common pattern and many close similarities, but the lexical roots seem to have little in common except a large number, possibly Vol. 6] LANGUAGES—MASON 197 as much as a quarter of the whole, obviously related and probably borrowed by one or the other language. They have been in close contact for probably several thousand years. Aymara is generally termed the “older” language, that is, that of wider extent in pre-Inca days, and one that has yielded ground to the Quechua. The two will probably eventually be found to be members of a large phylum; the Hokan-like traits that have been claimed for Quechua probably apply also to Aymara. Jij6n y Caamafio (1941-43) presents cogent arguments for his adoption of a Quechua-Aymara phylum. Both may possibly be mem- bers of Hokan-Siowan, one of the great phyla of North America; J. P. Harrington (1943) is convinced of the connection. This would not be entirely unexpected since scattered Hokan enclaves are found as far south as Nicaragua (Subtiaba), and Rivet has considered the possibility that Yurumangut (q. v.) may also be Hokan. QUECHUA Quechua (Kechua, Quichua, Keshwa, etc.) is the South American analogue of Aztec. That is, it was the language of a relatively small group, the so-called Inca, who established a great military empire, conquered surrounding peoples, and to some extent imposed their language upon the latter. In Colonial days it became a lingua franca over an even wider area, displacing still other aboriginal languages, and this process has continued until the present. Today probably several millions of Indians in Pert, southwestern Ecuador, western Bolivia, and northwestern Argentina speak Quechua, and most of them nothing else. As many Peruvians speak Quechua as Spanish. Prac- tically the entire population of the provinces of Cuzco and Ayacucho can speak Quechua. Of course, it is slowly losing ground to Spanish. Quechua probably occupied a comparatively small area in the upper Apurimac and Urubamba drainage until the era of the great Inca conquests under Pachacuti about 1450; it was then merely one of many possibly unrelated languages in the Andean region. It over- whelmed and supplanted many of these other languages, which prob- ably survived in local use until after the Spanish Conquest and then became extinct during the Colonial Period. (Personal letter from John Rowe; see also Handbook, vol. 2, pp. 183-470.) In 1530, although Jnca military sway extended from Ecuador to Chile and Argentina, the native languages had not yet been replaced by Quechua, which apparently occupied only a small region in the Cuzco region, represented by the groups Cavina, Cuzco, Chilque, Lare, Quechua, Paucartampo, Vilcapampa, and Yanahuara (see Handbook, vol. 2, map 3, facing p. 185). In a few years, however, Quechua 198 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 replaced “the native languages. throughout the northern‘highlands to Ecuador “and even to*southern Colombia, those of the’ central and southern coasts, and those of a considerable part of the highlands to the south. Jijén y Caamafio (1941-43) believes and presents some evidence that the former languages of the northern highlands and the coast were related to Yunca (q. v.). The Quechua dialects that replaced these are known as Chinchaisuyo, the autochthonous dialects of the Cuzco region as Tahuantisuyo. The Huanca seem to have been a group apart. To the south, Quechua replaced many Aymara groups. The extension of Quechua to the central coast was apparently a rather early one and many authorities accord it some littoral in earliest days. The following regions or groups in the Peruvian highlands speak or spoke (at least in part) Quechua at some post-Conquest period but were presumably originally of other linguistic affiliations, possibly many of them Yunga, and are, therefore, left unclassified on the linguistic map: Calva, Ayavaca, Huancapampa, Huambo, Chachapoya, Cajamarca, Huamachuco, Conchuco, Huacrachuco, Huayla, Pvznco, Huamali, Ocro, Huanuco, Cajatampo, Chinchaycocha, Atavillo, Tarma, Yauyo, Huanco, Angard, Chocorvo, Choclococha, Vilcas, Rucana, Chanca, Sora, Parinacocha, Aymara (distinguish from Aymara family), Contisuyo, Omasuyo (distinguish from Aymara Omasuyo), Cotapampa, Cavana, Chumpivilca, and Arequipa; also Cochapampa and Yampara to the east, and Chicha and Lipe to the south. The Quechua languages do not differ greatly, and none varies much from the norm—additional evidence of the relatively recent spread. There are a great number of dialects, probably a slightly variant one for each of the many Quechua-speaking villages, and these form regional groups, but probably none is absolutely unintelligible to any other. That of Cuzco was and is the standard. Those of the Ayacu- cho group are the most diversified, individualized, and in some respects most archaic. The list of Quechua-speaking tribes and groups depends greatly on the temporal period; ever since about 1450, Indian groups on the peripheries of the Quechua region in Ecuador, Pert, Bolivia, and Argentina have been abandoning their native languages in favor of Quechua. This presumably produces somewhat mixed languages but not true dialects. Among these Quechuaized groups the most promi- nent are the Cara or Quito (Kito) of Ecuador, the Chicha and Lipe of Bolivia, the Allentiac, the Sanavirén, and the Vilela-Chulupi of Argen- tina. (See also Handbook, vol. 2, map 3.) The Quechua dialects are known only by the names of the villages where they are spoken; the groups of dialects, by the names of the provinces in which they center. Vol. 6] LANGUAGES—MASON 199 MopERN QUECHUA CLASSIFICATION I. Northern (Chinchaysuyu) . Ayacucho B. Junin C. Hudnuco D. Ancash E. Huamachuco (Cajamarca) F. Chachapoya II. Southern (Tahuantisuyo) A. Cuzco B. Puno III. Coastal A. Arequipa > BIBLIOGRAPHY Like that of Aztec, the bibliography of Quechua is very large, both early and recent. The best grammar seems to be that of Middendorf (1890). Adam, 1878; Adelung and Vater, 1806-17; Aguilar, 1939; Anchorena, 1874; Anonymous, 1905 (Chinchaysuyo), 1914, 1919 c, 1919 d, 1927 b, 1928 c; Astete, 1936-37, 1937 a; Azpilcueta, 1938; Barcena, MS.; Barranca, S., 1876 (Caukz); Barranea, J. S., 1915-20; Berrfos, 1919 a, 1919 b; Brinton, 1892 a, pp. 52-88; Cordero Palacios, 1924; Ctineo-Vidal, 1914, 1915 a, 1915 b, 1915 c, 1920 a, 1920 b, 1922; Dangel, 1931 a, 1931 b; Dijour, 1931-32; Dorsey, 1898; Durand, 1915-18, 1921; Farfain, 1939, 1941-42, 1943; Ferrario, 1934; Figueredo, 1701; Fuhrmann, 1922; Garro, 1939, 1942; Gonzdlez Holguin, 1607, 1608, 1901; Grigérieff, 1935; Guzm4n, 1920 a, 1920 b; Harrington, J. P., 1943; Harrington, J. P., and Valedrcel, 1941; Herrera, 1941 b; JAuregui Rosquellas, 1937; Le6n, A. M., 1939; Lizondo Borda, 1927; Markham, 1864 a; Martinez, B. T., 1917; Martius, 1867, 2: 289-296; Medina, J. T., 1930 a; Middendorf, 1890; Navarro, 1903; Ortiz, 1940 c; Paris, 1924; Paz Solddn, 1877; Pérez Guerrero, 1934; Pulgar Vidal, 1937; Rivet, 1912 a (Lama); Rodriguez, M. C., 1921; Rowe and Escobar, 1943; Sala, G., 1905-06; Santo Thomds, 1891; Steinthal, 1890; Swadesh, 1939; Tello, 1931; Tessmann, 1930, p. 221 (Lamisto), p. 235 ff. (Chasutino); Torres Rubio, 1603, 1619, 1754; Tschudi, 1853, 1884; Valcarcel, 1933. Also: Agiiero, 1929; Anonymous, 1917, 1918-19, 1925, 19380 b, 1936; Barranca, S, 1868; Barranca, J. S, 1922; Basadre, J., 1939; Briining, 1913; Christian, 1932; Chuqiwanga, 1928; Cook, O. F., 1916; Cosio, 1916, 1919, 1923, 1924;Cuneo-Vidal, 1930 a, 1930 b; Dangel, 1930; Englert, 1934; Espinoza, 1938; Farfan, 1942, 1944; Feijoo Reyna, 1924; Ferrario, 1933; Gonzdles Sudrez, 1904; Guerrero y Sosa, 1932; Herrera, 1916, 1919, 1923, 1933, 1933-34, 1939 a, 1939 b, 1941 a; Hocquart, 1916; Imbelloni, 1928 b; Jiménez Borja, 1937; Jorge, 1924; Lemos, 1920; Leén, A. M., 1922, 1927, 1928-29, 1929, 1929-31, 1932-33, 1939-40; Lizondo Borda, 1928; Macedo y Pastor, 1931-35, 1936, 1939; Matto de Turner, 1926; Mercante, 1924; Morales, 1929; Mossi, 1916; Mostajo, 1923; Murrieta, 1936; Ollanta, 1878; Palavecino, 1926, 1928; Patrén, 1918; Patrén and Romero, 1923; Rojas, 1942; Ruiz Palazuela, 1927; Schuller, 1917-18; Soliz Rodriguez, 1926; Sudrez, 1930; Talbot, 1931-32; Tascén, 1934; Tola Mendoza, 1939; Torres, A. M., 1931; Tulcdén, 1934 a, 1934 b; Vara Cadillo, 1931; Vazquez, 1921-24; Velazco Aragén, 1923; Wechsler, 1917. 200 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 AYMARA Aymara is one of the great living languages of South America; there may be half a million speakers in Bolivia and Peri. A number of Aymara “dialects” are distinguished. The differentiation of the mod- ern ones is apparently not great, and none seems to vary greatly from the norm. Those most different are around Lake Titicaca. No sug- gestions have been made of the grouping of these dialects into major divisions. The most important ones are apparently Collao and Lupaca. The Aymara region was certainly originally larger than at present, and probably many Aymara dialects in addition to the few recorded have been replaced by Quechua. In many towns Aymara and Quechua are both spoken, and occasionally Aymara enclaves have been left in a present-day Quechua-speaking region. Similarly Uro groups are surrounded by Aymara. Apparently, however, Aymara was always limited to the Highlands of Bolivia and Pert, and its former extension to the Pacific seaboard in the Tacna-Arica-Arequipa region is no longer credited, nor the Aymara affinities of the Cauki (Cauqui, Huarochirt) group in the neighborhood of Lima, Perit. Aymara is spoken today by the historic subtribes Colla, Collagua, Cana, Canchi, Ubina, and parts of the Charca and Collahuaya (Hand- book, vol. 2, p. 503). The Caranga, Lupaca, Quillaca, Omasuyo, Pacasa, Paria, and Sicasica have given it up in favor of Quechua or Spanish. It was also spoken, together with Quechua, in Sora, Chanca, Arequipa, Chicha, Lipe, Chumpivilca, and Vilcas. BIBLIOGRAPHY Adelung and Vater, 1806-17; Anonymous, 1914, 1921, 1928 c; Ayllon, 1926; Barcena, MS.; Barranca, J. 8., 1922; Bertonio, 1879 a (1603), 1879 b (1612); Brand, 1941 ¢; Brinton, 1892 a; Diaz Romero, 1918; Dorsey, 1898; Englert, 1934; Farfin, 1939; Feijoo Reyna, 1924; Franco Inojosa, 1937; Garcia, J. A., 1917; Iturry Nufiez, 1939; Markham, 1871; McKinney, Medina, and Pefiaranda, 1930; Medina, J. T., 1930 a; Middendorf, 1891; Mitre, 1909-10, 2: 236-262; Paz Solddn, 1877; Pefiaranda and Medina, 1923; Ripalda, 1923; Saavedra, 1931; Solfs, 1923, 1928; Soliz Rodrfguez, 1926; Steinthal, 1890; Swadesh, 1939; Torres Rubio, 1603 a, 1603 b, 1616; Tschudi, 1891. CHIQUITOAN The Chiquito (Chikito) form a solid small group in southeastern Bolivia. This Spanish word, meaning “very small,” has always been applied to the family; Tarapecosi may be a synonym. It has been accepted as independent since earliest writers, but not unlikely may later be found to tie with other groups into a major phylum. Lafone- Quevedo (1910) notes many resemblances to Guaycurt (q. v.) and believes them related, Mbayd being the closest of the Guaycurt languages to Chiquito both geographically and pronominally. He Vol. 6] LANGUAGES—MASON 201 notes resemblances also with many other important families: Quechua, Mataco, Macd, Araucaman, Tupi-Guarani, Arawak, and Carib, and apparently believes that all these and others are related. As his de- ductions are based mainly on resemblances in the pronominal systems they cannot be accepted as more than suggestions at present. A connection with Bororo has also been suggested. Hervas y Panduro (1800) gives the names of some 35 Chiquito bands divided into 4 dialects; most of these are presumably extinct. (See Métraux, Handbook, vol. 3, p. 383.) Modern writers mention up to seven groups in two main divisions. There is general agreement regarding the modern divisions. Loukotka (1935) and Jijén y Caa- mafio (1941-43) place the Sansimoniano, generally regarded as Carib, with Chiquito; Rivet calls it Chapacuran. Of the extinct Manacica, Métraux (Handbook, vol. 3, p. 388) says that Lucas Caballero (1933) identifies them with Tapacura and Quitemoca, which, if true, would make them Chapacuran. CHIQUITO I. North: Chiquito A. Manasi (Manacica) B. Penoki (Penokikia) C. Pinyoca: 1. Kusikia! D. Tao: 1. Tabiica 2 II. South: Churapa 1 Métraux (Handbook, vol. 3, p. 383) says that D’Orbigny (1839) reported that the Kusikia dialects were full of foreign words, mainly Arawakan Paiconeca. 2 Possibly the same as the Tapii, who also may have spoken either Zamucan or Otukean. BIBLIOGRAPHY Adam and Henry, 1880; Adelung and Vater, 1806-17; Caballero, 1933; Cardus, 1886; Lafone-Quevedo, 1910 (1912); Métraux, 1942 a, pp. 114-120; Mitre, 1909-10, 2: 279-280; Nordenskiéld, 1911 b, pp. 231-241 (Churapa); Pauly, 1928, pp. 184-185 (Churapa); Tagliavini, 1928. MACRO-GUAICURUAN Macro-Guaicurté is a name here proposed for the first time for a phylum that includes several families, heretofore considered independ- ent, in the general region of the Gran Chaco. As at present consti- tuted it consists of Mataco, Macd (Enimagd, Cochaboth) (see Mataco- Macd), and Guaicuri. The latter, probably the most important of the three, has been taken as the basis for the name. Doubtless other families in this region, at present regarded as independent, will even- tually be joined to it; one of the first may be Chiquito (Lafone-Que- vedo, 1910); Lule-Vilela is a possibility. Evidence for the connections will be given in the family articles. That for Mataco-Macd is mainly 202 SOUTH AMERICAN INDIANS [B. A. B. Bull, 143 lexical; that for Guaicurié (and Chiquito) morphological. The mor- phologies have a Hokan-like aspect. MATACO-MACA Matako-Makdé was first suggested as a name for the combined Mataco-Mataguayo and Enimagdé (Cochaboth, Makkd) “families” by Métraux (1942 b). No thorough linguistic proof of this connection has yet been presented, but it is herein accepted as probable, though not as certain or proved. A comparison of Vejoz and Towothl, vocabularies shows a large number of correspondencies, many of them practically identical, but not a large proportion of the entire vocabularies. The possibility of extensive borrowing cannot be discounted, but the resemblances are mainly in common and fundamental words. No sound-shifts were noted with enough examples to warrant any suggestion of rules, but a number of cases of Vejoz 7 to Towothli k, ch to k, s to ts, e to ai, e to%, u to o were noted. At the same time vocabularies of Suhin-Chunupt and Chorott were compared. These seem to be about equidistant from Vejoz (Mataco) and from Towothli (Macd), a little closer, as would be expected, to Vejoz. MATACO This family has always been accepted as independent under the name Mataco or Mataco-Mataguayo. It is herein considered a member of the Macro-Guaicurt (q. v.) phylum, which includes also Macd (q. v.; also Mataco-Macd) and Guaicurti. The evidence of the rela- tionship of Mataco and Guaicurté is outlined by Henry (1939), who stated that the grammatical structures of Ashluslay and Pilagd are so similar that an ancient historical relationship should be posited. He decided, however, not to place Ashluslay in the Guaicurté stock since the lexical difference is so great. There seem to be no doubts of the Mataco affinities of Ashluslay. Suggestions of relationships between Mataco and Guaicurt had previously been made by D’Orbigny (1839), Lafone-Quevedo (1893), Hunt (1913 a), and W. B. Grubb (1913), but had not met with general acceptance. Several Mataco languages are still spoken by considerable numbers of Indians in the Gran Chaco; others are extinct. Mataco is considered by some * the oldest linguistic family in the Chaco, and as having had great influence on “newer” groups. Lafone- Quevedo thought it a very mixed language, with grammar from one stock and lexicon from another. 14 Brinton, 1891 a; Hunt, 1915 b. Vol. 6] LANGUAGES—MASON 203 There is no great disagreement regarding the component languages of the family. All the Mataguayo are now known as Vejoz. The north- west Mataco were called Nocten in the 18th century. The Pilcomayo Mataco are known as Guisnay today. Probably each of the bands mentioned by Lozano (1941, p. 81) had a slightly divergent and characteristic dialect; their names are not repeated here. The Ash- luslay have many synonyms, some of which must be distinguished from similar names of other groups; one, Chunupi or Choropi may be confused with the Lule-Vilela Chunupi; they are also incorrectly given the Tupi name Tapieté. Loukotka (1935) puts the extinct Guentuse with Mataco; most authorities place them with Macd (Enimagd). W. Schmidt (1926) includes the extinct Matard (Amulalé) (q. v.) and Malbald; Rivet (1924 a) agrees as to the latter, but Matard he®considers Lule-Vilela; Métraux (Handbook, vol. 1, pp. 231-232) and Nimuendaju (map and index) think it best to consider both of uncertain affiliation. The Matard were related to and understood Tonocote (q. v.), which also W. Schmidt (1926) and Nimuendajti (map and index) place with Mataco. Brinton (1891 a) adds Akssek, a group nowhere else men- tioned. MACA (ENIMAGA, COCHABOTH)~ Macdé is herein postulated as a member of the Mataco-Macd family of the Macro-Guaicurté phylum (q. v.). The history of the stock and of its nomenclature is most confusing. It was first called Guand, causing confusion with Arawak Guand. Later it was termed Ennimd or Enimagd, but most of the languages included therein differed greatly from Enimagd proper. Rivet (1924 a) split these off to form his Mascot family, retaining the name Hnimagd for the present group. Probably to avoid this confusion, W. Schmidt (1926) adopted the term Oochabot, the Enimagd self-name, which is preferred also by Métraux herein; most of the others stick to Enimagd. Of recent years the name Macd or Makkd has had some vogue. Max Schmidt (1936 a) demonstrated that the modern Macé or Towothli speak a language related to the old Enimagd and are probably the descendants of the latter (Enimagd-Macd). Nimuendaji (map and index), however, although admitting an Hnimagd family, puts Macd with Mataco, Toosle (Towothli) with Enimagd. Much of the confusion is due to the Lengua, a name applied to several different groups. The “old” Lengua are Cochaboth; the ‘‘new,’ Lengua Mascoi. (See fuller dis- cussion in Métraux, Handbook, vol. 1, pp. 236-237.) 204 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Maraco-Maca I. Mataco A. Mataco-Mataguayo 1. Mataco a. Guisnay b. Nocten (Octenat) 2. Mataguayo a. Northern: Hueshuo, Pesatupe, Abucheta b. Southern: Vejoz B. Choroti-Ashluslay 1. Choroti (Yofuaha) 2. Ashluslay (Chulupt, Chonopi, Suhin, Sotiagay, Tapieté) II. Macé (Enimagd, Cochaboth, Guand, Lengua) A. Enimagé 1. Macé (Towothli, Toosle) B. Guentusé C. Cochaboth-Lengua BIBLIOGRAPHY Mataco.—Anonymous, 1919 b (Vejoz), 1930 c, 1931, 1933 b, 1933 ¢ (Vejoz); Brinton, 1898 a; Cardus, 1886, pp. 390-391; Grubb, W. B., 1913; Hunt, 1913 a, 1913 b (Veroz), 1937, 1940; Huonder, 1902; Kersten, 1905; Lafone-Quevedo, 1893, 1895 b (Nocten), 1896 a, 1896 b (Vejoz), 1910; Lehmann-Nitsche, 1926; Métraux, 1942 b; Nusser-Asport, 1897; Orbigny, 1839; 1896 (Vejoz); Pelleschi, 1881, pp. 359-423, 1896; Remedi, 1896, 1904; Schmidt, M, 1937 a (Guisnay) ; Schuller, 1906. Choroti-Ashluslay.—Henry, 1939; Hunt, 1915 a; Karsten, 1932, pp. 225-230; Lehmann-Nitsche, 1910-11, 1936-37; Nordenskiédld, 1912, pp. 28-31; Pape, 1935; Rosen, 1904, p. 13. Macaé.—Belaieff, 1931-34, 1940; Brinton, 1898 a (Enimagdé); Hunt, 1915 b (Towothli); Huonder, 1902 (Enimagé); Kersten, 1905 (Lengua, Enimagdé, Guen- tuse) ; Koch-Griinberg, 1902 b (Enimagd) ; Kysela, 1931; Métraux, 1942 b; Schmidt, M., 1936 a, 1937 b. GUAICURU (WAICURU) Guaicuré was an important linguistic family of the Chaco region, but most of the languages are now extinct, and the surviving groups reduced to three or four with relatively few speakers. The family has always been accepted as independent, though several arguments for wider relationships have been made. When more care- ful linguistic studies are made it is not unlikely that Guaicurté and Mataco will fall together into a larger phylum to which Chiquito may also be added. This is the opinion of Lafone-Quevedo (1910), who considers Mataco a subgroup of Guaicurt, and both related to Chiguito; he also believes Quechua related to Guaicurt. All these languages have a superficial Hokan-like aspect which is not borne out by a Vol. 6] LANGUAGES—MASON 205 hasty comparison of vocabularies; phonetics, morphology, and pro- nominal systems are somewhat similar.” Guaicura, of course, must be distinguished from the Baja California language of identical or similar name. The languages fall into two, possibly into three, main groups. There is little disagreement among the various authorities regarding the relationships, and the adjoined table, compiled from these, varies little from any. Names of small groups or bands, ignored here, may be found elsewhere (Lozano, 1941, p. 62). The affinities between the various ‘‘dialects’”’ are said to be very close. Possible or doubtful members of the family are: Guachi.—Traditionally included but of doubtful affiliation. They may originally have had their own language, later abandoned for Mbaydé. Loukotka (1935) considers it a language mixed with Chiquito. Omitted by W. Schmidt (1926). Layana.—Generally considered Arawak, but placed by Nimuendaji (map and index) in Guaicurt. Juri (Suri)—Perhaps Guaicuri, probably sedentary Tonocoté. Querandi (q. v.).—Placed by Rivet (1924 a) in Guaicurté without any certainty. Others include Charria (q. v.). Mahoma or Hohoma.—Judging by linguistic position, according to Metraux (Handbook, vol. 1, p. 225), they may have been related to Toba or Mocovi. The relationships of Aguwilot and of Cocolot are based on historical, not on linguistic, evidence. Brinton (1891 a, p. 315) adds to his Guaicurté Family: Chica, Orején, Churumata, Malbalai, Matagayo-Churumata, Quiniquinauz, Tereno, and Yapitilagua or Pitilaga. Some of these are probably synonyms, others generally placed in other families. Loukotka (1935) lists the language Karraim, apparently mentioned by no other of the authorities consulted. A number of the tribes in this region seem to have adopted Guaicurt relatively recently. Prominent among these are the Tereno, Kini- kinao, Layand, and some of the scattered groups of Guand (q. v.), who apparently originally spoke Arawakan. They might therefore be classified in either of these ‘families,’ and are often differently classified by different authorities. On the accompanying linguistic map they are given as Arawakan. 18 See especially J. P. Harrington’s opinions (1943) on Quechua. 206 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 GvUAIcURG I. Guaicurt A. Northern 1. Mbaydé-Guaicurt% a. West: Caduveo (Cadiguegodi), Guetiadegodi (Guetiadebo) b. East: Apacachodegodegi (Mbayé Mirim), Lichagotegodi (Icachodeguo ?), Eytbogodegi, Gotocogegodegi (Ocoteguebo ?) ce. Payagué (Lengua): a. North: Sarigué (Cadigué) b. South: Magach (Agacé, Stacuds, Tacumbi) II. Frentones A. Middle 1. Toba (Tocowit) a. Toba: Guazt, Komlék, Michi (Mirt), Cocolot, Lanyaga- chek, Mogosma, Chirokina, Natica b. Pilagé c. Aguilot B. South 1. Abtpén (Callaga) a. Mapenuss (Yaukanigd) b. Mepene c. Gulgaissen (Kilvasa) 2. Mocovi (Mbocobi) BIBLIOGRAPHY Guaicuri.—Brinton, 1898 a; Hermann, 1908; Huonder, 1902; Imbelloni, 1936; Koch-Griinberg, 1902 b; Lafone-Quevedo, 1910 (1912); Lozano, 1941; Martius, 1867, 2:127—129; Schuller, 1906. Mbaya—Caduvéo—Payagua.—Adam, 1899; Aguirre, 1898, pp. 490-501 (Lengua); Boggiani, 1895, 1900; Castelnau, 1852, pp. 280-282; Gilij, 1780-84, pp. 367-371; Kersten, 1905; Koch-Griinberg, 1903, pp. 45-70; Lafone-Quevedo, 1892 b, 1896 c, 1897 b; Loukotka, 1929-30, pp. 99-106, 1933; SAnchez Labrador, 1896; Vellard, 1937; Vellard and Osuna, 1934. Toba.— Adam, 1899; Aguirre, 1898; Anonymous, 1933 a; Barcena, 1893; Cardus, 1886, p. 321; Ducci, 1904, 1905, 1911-12; Karsten, 1923, 1932, pp. 127-223, Ker- sten, 1905 (Toba, Pilagé, Aguilot); Koch-Griinberg, 1903, pp. 70-82; Lafone- Quevedo, 1893; Lehmann-Nitsche, 1925 a; Loukotka, 1929-30 (Toba, Pilagd); Nusser-Asport, 1897; Palavecino, 1931-33 (Pilagé); Tebboth, 1943. Abip6n—Mocovi.—Adam, 1899; Adelung and Vater, 1806-17; Aguirre, 1898, pp. 491-504; Dobrizhoffer, 1784; Ducci, 1911-12; Kersten, 1905; Lafone-Quevedo, 1892 a, 1892 b, 1892-93, 1893 a, 1893 b, 1896-97; Larrafiaga, 1924 a; Tavolini, 1856. Guachi.—Castelnau, 1852, pp. 278-280; Kersten, 1905; Martius, 1867, 2:131- 133. Tereno.—Baldus, 1937. LULE-VILELAN “Lulela” would be a good mellifluous hybrid term for this “family” if its validity is finally definitely established. The two groups have been linked in classifications since earliest days, but Loukotka (1935) separates them into two families. This suggests that they differ Vol. 6] LANGUAGES—MASON 207 greatly, with a possibility of nonrelationship. The terms applied to the joint group, however, have been many: Brinton (1891 a) and Chamberlain (1913 a) called it Zule, Nimuendajii (map, index) prefers Vilela; Loukotka (1935) uses both Lule and Vilela. Rivet (1924 a) and Pericot (1936) term it Vilela-Chunupt; W. Schmidt (1926) and Métraux (Handbook, vol. 1, p. 227) prefer Lule-Vilela, here adopted. Though it may be possible that a few Vi#lela-speaking Indians remain, the languages of the group are practically extinct. The linguistic data are relatively few. There is so much disagreement regarding the affiliations of languages in this region that it is not unlikely that many ‘‘families”’ will eventually be found to be related. Métraux suggests that a careful comparison with Mataco might prove significant. Other possible distant relatives are Diaguita, Macd, Sanavirén, Comechingén, Charria, etc. Even for this region there is an unusual amount of disagreement and question regarding the component languages of the group. Some authorities place Tonocoté (q. v.) with Lule; others put this language under Diaguita, but most consider it related to Mataco. There were two groups of Lule; the sedentary mountain Lule, the Lule of Barcena, spoke Quechua, Tonocoté, and Diaguita; the Lule of Machoni spoke Lule-Vilela. The Lule-Vilelan Chunupt (Chulupi, Sunupt) of the Bermejo River must be distinguished from the Mataco Chunupi (Choropt) of the Pilecomayo River. Loukotka (1935) includes Cacdn (Diaguita) and Sanavirén (q. v.) with Vilela; Jijén y Caamafio includes Sanavirén. Nimuendaji (map and index) apparently includes Giienoa, which all others consider as Charrta. Possible members of the family, according to Métraux, are Matard (q. v.) (Rivet, 1924 a: Vulela-Chunupi; Nimuendaji: unclassified), who were probably related to the Tonocoté (q. v.); Malbald (Rivet, 1924 a: Mataco; Nimuendajt: unclassified), who were associated with the Vilela; Palomo. LULE-VILELA I. Lule A. Great Lule (of Miraflores, of Machoni) B. Small Lule 1. Isistiné 2. Tokistiné 3. Oristiné II. Vilela A. Atalala B. Chunupi (Sinipé, Chulupi) 1. Yooc (Yoo, Wamalca) 2. Ocolé 3. Yecoanita 208 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 LuLe-VILELA—Continued II. Vilela—Continued . Pasain (Pazaine) . Omoampa (Umuapa) . Vacaa . Vilela Ipa . Takete . Yoconoampa (Yecunampa) - Wamalca . Malbalé ?) Aantorsoa -_- BIBLIOGRAPHY Lule.—Adelung and Vater, 1806-17; Brinton, 1898 a; Calandrelli, 1896; Huonder, 1902; Kersten, 1905; Lafone-Quevedo, 1894; Machoni de Cerdefia, 1732, 1894; Techo, 1673. Vilela-Chunupi-Choropi.—Adelung and Vater, 1806-17; Lafone-Quevedo, 1895 a; Lizondo Borda, 1938. TONOCOTE, MATARA, AND GUACARA These three extinct languages had best be left unclassified. All may be related. Rivet (1924 a) places them under Vilela-Chunupi, others with Mataco. Tonocoté is especially in dispute. Métraux (Handbook, vol. 1, p. 232) believes that the Matard spoke Tonocoté, which is included in the Lule region in the linguistic map herewith. Nimuendajt places Tonocoté with Mataco; the resemblance between the terms Tonocoté and Nocten is suggestive. They might also have been related to Diaguita, as Schmidt (1926) suggests. (See also Handbook, vol. 2, p. 657.) Bibliography.—Calandrelli, 1896; Lizondo Borda, 1938; Machoni de Cerdefia, 1732. ARAWAKAN Arawak is probably the largest and most important linguistic family in South America, both in extent and in number of component languages and dialects. It extends, or extended, from Cuba and the Bahamas, perhaps even from Florida, to the Gran Chaco and the sources of the Xingu, possibly even to Uruguay (Chand), and from the mouth of the Amazon to the eastern foothills of the Andes, possibly to the highlands (Uru), or even to the Pacific (Chango). In various groups, sometimes continuous, sometimes isolated, it ranges through- out this area. The distribution is very similar to that of the other great family of the tropical lowlands, the Carib. ‘The original home and point of distribution is supposed to have been the Orinoco and Rio Negro region of the borders of Guiana, Venezuela, and Brazil. If the Uru-Puquina languages are actually related to Arawak, that may Vol. 6] LANGUAGES—MASON 209 have been the first migration. Arawak languages seem to have been supplanted in places by Carib tongues, in other parts by Highland languages, Aymara and Quechua. The numbers of Arawak-speaking peoples are rapidly diminishing, and many tribes and languages are now extinct. Other names applied to the family have been Maipure (Gilij, 1780-84) and Nu-Aruac (Steinen, 1886). Several suggestions for wider relations have been made. If Arawak is ever linked in a phylum with other recognized families other than with small groups of present questionable independence, it will probably be with the Carib. A suggested tie-up with T'upt is less likely. Schuller (1919-20 a, 1928) believes in a great phylum including at least Arawak, Carib, Chibcha, and Maya, but he never presented cogent proof; his opinion has been accorded little consideration. ; A typical Arawakan language (Campa) shows absence of nominal incorporation. The pronominal subject is prefixed, the object suf- fixed. ‘There are temporal suffixes and modal prefixes. Verbal suffixes precede the pronominal object. The nominal plural is ex- pressed by a suffix. The same stem is generally employed for verb, noun, and adjective, the distinctions made by affixes. Arawakan languages generally have gender distinctions. The first person pronoun is usually nu, whence the generic name Nu-Arawak; the second person is generally p or pv. The correct grouping of the hundred-odd Arawak languages is an impossible task. Many of the extinct ones will never be classified with certainty, and the data on most of the living tongues are insuffi- cient. No comprehensive classification on a linguistic basis ac- companied by evidence has ever been attempted." Probably because of the large number of Arawak languages, and the poverty, both quantitative and qualitative, of the data upon them, no comparative Arawak grammar has yet been published. Rivet (1924 a), W. Schmidt (1926), and Loukotka (1935) have pre- sented classifications. These vary greatly; each contains certain languages considered independent by the others. Schmidt’s is the most detailed, with 7 main divisions and 16 subgroups. Loukotka has 14, 4 of which consist of a single ‘‘mixed” language. Rivet makes seven principal divisions. The main points of difference are: One of Schmidt’s groups is the Jivaro (q. v.), generally accepted as independent. Loukotka makes an independent family, the Araud, of some of the languages of the Araud or Jurud-Puris group. Schmidt considers the Tacana group as an independent family. Loukotka includes the Chamicuro, generally considered as Pano or Aguano. Rivet links the Goajiro and the languages of the Orinoco and the “i One may be expected in one of the promised volumes by Perea y Alonso (1942 ef seq.). 210 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 northern branches of the Amazon to those of the upper Xingi and the Paresst and Saraveca of Bolivia. In another division he joins the Arua group of the Jurudé-Purtis region with the Guand Group of the Paraguayan Gran Chaco. Since Rivet seems not to have presented the evidence for these unexpected groupings, and since they were not accepted in the later classifications of Schmidt and Loukotka, the more common geographical grouping has been herein accepted as the basis for classification, using the more detailed and less radical divisions of Schmidt as a base. There is general, but far from com- plete, agreement on the composition of the minor subdivisions. ARAWAKAN CLASSIFICATION I. Northern A. Insular ! 1. Lesser Antilles a. Igneri b. Cabre ? 2. Greater Antilles a. Taino b. Sub-Taino ce. Ciquayo d. Lucayo B. Northwestern 1. Goajiro * a. Goajiro: Cosina(?),4 Goburegual, Gimbuxegual b. Guanebucan c. Parauhano: Toa, Alile d. Tairona(?) > e. Chimila(?) § 2. Caquetio a. Caquetio:® Guaicari b. Achagua: Tayaga, Yaguai, Chucuna, Amarizana, ? Caourt ce. Tecua(?) d. Motilén of Catatumbo and Rio de Oro (?) 8 3. Guayupé a. Guayupé b. Eperigua c. Sae 1 Rivet (1924 a, pp. 249-250) does not mention this group in his classification of Arawak languages. 2 Probably identical with the Cabre or Caberre of the Orinoco. 3 Rivet puts Goajiro, Paresst, and Saraveca of Bolivia, the languages of the upper Xingt, and those of the Orinoco and northern Amazon in the same group. 4 Reichel-Dolmatoff (personal communication) says that, although located in the middle of the Goajira Peninsula, surrounded by Goajiros and always considered as Goajiran, the Cosina are not Goajiro and do not speak Arawakan. 5 See separate article on “Tairona and Chimila” in the Chibchan section. Reichel-Dolmatoff (personal communication) believes that the Chimila are Arawakan, if so, the Tairona probably were also. 6 W. Lehmann (1920) considers Caquetio as Chibchan, related to Betoi. 7 W. Schmidt (1926) classifies Amarizana as Carib. 8 Reichel-Dolmatoff (personal communication) says that, although the Motildn of the Sierra de Perijé are pure Cariban, those of Catatumbo and Rio de Oro are very different and seem to be Arawakan, though the linguistic materials are very scarce. Vol. 6] LANGUAGES—MASON 211 ARAWAKAN CLASSIFICATION—Continued I. Northern—Continued B. Northwestern—Continued 4. Piapoco (Dzase) a. Piapoco b. Cabre (Caberre) ® ce. Mitua II. Northern Amazon 1° A. Arawak 1. Arawak 2. Araua (n) B. Palicur 1. Palicur 2. Marawan C. Rio Branco 1. Wapishana (Wapiana, Wapityan) a. Wapishana b. Amariba 2. Atorai (Daur) a. Atorat b. Mapidian (Mayopityan) D. Orinoco Group 1. Guinau (Quinhao, Inao) a. Guaniare 2. Maipure 38 3. Mawacud 4. Yavitero (Paraene, Yavita) E. Indeterminate Group "4 1. Baniva a. Avant b. Quirruba 2. Baré 16 a. Baré b. Baratina 3. Arekena (Warekena) 4. Cariaya ® Related to and probably identical with the Cabre of the Insular Group. 10 Most of the languages below are listed by Gillin (Handbook, vol. 3, pp. 801-804). A few are added from other sources. Quite a number given by Gillin are here omitted. Tarwmd and Parauien are considered later herein. In addition to those that Gillin admits to be of questionable Arawak affinities—Apirua, Aramisho, Macapa, Marouriouz, Pino, Purui, Tocoyen—other authorities doubt three more. Nimuendajai leaves Arekena unclassified, Rivet considers Parauana as Cariban, and NimuendajG believes Pawishana to be Cariban. il The Araud migrated from Marajé Island to Guiana. (See Nimuendaji, Handbook, vol. 3, p. 195.) 12 This group contains only those languages that Rivet (1924 a) and W. Schmidt (1926) place in their Orinoco Group and Loukotka (1935) in his Guiana Group, except for Guinau which Loukotka places in the present Group C, the Rio Branco languages. 13 Gilij (1780-84) applied the name Maipure to the Arawak family. 14 Consisting of languages placed by Schmidt (1926) in his Orinoco Group, by Rivet (1924 a) in his Northern Amazon Group. 18 Baniva is a generic term employed for all Arawak-speaking groups in the Northwest Amazon region. The larger number of so-called Baniva languages are listed in the Rio Negro Group and the entire bibliog- raphy is therein. 16 Distinguish Baré from Bolivian Bauré. 794711—50——_15 212 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 ARAWAKAN CLASSIFICATION—Continued II. Northern Amazon—Continued F. Rio Negro Group 1. Izaneni (Baniva) Division a. Cartitana (Cazuzana): Yawarete-Tapuya, Baniva do Icana, Wadzoli dakenai, Mapache dakenet, Urubu- Tapuya, Dzawi-minanet, Adaru-Minanei, Arara- Tapuya, Yurupari-Tapuya. b. Catapolitant (Kadaupuritana) c. Caua- Tapuya (Maulieni) d. Cuati (Costi-Tapuya, Capité-Minanei) e. Huhutent (Hohodene) f. Mapanai (Ira-Tapuya) g. Moriwene (Sucuriyiu-Tapuya) h. Payualiene (Payoarini, Pacti-Tapuya) i. Siust (Walipéri-Tapuya): Ipeca-Tapuya (Cumata- Minanei) j. Tapiira 2. Miritiparané Division a. Cauyari (Karyart) b. Matapi ec. Yucuna d. Menimehe 3. Mawaca Division a. Adzaneni (Tatu-Tapuya) b. Mandawaca c. Masaca d. Yabaana 4. Tariana Division a. Tariana b. Itayaine (Iyaine) 5. Yapurd Division A a. Wainumd (Uainumd) b. Mariaté 6. Yapurd Division B ”” a. Cayuishana (Cawishana) b. Pasé (Passé) c. Yumana (Chimana) d. Manao e. Aruaki 7. Wirind 8 (Uirina) III. Pre-Andine A. Amazonian 1. Marawa * 2. Waraicté (Araiku, Uraicu, Wareku) 1” Loukotka, (1935) separates the Yapura Group as generally accepted, and places the last three languages in a separate group as ‘‘Languages mixed with Macia,”’ 18 Schmidt places Wirind in a group by itself. 1” W. Schmidt (1926) distinguishes between the Pre-Andine (Montafia) and the Jurué-Purts languages, but his division of these is greatly at variance from that of Métraux and Steward (q. v.) generally accepted herein. Loukotka (1935) considers them all as Pre-Andine. The division is probably purely a geographical one, with border-line instances; linguistically probably all fall together. The Pre-Andine languages are said to differ little from those of the North Amazon. (See Rivet and Tastevin, 1919-24.) 2% Distinguish from Marawan of Guiana. > Vol. 6] ARAWAKAN CLASSIFICATION—Continued LANGUAGES—MASON III, Pre-Andine—Continued B. Cutinana Group 1. Cutinana 2. Cuniba 21 3. Cujisenayert (Cujigeneri, Cushitineri) C. Jurudé-Purts 1. Canamari 22 2. Catukina 3. Catiana 4, Inapari 5. Ipurind 4 (Hypurina) a. b. Cangutu Casharart 6. Maniteneri 7. Wainamari (Uainamari) D. Montafia (Chuncho) 1. Campa COND or A Oo IV. South A. Bolivia 1. Bolivia a. Mojo (Moxo): Muchojeone b. . Anti . Antaniri (Unconino) . Camatica . Campa (Atirz) Catongo . Chicheren Chonta . Kimbiri . Kirinairi . Pangoa . Tampa . Ugunichiri Unint . Manatinavo . Chontakiro . Simirinch Upatarinavo - Machiguenga (Amachengue) Masco * . Sirineri . Wachipairi (Huachipari) . Puncuri . Pucpacurit Bauré 2. Chiquito a. Paiconeca, Paunaca 41 Distinguish from Panoan Conibo. % Distinguish from Panoan and from Catukinan Canamari or Canamare. % Distinguish from Catukina “family.” 4 Formerly considered an independent family by Chamberlain (1913 a) and Brinton (1891 a). 35 Aza (1935) writes of the “‘Arasaire or Mashco.”” The former are generally regarded as Panoan. 213 214 SOUTH AMERICAN INDIANS LB. A. B. Bull. 143 ARAWAKAN CLASSIFICATION—Continued IV. South—Continued B. Paressi * (Arittz) ‘1. Cashiniti a. Waimaré 2. Iranché 27 a. Sacurit-ind b. Tahuru-ind ce. Timaltiéd 3. Cozdrini a. Wild Cabishi b. Paressi-Cabishi c. Mahibarez C. Saraveca D. Parané %8 1. East: Guand * (Chuala, Chand) a. Layand (Niguecactemigi) b. Tereno c. Echoaladi (Echenoana, Chararana) d. Kinikinao (Equiniquindo) 2. West: Chané?® a. Izocento EK. Xingt 1. Xingi a. Mehinacti b. Yaulapiti (Jaulapiti) ce. Custenau (Kustenahi) d. Waurd (Uaure) % Métraux (Handbook, vol. 3, p. 349, ftn. 1) says that Paresst is closer to Mehinact than to Mojo. 27M. Schmidt (1942) claims that Iranché is not Arawakan (Métraux, Handbook, vol. 3, ftn. 2, p. 349). 28 Many of these groups, such as the Layand, Tereno, Kinikinao, and probably some others have abandoned their former Arawak speech and now speak Guaicurti. They are, therefore, properly placed under Guaicurt in some classificatory systems. 29 See following article on ‘‘Chané and Chana.” BIBLIOGRAPHY General and unidentified—Adelung and Vater, 1806-17; Brinton, 1869 a, 1871; Chamberlain, 1913 b; Farabee, 1918 b; Goeje, 1928 b, 1929-30; Gumilla, 1745; Koch-Grinberg, 1911, pp. 38-158, 203-282; Perea y Alonso, 1937, 1938 a, 1942; Quandt, 1807, 1900; Roth, 1924; Schmidt, M., 1917; Schuller, 1919-20 a; Schultz, 1850 T.; 1850; Tello, 1913 b; Williams, J., 1924. Insular.—Adelung and Vater, 1806-17 (Haiti); Goeje, 1939; Martius, 1867, 2:314-319 (Cuba, Haiti); Tastevin, 1919. Goajiro.—Candelier, 1893; Celed6n, 1878; Isaacs, 1884; Jahn, 1914, 1927, pp. 355-376; Marcano, 1890 b; Oramas, 1913 a, 1918 a; Tello, 1913 b; Uteaga, 1895. Parauhano.—Jahn, 1914, 1927, pp. 190-197; Oramas, 1918 a, 1918 b. Caquetio.—Jahn, 1927, pp. 199-223. Achagua.—Adelung and Vater, 1806-17; Alemany y Bolufer, 1929 a, 1929 b; Gilij, 1780-84, p. 346; Jahn, 1927, pp. 377-378; Oramas, 1916. Piapoco.—Crévaux, Sagot, and Adam, 1882, pp. 242-249; Koch-Griinberg, 1928, pp. 287-301; Tavera-Acosta, 1907, pp. 85-95. Cabre.—Gumilla, 1745; Roth, 1924. Vol. 6] LANGUAGES—MASON 215 Arawak.—Brinton, 1871; Crévaux, Sagot, and Adam, 1882; Ferreira-Penna, 1881 (Aruan); Goeje, 1928 a; Martius, 1867, 2:307-311; Mordini, 1935 (Aura’an); Penard, T. E., 1926-27; Stahel, 1944; Tavera-Acosta, 1907, pp. 333-335. Palicur-Marawan.— Martius, 1867, 2:324; Nimuendaji, 1926; Rivet and Rein- burg, 1921. Wapishana.—Coudreau, H., 1887, pp. 477-87; Farabee, 1918 b, pp. 13-132, 183-274; Grupe y Thode, 1890, p. 254; Koch-Griinberg, 1922, p. 220 (Amariba) ; Koch-Griinberg and Hibner, 1908, pp. 35-39, 44; Schomburgk, 1847-48, (Amariba). Atorai-Mapidian.—Farabee, 1918 b, pp. 158-164, 274-277, 283-286; Schom- burgk, 1847-48, pp. 515-523. Yavitero and Orinoco Group.—Adelung and Vater, 1806-17 (Maipure); Chaff- anjon, 1889 (Guinao); Crévaux, Sagot, and Adam, 1882; Gilij, 1780-84, pp. 185-90, 202-13, 375-82 (Maipure); Koch-Grinberg, 1909-10, 1928 (Guinao); Loukotka, 1929-30, p. 85; Martius, 1867, 2:312-313 (Guinao); Montolieu, 1882, pp. 281-284; Tavera-Acosta, 1907, pp. 63-74. Baré Group.—Chaffanjon, 1889, pp. 330-333; Chamberlain, 1910 a (Cartaya) ; Crévaux, Sagot, and Adam, 1882, pp. 251-252; Koch-Griinberg, 1909-10, pp. 56- 153 (Baré, Warekena), 1913, p. 455 (Arecuna), 1928, pp. 246-257 (Arecuna), pp. 272-278; Martius, 1867, 2:230-231, 285-286, 231-232 (Cariay); Montolieu, 1882, 1895; Nimuendaji, 1931-32, pp. 592-595 (Baré, Uarekena); Tavera-Acosta, 1907, pp. 63-84 (Baré, Uarekena); Tello, 1913 b. Baniva.—Chaffanjon, 1889, pp. 337-341; Crévaux, Sagot, and Adam, 1882, pp. 253-255; Gumilla, 1745; Koch-Griinberg, 1909-10; La Grasserie, 1892; Martius, 1867, 2: 261-263; Montolieu, 1882, pp. 276-280; Nimuendaju, 1931-32, pp. 590- 592; Tavera-Acosta, 1907, pp. 53-62. Baniva-Tapuya Group.—Cardona Puiz, 1945 (Karro); Koch-Griinberg, 1909- 10, 1911, pp. 56-153, 203 (Carutana, Katapolitani, Siusi); Nimuendaji, 1931-382, pp. 596-618 (Karwitana, Kadaupuritana, Moriwene, Waliperi-Dakenai, Hohodene, Mapanai, Mazilieni, Payualiene, Kumada-Mnanai, Kapité-Mnanet); Tavera- Acosta, 1907, pp. 76-84 (Carvitana); Wallace, 1853. Cauyari-Yucuna.—Crévaux, Sagot, and Adam, 1882; Koch-Grinberg, 1909-10, 1911, pp. 56-153, 203-257; Martius, 1867, 2: 253. Adzaneni-Mandawaca.— Koch-Griinberg, 1928, pp. 288-301; Loukotka, 1929- 30, p. 85; Nimuendaju, 1931-32, pp. 613-614; Tavera-Acosta, 1907, pp. 63-74. Tariana.—Coudreau, H., 1887; Koch-Griinberg, 1909-10, 1911, pp. 268-281; Martius, 1867, 1: 628-629 ( Yaboana), 2: 260; Wallace, 1853. Wainuma-Mariaté.— Martius, 1867, 2: 245-249, 266-268; Wallace, 1853. Yapura Group.—Brinton, 1892 a (Manao); Martius, 1867, 2: 221-222 (Manao), 229 (Uirina), 250-252 (Jumana), 254-256 (Passé), 257-260 (Cauixana). Marawa-Waraici.— Martius, 1867, 2: 223-225, 233-234. Cuniba.—Nimuendajt and Valle Bentes, 1923, pp. 215-217. Juru4-Puris.—Chandless, 1866 (Canamari, Maniteneri) ; Martius, 1867, 2: 161- 163 (Catukina), 235-236 (Canamari); Rivet, 1920 b (Catukina); Rivet and Taste- vin, 1919-24; Stiglich, 1908 (Inapart). Ipurina.— Chamberlain, 1910 a, p. 188; Chandless, 1866, p. 118; Ehrenreich, 1897 b; Koch-Griinberg, 1914-19; Nusser-Asport, 1890, p. 795; Polak, 1894; Steere, 1903, pp. 378-380; Tello, 1913 b. Campa-Anti.— Adam, 1890 b; Cardus, 1886, p. 325; Carrasco, 1901, pp. 205-211; Castelnau, 1852, pp. 290-291; Delgado, E., 1896-97; Farabee, 1922, pp. 21-52; Marcoy, 1875, 1: 548; Pauly, 1928, p. 151; Reich, 1903, p. 135; Sala, G., 1905-06; Steinen, 1906; Tello, 1913 b; Tessmann, 1930, p. 83; Touchaux, 1908; Weiner, 1880. 216 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Piro-Chontakiro.—Alemany, 1906 a; Carrasco, 1901, pp. 205-211; Castelnau, 1852, pp. 291-292; Farabee, 1922, pp. 62-76; Marcoy, 1875, pp. 579-580; Reich, 1903, p. 1385; Tessmann, 1930, p. 366. Machiguenga.—Anonymous, 1933 d; Aza, 1923, 1924 a, 1924 b, 1924 c, 1933 a; Rosell, 1916. Masco.—Aza, 1935; Farabee, 1922, pp. 77-78. Mojo-Bauré.—Adam and Leclerc, 1880 (Bauré); Adelung and Vater, 1806-17 (Mozo) ; Cardus, 1886, pp. 317-318; Magio, 1880 (Bauré); Marban, 1894 (Mojo); Métraux, 1942 a, pp. 53-80; Pauly, 1928, pp. 157-158; Tello, 1913 b. Paiconeca-Paunaca.—Cardus, 1886, pp. 319-320, 327; Pauly, 1928, pp. 164-166. Paressi.—Rondén, 1910, pp. 19-28; Schmidt, M., 1914 a, pp. 242-250; Steinen, 1894, pp. 542-547. Saraveca.—Cardus, 1886, p. 327; Pauly, 1928, pp. 164-166; Créqui-Montfort and Rivet, 1913 c. Parana Region.—Bach, 1916; Baldus, 1937 (both Tereno); Boggiani, 1896; Castelnau, 1852, pp. 274-276; Martius, 1867, 2: 129-131; Schmidt, M., 1903 (all Guand); Taunay, 1868, pp. 181-148. Xingi Group.—Steinen, 1886, pp. 357-360 (Custenau); 1894, pp. 523-532 (Mehinaci, Yaulapiti, Custenau Waura). CHANE AND CHANA The name Chané is applied especially to several small isolated enclaves of Arawak-speaking peoples, the southernmost Arawak groups. It is, however, unfortunately, frequently confused with Chand. Thus Brinton (1891 a) lists the Chané among the Charrian (q. v.) tribes of Uruguay; these are today known as Chand (q. v.). It was probably this analogy that led Perea y Alonso (1942) to claim the Charria to be Arawak. On the other hand, certain Arawak groups, especially the Layand, seem to be known as Chand. Guand is probably a term related to Chand. LANGUAGES OF PROBABLE ARAWAKAN AFFINITIES ARAUA GROUP !” The nature and composition of the group of Araud languages are much disputed. Brinton (1891 a, p. 293) made an Araud stock, composed of Araud, Pama, Pammary, and Purupuri. Loukotka, in his 1935 classification, also proposed an Araud family, but made it composed of Araud, Yamamadi, and Pammart; however, in 1939 he put the group back under Arawak and added the languages Kulina and Madiha. Nimuendaji (map) accepts Yamamadi, Pammary- Purupuri, Yuberi, and Culino as Arawak but refuses to classify Araud, Sewacu, Pama, and Pamana. Rivet (1924 a) includes all these in his Araud group of Arawak, and considers the languages to fall with 17 The Arawakan Araudé must be distinguished from a small Panoan group on the Madre de Dios River and from several other groups with somewhat similar names. Vol. 6] LANGUAGES—MASON 217 the Guand-Tereno-Layand group of Paraguay. The following classifi- cation is, therefore, very tentative: AravA Group 1. Araud 2. Culino a. Culina b. Curia ce. Curiana d. Culitia 3. Pama a. Pama b. Pamana 4. Yamamadi a. Yamamadi: Capandé, Capinamari, Colo b. Purupurté: Paumart (Pammart) ce. Yubert 5. Madiha 6. Sewacu 7. Sipd Bibliography.—Carvalho, 1929, 1931, pp. 246-248 (Culina); Chandless, 1866, p. 118 (Paumart), 1869, p. 311 (Araud) ; Ehrenreich, 1897 b (Pammari, Yamamadi) ; Rivet and Tastevin, 1938-40; Steere, 1903, pp. 386-387 (Yamamadz), 390-393 (Paumart). APOLISTA OR LAPACHU Chamberlain (1910 a) established an independent Apolistan family, based on early data. Crequi-Montfort and Rivet (1913 d) joined this to the Pre-Andine group of Arawak, mainly on the basis of a small vocabulary collected by Nordenskidld from one of the last speakers. This classification has been accepted by all recent authorities. Lou- kotka (1935) finds vestiges of Leco in the language which was known as Lapachu (Lapatu, Lapatsu) and has now been replaced by Quechua. The Apolista may be descendants of the Aguachile (Métraux, Hand- book, vol. 3, p. 506). Bibliography.— Cardus, 1886; Chamberlain, 1910 a, pp. 179-180; Créqui-Mont- fort and Rivet, 1913 d; Métraux, 1942 a, pp. 29-30; Nordenskidld, 1905; d’Orbigny, 1839, pp. 173-174. AMUESHA Synonyms: Amagues, Amage, Amaje, Amoeshe, AmoiSe, Amoische, Amoiza, Amueiza, Amuesa, AmueSa, Amuese, Armueshe, Lorenzo, Omaje, San Lorenzo. The Arawak affiliations of Amuesha are questioned. Chamberlain (1913 a) considered it an independent family, the Lorenzan. Tello (1913 b) first suggested its Arawak affinities, but despite this Rivet (1924 a) preferred to classify it as independent. Loukotka (1935) places Amoige with the Pre-Andine Arawak; Tessman (1930) sees Tupi elements in a mainly Arawakan language. Jijén y Caamafio 218 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 (1941-43) gives it independent statusas a phylum. Steward and Mé- traux herein (Handbook, vol. 3, p. 536) call the Amuesha “‘linguistically similar to the Campa’’; this statement does not seem to be borne out by the evidence. Mr. Louis Rankin writes (personal correspondence) from personal acquaintance, that, ‘‘The Amuexias to the west of the Campa are said to be a subtribe, but their language is quite different.” They have for some time spoken Quechua. The Lorenzo and Pana- tawa are, or were probably related. Bibliography.—Chamberlain, 1910 a, p. 191 (Lorenzan); Farabee, 1922; Iza- guirre, 1927-29; Sala, G., 1897, 1905-06; Tello, 1913 b; Tessmann, 1930, pp. 367- 368, 617. TUCUNA (TIKUNA) Nimuendajti (Handbook, vol. 3, p. 713) advances arguments for his opinion that, following Chamberlain (1910 a) and Tessmann’ (1930), Tucuna should be considered independent or isolated, not placed under Arawak, following Rivet (1912 b, 1924 a), who thinks it a very altered Arawakan tongue. However, W. Schmidt (1926), Krickeberg (1922), Loukotka (1935), and Igualada and Castellvi (1940) accept the Arawakan connection. Loukotka thinks it is mixed with Mura and Tucano. Bibliography.—Brinton, 1892 a, pp. 7-20; Castelnau, 1852, pp. 298-299; Chamberlain, 1910 a, p. 198; Marcoy, 1875, p. 379; Martius, 1867, 2:159-161 (Tecuna) ; Nimuendaji, 1931-32, pp. 573-580 (Tikuna) ; Rivet, 1912 b; Tessmann, 1930, pp. 564-565, 617 (Tikuna). TARUMA Tarumd has been generally classed as an Arawakan language (Rivet, 1924 a; Loukotka, 1935; W. Schmidt, 1926; Gillin, Handbook, vol. 3, p. 803), but Nimuendajii (map) places it among his isolated lan- guages; this opinion is apparently based on no new published data. If Arawakan, it is apparently an unusually variant form, since Lou- kotka (1935) puts it in a subgroup of its own as a mixed language (other element not stated), and with vestiges of Camacdn; the latter is most doubtful. Rivet (1924 a) states that it was related to the extinct Parawien. Bibliography.—Farabee, 1918 b, pp. 185-138, 277-283. TACANA Synonyms: Takana, Tecand. There are three linguistic groups in northwestern South America known by variations of the ¢-k-n phonetic combination; with the inevitable vowel modifications they are, therefore, liable to confusion. The standard spellings of these three tribes are Tacana, Tucuna or Vol. 6] LANGUAGES—MASON 219 Tikuna, and Tucano; each has been formerly accorded mdependent position. The linguistic position of the Jacana group is a most uncertain and controversial question, and one that will require much intensive study for a definitive opinion. Tacana was accorded independent status by the early authorities, Brinton (1891 a) and Chamberlain (1913 a). As a result of an exhaustive comparative study, Créqui-Montfort and Rivet (1921-23) put it under Arawak, in which opinion they are followed by Rivet (1924 a), Pericot y Garcia (1936), and Loukotka (1935). W. Schmidt (1926), Krickeberg (1922), K. G. Grubb (1927), and the authors of the monographs in this Handbook wisely prefer to leave it as independent, or at least unclassified and doubtful. Coterminous with both Panoan and Arawakan languages, the Tacanan languages show resemblances to both; the resemblance to one should be genetic, to the other the result of borrowing. Morpholog- ically, the resemblance is much greater with Panoan, a fact that should carry great weight for genetic connection. Some 65 of the 101 words compared by Rivet are either identical or very similar in Tacanan and Panoan, so similar that the presumption is for recent borrowing, although the words are mainly basic ones, and few are in modernistic categories. Of the 101 words compared, 60 occur in only one language, or in one small group of languages, either Panoan or Tacanan, and are, therefore, presumably not original in these stocks; another 17 seem to be common also to Arawakan, leaving only 24 really pertinent cases. Regarding the Arawakan resemblances, since 25 Tacanan vocabu- laries are compared with 65 Arawakan ones, a large number of fortui- tous apparent resemblances would be expected; many of them occur in only one language; in many others the meaning is greatly changed. Of the 178 examples only a dozen or so would qualify as apparent certainties, and half of these are of domesticated plants or animals, such as dog, cotton, maize, manioc, and tobacco. No rules of sound change are suggested and none are apparent. The genetic relation- ship of Tacanan to Arawakan requires much more careful study before it can be accepted. Tacanan has also many words in common with Aymara and Quechua, but these are almost certainly borrowings, mainly from Aymara. Armentia (1902) gives the names of some 40 subtribes or dialects of Araona, some of which are also found in the table below. Araona and Cavina are inextricably mixed, but some groups are pure Araona, and some pure Cavifa. Cavifia and Cavinefio are not synonymous, ac- cording to Rivet, and the latter not a subdivision of Araona. Rivet also does not group Guacanagua, Sapibocona, or Maropa with any other languages. He distinguishes between Toromona and Turamona, the 220 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 latter a Tacana subgroup. There are no data on Guacanahua, but the Tacana affinities are vouched for by Cardus (1886) and Norden- skidld (1905). The extinct Sapibocona are probably the same as the Maropa. Rivet considers Chiragua a subgroup of Tacana. Some of the Arasa speak Tacanan, but the group is really Southwestern Pano (Arasaire), and is also classified under Pano; the habitat is the same. Nordenskidld’s (1905) Arasa vocabulary is Tacanan; Llosa’s (1906) Arasaire vocabulary, Pano. Brinton (1891 a) also gives as subtribes Hguari, Samachuane, Carangue, Hucumano, and Torococy, which Rivet claims cannot with certainty be identified with Tacanan, as being extinct without re- corded data, or known by other names. No one has attempted to subdivide the Tacana group or to classify the component languages on a scientific linguistic basis. The following table incorporates the opinions of all authorities consulted, and greatly contravenes none. TACANA A. Araona Arauna, Arahuna) 1. Capachene (Kapaheniz) 2. Cavitia (Kavina) 3. Cavineiio 4. Mabenaro 5. Machui (Machuvi) B. Arasa C. Chirigua (Chiriba, TSirigua, TSiriba) 1. Chumana 2. Maropa 8. Sapibocona (Sapiboka) . Guariza (Guaziza) Tacana (Takana, Tucana) . Ayaychuna . Babayana . Chiliuvo . Chivamona . Idiama, Isiama . Pamaino . Pasaramona . Saparuna . Stliama . Tumapasa or Maracani 11. Turamona (Toromona) 12. Uchupiamona 13. Yabaypura 14. Yubamona F. Tiatinagua (Tambopata-Guarayo) 1. Guacanahua (Guanacanahua, Guarayo 1) ole bee SMVANADAD &OWOM 1 Distinguish from Tup{-Guarani Guarayo (Huaraya, Guarayu, etc.; some of the bibliographical references there noted possibly apply here instead, or vice versa). Vol. 6] LANGUAGES—MASON 221 Tacana—Continued F. Tiatinagua (Tambopata-Guarayo)—Continued . Chama . Baguaja (Baguajairi) . Chunchu . Echoja . Huanayo . Kinaki . Mohino G. Yamaluba BrNADAaAR & & BIBLIOGRAPHY (For full bibliography to that date, see Créqui-Montfort and Rivet, 1921-23.) Adelung and Vater, 1806-17 (Sapibocona) ; Armentia, 1888, 1902, 1904; Aza, 1928, 1930-32 (Huaraya); Brinton, 1892 a, pp. 7-21; Cardus, 1886, pp. 311 (Tacana), 414-415 (Maropa); Farabee, 1922, pp. 154-161 (Tiatinagua), 163-164 (Mabenaro) ; Giacone, 1939; Gilij, 1902; Groeteken, 1907; Heath, E., 1883 (Tacana, Maropa); Hoeller, 1932 a, 19382 b (Guarayo); Métraux, 1942 a, pp. 30-45; Nordenskidld, 1905, pp. 275-276 (Arasa), 1911 b, pp. 235-239 (Maropa) ; Pauly, 1928, pp. 121-124 (Tacana), 130-131 (Maropa), 133-134 (Cavina), 147 (Guacanagua); Schuller, 1933; Teza, 1868, pp. 117-143 (Guarisa). LANGUAGES OF POSSIBLE ARAWAKAN RELATIONSHIPS TUYUNERI The most recent compilers, Nimuendaji (map and index) and Loukotka (1935), prefer the spelling Tuyonerit to the standard Tuyuneri. This group is of later and less generally accepted standing than Itonama, Canichana, Cayuvava, Movima, and Yurucare (q. v.) in this region, and distant from them; it was discovered by Nordenskidld (1906) in the early years of thiscentury. Tuwywmiri, assigned by Brinton (1891 a) to Tacanan, is probably an orthographical error; it is not mentioned by Chamberlain (1910 a, 1913 a). Markham (1910) identi- fies the Tuyuneri with the Chunchos, a generic name for Indians of the Montafia and hence a meaningless association. Rivet (1924 a), Pericot y Garéia (1936), Loukotka (1935), and Nimuendajt (map and index) accept it as an independent family or as isolated; Loukotka sees vestiges of Panoan in it. However, Steward and Métraux (Hand- book, vol. 3, p. 541) place it unequivocally among the Arawakan groups. Bibliography.— Nordenskiéld, 1905, pp. 275-276. JIRAJARA Not mentioned by earlier writers, Jirajara has been accorded independent position by Rivet (1924 a), Loukotka (1935), and some other recent authorities. W. Schmidt (1926) follows Oramas (1916) in considering it related to Arawak, which may well be found to be 222 SOUTH AMERICAN INDIANS [B. A. B. Bull. 148 the case when more scientific studies are made on its vocabulary and grammar. Herndndez de Alba (Handbook, vol. 4, p. 469) dogmatically states that ‘‘the Jirajara . . . speak an Arawakan language (Oramas, 1916).”” The most recent opinion, however, that of Febres Cordero (1942) is that it is not Arawakan, though containing many Arawak words, probably borrowed. Also about 10 percent of the words seem to show Chibchan connections. The Ajdgua, given as a component language, may be synonymous with the Achagua, generally considered as Arawak. They may, however, be a separate group. The Cwiba, probably distinct from the Guahibo group of the same name, may be an Ajagua dialect. JIRAJARA 1. Gayén (Cayon) 2. Ayoman 3. Xagua a. Cuiba (?) 4. Jirajara Bibliography.—Febres Cordero, 1942; Jahn, 1927, pp. 379-395; Oramas, 1916. JIVARO The Jivaro family has always been known by orthographic variants of this name, such as Xivaro and Chiwaro; it is probably a corruption of Shuara or Shiwora, their own term. The resemblance to the name of a neighboring family, Zdparo, may be significant, but no genetic relationship with the latter has been suggested. The name apparently became used to imply a wild rustic person and is applied in Puerto Rico to the native countryfolk of the interior mountains. They must be distinguished from the Cawapanan Chébero (Xébero) and from the Hibito. The language is still spoken by some thousands of Indians, but several groups have adopted Quechua. Except for a few borrowed words, Jivaro seems to have nothing in common with Quechua, Tupian, Cawapanan, Zdéparoan, or Panoan. There are, however, a large number of apparent correspondences with Arawakan, the resemblance with Campa being especially strong. This may possibly be due to borrowing, especially since there are some important morphological differences. Beuchat and Rivet (1909-10) hesitatingly decided to place Jivaro in the Arawakan family, but in his later classification (1924 a) Rivet again gave it independent status, in which he has been followed by all other authorities except W. Schmidt (1926). J. P. Harrington (personal correspondence), how- ever, believes that the Arawak resemblances are genetic and that Jivaro is a very divergent form of Arawak. Jivaro is said to be clear and harmonious. The phonetic pattern is more like that of Amazonian than that of Andean languages. There Vol. 6] LANGUAGES—-MASON 223 is a quasi-inflection, that is, terminal changes or suffixes for person and tense. Mechanism for pluralization is absent, and there is no trace of gender. Both classificatory prefixes and suffixes are found, and post- positions. Monosyllables are rare, and accent unimportant. There may be said to be but one Jivaro language, relatively homo- geneous, but very many dialects. Apparently no attempt has ever been made to subdivide the language, or to group the dialects. The subdivisions as generally given are presumably political and geo- graphic, but the presumption is that the linguistic division would be roughly similar. JivARO I. Jivaro A. Shuara 1. Aguaruna . Alapico . Indanza . Iransa . Maranza . Santiago Patocuma Chiguasa . Yuganza 2. Wambisa Uambisa . Cherembo Chirapa . Chiwando Candoa Cangaime Mangosisa huale . Capawari . Copatasa Machine . Pindu e. Wampoya 4. Antipa 5. Maca a. Walakisa b. Zamora ce. Pintuc d e ram o ao of RMD Ao oP Ned pa 8 aoe . Ayuli . Morona . Miazal 6. Upano 7. Bolona 8. Bracamoro (Pacamuru) B. Palta! 1. Malacata ' See Murra, Handbook, vol. 2, p. 80. eh 224. SOUTH AMERICAN INDIANS [B. A. B. Bull, 143 BIBLIOGRAPHY Anonymous, 1918 b, 1924, 1939, 1941; Beuchat and Rivet, 1909-10; Briining, E., 1904; Briining, H., 1928 (Aguaruna); Caillet, 1930-33; Cordeiro, 1875; Delgado, A., and Vacas Galindo, 1929; Domville-Fife, 1924, pp. 209-218; Duroni, 1924, 1928; Dyott, 1926, pp. 153-212; Farabee, 1922, pp. 125-135; Flornoy 1938; Ghinassi, 1938, 1939; Gonzdlez Sudrez, 1904, pp. 51-62; Hassel, 1902 fAguaruna); Jij6n y Caamafio, 1919, pp. 380-388; José de Marfa, 1918-19; Karsten, 1919, 1920 a, 1920 b, 1921-22, 1922, 1935; Leén, A. M., 1928-29; Magalli, 1890, 1891 (1912); Prieto, 1885, pp. 63-68; Rimbach, 1897, pp. 360-409; Rivet, 1907-08, 1912 a, Romero y Cordero, 1930; Simson, 1886, pp. 87-100, 1899; Tessmann, 1930, pp. 338-365 (Chiwaro); Vacas Galindo, 1895, 1903 a, 1903 b, pp. 402-418; Verneau and Rivet, 1912. URU-CHIPAYA-PUKINA The relationship of Uru (Uro) and Pukina (Puquina) to Arawakan is quite illogical. The Uru-Puquina inhabit the region of Lakes Titicaca and Poopé in Bolivia, about the highest, coldest, and most inhospitable area in South America; the majority of the Arawak languages are in the Tropic lowland forested regions. The evidence advanced for the affiliation (Tello, 1913 b; Créqui-Montfort and Rivet, 1921, 1925-27) falls far short of proof, but it has been accepted by practically all the recent authorities on classification: W. Schmidt (1926), Pericot y Garcia (1936), Loukotka (1935), Jijén y Caamajfio (1943), etc. The relationship was first suggested by Tello (1913 b), the data for proof presented by Créqui-Montfort and Rivet (1925-27). Several of the “Handbook” authors (see La Barre, Handbook, vol. 2, p. 575), including the present one, consider the evidence advanced insufficient, doubt the connection, and think that the data should be reviewed. Dr. J. P. Harrington, however, is convinced of its validity. Uhle (1896) suggested a relationship to Yunca-Mochica, and Loukotka (1935) calls them mixed languages, with vestiges of Pano and Mose- tene. Many writers believe that the present Uru group is but a tiny remnant of a very early or autochthonous population that once occu- pied a much larger region, extending to, and including a large area on, the Pacific Coast. (See Jij6n y Caamafio, 1941-43, map 3.) If the result of an Arawak migration, it was probably the first of these. Three languages, Uru, Pukina, and Chipaya, are ordinarily placed in this group. The published vocabularies, however, show such differentiation that even the interrelationship of these is not beyond question. Uhle (1896), Polo (1901), and Boman (1908) believed Pukina and Uru distinct, and Chamberlain (1910 a, 1913 a) distin- guished Puquinan and Uran families. Posnansky (1915) considers Chipayan an independent family distinct from the others. La Barre (Handbook, vol. 2, p. 575) says that the Urw “call their language Vol. 6] LANGUAGES—MASON 225 Puquina,” but that the Uru language ‘“‘is not the same as the Puquina- Uro of La Grasserie (1894).” The data on the Uru group of languages seem to be exclusively lexical; grammatical material is a great desideratum. BIBLIOGRAPHY Adelung and Vater, 1806-17; Bacarreza, 1910, pp. 477-480; Barcena, MS. Basadre, M., 1884, pp. 196-205; Brinton, 1890 (all Puquina); Chamberlain, 1910 a, pp. 196-197, 200, 1910 b; Créqui-Montfort and Rivet, 1921, 1925-27; Franco Inojosa, 1937; La Grasserie, 1894 (Puquina); Métraux, 1936 a, 1936 b; Polo, 1901, p. 456; Posnansky, 1915, 1918 (both Chipaya), 1931, 1934; Tello, 1913 b; Uhle, 1896. OCHOSUMA Ochosuma or Uchuzuma may be a dialect of Uru, but had best be left with the unclassified languages. CHANGO AND COAST URU Insufficient data are available to classify Chango, an extinct lan- guage of miserable fishermen on the Chilean coast. As probable remnants of an early archaic population, an independent language is not unlikely, but this possibility is insufficient to justify the establish- ment of a separate family for them as Chamberlain (1913 a) did. The only data seem to be place and personal names, and the state- ment that they spoke a language different from their neighbors. Different opinions have placed them with the Atacamefio, Chono, and Alacaluf. ‘The most recent and thorough studies link them with the Uru (q. v.) of the Bolivian lakes, which linguistic group Rivet believes to be of Arawakan affinities. The argument is apparently based mainly on the fact that some groups adjacent to the Chango were known as Uru, and on a comparison of Chango names with Bolivian Uru. It is probable that the name Uru was applied to a number of nonrelated linguistic groups, just as the Lacandén in Chiapas are locally called ‘‘Caribs,” and Puerto Rican countryfolk “Jivaros,” and the existence of a group of true Uru on the Chilean coast is unlikely. At any rate the sources do not equate Chango and Uru. The suggestion that the Bolivian Uru had seasonal fishing colonies on the coast is improbable. Brand (1941 c) distinguishes between the Northern Chango or Uru, whom he believes to be linguistically Uran, and the Southern or True Chango, sometimes wrongly termed Uru, who were of unknown language. (See Handbook, vol. 2, pp. 575, 595-597.) Bibliography.—Boman, 1908; Brand, 1941 ¢; Chamberlain, 1910 a; Ciineo Vidal, 1913; Garcilaso de la Vega, 1723; Knoche, 1931; Latcham, 1910; Lozano- Machuca, 1885; Santa Cruz, 1913. 226 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 CARIBAN The Carib is one of the great linguistic families of South America, both in number of component languages and dialects and in extent, which is only less than that of the Arawak and Tupt. Carib languages are (or were) found from the Greater Antilles to central Mato Grosso, and from eastern Pert to central Para. Cariban and Arawakan groups have much the same distribution, but isolated Carib groups are much fewer. The great mass of the Cariban are north of the Amazon, occu- pying a great area that includes much of the Guianas, Venezuela, northern Brazil, and lowland Colombia. Nevertheless, the point of origin and dispersion is claimed to have been the region between the upper Xingt and the Tapajéz. Suggestions have been made that Carib and Arawak may eventually be tied up in one great phylum. Schuller (1919-20 a, 1928) proposed the further inclusion of Chibcha and Maya. 'Though comparative studies on the Carib languages have been made by Adam (1893) and De Goeje (1910), the classification of the many Carib languages is still to be done on a thorough linguistic basis, and those proposed are mainly arranged geographically. Rivet (1924 a), W. Schmidt (1926), Loukotka (1935), and Simpson (1940) have offered such classifications, with major and minor subdivisions. Those of Schmidt and Simpson are the most detailed and have been here adopted as a basis, incorpo- rating also some of the opinions of the others as well as those of Gillin and the other Handbook authors. Disagreements are, on the whole, few and minor. In addition to many languages, mostly extinct, on which data are insufficient and the classification, therefore, in doubt, there are several large groups whose Carib affiliation is questioned. One such is the Yagua-Peba group (q. v.), long considered independent and so still regarded by Loukotka (1935) and Nimuendajti (index) but accepted as Cariban by W. Schmidt (1926) and Simpson (1940) on the basis of Rivet (1911 b). Rivet (1943) has also presented cogent arguments for the inclusion of Chocé (q. v.) and many other languages of Colom- bia formerly considered as affiliated with Chibcha (q. v.). In the Guiana-Venezuela region, Gillin (Handbook, vol. 3, pp. 804— 813) lists some 80 tribes—and presumably dialects—that he considers of Cariban affinity, as well as some 30 more, probably all extinct, that are questionably Carib. Most of these are small groups, many of them mentioned by no other authority except Nimuendajt, who includes them on his map. Not all of these groups will be listed here again. Among those considered as Cariban by Gillin, and this affiliation not disputed by others, are: Acuria, Cashuena, Chikena, Cuacua (Mapoyo), Gabinairi, Heurd, Kirikiripe, Panare, Paraviyana, Puricoto (Catawian), Saluma, Terecumd, Tivericoto, Tonayena, Waiwai. Vol. 6] LANGUAGES—MASON Dot Other Cariban groups of undisputed relationship mentioned by Nimuendaji, Loukotka, and others, but not by Gillin, are: Azumara, Carib of Maturin, Mutuan, Wayewe '8, Zurumata. A number of Guiana groups, considered as Cariban by Gillin and others, are left unclassified by Nimuendaji, probably for lack of suf- ficient linguistic data. Among these are: Acokwa, Aracaret (Racalet), Ichu, Nourage (Norak), Pariki, Pirio (Apourout), Pishauc6, Sapai (Suppaye), Taira, Wai (Ouaye), Waikeri (Guaiqueri), Waya- culé (Oyaricoulet, Amibouane (?)), and Yapacoye. The Carib affinities of the following groups are disputed, mainly by Nimuendaju: Attaraya.—Given by Gillin both as Cariban and as a synonym of Arawakan Atorai. Asepangong.—Nimuendajii apparently considers Arawakan. Cariniaco.—Remarks same as for Seregong. Pawishana (Paushiana).—Cariban according § to Nimuendaji and? Loukotka; Arawakan according to Gillin and Rivet. Serecong.—Arawakan according to Nimuendaji; generally considered Cariban. Yao.—Cariban according to most; Nimuendajt believes Arawakan or unclassified. Taparito.—Nimuendajtii and Rivet agree with Gillin in considering Taparito as Cariban. Kirchhoff (Handbook, vol. 4, p. 439) makes Taparita a variety of Otomacan (q. v.). W. Schmidt (1926) considers Taparito as isolated. Caris CLASSIFICATION I. Northern A. Coastal 1. Insular Carib, Calino 2. Mainland a. Carib: Caribisi, Calinya, Galibi b. Cumanagoto ce. Palank (Palenque, Guarine) d. Pariagoto (Paria, Guayuno) e. Oyana (Upurut, Wayana): Rucuyen, Urucuiana f. Chacopata g. Piritu h. Cunewara i. Shiparicot, Chipa j. Core k. Chaima (Sayma, Warapiche): Tagare, Cuaga 1. Carinapagoto B. Central 1. Roraima Group a. Acawai: Patamona b. Purucoto 1 (Porocoto) c. Arecuna? (Jaricuna, Pemon): Camaracoto, Taulipang 1 Loukotka (1939 a) distinguishes Parukatu from Purukoto. 2 Nimuendaja lists an Arawakan Arecund in the same region. 18 Loukotka distinguishes Wayewe, Wayaway, and Vayamar. 794711—_50——16 228 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Caris CLASssIFICATION—Continued I. Northern—Continued B. Central—Continued 1. Roraima Group—Continued Arinagoto Macushi (Macusi): Teweya Waica * Ingaricé Saparé Wayumaré Paraviyana Kendéloco Monoicé . Azumara Paushiana Mapoyo . Taparito 2. Ventuart Group a. Makiritare: Yecuané (Mayongong), Maitsi, Ihuruand, Decuané (Wainungomo), Cunuand b. Yabarana: Curasicana, Wokiare VOPR rotrem ooo C. Amazon 1. Eastern a. Pianocoté b. Apalai: Aracuayu 4 ce. Waiwai (Ouayeoné) d. Pauzi e. Trio f. Diau g. Shikiana (Chikena) h. Tivericoto i. Catawian (Parucutu) j. Cumayena k. Urucuena 2. Western a. Carijona (Umawa, Omagua):* Hianacoto, Guake, Tsahatsaha (Saha), Guagua, Riama (?), Caicushana, Mahotéyana, Yacaoyana (?) D. Bonari 1. Bonari 2. Yauaperi (Crishand) a. Atroahy 3. Waimiry 4. Mutuan 3 Distinguish from Shiriandn Waica. 4 Rivet (1924 a) believes that the Apalaf are identical with the extinct Aracuaji, but the language of the latter seems to be mixed with Tupf, and Loukotka (1935) has put it in an independent subgroup for that reason. 5 Distinguish from Tupian Omagua. Vol. 6] LANGUAGES—MASON 229 Carisp CLassiFICATION—Continued II. Southern A. South 1. Arara a. Arara (Ajujure) ® b. Apiacé (Apingut) 7 ce. Pariri (Timirem) B. Xingt 1. Bacairz 2. Nahucua (Anauqua) a. Guicurt (Cuicutl) b. Apalakiri (Calapalo) ce. Mariape-Nahuqua d. Naravute e. Yaruma f. Yamarikuma g. Akuku IlI. Northwestern A. Maracaibo-Magdalena 1. “Motilones’’ ® a. Chaké: Macoa, Tucuco, Parirt, Chaké b. Mapé: Macoa, Macoita, Manastara, Yasa, Chapara, Sicacao, Tucuco, Cunaguasata, Maraca, Aguas Blancas, Aricuatsé, Catatumbo, Irapeno ce. Carate d. Zapara ® 2. Bubure (Coronado) 3. Yarigui a. Qutrtquire (Kirikire): Topocoro, Topoyo, Chiracota, Araya, Guamaca, Tholomeo 4. Opén 5. Carare 10 a. Colima (Tapas): Murca, Marpapt, Curipa b. Naura c. Nauracote 6. Muso (Muzo) 7. Burede 8. Guanao 9. Pemeno 10. Patagén 11. Camaniba 6 Distinguish between Panoan, Chapacuran, and Cariban Arara. Nimuendajii (Handbook, vol. 3, p. 214) states that the speech of the Arara is very close to that of the Yarumd (vide infra). 7 Distinguish from Tupian Apiacd of the Tapajéz. 8 Motilén classification according to Jahn, 1927, p.80. Reichel-Dolmatoff (personal communication) states that the Motildn of PerijA, of Bolinder and de Booy, are pure Cariban of the Chaima-Cumanagoto group, but those of Catatumbo and Rio de Oro are very different and seem to be Arawakan, though the linguistic data are scarce, 9 Hernandez de Alba (Handbook, vol. 4, p. 469) calls Zapara Cariban; Rivet (1924 a) considers it Arawakan. 10 Nimuendajii (map) leaves Carare unclassified. W. Schmidt (1926) places Amarizano in this North- western Group; most other authorities consider this language Arawakan (q. v.). 230 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Caris CLassiFICATION—Continued B. Chocoan 1. Chocé a. North a. Empera: Funucund, Dabeibe, Urubé b. Catio: Ibexico, Pequi, Norisco, Ituango, Teco, Peneco, Cararita, Cuisco, Araque, Pubio, Guacuseco, Tuin, Nitana, Pevere b. South a. Nonamé (Noanum4d): Chanco 2. Cent a. Nutabare (Nutabé): Tahami b. Cenufana 3. Cauca a. Quimbaya: Quimbaya, Carrapa, Picara, Paucura b. Ancerma: Ancerma, Caramanta, Cartama, Nort, Guaca ce. Antioquia: Antioquia, Buritica, Corome, Evéjico d. Arma: Arma, Pozo C. Southwest 1. Gorrén 4 2. Buga 3. Chanco "! D. Southeast 1. Arvi 2. Patdéngoro (Palenque): # a. Tamana b. Guarino c. Guagua d. Zamana e. Doyma 3. Panche a. Guazquia b. Guali ce. Marqueton 4. Pijao a. Quindio b. Cutiba c. Trico d. Toche e. Cacataima 11 Rivet, 1943, excludes these from his Chocé group. 12 Kirchhoff (Handbook, vol. 4, p. 339) groups Amant, Palenque, Zamand, Punchina, and Marquesote with Patdéngoro, and considers them of Chibchan relationship. BIBLIOGRAPHY General and unidentified Adam, 1878, 1879, 1893; Adelung and Vater, 1806— 17; Ahlbrinck and Vinken, 1923-24; Alvarado, 1919 a; Bertoni, M. S., 1921; Farabee, 1924; Gillin, 1936; Goeje, 1906, 1908, 1910, 1924, 1928 b, 1932-33; Mitre, 1909-10, pp. 280-314; Penard, A. P., 1928-29; Quandt, 1807; Schuller, 1919-20 a. Insular and Honduras.—Adam, 1906; Bererdt, 1874 a; Breton, 1665, 1666, 1669, 1877; Conzemius, 1930 b; Galindo, 1834; Henderson, A., 1847, 1872; Vol. 6] LANGUAGES—MASON 231 Lehmann, 1920; Rat, 1897-98; Rochefort, 1658 (1667), pp. 652-680; Sapper, 1897; Stoll, 1884, pp. 35-36; Taylor, 1938, 1946. Coastal Mainland.—Adelung and Vater, 1806-17 (Tama aco); Coudreau, H., 1887, 2:491-492 (Rucuyen), 1892; Coudreau, O., 1903 a (Ouayana, Rucuyen); Crévaux, Sagot, and Adam, 1882 (Rucuyen, Galibi); Gilij, 1780-84, pp. 375-389 (Tamanaco); Goeje, 1906 (Upurut); Martius, 1867, 2:327-370 (Galibi); Tastevin, 1919 (Galibi); Tavera-Acoste, 1907, pp. 325-332, 1921-22 (Tamanaco, Palenke, Guaykeri, Chaima, Chacopata, Piritu). Cumanagoto.—Platzmann, 1888; Ruiz Blanco, 1888 a, 1888 b, 1892, 7:161-168, 191-228; Tapia, D., 1888; Tauste, 1888 (Chayma, Core, Paria); Yangues, 1888. Roraima group.—Adam, 1905 (Acawat); Armellada, 1936; Armellada and Matallana, 1942 (Arecuna, Pemon) ; Barboza-Rodriguez, 1885 (Purucoto) ; Farabee, 1924 (Purucoto, Azumara); Koch-Griinberg, 1913, 1915, 1928 (Purucoto, Tauli- pang, Ingarico, Sapara, Wayumara); Martius, 1867, 2:227-228 (Paraviyana); Salathé, 1931-32 (Paushiana, Carime); Simpson, 1940 (Camaracoto); Tavera- Acosta, 1907, 1921-22 (Mapoyo, Arecuna, Waica). Macushi.— Barboza-Rodriguez, 1885; Coudreau, H., 1887, pp. 487-491; Farabee, 1924, pp. 121-152; Grupe y Thode, 1890; Koch-Griinberg and Hiibner, 1908, pp. 15-35; Martius, 1867, 2:225-227; Schomburgk, 1847-48, 2:515-523; Williams, J., 1932. Ventuari group.— Makiritare: Chaffanjon, 1889, pp. 342-344; Oramas, 1913 b; Tavera-Acosta, 1907, pp. 109-119, 1921-22, pp. 226-227. Yabarana: Koch- Grinberg, 1928, 4:233-242. Eastern Amazon.—Coudreau, H., 1887, pp. 491-492 (Uayeue); Coudreau, O., 1901, pp. 165-168 (Pzanacoto); Crévaux, Sagot, and Adam, 1882, pp. 39-40 (Trio); Farabee, 1924 (Watwai, Urucuena, Trio, Diau, Chikena, Cumayena, Catawian-Parucutu); Martius, 1867, 2:17-18 (Aracayu), 312-313 (Pianacoto, Waiwai, Tivericoto). Apalai (Aparai, Yauapiri).—Coudreau, H., 1887, 1892, pp. 60-75; Coudreau, 0, 1903 b, pp. 41-51; Crévaux, Sagot, and Adam, 1882, pp. 32-34; Farabee, 1924, pp. 229-241; Hiibner, 1907; Payer, 1906; Rice, 1931. Western Amazon.—Crévaux, Sagot, and Adam, 1882, pp. 35-38 (Carijona); Koch-Griinberg, 1906 c, p. 203 (Carijona), 1908 c, 1909 (Hianacoto-Umaua). Bonari group.—Barboza-Rodriguez, 1885, pp. 247-260 (Crishana); Brinton, 1892 a, p. 44 (Bonarz); Hiibner, 1907, pp. 238-246 (Yauwaperi); Payer, 1906, p. 222 (Waimiry); Pompeu Sobrinho, 1936 (Mutuan); Souza, A, 1916 a, pp. 77-78 (Bonari). Arara group.—Coudreau, H., 1897 c, pp. 199-210 (Arara); Ehrenreich, 1888, 1894-95, pp. 168-176 (Apingz); Krause, 1936, pp. 39-41 (Apingi); Nimuendaji, 1914 b (Pariri), 1931-82, pp. 549-551 (Arara), 1932 a, pp. 116-119 (Pariri). Xingi group.—Abreu, 1895, 1938 a (Bacairi); Krause, 1936 (Bacairi, Nahucua, Yaruma) ; Souza, A, 1916 b, pp. 71-73 (Bacairz) ; Steinen, 1892, pp. 1-1€0 (Bacairz), 1894, pp. 524-527 (Nahucva). Northwestern group.— Motilén-Macoa-Chake: Bolinder, 1917, 1925; Booy,1929; Ernst, 1887 b; Goeje, 1929-30; Isaacs, 1884, pp. 213-216; Jahn, 1927, pp. 340-354; Reichel-Dolmatoff, 1945 b (Motilén); Tavera-Acosta, 1921-22, pp. 221-230. Jahn, 1927 (Kirikire, Bubure) ; Jiménez de la Espada, 1897, pp. 28-30 (Patagén); Lengerke, 1878 a (Opén), 1878 b (Carare); Oramas, 1918 a (Kirtkiro). LANGUAGES OF PROBABLE CARIBAN AFFILIATIONS Naturally, Cariban relationships have been proposed for several other important linguistic groups and smaller languages by certain 232 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 scholars, whose opinions have been accepted by some of their col- leagues, rejected by others. Among these are the large Chocoan and Peba-Yagua groups, and the smaller languages Yuma, Palmella, Yuri, Pimenteira, and Ochucayana. For discussion of Ochucayana or Tarairit, see ‘Small Unclassified Languages of the Pernambuco Region.” cHocéd AND CARIBAN OF COLOMBIA Recent researches of Rivet (1943, 1944) and Jijén y Caamajio (1941-43) have advanced considerable evidence that many of the languages of Colombia formerly considered as Chibchan are (or were, since many of them are extinct) actually Cariban. These include Chocé and Pijao. They form a relatively solid group in north- western Colombia, separated from the main mass of Carib in eastern Venezuela and Guiana by belts of Arawakan and Chibchan peoples paralleling the cordillera to the Caribbean Sea. Rivet divides these into Eastern (Motilén, etc.) and Western (Chocod-Quimbaya) groups, separated by the Prjao-Panche-Paténgora. The Cariban affinities of Motilén (q. v.) have always been accepted. The Carib migration here is presumed to have been relatively late and to have supplanted former peoples of Chibchan speech. The Cariban affinities of Chocé are apparently more obvious and generally accepted than those of the Pijao-Panche-Patdéngoran, and the Quimbaya. Hernandez de Alba (Handbook, vol. 2, p. 922) places the Piao, Panche, Quimbaya, and Patdéngora in the Pdez subgroup, Talamanca-Barbacoa group of Chibchan. He also states (ibid., p. 923) that the “‘dialects of Pijao, Péez, Timana, and Yalcon were classed together.”’ Reichel-Dolmatoff (personal communication) considers the Cariban relationship of Chocé (Chami, Catio, Nonoama) as proved, but is less convinced of those of Pijao, Quimbaya, and the other former inhabitants of the Magdalena and Cauca Valleys. Cuna and Chocé are linked culturally and by inference linguistically in the Handbook (vol. 4, pp. 257-276). Chocé has generally been considered an independent family (Brinton, 1891 a; Chamberlain, 1913 a; Loukotka, 1935; Pericot y Garcia, 1936; Rivet, 1924 a; Ortiz, 1940 b). Mainly on account of the large number of Chibcha words, W. Lebmann (1920), followed by W. Schmidt (1926), believed it to be related to Chibcha. W. Lehmann (1920) thought it intermediate between the Barbacoan and the Central American groups of Chibcha. Jiménez Moreno (map, 1936) left it unclassified. The various dialects seem to be slightly differentiated. Chocé has adopted a large number of words from Chibcha and, like many Carib languages, from Arawak. Vol. 6] LANGUAGES—MASON Zoe Chocé languages or dialects mentioned by authorities, other than those given on the preceding chart, are Citard, Andagueda, Bandé, Chami, and Tadd or Tado. Cholo, Paparo, and Tucura are placed by some in the Citard subgroup. Other groups mentioned by only one writer, Brinton (1891 a) in particular, are Caviasgordas, Chiamu (Chocamu), Chochama, Murindo, Necodade, Pato, Rio Verde, and Sambo. In the accompanying linguistic map the following groups appear in the area that is presumably Colombian Carib, probably Chocé or Seni: Caramari, Fincenti, Guamoco, Malambo, Mompoz, Pacabueye, Pancent, Tamalamequi, Tolti, Turbaco, Yamici, Zamba, and Zondagua. Other Piao subtribes given by Rivet are Aype, Paloma, Ambeina, Amoya, Tumbo, Coyarma, Poina (Yaporoge), Mayto (Maito, Marto), Mola, Atayma (Otaima), Tuamo, Bulira, Ocaima, Behuni (Beuni, Biuni), Ombecho, Anaitoma, Totumo, Natagaima, Pana (Pamao), Guarro, Hamay, Zeraco, Lucira, and Tonuro. BIBLIOGRAPHY (All the important sources may be found in Reichel-Dolmatoff, 1945 a; Ortfz, 1940 b; W. Lehmann, 1920; Rivet, 1943-44.) Adelung and Vater, 1806-17; Anonymous, 1918 a (Catio); Bancroft, 1875; Bastian, 1876, 1878-89; Berendt, 1874 b; Beuchat and Rivet, 1910; Bollaert, 1860, pp. 65-67; Brinton, 1895 a (Andagueda), 1896 (Noanama); Cieza de Leén, 1881, p. 26 (Arma, Pozo); Collins, F., 1879; Cullen, 1851 a, 1868, 1875; Ernst, 1887 a; Greiffenstein, 1878 (Chami); Hurtado, G. O., 1924 (Noanama); Jij6n y Caamaiio, 1938, 1941-43; Latham, 1851; Lorenz, 1939 (Catio) ; Merizalde del Carmen, 1921, pp. 85, 89, 150; Mollien, 1824, 2:300, 1924, p. 450; Nordenskidld, 1928 a, 1929 b; Pablo del Santisimo Sacramento, 1933; Pinart, 1887, 1897; Rivet, 1912 a, 1943-44; Robledo, 1922 (Chamz); R6thlisberger, 1883-84; Saffray, 1872; Santa Teresa, 1924 (Catio); Seeman, 1853 (Cholo); Simon, 1887 (Tucura); Tessmann, 1930, pp. 472-475: Uribe, 1883; Uribe Angel, 1885, pp. 525-546; Vallejo, 1910 (Baudo); Vel4zquez, 1916 (Cham); Wassén, 1933, 1934 b, 1935; White, 1884 a, 1884 b, PEBA-YAGUA The classification of the Yagua or Peba group, generally agreed to consist of Yagua, Peba, and Yameo, has seen a recent return to belief in its independence. NHervas y Panduro (1800) had proposed a Yamea family, composed of Amaono, Nahuapo, Napeano, and Masamae. Brinton (1891 a) called the family Peban, the component languages Caumari, Cauwachi, Pacaya, Peba, and Yagua. Rivet (1911 b) then published his thesis that the group is affiliated with Carib; this opinion has been accepted in the classifications of Pericot y Garcia (1936), Krickeberg (1922), W. Schmidt (1926), Simpson (1940), and Jijén y Caamafio (1941-43). Much earlier, however, Chamberlain (1913 a) decided that more proof of this relationship is needed, and continued the use of Peban as an independent family. The more recent authori- 234 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 ties agree with this conclusion; Nimuendajii (map), Loukotka (1935) and Igualada and Castellvi (1940) accord it independent family status, the first terming it Peba, the second Yagua. Métraux also doubts the Carib affinities. Loukotka (1935) calls Yagua (Yegua, Yahua) a “pure” language, Peba mixed with Carib, Yameo mixed with Arawak and Carib. Tessmann (1930) calls them both ‘‘mixed-stem languages,’ Yagua mixed Pano-Carib, Yameo mixed Arawak-Pano; Peba he seems to consider a.synonym of Yagua. The group had best be left unclassified until further linguistic researches are made upon it. A number of component languages and subdivisions of Peba-Yagua are mentioned in literature. Most of these are probably extinct, and the whole Yameo group is on the verge of extinction if not already gone, A. Yagua 1. Yagua 2. Peba a. Cauwachi b. Caumari ce. Pacaya B. Yameo 1. Yameo . Napeano . Masamai . Nahuapo . Amaona . Mikeano . Parrano . Yarrapo . Alabono San Regino (?) . Mazan (?) . Camuchivo (?) Tore tee =e} hw Gn jes to) ont BIBLIOGRAPHY Peba-Yagua.—Castelnau, 1852, pp. 296-298; Chamberlain, 1910 a, pp. 195, 200-201; Fejos, 1943; Marcoy, 1867, pp. 131-132; Martius, 1867, 2:296—297, 300-301; Orton, 1871; Rivet, 1911 b, 1912 a; Tessmann, 1930, pp. 459, 577-580. Yameo.—Adelung and Vater, 1806-17; Gonzdlezz Sué4rez, 1904, pp. 67-68; Tessmann, 19380, p. 565. ARDA Arda was accepted as an independent linguistic family by all author- ities from 1858 to 1924, including Rivet (1924 a) and Schmidt (1926). This opinion was based upon a Doctrina in a language of this name, the Lord’s Prayer of which was published by Ludewig in 1858. This obviously bore ro relationship whatever to any adjacent language. Paul Rivet (1925 e) examined the original manuscript in Madrid and found that it made no reference to the country in which Arda was Vol. 6] LANGUAGES—MASON 935 spoken. Following some suspicions, he compared the words with modern Dahomean in Africa and determined their close relationship, especially to the Popo dialect. The text was evidently taken in the Slave Coast Kingdom of Arda, and the language has therefore no relation to that of the Arda tribe of southeastern Colombia, an extinct group probably related to the Peba, Yagua, and Yameo. Nimuendajt (map) continues to regard Arda as an isolated language. Bibliography.—Chamberlain, 1910 a; Ludewig, 1858; Rivet, 1912 a, 1925 e. YUMA The Yuma, with one relatively large group and a small enclave in the state of Amazonas, ere rather isolated from any other Carib groups. Accepted as of Carib affiliation by all other authorities, Nimuendajti leaves them unclassified, a conservative opinion herein followed. PALMELLA No authority, not even Nimuendaji, doubts the Carib affinity of Palmella, but as the linguistic data are very poor, as the Palmella are a tiny group, and far removed from any other Carib people, even much farther south than the doubtfully Carib Yuma (q. v.), they might well be left unclassified. If of Carib affiliation, they form the southwesternmost Carib group, near the Brazil-Bolivia border. Bibliography.—Fonseca, J. S. da, 1880-81, pp. 193-196. YURI (JURI) Opinions regarding the relationship of the small Yuri (Chamberlain, 1913 a, and W. Schmidt, 1926, prefer the spelling Juri) group are very contradictory. Markham (1910) claimed a linguistic connection with the Arawakan Passé; Brinton (1891 a) accepted this classification. Loukotka (1935) and Igualada and Castellvi (1940) consider it Carib. The more conservative recent opinions, Rivet (1924 a), Nimuendaji (map), W. Schmidt (1926), Tessmann (1930), Krickeberg (1922), follow Chamberlain’s (1913 a) classification as independent or isolated. Possibly several Juri or Yuri languages are here confused. Métraux (Handbook, vol. 3, p. 708) describes one as one of the ‘“‘Arawakan tribes of the left middle Amazon.’ The data seem to be limited to the vocabulary in Martius (1867). There is apparently only one language, but there are said to have been 10 dialects. As the language is almost extinct, spoken today by a very few individuals, a modern grammar of Yuri is a great desideratum. It is a reasonable guess that if such a grammar is ever prepared, Yuri will be found to fall with either Arawak or with Carib. This Yuri must not be con- 236 SOUTH AMERICAN INDIANS [B. A. B. Bull. 148 fused with an unclassified Juri or Surilanguage of the Gran Chacoregion. Bibliography.—Cabrera, P., 1924; Chamberlain, 1910 a; Martius, 1867, 2:268-272; Rivet and Tastevin, 1921; Tessmann, 1930, p. 584; Wallace, 1853, pp. 528-529. PIMENTEIRA All the older standard authorities consider Pimenteira a Cariban language. Nimuendaji (map) places it with Botocudo (q. v.) and Lowie (Handbook, vol. 1, p. 381) calls it a separate family. It is far to the east of any other Carib group. Bibliography.— Martius, 1867, 2:219-220. MACRO-TUPI-GUARANIAN A Macro-Tupi-Guarani phylum is here diffidently proposed for the first time. It consists of Tupt-Guarani, Miranya (Bora), Witoto, Zdparo, and a number of less important languages which are generally placed in one or another of these ‘families.’ It is not advanced with any claim to certainty or with any evidence of proof, but as a result of opinions, deductions, and intuitions of the several authorities and of the present writer, plus the fact that there is great difference of opinion concerning into which of these families many of the small languages fall. Rivet (1911 a) has presented evidence for the inclu- sion of Miranya (Bora) in Tupi-Guarant. 'This has been accepted by some, rejected by others. Dr. J. P. Harrington is convinced that Witoto also belongs with Tupi-Guarani. Jijé6n y Caamafio (1941-43) establishes a W2toto-Bora-Zdparo phylum separate from Tupi-Guarani. Zdéparo is the most doubtful member of the phylum. Nimuendajt (map) and Loukotka (1935) keep all separate. As these families are contiguous a genetic connection is not unreasonable. TUP{-GUARANIAN Tupt-Guarani, like Arawak and Carib, is one of the great wide- spread linguistic families of South America. The languages were, or are, spoken from easternmost Brazil to the foot of the Andes in Pert, and from Guiana to Uruguay. Though in many isolated groups, the bulk is in eastern Brazil. The distribution is mainly fluvial and mari- time, most of the groups restricted to the coast or the river valleys. The original home seems to have been in the region of the Paraguay- Parana, from which they spread, following the rivers. Soon after the time of the Conquest they held the entire Brazilian coast from the Amazon nearly to Buenos Aires. Much of this migration was recent and probably even post-Conquest, and largely during the sixteenth century. Other migrations up to and including the present century are of historical record. (See Métraux, Handbook, vol. 3, pp. 97-99.) Many of these migrations were at the expense of Ge groups, especially on the Brazilian coast. The distribution of languages at the time Vol. 6] LANGUAGES—MASON 237 of the Conquest is, therefore, most difficult to determine; branches of the same group, sometimes bearing the same or similar names, are found in very widely separated regions, and present geographical propinquity carries no presumption of close linguistic relationship. Tribes of other linguistic affinities frequently adopted Tupi-Guarant tongues, especially after the Conquest. Northern Tupi, Tupi Proper or Nhengatu, was adopted by the Spanish missionaries and traders as the lingua geral, which aided its spread and vogue. The use was probably largely due to the fact that it is said to be a relatively simple language morphologically, and easy to learn. Though Tupi-Guarani may eventually form the body of one of the great phyla into which South American languages may be grouped, it is unlikely that it will be found tobe related_to any of the other major families. No documented study of the divisions of Tupi-Guarani on a lin- guistic basis has ever been made. Rivet (1924 a) lists them geo- graphically. W. Schmidt (1926) and Loukotka (1935) group them into subdivisions with geographic terminologies, but with great mutual disagreement, since many isolated intermediate groups may be placed in any one of several subdivisions. The present classification is based on all of them, modified by opinions of more recent observers regarding the relationship of certain groups. It makes no claim to correctness or finality, and will doubtless be modi- fied greatly by future researches. The two main divisions are into Tupi and Guarani, but, even in these, apparently the relationship is relatively not very distant. Tupfi-GUARAN{ CLASSIFICATION I. Guarani A. Paran4 1. Guarani (Carijé6, Cari6, Chandule) a. Arechane, Itatin, Tapé, Tobatine, Guarambaré, Taioba 2. Caingué (Kaigudéd, Montese) a. Apapocuva, Caingud, Carima, Chiripd, Guayand,}! Mbya2? (Mbiihé, Apiteré, Baticola, Boaberd), Oguana (Oguaiuva), Pan’, Tanyguéd, Taruma,* Cheiru, Avahugudi, Paiguagu, Yvytyigud, Avachiripd, Catanduva Jaiahy 8. Paranaé 4. Guayaki 4 5. Aré (Seté, Ivaparé, Shocleng, Notobotocudo,’ Pihtadyovac) 1 Métraux (Handbook, vol. 3, p. 69) uses Caingud to distinguish the modern primitive from the civi- lized Guarani. Distinguish from two other Guayand in the same general region, one Ge and one unclassi- fied, according to Nimuendaji (map). (See Métraux, Handbook, vol. 3, pp. 70-71.) 2 Distinguish from Guaicurt Mbayd. § Distinguish from Arawak (?) Tarumd. 4 Closely related to G@uaranfé linguistically, according to Métraux and Baldus (Handbook, vol. 1, p. 435). Most modern authorities agree, but a few consider it independent or refuse to classify it. +The Aré or Seté are erroneously called Botocudo or Notobotocudo, but have no connection with Botocudo proper (Macro-Ge). 238 SOUTH AMERICAN INDIANS [B. A. B. Bull, 143 Tupf-GuaRAN{ CLASSIFICATION—Continued I. Guarani—Continued B. Bolivia 1. Chiriguano (Aba, Camba, Tembeta) a. Guarayt% a. Pauserna: Itatin, Carabere, Araibayba, Motere- quoa (Moperacoa), Varai, Pirataguari (Pita- quart), Cario, Kiriticoci, Guarayt-Té b. Sirioné ® (Chori) a. Nyeoze-Née, Tirinié, Jandé, Qurungia 2. Tora? 3. Porokicoa 4. Palmares (?) 8 5. Tapieté® (Tirumbae) a. Yana (Yanaigua) 6. Ubegua (?) 7. Chané "! C. Araguaya ” 1. Tapirapé * 2. Canoeiro (Avéd) II. Tupi A. Coastal (Nyeengatu) 1. Tupina (Tupiguae) a. Aricobé, Amoipira 2. Tupinamba a. Apigapigtanga, Araboyara, Caeté (Caheté), Guaracaio (Itati), Muriapigtanga, Potiguara (Pitonara), Rari- guara, Tamoyé, Timimino, Tabayara,4 Tupinikin, Viatan-Pernambuco B. Guiana 1. Apoto (Aponto) 2. Caliandé % 3. Oyampi a. Camacom 4. Wayapi (Guayapi) 5. Emerillon 6. Paikipiranga (Parichy) 6 The former language of the Sirioné is unknown. 7 A number of the tribes mentioned below have adopted Guarani in recent centuries. Tord was formerly classified as Chapacuran; a large vocabulary gathered by Nimuendaji shows beyond doubt that it is impure Tupi (Nimuendaji, 1925; Nimuendaja and Valle Bentes, 1923). 8 Palmares and Ubegua were placed by Brinton (1898 a) among the Guarani groups of the Gran Chaco; other authorities do not mention them. 9 The former language of the Tapieté may have been Arawak. 10 See footnote 9. 11 The Chané (q. v.) formerly spoke Arawak. 12 These two groups (Tapirapé, Canoeiro) apparently migrated here from the south in historical times. Both are slightly known. The Canoeiro are said to be descendants of the Carijé (Guarani). Lipkind (personal communication) thinks that they may not be extinct, and may not be Tupi. 13 Wagley (personal letter) says that the Tapirapé language resembles Guarani more than northern Tupi in its use of specific pronominal prefixes, suffixes of time, place, and condition, and in high develop- ment of nasalization. 14 See Métraux, Handbook, vol. 3, page 96. 16 Distinguish from independent Calidna (q. v.). Vol. 6] LANGUAGES—MASON 239 Tupf-GuaRANi CLASsIFICATION—Continued II. Tupi—Continued B. Guiana—Continued 7. Cusari 16 8. Wara-Guaju (Araguaji%) " (?) C. Southern Amazon 1. Araguaya Division (He-group) a. Nyengahiba b. Ararandewara c. Miranyo (?) 8 d. Amanayé (Manazo): Anambé, Paracana e. Tenetéhara: Guajajéra, Tembé, Guaja f. Urubé: Turiwara (Turuara) g. Pacajé: Pacajé, Jacundé (Amiranha?), Anta (Tapiraua) h. Cubenepre: Kupé-ré6b (Jandiahi) i. Asurini (?) 2 2. Xingi Division a. Yuruna Group #4 a. Yuruna: Yuruna, Shipaya (Ashipaye) b. Manitsawé c. Arupai (Urupaya) * b. Tacunyapé (Pewa) ce. Upper Xingi Group a. Aueté: Arawiti 8 b. Arawine 3. Tapajéz Division a. Cawahib Group a. Cabahyba: Cawahib,*4 Parintintin, Apatrande, Odyahutibe b. Apiacé % a. Tapanyuna (Arino) ce. Munduruct 6 a. Curuaya (Kuruahe) 16 Probably originally Carib, acculturated to Tupi, according to Gillin (Handbook, vol. 3, p. 814). Sub- tribe of Oyampi, according to Rivet (1924a). Nimuendaji (map) leaves Kussari unclassified. 17 Nimuendajii (map) distinguishes two groups of Aracajz in this region, a Carib group on the Rio Paru and an unclassified group on the Rio Pacaja, (See also NimuendajG, Handbook, vol. 3, p. 209.) 18 See Nimuendaji, Handbook, vol. 3, page 208. 19 According to Wagley (personal correspondence), the Guajajdra and Tembé form one tribe and call them- selves Tenetéhara. The Guajé are neighbors of the Guajajdra who say that they speak “our’ language Urubt is grammatically like Guajajéra, with a slight phonetic difference governed by regular sound shifts. 20 Nimuendajii (Handbook, vol. 3, p. 214) says that Asurinf is reported to be like Guajajara; as there is no record of the speech, most of the other authorities who have mentioned the language have left it unclassified. 21 According to Nimuendajii (Handbook, vol. 3, p. 214), Yuruna, Shipaya, and Manitsawd form a group of impure Tupi, to which Arupaf may also belong. It differs considerably from Tupi proper by reason of Arawak and Carib influences. Yuruna and Shipaya are almost mutually intelligible, differing by regular sound shifts. Martius (1867) and Adam (1896) doubt the Twpi relationship generally accepted. Loukotka (1935) considers Manitsawd as mixed with Ge; Yuruna and Shipaya as mixed with Arawak. Lévi-Strauss (Handbook, vol. 3, p. 322) says that Manitsawd includes much from Suya (Ge). 22 Distinguish Arwpai from Gurupé of Tocantins, and from Urupéd of Gy-Parana. % Arawiti is a mixture of Aweto and Yawalapiti, according to Lévi-Strauss (Handbook, vol. 3, p. 322). % Lévi-Strauss has a large unpublished vocabulary of Cawahib. Cawahib and Parintintin are very similar and both have strong affinities with Apiacd. They are remnants of the ancient Cabahyba (Nimuendaja, Handbook, vol. 3, p. 283). 35 See Nimuendaji, Handbook, vol. 3, page 313. 2% According to Loukotka (1935, 1939 a), Munduruct and Kuruaya are mixed with Arawak, Maué mixed with Carib and Arawak. Nimuendaji (Handbook, vol. 3, p. 246) says that Maué contains non- Tupi elements that cannot be traced to any other linguistic family. 240 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Tupf-Guaranf CuassiFication—Continued II. Tupi—Continued C. Southern Amazon—Continued 3. Tapaj6z Division—Continued d. Maué *6 . Wirafed . Cayabi: 27 Camayurd . Tupinambarana . Paranawét . Ramarama (Itanga) *% a. Itogapuc (Ntogapid) j. Catukinaré D. Upper Amazon 1. Cocama Group * a. Cocama: Cocamilla, Xibitaona b. Omagua (Campeva) 2. Aizuare Group *° a. Aysuart: Curuzicari b. Bonama (Ibanoma) c. Pawana d. Soliman (Yoriman) ~ FOO eh Oo % For footnote 26, see page 239. 37 Rivet considers Cayabi as Carib. (See Nimuendaji, Handbook, vol. 3, p. 307.) 38 Ramarama and Néogapid are mixed with Arawak and Arikem, according to Loukotka (1935, 1939 a). 29 Cocama and Omagua are not mutually intelligible; they also speak Quechua and Spanish. Cocamillais the southern or Ucayali dialect of Cocama, the difference negligible. 30 These languages are all extinct. Nimuendaji (map) leaves the first three, the only ones there listed, unclassified; he considers Curacicari a synonym of Aysuari. Nimuendajti lists a number of groups under the Tupi designation that are not found in the above outline. Some of these may be synonyms; a few others are put in other families by other authorities, or left unclassified. Most of these are in the southern Amazon region. Of these languages, found in a prevailingly Tupi region, the following are left unclassified by Handbook authors for lack of suffi- cient information: Ammniapé, Guaratagaja, Kepkiriwat, Macurap, Tupari, Arikem (q. v., infra), and Buruburd (Purubord, generally considered Huari or independent). Others in this general region, considered Tupi by Nimuendaji, are: Arud, Guardyo, Ipotwat, Jabotifed, Mialat, Paranawat, Sanamaica, Taipé-shishi, Takwatib, Tucumafed, and Wayoré. YURIMAGUA (ZURIMAGUA) Yurimagua is generally considered Tupian; Métraux (Handbook, vol. 3, p. 704) gives his evidence for believing that it is not related to Omagua and possibly independent; Nimuendajt leaves it unclassified. ARIKEM Though generally placed in the Chapacuran family, Loukotka (1935, 1939 a,) considers the Arikem (Ariqueme, Arikéme, Arikeni, Vol. 6] LANGUAGES—MASON 241 Ahépovo) language sufficiently distinct to form a separate family, with intrusions of Zupi and Arawak. Nimuendaji (map) classifies it under Tupi, and Métraux (Handbook, vol. 3, p. 406) believes that he (Nimuendaji) has proved the Tupian relationship. BIBLIOGRAPHY Guarani.—As an important living language of the Paraguayan region the bibliography of Guarani is large. It is, therefore, divided into two parts. Many of the works listed under Tupi refer to the Tupi-Guarani “family”? and some therein may be specifically on Guarani. The more important works are: Adelung and Vater, 1806-17; Bertoni, M.S., 1916, 1921, 1922; Bottignoli, n. d. (1938, 1940), 1926; Dominguez, 1912; Gez, 1915; Lafone-Quevedo, 1919 b; Medina, J. T., 1930 b; Osuna, 1923 a, 1923 b, 1924 a, 1924 b, 1924 c, 1925, 1926 a, 1926 b; Ruiz de Montoya, 1876 a, 1876 b, 1876 c, 1876 d, 1892; Teschauer, 1921. Also: Bertoni, A. de W., 1924; Bertoni, G. T., 1926; Bertoni, M. S., 1914, 1920, 1927 b, 1932; Branco, 1937; Colman, 1917, 1921, 1929 a, 1929 b, 1932, 1936; Fiebrig-Gertz, 1927, 1932; Gandia, 1931; Martinez, T. A., 1916; Morinigo, 1931, 1935; Osuna, 1921; Recalde, 1924, 1937 b; Schuller, 1913 b; Solari, 1928; Storni, 1944. Caingu4.—Ambrosetti, 1895 a; Borba, 1908, pp. 73-76, 188-139 (Guayand); Martius, 1867, 2:18-14; Miller, Franz, 1934-35 (Mbu&hd); Nimuendaji, 1914 a (Apopocuvdé); Sampaio, T., 1890, pp. 133-148; Vellard, 1937; Vellard and Osuna; 1934; Vogt, 1904, pp. 207-214. Guayaqui and Setaé.—Bertoni, G. T., 1924, 1926-27, 1927, 1939; Borba, 1904, p. 57 (Aré); Ihering, 1907, p. 282 (Shocleng); Loukotka, 1929 (Seid); Maynt- zhusen, 1919-20; Panconcelli-Calzia, 1921; Steinen, 1901; Vellard, 1934-35; Vogt, 1902-03. Chiriguano.— Campana, 1902, pp. 17-144, 283-289; Cardus, 1886, pp. 309-310; Giannecchini, 1896; Kersten, 1905; Nino, 1917; Nusser-Asport, 1897; Romano and Cattunar, 1916; Schmidt, M., 1938. Guarayi.—Cardus, 1916; Métraux, 1942 a, pp. 95-110; Pauly, 1928, pp. 189— 190; Pierini, 1908; Schmidt, M., 1936 b; Snethlage, E. H., 1936. Pauserna.—Fonseca, J. 8. da, 1880-81; Métraux, 1942 a, pp. 95-110; Schmidt, M., 1936 b; Snethlage, E. H., 1936. Sirioné.—Cardus, 1886, p. 280; Krause, 1911; Métraux, 1942 a, pp. 110-114; Nordenskiéld, 1911 a, 1911 b; Nusser-Asport, 1897; Pauly, 1928, p. 193; Radwan, 1929; Rydén, 1941; Schermair, 1934; Snethlage, E. H., 1936; Wegner, 1934 a, 1934 b, 1934 c, pp. 5-54. Tapieté.— Nordenskidld, 1910 a; Palavecino, 1930; Schmidt, M., 1937 c. Araguaya Group.—Tapirapé: Kissenberth, 1916, pp. 52-64; Krause, 1911, pp. 405-406; 1936, p. 48. Canoeiro: Couto de Magalhaes, J., 1902, p. 119; Rivet, 1924 d. Tupi (Neengataéi).— Eastern Tupt is, or is the basis of, the lingua geral of Brazil. The bibliography is, therefore, very large and has herein been divided into two parts, the first containing the more important works and those on the Tupi- Guarani family; the second, the works of lesser importance and those on the lingua geral. Adam, 1878, 1896; Ayrosa, 1934 a, 1935 a, 1937 a, 1939, 1941, 1943; Barbosa da Faria, 1925; Barbosa-Rodrigues, 1892-94; Branco, 1935; Brinton, 1898 a; Drum- ond, 1944, 1946; Eckart, 1890; Figueira, 1878; Garcia, R., 1927; Garraux, 1898; Goeje, 1928 b; Gongalves, Dias, 1855; Hestermann, 1925; Ihering, 1907; Koch- 242 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Griinberg, 1932; Leeris, 1592, pp. 250-266; Martius, 1867, 2: 99-122, 373-544; Nimuendaji, 1925, pp. 143-144; Nimuendaji and Valle Bentes, 1923, p. 222; Paula Martins, 1941, 1945 a, 1945 b; Philipson, 1946 a, 1946 b; Platzmann, 1901; Schmidt, M., 1905, pp. 419-424, 440-446; Silveira, E. do, 1935; Tastevin, 1908, 1923 b, 1923 c, 1923 d; Valle Cabral, 1880. Also: Adelung and Vater, 1806-17; Almeida, 1931; Anonymous, 1937; Ayrosa, 1933, 1934 b, 1934 c, 1934 d, 1934 e, 1934 f, 1934 g, 1935 b, 1935 c, 1935 d, 1935 e, 1935 f, 1935 g, 1935 h, 1935 i, 1935 j, 1935 k, 1936 a, 1936 b, 1936 c, 1936 d, 1936 e, 1936 f, 1937 b, 1937 c, 1938; Brandéo de Amorém, 1928; Camara Cascudo, 1934; Coqueiro, 1935 a, 1935 b; Ferreira-Franca, 1859; Friederici, 1930; Guimaraes, J. J. da S., 1854; Lemos Barbosa, 1937; Machado d’Oliveira, 1936; Martius, 1867, 2: 7-11, 23-97; Nimuendaji, 1914 c; Pombo, 1931; Recalde, 1937 a, 1937 b; Stradelli, 1929; Studart, 1926; Sympson, 1926; Tasso Yatahy, 1918; Tastevin, 1919. Tupinamba.—Lery, 1599; Platzmann, 1901. Guiana Group.—Coudreau, H., 1892, pp. 76-140; Crévaux, Sagot, and Adam, 1882; Martius, 1867, 2: 17-18, 320-323; Perret, 1933 (Emerillon). Araguaya Division.—Ehrenreich, 1894-95, pp. 163-168 (Anambé); Lange, 1914, pp. 445-446 (Ararandewara); Lopes, 1934, pp. 167-170 (Urubu); Nimuendaji, 1914 ¢ (Amanajé, Turiwara); Rice, 1930 (Urubi). Guajajara.—Ehrenreich, 1894-95, pp. 163-168; Froes de Abreu, 1931, pp. 157-163; Roberts, F. J., and Symes, 1936; Snethlage, E. H., 1931, pp. 117-139. Tembé.—Baptista, 1931-32; Hurley, 1981 a; Nimuendajd, 1914 b; Rice, 1930, pp. 314-315; 1934, pp. 152-180; Snethlage, E. H., 1932 b, pp. 351-362. Yuruna.—Coudreau, H., 1897 a, pp. 165-198; Krause, 1936, pp. 39-41; Ni- muendaju, 1923-24, 1928-29 (both Shipaya), 1931-32, pp. 580-589; Snethlage, E, 1910; Snethlage, E. H., 1932 a (both Shipaya); Steinen, 1886, pp. 360-363 (and Manitsawa). Tacunyapé.—Nimuendaji, 1931-32, pp. 5438-547. Upper Xinga.— Aueté: Schmidt, M, 1905; Steinen, 1894, pp. 535-537. Arawine: Krause, 1936. Cawahib.—Nimuendaju, 1924 (and Parintintin);. Apiaca.—Castelnau, 1852, pp. 277-278; Coudreau, H., 1897 a, pp. 184-191; Guimaraes, C. J. da 8., 1865; Hoehne, 1915; Katzer, 1901, p. 41; Koch-Griinberg, 1902 c; Martius, 1867, 2:11-13. Munduruca.—Coudreau, H., 1897 c, pp. 192-202; Krause, 1936; Martius, 1867, 2: 18-20; Nimuendaji, 1931 (Curuaya), 1932 a, pp. 106-108; Rondén, 1915, pp. 179-183; Snethlage, E., 1910; Snethlage, E. H., 1932 a (both Curuaya) ; Strémer, 1932; Tocatins, 1877, pp. 73-161. Maué.—Coudreau, H., 1897 c, pp. 173-181; Katzer, 1901, p. 37; Koch-Griin- berg, 1932; Nimuendajt, 1929 b. Camayura.—Krause, 1936, p. 43; Schmidt, M., 1905, pp. 446-447; Steinen, 1894, pp. 537-540. Tapajéz Miscellaneous.—Church, 1898, p. 64 (Catukinari); Horta Barbosa, 1922, p. 25 (Ramarama); Koch-Griinberg, 1932 (Wirafed); Nimuendaji, 1924, pp. 275-276 (Wirafed), 1925, pp. 144-145, 172 (Ntogapid, Ramarama); Schmidt, M., 1929 a, (Cayabi). Omagua and Cocama.—Adelung and Vater, 1806-17 (Omagua); Castelnau, 1852, pp. 293-294 (Cocama); Espinosa, 1935 (Cocama); Gilij, 1780-84, pp. 371- 375 (Omagua); Gonzdlez Sudrez, 1904, pp. 65-66 (Omagua); Marcoy, 1875, 2: 296, 402; Martius, 1867, 2:16-17, 299-300; Orton, 1871, p. 473 (Omagua); Rivet, 1910 a; Tessmann, 1930, pp. 47, 65 (Omagua), 66 (Cocama), 82 (Cocamilla). Yurimagua.—Veigl, 1785 a, p. 54. Vol. 6] LANGUAGES—MASON 243 Arikem.—Lopes, 1925; Nimuendaji, 1932 a, pp. 109-116. MIRANYAN OR BORAN Synonyms: Miranhan, Miraiia, Bora, Boro. Not recognized by Brinton’s (1891 a) classification, Miranyan was first proposed as a separate family by Chamberlain (1913 a), on the basis of publications by Mochi (1902-03), Koch-Griinberg (1906 c, 1909-10), and Rivet (1911 a). Although Rivet had already pub- lished his study with his conclusion that Miranya is a very much modified and differentiated Tupi-Guarant dialect, Chamberlain, with the comment that ‘‘more evidence is needed,” preferred to leave it unaffiliated with any other large group. Krickeberg (1922) and Pericot y Garcia (1936) accept Rivet’s classification; Loukotka (1935) and Igualada and Castellvi (1940), preferring to call the family Bora, class it as independent. W. Schmidt (1926) outlines the problem and begs the question. Tessmann (1930) calls it a mixture of Ge- Tupt; Nimuendajii (map) considers it independent. Steward (Hand- book, vol. 3, p. 749), considers it under Witotoan, which latter he accepts as Tupian. Jijén y Caamafio (1941-43) makes it a member of his phylum Witoto-Bora-Zdparo, distinct from Tupi. A grammar of some Miranyan language is a great desideratum; the data available for comparison are purely lexical. While the evidence of Zupt relationship presented by Rivet (1911 a) is not entirely convincing—witness its nonacceptance by many authorities— it is nonetheless highly probable and is herein accepted. Dr. J. P. Harrington has made an independent comparison and is convinced of the relationship. Miranya seems to be most closely related to Witoto, and several languages, such as Muinane, Coeruna, Nonuya, and Imihita, are placed by some under Miranya, by others under Witoto. Harrington’s conclusions (personal communication) are: The Witotoan probably includes Miranya and is certainly an outlying member of Tupi-Guaranf. Miranya and Witoto are distantly related and both are related to Tupi. The Tupi affinity is not one of admixture. Miranya shares a very considerable number of etyma with both Tupi-Guaranf and Witoto. The percentage is larger than Rivet thought and extends to fundamental words. Rivet (1911 a) believes that Miranya is the northwesternmost of the Tupt-Guarant languages, the remains of a very early invasion, before that of the not-far-distant and more purely Tupi languages such as Omagua and Cocama, and, therefore, more affected by borrow- ings from adjacent languages. The same remarks would doubtless apply to Witoto, and possibly even to Zdparo. There is great difference of opinion regarding the component languages of the Miranya-Bora group. Igualada and Castellvi (1940) subdivide the Bora into the True Bora (Bora, Mira7ia), and the False 794711—50—17 244 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Bora (Imthita, Nonuya-Bora, Muinane-Bora); these last languages are considered Witotoan by several authorities. Tessmann (1930) lists 20 bands, Whiffen (1915) 41. Harrington (personal communica- tion) says that Koch-Griinberg’s Jmihita words are typically Miran- yan, controverting Igualada’s and Castellvi’s (1940) opinion of them as “false.”” Koch-Griinberg (1906 a) gives four Miranya languages: Imihita, Fa-ai, Miranya, and Miranya-Oird-Acu-Tapuya. Martius’ (1867) “Hawk” and “Mosquito” vocabularies are also Miranyan, according to Harrington (personal communication). Orején and Coeruna have also been considered as Miranyan languages. Martius’ (1867) Miranya-Carapana-Tapuya vocabulary is apparently closer to Witoto. Bibliography.—Igualada and Castellvf, 1940; Jiménez Seminario, 1924; Koch- Griinberg, 1906 a, 1906 c, 1909-10, 1910 a, 1910 b; Martius, 1867, 2: 279-281; Mochi, 1902-03; Ortiz, 1942; Rivet, 1911 a; Tessmann, 1930, pp. 267-280. WITOTOAN Though the independent position of Witoto has not to date been abandoned, certain languages—for instance, Nonuya, Minuane, Ocaina, and Miranya-Carapana-Tapuyo—have been placed by some authorities in the Witoto family, by others in Miranya or Tupi-Guarani, suggest- ing possible relationship with the latter. It was originally believed to be Cariban. Dr. J. P. Harrington has compared Witoto, Cocama, proved. Ortiz (1942) does not accept the relationship to Bora Miranya, and Tupi-Guarani, and is convinced of the relationship of all. Dr. Harrington’s unpublished treatise indicates a general resemblance in morphological type, and close resemblance of morphological ele- ments in position, meaning, and phonetic type; the lexical relation- ship, as presented, is not so convincing. The relationship of Witoto to Tupi-Guarani is accepted herein, though not as incontrovertibly proved. Ortiz (1942) does not accept the relationship to Bora (Miranya) but Jijén y Caamafio (1941-43) makes it a member of his Witoto-Bora-Zdparo phylum. The native name for the language is Komiuveido. Castellvi (1934 b) believes that the historical Quiyoya were the ancestors of the Witoto; Ortiz (1942) thinks they were the Cambeba or Omagua (Cafuane). Orellado and Orelludo are probably synonyms for Orején; their lan- guage was known as Mativitana. No methodical attempt seems to have been made to subdivide Witoto on a linguistic basis. There seems to be a group of true or proper Witoto, and one more closely related to Miranya. The classi- fication of Ortiz (1942) is adopted as the basis herein. The Miranya- Carapana-Tapuyo vocabulary of Martius (1867) is Witotoan. Vol. 6] LANGUAGES—MASON 245 Pinell (1928) mentioned 136 W2toto subdivisions, Tessmann (1930) 50-60, Ortiz (1942) 39, Farabee (1922) 16. Probably each of these had its particular dialect or variety; Pericot y Garcia (1936) copies Farabee’s subdivisions, names not mentioned, except as synonyms, in the classificatory table. The several authorities quoted, especially Ortiz (1942), mention a number of Witoto groups in addition to those given above, without indicating their relationships with other groups. Those that are not synonyms probably each has or had its own dialect. Among these are: Aefuye, Aipui, Ajayt, Bodydnisar, Gayafeno, Emenani, Eraye, Fayagene, Fusigene, Gibutie, Idekofo, Itchibuyene, Jetuye, Jidua, Joyone, Kanieni, Kotuene, Meresiene, Mobenidza, Nequerene, Nofuiqiie, Orotuya, Uitoto Piedra, Uiyokoe, Yane, Yari (Jémane, Neimade), Yusigene, and Yauyane. WirToTo ! I. Witoto A. Witoto 1. Witoto . Kaime (Caimo) Xura . Séuent . Jayruya . Mekka: Yaboyano . Menekka . Bie . Ifikuene-Caimito (?) B. Miranyan, Boran 1. Miranya-Carapana-Tapuyo 2. Nonuya (Achiote)? 3. Ocaina-Muenane a. Ocaina (Ducaiya); Fitita (?) ? b. Muenane ? C. Southeastern 1. Orején 2 2. Coeruna (?) 2 D. Andoke (?) ? 1. North a. Araracuara 2. South E. Resigero (?) 2 1 This classification is exceedingly and unusually controversia] and uncertain. 2 See independent short articles on these. PR mows TP NONUYA Nonuya is considered a Witotoan language by Nimuendajti (map). Tessmann (1930) leaves it as unknown or independent. Loukotka (1935) places it, with Muenane, in his Bora (Miranya) family. 246 SOUTH AMERICAN INDIANS [B. A. E. Bull. 143 MUENANE Steward (Handbook, vol. 3. p. 750) treats Muenane under Witoto. Nimuendajii (map) omits it. Few writers have mentioned them (Whiffen, 1915; Preuss, 1921-23). Loukotka (1935) places it under his Bora family (see Miranya), with Tupi intrusion and vestiges of Witoto. Tessmann (1930) gives a vocabulary of 38 words from which, by his comparative method, he deduces that it is a mixture of Ge and Carib, a doubtful conclusion. They should be considered as unclassified. FITITA Steward (Handbook, vol. 3, p. 750) treats the Fitita culturally under Witotoan, possibly a subdivision of Ocaina; the linguistic evidence has not been presented. Nimuendaji leaves them unclas- sified, a decision herein accepted. 'Tessmann (1930) could find almost no information on them and no clue to their linguistic affiliation. OREJON The Spanish word ‘‘orején,” “big ears,” was applied to native groups that wore large earplugs, distending the lobes, and, therefore, is applicable to several unrelated linguistic groups. The more im- portant of these are in the Witoto-Tucano region. Here, the name seems to be applied to two adjacent groups, resulting in inevitable confusion. One group, apparently known by no other synonym, is southern Witoto and apparently extinct. Another tribe known as Orején is the Coto (q. v.), which is generally agreed to be of Tucano linguistic affinity. COERUNA Though grouped with Witoto by all other authorities, including Rivet (1924 a) and Nimuendaji (map and index), Coeruna seems to be so different lexically that Loukotka (1935) makes it an independent family (with Witoto and Tucano intrusions). Apparently extinct, the lexical data seem restricted to the old vocabulary in Martius (1867). It is most often linked with Orején. ANDOKE Tessmann (1930), Castellvi (1934 b), and Igualada and Castellvi (1940) from first-hand acquaintance with the language of the Andoke consider it independent or of unknown affiliations. Rivet (1924 a) places it in the Witoto family and thinks that it may be identical with Miranya-Carapana-Tapuyo and Nonuya. Steward herein accepts 1t as Witotoan, but of unknown subclassification. Ortiz (1942) does not accept the Witoto relationship. Most of the other authorities do not mention it. Dr. J. P. Harrington (personal correspondence) Vol. 61 LANGUAGES—MASON 247 believes in the affiliation with Witoto. Though in the southern Colom- bian forests, it is not far distant from the extinct Andaki of the southern Colombian Highlands with which it is liable to be confused. RESIGERO The position and affiliations of Resigero are most uncertain. Tess- mann (1930) thinks it may be Bora (Miranya); Loukotka does not seem to mention it. Nimuendaji (map) leaves it unclassified. Igualada and Castellvi (1940) believe it related to Arawak. Ortiz (1942) doubts the Witoto relationship. BIBLIOGRAPHY Witoto.—Anonymous, 1919 e, 1930 a; Castellvi, 1934 b; Farabee, 1922, pp. 148-151; Hardenburg, 1910, 1912; Harrington, J. P., 1944; Igualada and Castellvi, 1940; Kinder, 1936; Koch-Griinberg, 1906 a, 1906 c, 1909-10, 1910 b; Martius, 1867, 2:277-279; Murdock, 1936; Ortiz, 1942; Pinell, 1928; Preuss, 1921-23; Rocha, 1905, pp. 202-206; Tessmann, 1930, pp. 311-329, 547-559 (Ocaina); Whiffen, 1915. Nonuya.—Tessmann, 1930, pp. 583, 617. Muenane.—Ortiz, 1942; Tessmann, 1930, pp. 329-337; Preuss, 1921-23; Whiffen, 1915. Fitita— Ortiz, 1942; Tessmann, 1930, pp. 583, 617; Whiffen, 1915. Orején.— Martius, 1867, 2:297-298. Coeruna.— Martius, 1867, 2:273-275. Andoke.—Castellvi, 1934 b; Igualada and Castellvi, 1940; Ortiz, 1942; Tess- mann, 19380, pp. 584, 617. Resigero.—Igualada and Castellvi, 1940; Ortiz, 1942; Tessmann, 1930, pp. 583, 617. ZAPAROAN Zdparoan is an important group that has been accepted as an independent family in all the major classifications since the earliest. However, while yet to be proved, it is not unlikely that it will be found to be related to Witoto, Miranya, and Tupi-Guarani, and it is tenta- tively accepted herein as a member of the Macro-Tupi-Guarani phylum. Rivet (1911 a) noted a large number of related words in these four languages and suggested the possibility of relationship but withheld final opinion. The most recent writer, Ortiz (1940 a), also continues to grant it independence, but Jijén y Caamaii, in his 1941- 43 classification, makes a Witote-Bora-Zdparo phylum, distinct from Tupi-Guarant. Loukotka (1935) finds a ‘‘Tupi intrusion” and Tess- mann (1930) considers the Zdéparo language a Carib-Tupi mixture. The linguistic data on Zdéparoan are very deficient, consisting of a few short vocabularies on a half dozen of the many component lan- guages, and a few short translated religious prayers. Almost nothing is known of the grammar and only a few points were deduced by Beuchat and Rivet (1908). The grammar is, probably mistakenly, said to be simple, the phonetics nasal and guttural, though the effect 248 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 is ‘‘agreeable.”” There are said to be pronominal possessive prefixes, pluralizing suffixes and pronominal “‘case’”’ suffixes. The classification of the component languages and dialects of the Zdparoan family is in utter confusion; no attempt has ever been made to do this on a scientific linguistic basis, and the available data are insufficient. Most compilers have merely given a list of names of groups, many of them geographical, and such suggestions as are made are based mainly on travelers’ published remarks regarding linguistic relationships. These, as well as the deductions based upon them, are highly equivocal and contradictory. The earliest authorities, Velasco (1840) and Hervis y Panduro (1800), were in practical agreement on three main divisions of the Zdparoans proper (excluding the Jquito, not included by Hervéas), and on the dialects composing these: Andoa: Simigae of Curaray (Ve- Simigae of Tigre (Velas- Araro (H.) lasco) ; Simigaecurari co); Jinori (Hervds): Bobonazo H.) (Hervas): Acamori or Acamaori Chudavina or Chuuda- Araza or Arazo Comacort vino Ginori (V.) Conejori (V.) Frascavina (V.) Iginori or Ijinori Iqueconejori Gae (H.) Incuri (V.) Itremojori (V.) Guazaga or Guazago Napotoa (V.) Panajori Macavina (V.) Nepa or Nevo Tremojort Murata or Murato Oa Pava or Pabo Yniri (V.) Pinche Zapa Quirivina (V.) Zapara or Zaparro Semigae (H.) Velasco divided the Jquito into two groups: Iquito of Tigre: Iquito of Nanay: Aicore Blanco Ayacore Huasimoa Eriteyne Himuetaca Neracamue Most of these groups are not mentioned, and probably are extinct or amalgamated today, as well as the dozens of small groups listed by other writers (Brinton, 1891 a, listed 62). No recent compiler has attempted to classify these small groups, but Rivet (1924 a), Pericot y Garcia (1936), and W. Schmidt (1926) recognize five main sub- divisions; Loukotka (1935), four, which more or less agree with the earlier divisions: Andoa Conambo ( Combo) Gae (not accepted by Loukotka) Iquito (Ikito) Zaparo Vol. 6] LANGUAGES—MASON 249 Ortiz (1940 a) gives the main dialects as Gae, Semigae, Iquito? Iginorri, and Panocarrv. Tessmann (1930) considers the Jquito independent, divided into two groups, the Iquito or Iquito Proper and the Cahuarano or Kawa- rano; he calls it a mixed Tukano-Pano language. Steward and Métraux (Handbook, vol. 3, p. 636) believe that this is erroneous, that the Jquito language is very similar to Gae, and that Tessmann’s Cahuarano may be Maracano. Nimuendajtii (map) places Jquito as Zdparoan. 'Tessmann believes the Gae to be strongly Carib: he calls the Zdparo Carib-Tupt and the Andoa Carib-Ge. He also considers the Murato to be a Candoshi (Maina) subtribe. According to other evidence (Handbook, vol. 3, p. 634), the Coronado were kinsmen of the Oa and must be distinguished from the totally unrelated (Zucano?) Coronado on the Aguarico River. ‘‘Auca’’ is the generic name for ‘‘pagan.” Andoa, Gae, and Semigae are closely related, but Andoa and Semigae are not synonymous, as Tessmann thought. The vocabulary supposed to be Murato (Anonymous, 1928 b) has no resemblance to Zdéparoan. Soronotoa may be a syno- nym for Semigae; the latter is very similar to Andoan Murato. Many of the Roamaina and Zdparo also speak Quechua. Comacor may be a subtribe of Semigae, of Roamaina, a synonym for Iquito, or a distinct tribe. Of doubtful affiliation with Zdparo are Aunale, Alabano, Curizeta, Sucumbio, and Neva. Steward and Métraux (Handbook, vol. 3, p. 634) make a Roamaina group of Zdparoan, apparently including Pinche, Zapa, Pava, Arazo, and some subsidiary languages. Roamavzna is generally placed with Omurano (q. v.), whose affiliations are so disputed that it probably should be considered unclassified. Tiputini (Tiwacuna) and Chiripuno are considered by Tessmann (1930) and Loukotka (1935) as languages akin to Sabela (q. v.), to which Loukotka accords independent status. ZAPARO I. Coronado Group A. Coronado (Ipapiza, Hichachapa, Kilinina) ! 1. Tarokeo 2. Chudavina (?) 3. Miscuara (?) B. Oa (Oaki, Deguaca, Santa Rosina) II. Andoa Group A. Andoa 1. Guallpayo 2. Guasaga 3. Murato ? 1 Beuchat and Rivet, 1909, classified the Coronado as Cahuapanan. 2 See separate article, Candoshi and Murato. 250 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 ZAparo—Continued II. Andoa Group—Continued B. Gae (Siavirz) C. Semigae . Aracohor . Mocosiohor . Usicohor . Ichocomohor . Itoromohor . Maithiore . Comacor (?) D. Iquito (Amacacora, Kiturran, Puca-Uma) 1. Iquito 2. Maracana (Cawarano ?) 3. Auve E. Asaruntoa (?) III. Zéparo Group A. Zadparo . Muegano . Curaray . Matagen Yasuni Manta Nushino . Rotuno Supinu NOP Whe DAA oP wpe BIBLIOGRAPHY Zaparoan.— Undifferentiated: Anonymous, 1928 b; Beuchat and Rivet, 1908, 1910; Igualada and Castellvi, 1940; Leén, A. M., 1928-29; Martius, 1867, 2: 302— 307; Ortiz, 1940 a; Orton, 1874; Osculati, 1854, pp. 281-301; Rivet, 1911 a, 1912 a, 1930 b, p. 696; Simson, 1879 b, 1886, pp. 263-266; Tessmann, 1930; Veigl, 1785 a, 1785 b; Velasco, 1840; Villavicencio, 1858. Andoa.—Brinton, 1898 b; Tessmann, 1930. pp. 530, 617. Semigae.—Anonymous, 1930 b; Leén, A. M., 1930 b; Tessmann, 1930. p. 534. Iquito.—Castelnau, 1852, pp. 295-296; Gonzdles Sudrez, 1904, pp. 69-74; Martius, 1867, 2: 302; Tessmann, 1930, pp. 512, 617. OMURANO (ROAMAINA?) Synonyms: Numurana, Hunurana, Roamayna (?). Tessmann (1930) gives an Omurano vocabulary of about 250 words, large enough for Loukotka (1935) to decide to give it a position as an independent family consisting of Omurana (with vestiges of Chimu) and the extinct Maina. To Tessmann it is a mixture of Arawak and Carib, rather strongly Arawak. He identifies them with the historical Roamaina (Roamayna). Brinton (1891 a) classifies Humurano (evi- dently the same as Hunurana) and Roamaina as belonging to the Maina; this classification follows Hervas. The Maina stock of Hervas y Panduro (1800), Brinton (1891 a) and Chamberlain (1913 a) is the Vol. 6] LANGUAGES—MASON 251 modern Cahuapana (q. v.), and Beuchat and Rivet (1909) classified Roamaina as Cahuapanan. Loukotka (1935) seems to have found little Cahuapana resemblance in Omurana. Steward and Métraux (Handbook, vol. 3, p. 634) consider the Roamaina as a division of Zaparoan (q. v.) and to be totally distinct from the Maina. It is believed to be represented by the extinct Zapa, and today by the Pinche, with the dialects Pava, Arasa, Uspa or Llepa, and Habitoa. (See Steward and Métraux, Handbook, vol. 3, p. 634.) With such difference of opinion, Roamaina-Omurano and Pinche had best be considered as unclassified. Bibliography.—Tessmann, 1930, pp. 444 (Omurana), 582 (Pinche). SABELA Sabela is a new group, discovered and named by Tessmann (1930) and mentioned by no other compilers except Loukotka (1935) and Jij6n y Caamajio (1941-43). 'Tessmann considers it a mixed “‘stem,”’ Ge-Arawak-Pano. ‘The data for linguistic classification seem to be limited to less than 30 words published by Tessmann, surely not enough on which to award it a status as an independent family, as Loukotka (1935) has done. ‘Uncertain affiliation” is the better decision for the present. The group now speaks Quechua. Jijén y Caamafio (1941-43) follows Loukotka in considering Sabela an independent ‘‘phylum.” The two divisions are Tihuacuno (Tibakuna and orthographic variants) and Chiripuno (Tschiripuno, Schiripuno); the degree of linguistic variation between them is not stated. TZiputini is close to or identical with Tihwacuno. Steward (Handbook, vol. 3, p. 747) mentions the Sabela, with the Tihuacuno and Chiripuno as tribes of uncertain affiliation. However, Steward and Métraux (Handbook, vol. 3, p. 629) list Shiripuno and Tiputini as dialects of Zéparo proper. In this region are many tribal names, Zdparo, Sabela, Jivaro, Chebero, Hibito, etc., that may well be phonetic variants of one root. Bibliography.—Tessmann, 1930, pp. 298, 617. CANELO Synonyms: Kanela, Napo, Santa Rosina, Loreto. Canelo was superseded by Quechua about 1580. The linguistic data are so few that its affiliations will probably never be certain; they may have been with Zdéparo, Jivaro, or Chibcha. Steward and Métraux (Handbook, vol. 3, p. 637), following Reinburg (1921), class it with Zdparoan. Karsten (1935) believes it a mixture of Jivaro, Zdparo, and Quechua; Rivet is (or was) convinced that it was related to Chibcha. PARA SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Jij6n y Caamaifio (1941-43) does not list it in his classification but apparently includes it in the Jéivaro area in his map. Early accounts indicate that it included Gae, generally considered a Zdparoan tongue, as well as three other unknown and extinct languages or dialects: Ymmunda or Ynmuda, Guallingo, and Sante or Santi. Other minor groups were Penday, Chontoa, and Canicha. Bibliography.—Karsten, 1935; Reinburg, 1921; Tessmann, 1980, p. 250. AWISHIRA The standard classification of Awishira (with many orthographic variants such as Abiira, Avixiri, Abira) in the Tucanoan family, accepted by Rivet (1924 a), Pericot y Garcia (1936), and Krickeberg (1922), has been doubted in recent years. Dr. J. P. Harrington has recently examined the data and concluded that the material has so far proved insufficient for definite inclusion with Tucanoan. Evidence cited in the Handbook (vol. 3, p. 635) suggests its relationship with Zdéparoan. W. Schmidt (1926) leaves it unmentioned. Nimuendaji (map) puts it with the unclassified languages. Neither Tessmann (1930) nor Loukotka (1935) apparently find anything Tucanoan in it in their comparison of vocabularies. Tessmann (1930) calls it a mixed- stem language, Pano-Arawak. Loukotka (1935) gives it an inde- pendent family, Auwsiri, finding vestiges of Ohibchan in it. This is probably too radical; as a living language it will probably be found before many years to belong to one of the larger groups, but as that one is uncertain it had best be left unclassified for the present. Bibliography.—Tessmann, 1930, pp. 495-489. NORTHERN TROPICAL LOWLAND FAMILIES OF PRESUMED INDEPENDENCE WARRAUAN Synonyms: Warau, Waraw, Worrow, Uarow, Guarau, Uarauno, Waraweti, Araote, Tivitiva. The independence of the Warrau linguistic family has been admitted by all authorities since earliest days. A littoral people of peculiar culture, the general feeling is that they preceded the Carib and Arawak in the Orinoco Delta and Guiana coast. No suggestions as to linguistic affiliations with other groups have been made; the Arawak vestiges that Loukotka (1935) finds are probably due to borrowing. Generally only one language is assigned to the family, but Nimuen- daji (map and index) places the Mariusa and Chaguan, tribes not mentioned by others, with them. The extinct Waikeri or Guayqueri may also be related; although Rivet (1924 a) considers the latter as Carib, Humboldt (1826) reports that they claimed that their language was related to that of the Guarauno. Vol. 6] LANGUAGES—MASON 253 WARRAU 1. Mariusa 2. Chaguan 3. Waikeri (Guaiquert) Bibliography.— Adam, 1897 b; Brett, 1868, pp. 176-199; Crévaux, Sagot, and Adam, 1882, pp. 263-266; Goeje, 1930, 1930-31; Hilhouse, 1834; Humboldt, 1822-27, 3: 216; Im Thurn, 1883; Olea, 1928; Quandt, 1807; Schomburgk, 1847-58, pp. 47-48; Tavera-Acosta, 1907, pp. 120-123, 1921-22; Williams, J., 1928-29. AUAKEAN Auaké (Auaqué, Auqué, Aoagui, Oewaku) is probably the smallest of all the South American linguistic “families” accepted as independent by all authorities. It may be now extinct, at any rate spoken by only a very small group, possibly of only one family. Unknown to Brinton and Chamberlain, the independent status is based on a vocabulary of some 300 words recorded by Koch-Griinberg (1928) in 1913, which apparently shows no connection with any language with which it has been compared, though Loukotka (1935) sees in it vestiges of Chibchan. Consonantal endings are few. Pronominal possession is expressed by prefixes. Armellada and Matallana (1942) call the language Arutanz. Bibliography.—Armellada and Matallana, 1942; Koch-Griinberg, 1913, 1922, 1928, 4: 308-313, 331, 332. CALIANAN Caliand (Kaliana) is one of the small linguistic groups which, un- known to earlier authorities, was discovered by Koch-Griinberg in 1913. The small vocabulary, collected by him, shows no resemblance to any of the surrounding languages, and the language, without any known subdivisions, is, therefore, accorded independent position by all subsequent compilers. No studies have been made, or at any rate published, upon it, and no suggestions regarding relationship, except that Loukotka (1935) claims vestiges of Chibcha in it. It should be distinguished from the Caliand of Tumuc-Humac. Many of the words are monosyllables with consonantal endings. Pronominal possession is expressed by prefixes. Bibliography.— Koch-Griinberg, 1913, p. 458; 1922, p. 227; 1928, 4: 313-317, 332-333. MACUAN The Macu (Mahku) “family” of the Uraricoera River must be dis- tinguished from several other groups of the same or similar name: the Sdlivan Macu, the Puinavean Macti (Loukotka uses Maku for the family name), and the Cofén Macu or Maco. The family has been ac- corded independent status since its discovery by Koch-Griinberg on the basis of the rather large vocabulary of nouns collected by him 254 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 (1928). Only the single language is known. No suggestions as to wider affiliations have been made, except that Loukotka (1935) finds Arawak intrusions. The language is said to be spoken rapidly. Pos- sessive pronominal elements are prefixed. Armellada and Matallana (1942) call the language Sope. Bibliography.—Armellada and Matallana, 1942; Koch-Griinberg, 1913, 1922, p. 227, 1928, 4: 317-324. SHIRIANAN Shiriand (Siriand) is one of the newer families, not recognized by Brinton or Chamberlain, but generally accepted since. Relationship with Carib has been suggested but uniformly rejected. The best lin- guistic source is Koch-Griinberg (1928); the languages are said to be “outtural,” the words often with consonantal endings. The principal languages of the family seem to be Shiriand and the less important Carimé (Karimé). Waikdé (Guaicd, Uaicd) appears to be closer to Shiriand, possibly a dialect; it must be distinguished from a Cariban Waika in British Guiana, and Shiriand must not be confused with the Arawakan Shiriana nearby. Most modern writers are agreed that Shiriand and Wavkd are de- scended from the historically important Guaharibo (Iaribu, Guahabo, Uariba, Uajaribo, Uaharibo), though Nimuendajii (map) leaves the latter unclassified; Brinton (1891 a) assumed that they were Carib. SHIRIANA A. Waharibo (Guaharibo) 1. Shiriand a. Waicdé (Guaica, Uaica) B. Carimé (Karimé) Bibliography.— Armellada and Matallana, 1942; Koch-Griinberg, 1913, pp. 454-456, 458, 1928, 4: 302-308, 331, 334; Salathé, 1931-32; Tavera-Acosta, 1921— 22, pp. 230-331. SALIVAN, MACU, AND PIAROA The independence of this stock has been long and universally recog- nized, but the names Sdliva(n) and Piaréa(n) are about equally em- ployed to distinguish it. No suggestions for wider relationships have been made, though Loukotka (1935) sees Arawak intrusions in Sdliva, and Carib vestiges in Piaréa. The majority of authorities agree that three languages compose the family: Sdliva (Sdliba, Sdlliba, Sdliua, Sdlivi), Piarda, and Macu (Maco). Loukotka (1935) groups Piaréda and Macu in an eastern group, Sdlivi and Pamigua (q. v. infra) in a western group. The Sdliwwan Macu must be distinguished from three other tribes of the Vol. 6] LANGUAGES—MASON 255 same or very similar names: the independent Macu of the Auari River, the Puinave Mact and the Cofan Macu (q. v.). Sdliva was one of the nine stocks of the Orinoco Valley established by Gilij (1780-84); the component languages he names as Ature, Piaréa, and Quaqua. 'The Ature (Adole) are extinct; the name is also given as asynonym of the Piaréa. Brinton (1891 a) says the modern Quaqua (Kuakua, Guagua) speak Arawak; Rivet (1924 a) puts them with the Carib. Brinton (1891 a) furthermore found no resemblance between Sdliva and Piaréa; the languages must be considerably different SAuiva (PraRr6éA) I. Western. Sdliva A. Saliva II. Eastern: Piaréa A. Piaréa 1. Ature (Adole) 2. Piaréa 3. Quaqua (Guagua) (?) B. Macu Bibliography.—Adelung and Vater, 1806-17; Chaffanjon, 1889, pp. 324-326; Chamberlain, 1910 a, pp. 195-198; Crévaux, Sagot, and Adam, 1882, pp. 257-258; Ernst, 1895, Gilij, 1780-84, pp. 383-384; Koch-Griinberg, 1913, pp. 469-472, 1922, p. 236, 1928, 4: 351-357; Loukotka, 1929-30; Marcano, 1890 a; Oramas, 1914; Rivet, 1920 a; Schuller, 1912 a; Tavera-Acosta, 1907 a, pp. 85-107. PAMIGUA AND TINIGUA Castellvi has recently (1940) published a vocabulary of Tinigua. He finds no resemblance to any other family, but an evident connec- tion with the extinct Pamigua. He therefore suggests a Pamiguan linguistic family, consisting of Pamigua and Tinigua. Rivet (1924 a) placed Pamigua in the Guahiban family, and Loukotka (1935) placed it with Sdliva in the western group of Sdlivan (Piaréan). Igualada and Castellvi (1940) considered Tinigua as composing a third or Southern group of Sdlivan. The language is still spoken in the Caqueta region of Colombia. Bibliography.—Castellvi, 1940; Jij6n y Caamajfio, 1941-43, 2:108-109; Ernst, 1891. OTOMACAN, GUAMO OR GUAMA, AND YARURAN The Otomaco and Yaruro, small groups that are close neighbors in southern Venezuela, are limited to one language each. The Otomaco are extinct, the Yaruro approaching it. Both have been accorded independent linguistic status since, at the least, Brinton (1891 a) and Chamberlain (1913 a). Otomaco (Otomac, Otomak).—-Rivet (1924 a) says the language 256 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 is called Tarapita.”. Loukotka (1935) sees Carib intrusions in it, a distinct possibility. Guamo or Guama.—Loukotka (1935) establishes an imdependent Guama family of one language, the Guama, in which he finds vestiges of Chibchan. This is apparently the small tribe of southwestern Venezuela termed Guamo by other writers. Petrullo (1939) says that their remnants became extinct in recent years. Most authorities do not mention them; Nimuendaji (map) leaves them unclassified. Loukotka (1935) does not mention the data on the basis of which he isolates them, but it is doubtless insufficient evidence on which to establish a new family; they had better be left with the many on which we lack sufficient information for classification. Kirchhoff (Hand- book, vol. 4, p. 439) believes that Guamo was very closely related to Otomaco. Yaruro (Yarura, Jarun, Jaruro, Sayuro, Zavuri, Saururi, Japaria, Yapin, Yuapin, Pume, Pumeh).—Pumeh is said to be their name for themselves. Loukotka’s opinion (1935) regarding ‘“‘Coroado intru- sions” is to be doubted. Jijén y Caamajfio (1941-43) considers Yaruor as related to Chibcha, forming, with Esmeralda, his first or ‘‘Paleo- Chibcha”’ group of his Macro-Chibcha phylum. BIBLIOGRAPHY Otomaco.—Chamberlain, 1910 a, p. 194; Crévaux, Sagot, and Adam, 1882, pp. 262-263; Gilij, 1780-84, 3: 262-263; Rosenblat, 1936; Tavera-Acosta, 1907, pp. 332-374. Yaruro.—Adelung and Vater, 1806-17; Chaffanjon, 1889, pp. 319-320; Cham- berlain, 1910 a, p. 201; Crévaux, Sagot, and Adam, 1882, pp. 260-261; Miiller, Frederich, 1882; 1: 360-363, 2; Oramas, 1909; Petrullo, 1939, pp. 265-289; Tavera-Acosta, 1907, pp. 120-133. GUAHIBAN Guahibo (Wahibo, Goahivo, Goahiva, Guayba, etc.) has been univer- sally recognized as an independent family since earliest days (Brinton, 1891 a; Chamberlain, 1913 a), and no suggestions as to larger rela- tionships have ever been made. Its true status should not be difficult to determine since, in addition to a published grammar (Fernandez and Bartolomé, 1895), they form a large living group not very imac- cessible. Churoya, generally now recognized as one of the languages forming the family, was formerly considered as the type member of another family, the Churoyan, accepted by Brinton (1891 a) and at first by Chamberlain (1907). Not all of Brinton’s Churoyan lan- guages, however, are now considered Guahiban; Cofén and Mako are 21 Rivet (1924 a) distinguishes Tarapita and Taparito, which latter Gillin and Nimuendaji agree with him in considering as Cariban (q. v.). Kirchhoff (Handbook, vol. 4, p. 464) calls Taparita a variety of Otomaco; possibly some confusion in names exists. W. Schmidt (1926) considers Taparito as isolated . Vol. 6] LANGUAGES—MASON 257 generally put with Cofdn (q. v.). A careful comparison of Guahibo and Oofdn is, therefore, indicated; Ortiz (1943, 1944) doubts the relationship. GUAHIBO I. Guahibo A. Wahibo 1. Chiricoa a. Sicuane 2. Cuiba a. Mella b. Ptamo 3. Yamu B. Churuya 1. Bisanigua C. Cunimia 1. Guayabero Possibly: Amorua (generally placed with Arawak), Catarro, Cuiloto, Mazba. Bibliography.—Chaffanjon, 1889, pp. 320-323; Chamberlain, 1910 a, p. 186 (Churoya) ; Crévaux, Sagot, and Adam, 1882, pp. 258-260; Ernst, 1891; Fernandez de San José and Bartolomé, 1895; Koch-Griinberg, 1913, p. 472, 1928, 4: 443-451; Loukotka, 1929-30, 1938; Marcano, 1890; Ortiz, 1943, 1944; Ossa. V., 1938; Pérez, 1935; Reichel-Dolmatoff, 1944; Rivet, 1912 a, pp. 128-131; Sdenz, 1876, pp. 341- 342 (Churoya); Schomburgk, 1849; Tavera-Acosta, 1907, pp. 85-95. PUINAVEAN OR MACU Puinave, a family of slight present extent, may have been formerly much more important and widespread. Rivet, Kok, and Tastevin (1924-25) suggest that it may be a substratum common to many adja- cent languages, now so differentiated that a common origin is difficult to prove. It has been recognized as independent since at least the days of Chamberlain (1913 a). W. Lehmann (1920) suggests a possible relationship with Witoto (q. v.), which might tie it to Tupi. Loukotka (1935) terms the family Maci from its most important tribe, and Jijon y Caamajfio (1941-43) distinguishes a ‘‘Phylum Maki,” but the older and more generally accepted name Puinave is preferable to avoid confusion; Maci% must be distinguished not only from the Macu family, but also from the Praréan Macu and the Cofdn Macu. W. Schmidt (1926) includes in the family a Cabere language, generally regarded as an Arawak dialect. There is said to be considerable differentiation between the dialects, and much borrowing from Tucanoan languages. Some of the linguis- tic characteristics are: Accent generally on the ultima; vocalic har- mony; noun-verb composition; postpositions; compound nouns; pronominal and reciprocal prefixes; gender, pluralizing, tense and mode suffixes. 258 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 PuINAvVE or Mact I. Puinave (Puinabe, Puinavis, Uaipunabis, Guaipunavos, Uaipis) A. Puinave (Epined) 1. Western a. Bravos, Guaripa 2. Eastern a. Mansos B. Mact . Maci Tikié . Kerari . Papuri . Nadébo ore whe BIBLIOGRAPHY Puinave.—Chamberlain, 1910 a, p. 196; Crévaux, Sagot, and Adam, 1882, pp. 255-256; Ernst, 1895, pp. 396-398; Koch-Griinberg, 1928, 4: 335-343; Oramas, 1913 b, pp. 20-25; Tavera-Acosta, 1907, pp. 97-107. Maci.—Koch-Griinberg, 1906 b, pp. 885-899, 1909-10, 19138, 1922, 1928, vol. 4; Rivet, Kok, and Tastevin, 1924-25, pp. 148-185; Rivet and Tastevin, 1920, pp. 76-81; Tastevin, 1923 a. TUCANOAN (BETOYAN) Tucano is one of the important, though not one of the great, lin- guistic families of South America. It occupies two adjacent homo- geneous areas in the northwestern central part of the continent, separated by Carib and Witoto groups. A small isolated third group to the north is no longer credited. The name refers to the toucan bird, a translation of the native name, Daxsea. There may be 10,000 speakers of the Tucano languages today. The name Tucano or Tukano, generally accepted today, was first proposed by Beuchat and Rivet (1911). Brinton (1891 a) and Chamberlain (1913 a) termed the family Betoya(n) from the Betoi or Betoya tribe. The latter, however, Beuchat and Rivet demonstrated to belong to the Chibchan family, as apparently do also a number of the other languages given by Brinton (1891 a, p. 273) as members of the Betoyan stock: Anibali, Kilifaye, (Quilifaye), Situfa, Tunebo, and probably Jama. No suggestions have been made that would connect Tucano with other linguistic families. The languages are said to be characterized by consonantal clusters, nasalization, unclear pronunciation, and unusual phonemes. Stems are composed, verbal modifications are by means of suffixes, there being apparently no verbal prefixes and no reduplication. Nominal plural is expressed by suffixes or by modifica- tions of the ultima; prepositional relations are shown by postpositions. Indirect object is expressed by a suffix, as is feminine gender. No careful attempt has ever been made to subdivide the Tucano Vol. 6] LANGUAGES—MASON 259 languages on a linguistic basis. It is generally assumed that linguistic divisions follow the geographical ones, adjacent tribes speaking more closely related languages than distant ones, sibs of tribes speaking dialects. This divides the family into two main groups. This classi- fication is herein accepted as the only available one, with the realiza- tion that it may be without any basis. The following classification is based primarily on Steward (Hand- book, vol. 3, pp. 737-741) and Goldman (Handbook, vol. 3, pp. 764— 766), partly on W. Schmidt (1926). It is very tentative. TucaNo CLASSIFICATION I. Eastern Tucano A. Cubeo (Cobewa, Kobeua, Pamiwa) 1. Cubeo a. Héléna (Holowa? ) b. Hahdénana (Hehenawa) B. Tucano-Tuyuca 1. Tucano (Déchsea) a. Arapaso (Cored) b. Neenod c. Yohorod d. Uiua Tapuyo 2. Tuyuca (Dochkdéfuara) a. Bard } b. Tsola 3. Wanana (Ananas, Kotitia) a. Waiana (Yuruti Tapuya) b. Piratapuyo (Waickea, Urubi Tapuya) c. Uaicana d. Uainana 4. Carapané? (Méchda) 5. Wdséna (Pisé Tapuya) 6. Paméa (Tati Tapuya) C. Buhdgana (Carawatana-mira) 1. Macuna a. Hobacana (Japuana) 2. Buhdgana a. Oméa b. Sdra ce. Dod d. Tsaina e. Tsolod f. Ydba D. Desana (Wind) 1. Yupud E. Cueretu 1. Cashiita (Kusziita) 1W. Schmidt (1926) considers Bara, Uaiana, Uaséna, Méchdad, and Pamoa as subtribes of Uandna, 1 Loukotka (1935) considers Carapana as Witoto, mixed with Tucano 794711—50——_18 260 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Tucano CiLassIFICATION—Continued I. Eastern Tucano—Continued F. Yahuna 1. Opaina (Tanimboca) 2. Ddtuana G. Béléa H. Erulia II. Western Tucano 3 A. Piojé-Sioni 4 1. Encabellado (Icaguate, Angutera) Piojé (Pioche) Encabellado Secoya-Gai Campuya Santa Maria Guaciguaje Cieguaje Macaguaje . Amaguaje a ie oes 2. Sioni B. Correguaje-Tama ® 1. Correguaje a. Correguaje 2. Tama (Tamao) a. (Ayrico) C. Coto Other possible Western Tucano languages or dialects are: Tetete,6 Pasto,’ and Awishiri 8 (q. v.). 3 See Steward (Handbook, vol. 3, pp. 737-739) for smaller Encabellado divisions; Beuchat and Rivet (1911) for many small Tucano subdivisions and synonyms. 4 The terminology among the Piojé-Cioni is greatly confused. Steward (Handbook, vol 3, p. 738) states that the five principal names Encabellado, Icaguate, Piojé, Santa Maria, and Angutera are synonyms for the entire group as well as names of divisions used at different periods. 5 The Correguaje and Tama are said to be closely linked, but Rivet (1924 a) places Tama, together with Ayrico, the latter not mentioned by other modern sources, in a third or Northern Tucano group, leaving Correguaje with the Western Tucano. 6 Tetete is mentioned as an important Western Tucano group by Castellvi (Igualada and Castellvi, 1940) and Loukotka (1939 a). 7 On grounds of toponymy Rivet (1924 a) places the extinct and peripheral Pasto with the Western Tucano; Loukotka (1935) accepts this. It is herein classified with the Parbacoa group of Chibchan. 8 One of the doubtful questions is the inclusion of the Awishiri (Abijira, etc.) (q. v.), traditionally included with the Tucano. (See independent article supra under Zéparoan classification. Other groups, considered by one or more authorities as Tucano and not disputed by others, are: Bahukiwa, Corocoro Tapuya, Corewa, Palanoa, Patsoca, Usa Tapuya, Wantuya, Yi Tapuya, Yiboya Tapuya. Some of these may be synonyms of others given above. BIBLIOGRAPHY General.—Beuchat and Rivet, 1911; Brinton, 1892 a, pp. 62-67, 1892 b; Castellvi, 1939; Pfaff, 1890, pp. 603-606; Rivet, 1916, 1929 a. Eastern.—Adelung and Vater, 1806-17; Coudreau, H., 1887, pp. 464-474; Giacone, 1939; Koch-Griinberg, 1906 c, 1909-10, p. 324, 1912-16; Kok, 1921-22; Vol. 6] LANGUAGES—-MASON 261 Martius, 1867, 2:164—166, 275-276, 281-285; Stradelli, 1910, pp. 236-317; Wallace, 18538, pp. 520-521. Western.— Anonymous, 1909, 1919 e; Chantre y Herrera, 1901; Crevaux, 1882; Gonzd4lez Sudrez, 1904, pp. 63-64; Rivas, 1944; Rocha, 1905, pp. 199-201; Simson, 1879 a, pp. 210-222, 1886; Tessmann, 1930, pp. 205-221. COTO The Coto (Koto) is one of several groups known as Orejon, ‘‘big ears,”’ and sometimes Coto is considered the only synonym of Orején (Nimu- endajii, index). However, an extinct group of southern Witoto was also known by this name (q. v.). The Coto (as apart from Orej6n) are placed by all authorities under Tucano, and Tessmann (1930) reports that his 235-word vocabulary is nearly pure Tucano. He says that in addition to Orején they are known as Payagua and Tutapisho. There are said to be 500 living today. They must be distinguished from the Chibchan Coto of Costa Rica. Bibliography.— Marcoy, 1875; Simson, 1886; Tessmann, 1930, pp. 189-205. CAHUAPANAN The Cahuapana (Kahuapana, Cawapana, Maina) family is relatively unimportant, occupying a small region without enclaves, and spoken by a few thousand persons. Though it will probably eventually be found to belong to some greater family or phylum, no suggestions of larger relationships seem to have been made: Rivet, who has published the principal studies (Beuchat and Rivet, 1909; Rivet and Tastevin, 1931), made none. Within the family the classification is also difficult; little is known of many of the languages and dialects, and many, which would better be left unclassified pending fuller data, are generally included on purely geographical grounds. The name Cahua- pana, preferred by Beuchat and Rivet, supplants the former Mainan or Mayna of Brinton (1891 a) and Chamberlain (1913 a). The language is said to be harsh and difficult. A kind of true inflection with different suffixes for each tense and person is found. Modal relations are expressed by either suffixes or prefixes, also some infixes, but verbal and nominal suffixes predominate over prefixes. Pronominal (except subject) and gender relations are shown by suffixes, as are nominal and verbal pluralization. Loukotka adds an extinct Mikird to the family. Omurano (Humu- rano) and Roamaina (q. v.) were long considered prominent members of this family but are not mentioned by later authorities. Hervas y Panduro (1800) gave Chayavita as a separate stock, consisting of Cahuapano and Paranapuro. See Handbook (vol. 3, pp. 605-608) for comments on relationships and critique of former classifications. 962 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Steward and Métraux question the relationship of the entire Mayna group. CAHUAPANAN I. Cahuapana A. Cahuapana 1. Cahuapana 2. Concho (Chonzo) II. Chébero (Xevero) 1. Chébero ! 2. Chayawita (Chawt) a. Chayawita b. Yamorai (Balsapuertino) ? 3. Ataguate 3 (?) 1Chébero (Xévero) must not be confused with the Jivaro (Xivero) (q. v.) or the Hibito. 2 The Yamorai are mentioned only by Tessmann (1930), on whose opinion they are here included. 3 Ataguate is mentioned only by Beuchat and Rivet (1909, p. 619), who also include Cutinana and Tivilo, languages apparently related to Aguano (a. v.). Bibliography.—Adelung and Vater, 1806-17; Beuchat and Rivet, 1909; Brinton, 1892 a, pp. 21-29; Ortiz 1941 b; Rivet, 1912 a; Rivet and Tastevin, 1931; Schuller, 1912 d; Tessmann, 1930, pp. 378, 415, 440-444; Veigl, 1785 a. MUNICHE Synonyms: Munitschi, Muniti, Paranapura. Muniche was one of the old stocks of Hervds y Panduro (1800), quoted without comment by Brinton (1891 a). The component languages were Muchimo and Otanabe, doubtless extinct and not men- tioned by any recent writer. Neither was Muniche until Tessmann (1930) rediscovered it. He considers it a mixed-stem language, Ge-Pano, with 16 resemblances, in about equal proportion, out of the 33 that he used as a criterion. Loukotka (1935), the only compiler who mentions the group, gives it independent status. Steward and Métraux (Handbook, vol. 3, pp. 606-607) consider Muniche related to Chébero, of the Mayna branch of Cawapana. As Tessmann’s collected vocabulary amounts to only 38 words, this is hardly enough on which to base any opinion; it should be left unclassified pending further study. Tessmann (1930) appears to be the sole source of information. MUNICHE 1. Muniche 2. Muchimo 3. Otanabe 4. Churitana PANOAN Pano is one of the major linguistic families with many subdivisions. Tradition brings them from the north, but no relationship with any other family has ever been suggested, much less demonstrated. Vol. 6] LANGUAGES—-MASON 263 Rivet (1924 a) observes that the Tacana group (q. v.), which he believes to be Arawak, shows considerable grammatical resemblances to Pano. Neighboring languages seem to have affected the Panoan languages little. Comparatively few of the component languages are extinct and some groups number several thousands; the total number of speakers may be about 15,000. They occupy four isolated homo- geneous areas east of the Andes in far west Brazil and adjacent repub- lies; the central one is the largest in area, number of groups, and importance. Not even vocabularies exist from many, probably from most, of the tribes, and no classification other than into the three main areas, has ever been attempted. Such comparisons as have been made indicate that the linguistic groupings on the whole follow the geographical divisions. The languages of the Juru4-Purts seem to fall with the central group rather than with the southwestern or the southeastern. The main or central branch of the Pano comprises a very large number of tribes and subtribes, each of which doubtless spoke a more or less variant language or dialect. For lack of sufficient data, no one has attempted a classification of these languages, though some are said to be mutually intelligible, others not. On the presumption that adjacent and affiliated groups are linguistically more closely related than more distant groups some classification may be made, but this is a dangerous assumption, the groupings highly tentative. The two main groups are those on the Ucayali, the names of which typically end in -bo (a pluralizing suffix), and those on the Jurud and Purtis, with the majority of names ending in -nawa, meaning “strangers.” THE CHAMA LANGUAGES The Conibo, Setebo, and Shipibo of the Ucayali River are grouped under the Chama (Tschama, Cama), and are apparently mutually intelligible. While the groups are large and important, the speech forms would probably be classed as varieties or dialects of the Chama language. This must be distinguished from the Tacanan Chama on the Madidi River. Except for the three above-named main groups there is disagreement regarding the affiliation of certain other groups with Chama. According to Steward and Métraux (Handbook, vol. 3, pp. 559-560), Setibo is divided into two subgroups, Sensi and Panobo, which, perhaps with Puinawa, separated from the parent group in recent days. As regards Panobo there is general agreement on its close association with Setibo. Pano, Pelado, Manoa, and Cashiboyano are given as synonyms or subdivisions. In addition to the obvious synonyms (Sensvwo, Ssenssi, Senci, Sentr), Tenti, Barbudo, and Mananawa (Mananahua, 264 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Mananagua) are given as synonyms of Sensi. There is great diversity of opinion regarding Mananawa, which Rivet (1910 b) considers a branch of Shipibo, Tessmann (1930) a branch of Cashibo. Nimuendajt (index) refuses to classify Mananawa; K. G. Grubb (1927, p. 88), calls it a subtribe of Remo (q. v.) with the subdivisions Marubo and Pisabo (q. v.); the latter languages are not generally classed in the Chama group. Steward and Métraux (Handbook, vol. 3, p. 560) con- sider Ynubu, Runubu, and Casca as subdivisions of Sensi, inferentially dialects or varieties thereof. Both Tessmann (1930) and Loukotka (1935) claim that Sensi is not as pure Pano as the other Chama languages; Loukotka places it in a separate “mixed” group. Steward and Métraux (Handbook, vol. 3, p. 561) say that Shipibo is the same language as Setibo and the extinct Caliseca (Kaliseka). Rivet (1910 b) agrees with the latter and adds Mananawa as a sub- tribe. The name Caliseca has not been used since the 17th century (Steward and Métraux, Handbook, vol. 3, p. 561) and it is disputed whether they were the modern Shipibo or Cashibo. CASHIBO Cashibo (Kaschibo, Cachibo, Cahibo, Cacibo, Caxibo, Casibo, Cahivo, Kassibo, Kasibo), also given the synonyms of Mananawa (Mananagua, Managua, Mananabua), Carapacho, Haqueti (Hagueti), and San Lorenzo, is generally placed in the Chama language, but according to a local missionary (Rankin, personal correspondence) the speech of the Cashibo is unintelligible to the Conibo-Shipibo-Setibo. 'The sub- groups are given as Cacataibo (Kakataibo), Cashifio, and Runo by Steward and Métraux (Handbook, vol. 3, p. 564); as Buninawa (Bununahua, ete.), Carapacho (Karapatsa, etc.), and Puchanawa (Putsanahua, ete.) by Rivet (1910 b), and as Cashinéd (Kaschin6d) and Shiriné (Ssirinéd) by Tessmann (1930). K. G. Grubb (1927, p. 84) says that the Cashinawa are closely related. The differences of opinion regarding Mananawa and Caliseca, listed above as synonyms of Cashibo, have been already considered. Steward and Métraux (Hand- book, vol. 3, p. 564) consider Carapacho as of uncertain affiliation; P. Marcoy (1875) calls it a synonym of Caliseca. Nocoman (Nokamédn), recently identified by Tessmann (1930, p. 172), is probably now extinct. It was formerly confused with Cashibo (Handbook, vol. 3, p. 567). Nimuendajti (map) leaves it unclassified. Loukotka (1935) and Jijén y Caamafio (1941-43) place it with Pano. Other languages or dialects, generally agreed to be Panoan, whose closer affiliations are unknown but which probably belong in the Uca- yali group, are Pichobo (Pitsubo, Pichaba, Pitsobu, Pisabo, etc.) and Soboibo (Saboibo, Soboyo, Soyboibo, Bolbo); Mochobo (Mochovo, Uni- Vol. 6] LANGUAGES—MASON 265 vitza) and Comobo (Comavo, Comambo, Unimtza). K.G. Grubb (1927) calls Pisabo a subdivision of Mananawa, the latter a subtribe of Remo. Considered as Pano by Nimuendajt (map), but not noted in other lists: Mainawa (distinct from Marinawa), Yananawa, and Yumba- nawa. Considered as Pano by Nimuendajti (map); listed as unidentified tribes of Ucayali by Steward and Métraux (Handbook, vol. 3, p. 567): Puyamanawa (Punyamumanawa) and Camarinawa (Kamarinigua). Ruanawa (Ruanagua, Rununawa) and Puynawa (Poyanawa, Poianaua, Puynagua) are less often mentioned Panoan groups (Steward and Métraux, Handbook, vol. 3, p. 566). The following groups of the Jurué-Purts Basin are considered as Panoan, without any suggestion as to their relationships inter se or with neighboring groups. Probably each is entitled to the status of a dialect or variety, though some may be radically different. Canamari, (Kanawari, etc.)—These must be distinguished from the Arawak and Catukina Canamari. W. Schmidt (1926) places them with the Catukina in a separate Middle Group of Pano, a division of the Northern Pano not recognized by other authorities. Curina (Kulino, etc.).—These must be distinguished from the Arawak Culino or Kurina. Contanawa (Kuontanaua, etc.). Espino. Marinawa. Nana, probably same as Nawa. Nawa. Pacanawa (Pakanaua, etc.). Shanindawa (Shaninawa, Saninaua). Distinguish from Saninawa. Shipinawa (Sipznaua, etc.). Tushinawa (Tudsinaua, etc.). Yaminawa (Jaminaua, Yuminawa, etc.). Yawavo (Jauabo, Yawabu, etc.). Distinguish from Southeastern Pano Jau- navo or Caripund. Yura. Remo (Rhemo), Sacuya, Maspo (Impeniteri), Nucuini, Cuyanawa, and Niarawa (Niamagua) seem to form a related group. Possibly all belong with the Chama language, since Steward and Métraux (Handbook, vol. 3, p. 565) call Remo a Conibo dialect. Mananawa is mentioned as a subtribe of Remo, but this is not generally accepted. Nucuini is considered by Loukotka (1935) as a synonym of Remo. Sacuya (Sakuya) is generally linked with Remo and probably is a subgroup. Maspo or Impeniteri is also generally linked with Remo. Cuyanawa (Kuyanahua, etc.) is apparently associated with Nucuini. Possibly to the above group belongs Capanawa (Kapanagua, Kapahuana, etc.) with its subgroup Buskipani (Busquipani). ‘The 266 SOUTH AMERICAN INDIANS [B. A. B. Bull, 143 latter name has also sometimes been considered as a synonym of Amawaca. Various opinions seem to correlate the languages or dialects Catu- kina, Arara, Ararapina, Ararawa, Saninawa, Saninawacana, and Shawanawa. Catukina (Katoquino, etc.) must be distinguished from four other languages of the same name of other families. Schmidt (1926) places it with Canamari in a separate Middle Group of Pano. Pericot y Garcia (1936) considers Arara, Ararapina, and Ararawa as subgroups of Catukina, and Rivet (1924 a) states that these three are doubtless identical with Saninawa. Métraux (Handbook, vol. 3, p. 660) links Shawanawa with the Arara group, and believes that Saninawa is related to Saninawacana. According to Steward and Métraux (Handbook, vol. 3, p. 565), the Amawaca (Amaguaco, Amajuaka, Ameuhaque, etc.) are close relatives of the Cashinawa (Kachinaua, Cushinahua, Caxinagua, etc.). Rivet (1924 a) gives Maspo and Impetineri as synonyms of Amawaca but Steward and Métraux (Handbook, vol. 3, p. 566) place the former in another group. Sayaca (Sacuya ?) and Busquipani have also been given as synonyms (or subdivisions) of Amawaca, but these also seem to belong to other groups (vide supra); nevertheless, the possibility of the relationship of all these must be considered. Pericot y Garcia (1936) calls Arawa a subtribe of Amawaca; Arawa is generally placed with the Southwestern Pano group (q. v.). Steward and Métraux (Handbook, vol. 3, p. 566) believe that the extinct Amenwaca (Amen- guaca) is a synonym of Amawaca; it has many groups and subgroups, the best known of which are Inuvakeu and Viwivakeu (Viuivaqueu). W. Schmidt (1926) links Cashinawa with Sheminawa. Of the groups mentioned above, Steward and Métraux (Hand- book, vol. 3, p. 567) consider these Ucayali languages unidentified but probably Panoan: Camarinawa, Puyamanawa, Saninawaca, and Sinabu. Also these not before mentioned: Awanawa (Aguanagua), Chunti, Diabu, Isunawa (Ysunagua), Ormiga, Trompetero, and Viabu. There is general agreement as to the composition of the Southwestern Panoan group in the region of the basin of the Inambari. The two important groups seem to be Arasa (Arasaire, Arazaire, Aratsaira) and Atsawaca-Yamiaca. The Atsawaca (Atsahuaka, etc.) or Chaspa, and the Yamiaca (Yamiaka) or Hasufeiri dialects (?) must be very similar, as they are said to have separated in historic times. Arasa is probably more divergent. Some of the Arasa and Atsawaca also speak Tacana (Arawak ?) and are often classified in that group. Nordenskiéld’s Arasa vocabulary (1908) is Yacana, but Llosa’s Arasaire vocabulary (1906) is Panoan. A doubtful member of this group is Araua, which language must be considerably different from the others. Nimuendajti (map) leaves Vol. 6] LANGUAGES—MASON 267 it unclassified; others consider it a link between the Southwestern and Northern Pano. Stiglich (1908) considers it a small group of Amawaca (Northern Pano). It must be distinguished from the Arawak Araua. The Southeastern Pano group is generally known as Pacaguard (Pacawara, Pakavara). The consensus is that the Pacaguard group consists of four languages, the interrelations of which are not sug- gested. These four are Chacobo (Tschakobo), Caripund, Capuibo (Kapuibo), and Sinabo; the position of the latter alone is disputed. The Caripund (Karipund) or Jau-navo (Jatin Avé) are divided into subtribes, the Jacarid (Jakarid, Yacariae, Jacaré-Taputiya, Yacaré- Tapuuya) and the Pamd (Pamand4). The position of the Sinabo is uncertain. Two groups are men- tioned in the literature, the Sinabo and the Shenabu, with inter- mediate orthographical variants. Rivet (1924 a) classifies them both as members of the Pacaguard group, and probably synonymous. One or the other is also classified as a subtribe of Shipibo (Northern Pano), as related to Sensi, and as unidentified. Doubtful also is the position of Zurina; Métraux (Handbook, vol. 3, p. 450) apparently considers it a member of the Pacaragudé group while Nimuendajt (map) leaves it unclassified; the others fail to mention it. Pano I. Central A. Chama (Ucayali) 1. Conibo a. Conibo b. Shipibo a. Caliseca, Sinabo (?) b. Manamabobo, Manava c. Setebo a. Sensi: Casca, Runubu, Ynubu, Barbudo, Tentz, Mananawa (?) b. Panobo: Pano, Pelado, Manoa, Cashiboyano 2. Cashibo (Comabo) . Cacataibo . Cashitio . Rufio . Buninawa . Carapacho (?) . Puchanawa g. Shiriné B. Curina (Kulino)! C. Capanawa 1. Capanawa a. Buskipani 2. Remo a. Sacuya moaeoodp 1 Distinguish from neighboring Arawak Culino or Culina. 268 SOUTH AMERICAN INDIANS Pano—Continued I. Central—Continued C. Capanawa—Continued 3. Maspo a. Epetinerit (Impenitari)? 4. Nucuint a. Cuyanawa 5. Niarawa 6. Puyamanawa (?) D. Amawaca (Amenguaca ?) 1. Amawaca a. Cashinawa a. Sheminawa b. Inuvakeu c. Viwivakeu 2. Pichobo a. Pichobo (Pisobo) b. Soboibo a. Ruanawa ce. Mochobo a. Comobo E. Catukina 3 1. Arara a. Shawanawa 2. Ararapina 3. Ararawa 4. Saninawa a. Saninawacana F. Juru4-Purts Povanawa Shipinawa Ararawa Yauavo Yaminawa Rununawa Contanawa Yawanawa Pacanawa Yumbanawa 1 Yara, 12. Tushinawa 13. Marinawa 14. Espin6d 15. Manawa 16. Canamari 4 — ge I aid ie [B. A. B. Bull. 143 2 Steward and Métraux (Handbook, vol. 3, p. 565) list Ipilinere as a synonym of Amahuaca, but in an earlier section (Handbook, vol. 3, p. 541) give Hpetineri as a probably Arawakan group. 3 Distinguish from Arawak and from Catwkina Catukina. 4 Distinguish from Arawak and Catukina Canamari. Vol. 6] LANGUAGES—MASON 269 Pano—Continued II. Southwest A. Arasaire B. Atsawaca 1. Atsawaca 2. Yamiaca C. Araud (?) III. Southeast A. Pacawardé 1. Chacobo 2. Caripund (Jau-navo) a. Jacarid b. Pamé (Pamand) 3. Capuibo 4. Sinabo B. Zurina (?) BIBLIOGRAPHY Undifferentiated.—Durand, 1921, pp. 85-102; Grubb, K. G., 1927; Hester- mann, 1910, 1914-19; Ia Grasserie, 1890; Mitre, 1909-10, 2:316-17; Rivet and Tastevin, 1927-29, 1932; Schuller, 1911 a. Chama languages.—Carrasco, 1901; Marqués, 1903, 1931; Navarro, 1903; Orton, 1871; Schuller, 1912 b; Tessmann, 1929, 1930, p. 103. Conibo-Shipibo: Alemany, 1906 b; Armentia, 1888, 1898; Farabee, 1922, pp. 80-95; Marcoy, 1875, pp. 52-53; Métraux, 1942 a, pp. 45-53; Reich, 1903, p. 185; Schmidt, W., 1905; Steinen, 1904; Tessmann, 1928. Setibo-Sensi-Panobo-Pano: Cardus, 1886; Castelnau, 1852, pp. 292-302; Martius, 1867, 2: 298-299; Tessmann, 1930, pp. 105, 187-189. Cashibo: Tessmann, 1930, pp. 124-155. Curina-Kulino.—Carvalho, 1929, 1931; Martius, 1867, 2: 242-244. Capanawa Group.—Remo: Carvalho, 1929, 1931, pp. 254-256; Tessmann, 1930, pp. 580-582. Capanahua: Tessmann, 1930, p. 156. Amawaca-Cashinawa.—Abreu, 1914, pp. 524-546, 1938 b; Farabee, 1922, pp. 105-114; Stegelmann, 1903, p. 137; Steinen, 1904; Tessmann, 1930, pp. 158-170. Catukina.— Rivet, 1920 b, p. 87; Rivet and Tastevin, 1924, 1927-29, pp. 489-516. Jurua-Puris Group.—Carvalho, 1931, pp. 234-252; Chandless, 1866, p. 118 (Canamarz); Rivet and Tastevin, 1927-29; Stegelmann, 1903, pp. 135-137. Southwest.—Arasatire: Aza, 1933 b, 1935, 1937; Cipriani, C. A., 1902; Llosa, 1906; Rivet, 1910 b. Atsawaca-Yamiaca: Anonymous, 1902; Barranca, J. S., 1914; Cipriani, C. A., 1902; Créqui-Montfort and Rivet, 1913 e; Farabee, 1922; Nordenskidéld, 1905; Rivet, 1910 b. Arawa: Stiglich, 1908. Southeast.—Pacaward: Armentia, 1888, 1898; Heath, E., 1883; Orbigny, 1839, 1:164; Rivet, 1910 b; Chacobo: Cardus, 1886, p. 315; Nordenskiéld, 1911 b, pp. 230-240; Pauly, 1928, p. 138; Cartpuna: Keller, Fr., 1874, pp. 158-159; Martius, 1867, 2:240—-242; Pauly, 1928, p. 143. MAYORUNA Synonyms: Mujuruna, Maxuruna, Mashoruna, Mashobuna, Maioruna, Morike, Pelado. Mayoruna is generally accepted as a Panoan language (Brinton, 1891 a; Rivet, 1924 a; Krickeberg, 1922: Schmidt, W., 1926; Nimu- 270 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 endaju, map and index; Jijén y Caamafio, 1941-43). However, Tess- mann and Loukotka see nothing Panoan init. 'Tessmann (1930) con- siders it a mixture of Arawak and Tupi; Loukotka gives it inde- pendent status, at first (1935) terming it Mayoruna family, Morike language, and later (1939 a) Morike family; he sees Arawak intrusions init. Steward and Métraux (Handbook, vol. 3, p. 551) consider it a Panoan group. It had best be left unclassified; the linguistic data are poor. Two subtribes are generally recognized, Maruba and Chirabo (Tess- mann, 1930). However, the opinions regarding these and other possible subdivisions are very contradictory. The synonym Pelado is also given as a synonym for Panobo. Maruba (Marubo, Moruba, Marova, Marahua) is considered by Loukotka a synonym for Mayo- runa; 'Tessmann (1930) considers it of uncertain affiliation but prob- ably Panoan; K. G. Grubb (1927) calls Marubo and Pisabo sub- divisions of Mananawa (q. v.); others agree to the Pano affinities. Chirabo (Cirabo, TSirabo) is generally accepted as Panoan. Biblicgraphy.—Castelnau, 1852, pp. 299-300; Martius, 1867, 2:236—239; Tess- mann, 1930, pp. 368-378, 582. ITUCALE, SIMACU, AND URARINA Itucalean was considered as an independent family until at least the time of Chamberlain (1913 a), though he noted the suggestion of Beuchat and Rivet (1909) that it might be Panoan. The latter opinion has been accepted by most recent authorities. Jtucale, Urarina, Shimacu, Chambira, Singacuchusca, and Arucui have been placed in this group by various writers; it is not unlikely that not all are related, some Panoan, others not. Tessman (1930) finds no Panoan resemblances in Simacu, and considers it a mixture of Tucano and Arawak, with about equal lexical resemblance to Tucano, Arawak, and Ge. Loukotka (1935) agrees with him, finds only vestiges of Pano, and makes Simacu an independent family. The Jtucale and Urarina, at least, are probably extinct; the latter is generally classified as Panoan. According to Steward and Métraux (Handbook, vol. 3, p. 557), Urarina was once reported to be related to Mayoruna (q. v.), while the Jtucale were said to have spoken the same (T7upian) language as the Cocamilla. As the lexical resemblances reported by Tessmann (1930) total less than 10 percent of his collected vocabulary of over 300 words they may well all be borrowed words of no value for classificatory purposes. This vocabulary should be sufficient for a determination of the true relationship of Simacu, at least, without further field investigation. In the present status, all these languages had best be considered as unclassified. Vol. 6] LANGUAGES—MASON Qik Bibliography.—Beuchat and Rivet, 1909, p. 621; Chamberlain, 1910 a, pp. 189-190; Rivet, 1912 a; Tessmann, 1930, p. 486. AGUANO It is disputed whether the Aguano (Aguanu, Awano, Santa Crucino) group of languages was independent, Panoan, or Cahuapanan. Whatever the former language, it was given up in favor of Quechua, probably some time after the Spanish Conquest. The group appar- ently consisted of three divisions (see Handbook, vol. 3, pp. 557-558), the Aguano Proper, Cutinana, and Mapurina. Other tribes of un- certain affiliation and subclassification are Chamicura and possibly Sicluna. Chamicura and Maparina are considered Panoan by Rivet (1924 a). Cutinana Beuchat and Rivet (1909) consider as akin to Chebero, that is, belonging to the Cahuapanan family. But the Cutinana are reported as having spoken the same language as the Aguano and Maparina, and whatever affiliation can be proved for any of these languages would seemingly apply to the entire Aguano group. AGUANO A. Aguano Proper 1. Seculusepa a. Chilicawa 2. Melikine te a. Tivilo B. Cutinana C. Maparina Bibliography.—Tessmann, 1930, p. 253. CHAMICURO Chamicuro (Chamicura, Tschamikuro, etc.) is generally considered a Chama (Panoan) language, and Steward and Métraux (Handbook, vol. 3, p. 559) call it “closely related to Shipibo,’”’ with the subtribe Chicluna, but, unless an entirely different language of the same name is referred to, some other opinions are at entire variance. While Rivet (1924 a) agrees that it is Pano, both Tessmann (1930) and Loukotka (1935) find nothing Panoan in it, but consider it a mixture of Arawak and Tucano. Elsewhere (Handbook, vol. 3, p. 558) Steward and Métraux remark that if Chamicuro is Panoan, then all the Aguano group of languages is. Bibliography.—Tessmann, 1930, p. 397. SOUTHERN TROPICAL LOWLAND FAMILIES OF PRESUMED IN- DEPENDENCE UNCLASSIFIED LANGUAGES OF EASTERN PERU In the forests of eastern Pert are, or were, a large number of slightly known tribes, some now extinct, assimilated or deculturized. On AeA SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 most, there are little or no linguistic data. In addition to a few specifically discussed elsewhere, the more important of them are: Alon, Amasifuin, Carapacho, Cascoasoa, Chedua, Cholto, Chunanawa, Chusco, Cognomona, Chupacho, Huayana, Kikidcana (Quiquidcana), Moyo-pampa, Nindaso, Nomona, Pantahua, Payanso, Tepqui, Tingan, Tulumayo, and Zapazo. Most of these spoke Quechua when first discovered. SMALL “FAMILIES” OF BOLIVIA There are at least seven languages of northern and western Bolivia that are generally considered independent: Jtonama, Canichana, Cayuvava, Movima, Moseten, Leco, and Yuracare. Most of them are without known subdivisions, and no evidences for larger affiliations have ever been presented. Most of them have been the subject of special studies. None is known to be extinct, though the speakers are not numerous. Scientific linguistic field research and study upon them is greatly needed, and it will probably eventually be found that they are very variant members of better known families. All have always been known by minor orthographical variants of their stand- ard name. ITONAMAN All authorities consider Jtonama as independent, isolated, or un- classified. Machoto is given as a synonym. Markham (1910) con- sidered it a branch of Mojo. Loukotka (1935) believed he saw evi- dences of Arawak intrusion. The available linguistic data are limited to about 300 words in several collected vocabularies and prayers. The language seems to be both grammatically and lexically different from all its neighbors, rather complex in distinction to other Bolivian languages, which Rivet be- lieves to be rather simple. The pronominal subject is said to be different for men and for women in the second person. “‘Preposi- tions” are prefixed, as are possessive pronouns. The verb employs both prefixes and suffixes. Bibliography.— Adam, 1897-98; Adelung and Vater, 1806-17; Cardus, 1886; Chamberlain, 1910 a; Créqui-Montfort and Rivet, 1916-17; Fonseca, J. S. da, 1880-81; Gillin, 1940; Métraux, 1942 a, p. 83; Nordenskidld, 1915 a; Orbigny, 1839; Pauly, 1928, p. 173; Rivet, 1921 a. CANICHANAN Synonyms: Kanichana, Kanit3ana, Kaniéana, Kanitschana, Canisiana, Kanisiana, Canest, Canechi. Except that Markham (1910) considered it a branch of Mojo, there has never been any suggestion of wider affiliations for Canichana. The extinction of the language seems to be disputed. Vol. 6] LANGUAGES—MASON Sis The available linguistic data are limited to about 240 words in three vocabularies. These seem to show no resemblance to any surrounding language. Canichana shows less resemblance to Uro than does Chapacura. ‘The pronominal possessive is prefixed. There are classificatory nominal prefixes and also apparently nominal suffixes. Pluralization is by suffix. The verb employs both prefixes and suffixes. Bibliography.—Cardus, 1886; Chamberlain, 1910 a, pp. 182-83; Créqui-Mont- fort and Rivet, 1913 f; Gillin, 1940; Heath, E., 1883; Métraux, 1942 a, p. 81; Orbigny, 1839, p. 80; Pauly, 1928, p. 171; Teza, 1868. CAYUVAVAN Synonyms: Kayubaba, etc., Cayuaba, Chacobo. Loukotka (1935) saw evidences of Arawak intrusion in Cayuvava and Markham (1910) considered it a branch of Mojo, but there has been no evidence presented for its wider relationships. Créqui- Montfort and Rivet (1920) found some lexical resemblance to Guai- curt but not enough on which to assume any relationship; some, but even less resemblance was found to Yuyoneiri, Canichana, and Movima. A few hundred words are known in seven collected vocabularies, and a few short texts. The nominal plural is by means of a prefix, much similar to that in Movima. Also like Movima, there is an article or indefinite demonstrative. Pronominal possession is by prefixa- tion; prepositions are also prefixed. Both verbal and nominal pre- fixes and suffixes are employed. Bibliography. Adelung and Vater, 1806-17; Cardus, 1886, pp. 315-316; Chamberlain, 1910 a, pp. 182-184; Créqui-Montfort and Rivet, 1917-20; Heath, E., 1883; Métraux, 1942 a, p. 83; Nordenskiéld, 1911 b, pp. 231-232, 235-239, 241; Orbigny, 1839; Teza, 1868. MOVIMAN This small one-language family is of doubtful validity, but on the basis of the not-too-large available lexical data it has been accorded independent status by all from Chamberlain (1913 a) and Brinton (1891 a) down. No synonyms for the group except minor orthograph- ical variants have been used in the literature. Markham (1910) considers it a branch of the Mojo (Arawak); relationship with the neighboring Canichana might be suspected. There are few if any speakers left. Bibliography.—Adelung and Vater, 1806-17; Cardus, 1886, p. 516; Chamber- lain, 1910 a, pp. 192-193; Créqui-Montfort and Rivet, 1914-19; Métraux, 1942 a, p. 81; Nordenskiéld, 1922, pp. 76-77; Orbigny, 1839; Pauly, 1928, pp. 162-163; Rivet, 1929 b. 274 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 MOSETENAN A small “family” of doubtful validity that will probably be in- corporated in some larger group when more careful studies are made and/or more data available. Up until the present all authorities have granted it independent position and none has even hinted at broader relationships. Métraux suggests personally that a comparison with Yuracare might prove significant. It consists of two main languages, Moseten and Chimane. A number of affiliated groups are mentioned in the literature, but it is not certain how these are related, which ones are synonyms, or whether the linguistic divergences are on a dialectic level or greater. There are probably a few Moseten left, and several thousand Chimane. MosETEN 1. Moseten Amo Aparono Cunana . Chumpa Magdaleno Punnucana Rache . Muchanes i. Tucupi 2. Chimane a. Chimaniza b. Chumano c. Nawazi-Montji Bibliography.—Armentia, 1888, 1903; Bibolotti, 1917; Cardus, 1886, pp. 310-311; Chamberlain, 1910 a, p. 192; Groeteken, 1907; Métraux, 1942 a, pp. 15-17; Pauly, 1928, pp. 104-105; Schuller, 1916. ‘PROS Boop LECAN Synonyms: Leko, Leka, Lapalapa, Chuncho, Ateniano. The small Leco group has been accepted as independent at least since Chamberlain’s day (1910 a). Brinton (1892 a), lacking any lexical material, at first placed them with Tacanan on geographical grounds. The language is called Lapalapa. It seems to have been the language spoken at the mission of Aten, whence the natives were referred to as Atenianos; the latter are sometimes considered a sub- group of Leco; otherwise there are no subdivisions of the family. D’Orbigny (1839) classified the Ateniano as Tacanan. Markham (1910) called them Mositen (Moseten, q. v.). Probably a few hundred still survive. Bibliography.— Brinton, 1892 a; Cardus, 1886, p. 314; Chamberlain, 1910 a, p. 190; Lafone-Quevedo, 1905; Métraux, 1942 a, pp. 27-29; Orbigny, 1839. Vol. 6] LANGUAGES—MASON 275 YURACAREAN Synonyms: Yurukare, Yurujure, Yuruyure, ete. Modern writers prefer Yuracare to the older standard Yurucare. Métraux (personal conversation) suggests that a comparison with Moseten might prove fruitful. Markham (1910) considered them a branch of the Chiquito; Loukotka (1935) sees Pano vestiges. How- ever, no evidence for wider relationships has ever been presented though, like the other independent (?) small languages in this region, its isolated status is doubted and will probably eventually be dis- proved. Though most authorities mention no subdivisions, the Yuracare are said to be divided into two main subgroups, probably of the status of languages, the eastern and the western. Their components are probably of dialectic quality. YURACARE 1. East a. Soloto (Mage) 2. West a. Mansinyo b. Oromo Possibly: Coni, Cuchi, Enete (Brinton, 1891.) Bibliography.—Cardus, 1886, pp. 314-315; Castillo, 1906; Chamberlain, 1910 a; Cueva, 1893; Holter, 1877; Ludewig, 1858; Metraux, 1942 a; Nordenskidld, 1910 b, 1911 b; Orbigny, 1839; Pauly, 1928, p. 177; Richter, 1928. SMALL LANGUAGES OF THE BRAZIL-BOLIVIA BORDER (Havari, Maséca, Capishand, Puruboré, Mashibi, etc.) Huari, Masdca (Massaka), and Capishand (Kapigand) are located very close together on Nimuendajt’s map, Purubord (Burubord) near- by. The first three, Nimuendaju considers separate isolated families. He apparently had some notes on the Capishand, but it is practically unknown in ethnological literature. Huari has long been recognized as independent (Rivet, 1924 a; Pericot y Garcia, 1936; Schmidt, W., 1926; Loukotka, 1935), though it is too small and unimportant to have been known to Brinton (1891 a) and Chamberlain (1913 a). There are few data on it and apparently no component languages of the family. Gillin (1940) doubts its validity, and Métraux herein leaves it unclassified. Levi- Strauss (Handbook, vol. 3, p. 372) apparently considers Huari equiv- alent to Massaka (Masdca), and both linguistically related to Buru- bord (Purubord). ‘The latter, Purubord, Loukotka (1935) considers as an independent family, consisting of one language with Jupt in- trusions. Nimuendajti (map), however, places it in the Tupi family. 7947115019 276 SOUTH AMERICAN INDIANS [B. A. BD. Bull. 143 Koch-Griinberg (1932) published a vocabulary. With the exception of Huari and Masdca all these groups are too small and unimportant to be mentioned by Rivet (1924 a) and W. Schmidt (1926). Rivet considers Masdca Arawakan. Mashubi (Masubi), also in this general region, was given the status of an independent family by Rivet (1924 a) and accepted by Loukotka (1935, 1939 a). Rivet’s opinion is based on a short vocabulary collected by Colonel Fawcett in 1924. This apparently passed into the posses- sion of Nordenskidld and was examined by Rivet; it seems never to have been published. The “family’’ is, therefore, of doubtful validity; Nimuendajt does not list it or place it on his map. An independent family should not be posited on such slim evidence. Kepikiriwat, Sanamaicd, Tupari, Guaycart, Aricapu, Yaputi, Arua- shi, and Canoa are languages occupying tiny areas in a small region. The data on them are few, and they had best be left unclassified pend- ing further information. None is classified by Rivet. Nimuendaji considers Kepikiriwat, Sanamaicd, Tupari, and Guaycart as Tupi; Aruashi and Canoa he leaves unclassified. Bibliography.— Koch-Griinberg (Purubord), 1932. CATUKINAN Synonyms: Katokena, Catequina, etc. The Catukina family is one of rather recent acceptance; Brinton (1891 a) believed that it was a jargon, and Chamberlain (1913 a) did not include it among his families. The name seems to be a descriptive Tupt word, not originally a proper name, and thus has been used as the name for a number of different tribes that are, therefore, liable to confusion and must be distinguished. Rivet (1924 a) lists five tribes of this name, most of them of different linguistic affinities, one Tupi-Guarani, one Arawak, one Pano, and two belonging to the present group. Several of the subtribes, Catawishi and Canamari, also have duplicates in other families so that the possibility of confusion is very great. There are Pano and Arawak Canamari, and an unclas- sified Catawisht, according to Nimuendajti (map). The Catukina family area is a large one and the component tribes very numerous, but there seems to be no general agreement regarding the grouping of these into languages and dialects. The fullest list seems to be on Nimuendajt’s map, with little suggestion as to sub- eroupings. Catukina, Canamari, and Catawishi are the most important and best- known languages. The Katokina of Spix (Martius, 1867, 2:161—163) belongs with the Canamari of the Jurué. Brinton (1891 a) and Chamberlain (1913 a) mistakenly identified this with the Katukinaru of Bach (Church, 1898), a Tupi Guarani language, which probably Vol. 6] LANGUAGES—MASON yeh accounts for Brinton’s considering Catukina a jargon. If these Catu- kina and Canamari are Arawak, they belong to a new and very variant subgroup. The Katukino of Marcoy (1867) on the other hand go with the Catawishi, and may be identical with the latter. CATUKINA . Bert-Dyapd (Bendiapa) . Burue (Buruhe) . Canamari . Catawishi (Hewadie) . Catukina a. Pidé-Dyapa b. Kutid-Dyapaé . Catukino . Parawa . Tawart (Tauaré) a. Cadekili-Dyapa b. Wadyo-Paranint-Dyapé (Kairara) 9. Tucun-Dyapé (Tucano Dyapa, Mangeroma) 10. (Miscellaneous) . Amena-Dyapé . Cana-Dyapé . Hon-Dyapa . Maré-Dyapa . Ururu-Dyapa . Wiri-Dyapé Bibliography.—Brinton, 1898 b; Marcoy, 1867 (1869); Martius, 1867, 2:161— 163; Rivet, 1920 b; Tastevin, 1928 a. or WN Fe on m moaAeo® CHAPACURAN The Chapacuran (Capakura) family has always been known by orthographical variants of the above name though Chamberlain (1912 c) proposed, but never employed, the alternative Pawumwa(n). It now includes the former Itenean stock of Hervaés y Panduro (1800) and the former extinct Ocoronan stock; Chamberlain (1913 a) included the latter in his definitive classification, believing that Créqui-Mont- fort and Rivet’s identification (1913 b) of Ocoronan and Chapacuran needed more proof; this is now generally accepted. According to Rivet, the extinct Ocorono group had been strongly influenced by Arawak and especially by Uro (q. v.) (Créqui-Montfort and Rivet, 1921, p. 104). Some of the tribes had been previously classified as Pano. Loukotka (1935) also finds Arawak intrusions in all the languages except Tora which he considers mixed with Carib. Martius (1867) thought them connected with the Paraguayan Guache, a Guaicurt group, but Brinton (1891 a, p. 303) sees slight resemblance. Wanyam (Huafiam), the modern name for Pawumwa, has always been considered a member, and often the most typical member, of 278 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Chapacura, but it has been considered by some a separate linguistic stock (see below). No definite classification of the Chapacura languages and dialects has ever been proposed on a linguistic basis. As usual, it may be assumed that each group named, if not a synonym, had its own dia- lect, that affiliated groups had closely related languages, and that the linguistic divisions roughly followed the geographical ones. On this basis three main divisions might be proposed, the languages of the Guaporé River Basin, those of the Madeira River Basin, and the extinct Ocorono group. If they are Chapacuran, Wanyam, and Cabishi belong in the Guaporé group. The languages or dialects of the Madeira River group cannot be very variant, as they are said to have moved there in post-Colum- bian days. Arikém (Ariqueme, Ahopovo) (q. v.) is generally considered in this group, but the language seems to be very different from the others; Loukotka (1935) gives it an independent family, and Nimuendaji (map) and Jijén y Caamajfio (1941-43) apparently consider it Tupi. CHAPACURA I. Guaporé Division A. Chapacura Group (Huachi, Guarayo, etc.) 1. Chapacura 2. Kitemoca a. Napeca 3. Moré (Iten) a. Itoreauhip B. Wanyam ! (Pawumwa) 1. Wanyam 2. Cabishi II. Madeira Division 1. Toré (Toraz) 2. Jar% 3. Urupad 4. Pacas Novas III. Ocorono Division A. Ocorono (Rokorona) 1. Sansimoniano 2 (?) B. San Ignacio 1. Borja C. Herisobocona 1 See following separate section on these. 2 Jijon y Caamafio places Sansimoniano under Chiquitoan. BIBLIOGRAPHY Undifferentiated. Cardus, 1886, p. 320; Chamberlain, 1910 a, p. 184; Créqui- Montfort and Rivet, 1913 b; Métraux, 1940, 1942 a, pp. 86-95; Nordenskidld, 1915 b; Orbigny, 1839, p. 164; Pauly, 1928, pp. 168-169; Snethlage, E. H., 1931; Teza, 1868. Vol. 6] LANGUAGES—MASON 279 Guaporé division.—Cardus, 1886, p. 320; Chamberlain, 1910 a, pp. 188-189; Créqui-Montfort and Rivet, 1913 b; Orbigny, 1839; Pauly, 1928, pp. 168-169. Madeira division. Nimuendaji, 1925, pp. 187-140, 148-159; Nimuendaji and Valle Bentes, 1923, p. 217. Ocorono division.—Chamberlain, 1910 a, pp. 193-194; Créqui-Montfort and Rivet, 1913 b, pp. 169-171; Snethlage, E. H., 1936. WANYAM (HUANAM) AND CABISH{ (KABICI) Wanyam seems to be the rediscovered Pawumwa, and the name has displaced the latter. To date they have always been considered one of the Guaporé Chapacuran groups, and are accepted by Métraux (Handbook, vol. 3, p. 398) and by Nimuendajt (map) as such. The latter considers Abitana Wanyam a subdivision, the former a synonym. However, Lévi-Strauss who pursued studies in their region, believed in a Huanyam linguistic stock, consisting of Mataua Cujuna (Cuijana), Urunamakan, Cabishi, Cumand, Snethlage’s Abitana-Huanyam, and Haseman’s Pawumwa. Only the last four are at all known to science, and the last two probably constitute one group. All of these are generally considered as Chapacura, with the exception of Cabishi, which name seems to be applied to several groups of different linguistic affiliations. (See Nambicuara.) WaANYAM — Cabishi1 Cujuna 3. Cumand a. Cutiand 4. Matama (Matawa) 5. Urunamacan 6. Pawumwa a. Abitana Wanyam \Rivet (1924 a) considers Cabishi as Nambicuara; Schmidt (1926) (Paressi-Kabict), Arawak; Nimuendaja, Nambicuara or Chapacura. There may be several groups of Cabishi. io Bibliography.— Chamberlain, 1912 c; Créqui-Montfort and Rivet, 1913 b, pp. 141-68; Haseman, 1912; Snethlage, E. H., 1931. MASCOIAN A confusion of identities and of groups of the same names but of different linguistic affinities characterizes this family, which consists of one language. Ehrenreich first used the name Machicwi or Muscov; Koch-Griinberg (1902 a), Mascoi. Boggiani (1900) called them Enimagd, confusing them with the true Enimagd (Cochabot, q. v.), which name was applied to the Mascoi family by Chamberlain (1913 a) and by Brinton (1898 a). To avoid further confusion the Enimagd are now termed Macd (q. v.). One of the constituent dialects is the Lengua. These are the “new’’ Lengua, the Lengua of W. B. Grubb (1911); the “old” Lengua are a Macd (Enimagd) group. 280 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 D’Orbigny (1839) insisted that the Mascot were Guaicurt, like the Toba, but Lafone-Quevedo (1896-97) compared the vocabulary with Abipon with a negative conclusion. Brinton (1898 a) found a few similarities to Chon. There is practically no difference of opinion regarding the compo- nents of Mascoi, six scarcely differentiated dialects, except that W. Schmidt (1926) includes Suhin (Sujén, Suxén), a neighboring language, generally classed as Mataco and connected with Ashluslay, and Toosle (Towothli), now placed with Macdé (Enimagd). Nimuendaji (map) differentiates Angaité from Enenslet, Machicuy from Mascoi; by others they are considered synonyms. Métraux (Handbook, vol. 1, p. 227) mentions the following extinct groups that may have been either Mascoi, or Arawakan Guand: Guatata, Nohaague, Empiri, Yapert or Apiri, and Napert. Mascor 1. Angaité a. Enenslet 2. Caskihé (Guané') 3. Lengua ? (Gecoinlahaac) 4. Mascot a. Machicui (Tujetge) 5. Sanapand 6. Sapuki 1 Distinguish Mascoi and Arawak Guand. 2 Distinguish Mascoi and Maca Lengua. Bibliography.—Baldus, 1931; Boggiani, 1900; Brinton, 1898 a; Cardus, 1886, p. 271; Cominges, 1892, p. 245; Coryn, 1922; Ehrenreich, 1891; Grubb, W. B., 1911, pp. 318-321; Hunt, 1917; Kersten, 1905; Koch-Griinberg, 1902 a; Lafone- Quevedo, 1896-97; Loukotka, 1930; Orbigny, 1839. ZAMUCOAN The Zamuco (Samucan) ‘‘stock’’ occupies a relatively small area in the northern Chaco. The people and their languages are of slight cultural importance and little known, few in number and disappearing, though the names of a large number of groups are known, many of them extinct. No other name has ever been applied to the group. The differentiation between the component languages and dialects seems to be relatively slight. Brinton (1898 a) pointed out many lexical resemblances with Arawak but refused to posit any genetic relationship; the common elements may be largely due to borrowing. Métraux personally suggests that a comparison of Zamuco with Bororo-Otuque might prove significant, and, as in the case of most small groups at present considered independent, an intensive study will probably tie it to some large family or phylum. The grouping and subclassification of the languages and dialects Vol. 6] LANGUAGES—MASON 281 are uncertain since the various authorities, as usual, do not agree. That adopted below follows Métraux (Handbook, vol. 1, pp. 241-245) so far as possible. Even the major divisions, apparently, do not differ greatly, and some of the names may be synonyms, or merely bands without linguistic differentiation. The language is reported to be mellifluous. ZAMUCO I. North: Zamuco A. Zamuco 1. Zamuco (Samuca) 2. Satienyo (Zatieno, Ibiraya) B. Morotoco (Coroino) 1. Cucarate (Kukutade) 2. Orebate (Ororebate) a. Careraé 3. Panono (Panana) 4. Tomoeno C. Guaranyoca (Guarafioca) 1. Tstracua (Empelota) a. Moro (remnants of Morotoco and Guarafioca) D. Ugaranyo (Ugarono) 1. Ugaranyo E. Tapit (Tapio) (?) } F. Poturero (Ninaguilé, Ninaquiguila) II. South: Chamacoco A. Chamacoco (Tumanahd, Timinthd) 1. Ebidoso 2. Horio (Ishira) 3. Tumereha B. Imono C. Tunacho (Tunaca) D. Caipotorade Also (Nimuendaji map): Aguiteguedichagé, Laipisi. 1The Tapii may have been Otwkean. BIBLIOGRAPHY Zamuco.—Baldus, 1932; Brinton, 1898 a; Huonder, 1902; Kersten, 1905, pp. 64-66; Loukotka, 1931 a; Nordenskiéld, 1912, p. 324; Oefner, 1942; Orbigny, 1839, p. 164; Steinen, 1912. Chamacoco.—Baldus, 1927; Belaieff, 1937; Boggiani, 1894, 1929; Cardus, 1886, p. 327; Loukotka, 1929-30, 1930, pp. 560-572; Steinen, 1895. GUATOAN No question has ever been raised regarding the independence of the Guaté; no suggestion has been made of relationship with any other group. No linguistic subdivisions are known, but the Guazarapo (Guachi) are always associated with them. (See Métraux, Handbook, 282 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 vol. 1, pp. 214, 225, 409.) The Guwaté verge on extinction. The principal source is Max Schmidt (1914 b). Bibliography.—Castelnau, 1852, pp. 283-284; Chamberlain, 1913 c; Koslowsky, 1895; Martius, 1867, 2:209-210; Monoyer, 1905; Schmidt, M., 1902, 1905, 1912, 1914 b. BOROROAN AND OTUKE The Bororo and the Otuque (Otuké, Otuqui) were long considered separate and independent, and the former was once thought a Z7upi tongue (Brinton, 1891 a). Cogent arguments for their linkage were presented (Créqui-Montfort and Rivet, 1912, 1913 a) and all recent authorities have linked them, either under the name Bororo or Otuque. Métraux, however, is not convinced of the connection, and it needs further study and corroboration. The Otuque are probably extinct. Two, possibly four, extinct languages, formerly given independent status, Covareca and Curuminaca, and possibly Corabeca, Curave, and Curucaneca (q. v.), compose, with Otuque, the Otuquean group. Brinton (1891 a) noted a distant resemblance to Tacana. Bororo- Otuke will probably eventually be incorporated in some larger phylum. Métraux suggests that a comparison with Zamuco-Chamacoco might prove significant. The Bororo are sometimes known as Coroado, a name applied to a number of different groups, especially of Ge and Macro-Ge, from which they must be distinguished. Borotuke ! I. Bororo (Coroado) A. Bororo 1. Eastern a. Orarimugudoge 2. Western a. Cabasal b. Campanya . Acioné . Aravira . Biriuné . Coroa (?) . Coxipo (?) B. Umotina (Barbado) Il. Otuké A. Otuké . Covareca 2 . Curuminaca 2 B C D. Coraveca (?) 2 E SID Oe Ww 1. Curavé (?) 2 . Curucaneca (?) 2 F. Tapii (?) 3 ! Hybrid name suggested here for group of Bororo-Otuke. 2 See following section. * The Tapii may have been Olukean or Zamucoan. Vol. 6] LANGUAGES—MASON 283 BIBLIOGRAPHY Bororo.—Anonymous, 1908, 1919 a; Caldas, 1899, 1908; Castelnau, 1852, pp. 285-286; Chamberlain, 1912 b; Colbacchini, n. d., 1925; Colbacchini and Albisetti, 1942; Fri¢é and Radin, 1906; Magalhdes, 1919; Martius, 1867, 2: 14-15; Steinen, 1894, pp. 545-547; Tonelli, 1927, 1928; Trombetti, 1925. Umotina.—Schmidt, M., 1929 a, 1941. Otuke.—Chamberlain, 1910 a; Créqui-Montfort and Rivet, 1912, 1913 a; Orbigny, 1839. CORAVECA AND COVARECA, CURUCANECA AND CURUMINACA The similar names in these two groups are not metathesized syno- nyms. In close geographic propinquity, some close relationship would seem indicated, but the available vocabularies on these extinct languages indicate that the four are separate and very different. In the older classifications all four were considered as independent families. Some authorities, such as Créqui-Montfort and Rivet (1912, 1913 a), place all under Bororo or Otuké (q. v.). The inclusion of Covareca and Curuminaca is generally accepted, but Coraveca (Curave) and Curucaneca (Curucane), showing less resemblance to Otuké, are included with hesitation or reservations, left unclassified, or awarded independent status. By some, Curave and Coraveca are considered as separate, not as synonymous. The arguments for the inclusion of Covareca and Curuminaca in the Otuké family have been published (Créqui-Montfort and Rivet, 1913 a), but not those for Curave and Corabeca, Curucaneca and Tapti. The languages are all in the Chiquito region, but no Chiquito connections have been suggested, although all four groups are said to have spoken Chiquito a century ago (Métraux, Handbook, vol. 3, p. 381). The names of many groups in this region have a similar ring: Saraveca, Paiconeca, Paunaca (Arawak), Kitemoca (Chapacura), Waranyoca (Zamuco), etc. The four languages in question may be closely related, the recorded vocabu- laries at fault, but as they are extinct their degree of relationship will probably never be known. (See Otuké-Bororo for table.) Bibliography.— Chamberlain, 1910 a; Créqui-Montfort and Rivet, 1912, 1913 a. NAMBICUARAN The relatively recently discovered Nambicuara were unknown to and not classified by the earlier authorities; they and their languages have recently been studied carefully by Claude Lévi-Strauss. No detailed comparisons with other linguistic families have been made, but Lévi-Strauss states that the logical structure is much like Chibcha, with also some similarities in vocabulary, but nothing conclusive. Loukotka (1935) sees vestiges of several eastern Brazilian languages in several of the component Nambicuara languages but nothing that 284 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 characterizes the family as a whole. The outstanding characteristic of the group according to Lévi-Strauss (personal correspondence) is an extensive use of classificatory suffixes dividing things and beings into about 10 categories. According to Lévi-Strauss (personal correspondence), the classifi- cation of the Nambicuara languages by Roquette-Pinto (1917) is not good. There are two main groups, the Nambicuara Proper and the pseudo-Nambicuara. The latter, northern group, the Sabane, never before mentioned in literature, is considerably different from the other languages in structure and has a vocabulary rich in Arawakan ele- ments probably borrowed from the Paressi. It may be found to be- long to some other, or to be a new linguistic family. The Nambicuara Proper are divided into two languages and each of these into two dialects, as shown in the following table. The end- ings of words is the only difference between the Cocozi and Anunzé dialects. The Tamaindé vocabulary of Max Schmidt (1929 a, p. 102) belongs to the western dialect, characterized by a new form for the verbal suffix. In addition to orthographical synonyms, such as Nambikwara, they are known as Mambyuara, Mahibarez, and Uaikoakore; the dialect names have also many orthographical variants. A few groups not mentioned by Lévi-Strauss are included in the Nambicuara family by some of the other authorities, sometimes by several of them, and contradicted by none. These are Salumd, Néné, Congoré, and Navaité; some of them may be synonyms. Métraux suggests the possible inclusion of Guayuakuré and Tapanhuana, apparently little- known groups. NAMBICUARA I. Nambicuara Proper A. Northeastern 1. Eastern a. Cocozu 2. Northeastern b. Anunzé B. Southwestern 1. Central and Southern a. Uaintazu b. Kabishi ce. Tagnani d. Tauité e. Taruté f. Tashuité 2. Western a. Tamaindé II. Pseudo-Nambicuara A. Northern 1. Sabane Vol. 6] LANGUAGES—MASON 285 Bibliography.—Rondén, 1910, pp. 52-53; Roquette-Pinto, 1912, 1917, 1935; Schmidt, M., 1929 a; Schuller, 1921; Souza, A., 1920. CABISHI Cabisht is one of those names applied to a number of different tribes. Authorities such as Nimuendajtii and Rivet seem to agree that the true Cabishi are a branch of the Nambicuara, and Nimuendajt equates the term with Waintazi (Uaintacgié). Another Cabishi are a branch of the Wanyam. (See Chapacura.) The Paressi-Cabishi are Arawak (q. v.). The name Cabishiana (Kabixiana, Kapisana, q. v.) is prob- ably related. MURAN A moderately small group, Mura is considered as forming an inde- pendent linguistic family by all modern authorities except Brinton (1891 a). He states that the majority of its words are from Tupi roots; as his opinions—and often his guesses—are generally good, and as no other authority has noted this resemblance, we may suspect that the vocabulary that he used was unsuitable in this connection. No other suggestions regarding larger relationships have been made, though Loukotka (1935) finds vestiges or intrusions of Camacan and Caingang—an unlikely possibility. The Mura family may consist of two main divisions, Mura Proper and Matanawt or Matanauy (q. v.). But the latter is so divergent that Nimuendajt (map) considers it isolated. Mura Proper is divided into three ‘dialects,’ those of Bohurd, Pirahd, and Yahahi. (See Nimuendaji, Handbook, vol. 3, pp. 255-258.) Mura must not be confused with the Bolivian (Chapacuran ?) Mure. Mura A. Bohurad B. Pirahé C. Yahahi Bibliography.— Chamberlain, 1910 a; Martius, 1867, 2: 20-21; Nimuendajd, 1925, 1932 a; Nimuendaji and Valle Bentes, 1923. MATANAW{i Rivet (1924 a), who uses the spelling Matanauy, Loukotka (1935, 1939 a), and Jijén y Caamafio (1941-43) place Matanawi in the Mura family; Loukotka sees Caingang intrusions. But Nimuendaji (map and linguistic index), whose vocabulary (1925, pp. 161-171) seems to be the basic source, prefers to list it among his “isolated languages.” Bibliography. Nimuendaji, 1925, pp. 143, 166-171; Nimuendaji and Valle Bentes, 1923, p. 222. 286 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 TRUMAIAN Trumai has been accepted as independent ever since its discovery by Von den Steinen. No suggestion as to larger affiliations has ever been made. The linguistic data, however, are few, and when it is better known an attachment to some larger family is likely. A single language composes the “family.” Bibliography.—Chamberlain, 1910 a; Steinen, 1894, pp. 540-542. CARAJAN Synonyms: Carayé, Karayd, Karadza. Universally recognized as an independent family, at least since the days of Brinton (1891 a) and Chamberlain (1913 a), no other variant synonym has ever been proposed for the stock. Lipkind (vol. 3, p. 179), who has recently recorded and studied it, says it is unrelated to at least the four great families of Arawak, Carib, Tupt, and Ge. The speech of men and women differ. Brinton (1891 a, p. 261) found a little lexical resemblance to ‘‘Tapuya” (Ge), which may probably be ascribed to borrowing. While no classificatory subdivisions of the family have been pro- posed, there is general agreement that Carajdé (see same synonyms as for family) or Karayaki (Carajahi), Yavahé (Yavahai, Javahé, Javahai, Zavazé, Shavayé, Jawagé), and Shambiod (Schambiod, Sambiod, Chim- biod, Ximbiod) are the principal component languages. Linguistic differences in the three are slight and on a dialectic level (Lipkind, Handbook, vol. 3, p. 179). W. Schmidt (1926) also includes Asurini (probably Tupian). Brinton (1891 a) distinguishes Carajahi from Carajd. Nimuendaji (map) distinguishes from the Carajdé of the Araguaya an earlier unclassified extinct group of the same name in Minas Gerais, possibly ancestral. CarasA 1. Carajé (Karayd) a. Carayahi 2. Yavahé (Shavaye) 3. Shambiod Bibliography.—Castelnau, 1852, pp. 268-269; Chamberlain, 1913 c; Coudreau, H., 1897 b, pp. 259-270; Ehrenreich, 1888, 1894-95, pp. 20-37; Krause, 1911, pp. 458-469; Kunike, 1916, 1919; Martius, 1867, 2: 264-266. CARIRIAN Synonyms: Kariri, Kairirt, Cairirt, Kayriri, Ktiriri, Cayrirt. Carirt has always been recognized as an independent family. The suggestion has been made that it belongs with the great Carib stock (Gillin, 1940), but no evidence in support has been offered. As a Vol. 6] LANGUAGES—-MASON 287 grammar (Mamiani, 1877) and other studies on the language (Adam, 1897) have been published, the determination should not be difficult. The family is on the verge of extinction; the few remnants of Camurt probably do not use their language in its former purity. The linguistic subdivisions of the family seem to be: A. Cariri 1. Kipea 2. Camurt 3. Dzubucua 4. Pedra Branca B. Sapuya BIBLIOGRAPHY Cariri—Adam, 1878, 1897 a; Bernard de Nantes, 1896; Chamberlain, 1913 ¢; Gillin, 1940; de Goeje, 1932, 1934; Mamiani, 1877; Martius, 1867, 2: 215-217; Mitre, 1909-10; Pompeu Sobrinho, 1928, 1934. Sabuya.—Adelung and Vater, 1806-17; Martius, 1867, 2: 218-219. MACRO-GE Macro-Ge is a term here proposed for the first time as an equivalent for the Ge or ‘‘Tapuya stock” or “family” as constituted until recent years, synonymous with Rivet’s (1924 a) Ze and W. Schmidt’s (1926) Ges-Tapuya. As herein conceived, it consists of Ge and some eight other ‘‘families,”’ ‘‘stocks,’’ or languages, formerly considered as mem- bers of the great Ge family, which some recent authorities suggest as independent. Future research will have to give the final decision as to their independence; they are herein considered as far from proved. There is considerable lexical resemblance, which may or may not be a result of borrowing. Had these languages always been considered independent, articles would certainly have been written to prove their relationship with Ge, as cogent, and as convincing or uncon- vincing, as many others linking other groups, formerly considered independent, with larger entities. It may eventually be decided that some of these languages are independent, but it is more likely that other small languages will be added to make Macro-Ge an even more inclusive phylum. Rivet (1924 a, p. 697) summarized the Ge situation well in his remarks: This family, of all the South American families, is the one most artificially constituted. It is the caput mortuum of South American linguistics. Its careful and complete revision, on truly scientific grounds, is imperative. Rivet, who made researches on most of the South American families, left Ge severely alone. Loukotka took up the problem and, with his usual “radical conservatism,’’ split the old Ge into nine independent families: Zé (Ge); Kaingdn (Caingang); Kamakdén (Camacan); Masa- kali (Mashacalt); Coroado; Patasé (Patasho); Botocudo; Oparé (Opaye) ; 288 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Taté (Fornio, Fulnio). He sees Ge “intrusions’”’ in all but the first and last. In his earlier studies Loukotka retained Caingang under Ge Proper, but finally (1935, 1939 a) decided to separate this also. It was his intention to publish a monograph on each of these new fam- ilies, with large comparative vocabularies, written in the same pattern, but to date he has published on only Mashacali (1931-32 a), Camacdén (1931-32 b), Coroado (1937), and Patasho (1939 c). Nimuendajt and Lowie herein accept all these components as inde- pendent from Ge Proper, either as families or as isolated languages, with the exception of Caingang, which they still consider as a member of Ge Proper. Herein Caingang (q. v.) is considered as separate from Ge Proper as the other components of Macro-Ge. As conceived herein, Macro-Ge consists of nine groups or families as follows: (1) Ge; (2) Caingang; (3) Camacdn; (4) Mashacali; (5) Puri; (6) Patashé; (7) Malali; (8) Coropé; and (9) Botocudo. It will be noted that, in addition to orthographic variations, this list, while equal in number to Loukotka’s, differs slightly. The name Puri is preferred to Coroado. Malali and Coropé have been added because their classification in some one of the other groups is so uncertain. As regards Loukotka’s last two, Opayé and§Jaté (Fornio), the resemblance to any Ge language seems to be so slight that they are better considered for the present as independent and non-Ge. Of the nine above, Nimuendajt considers Ge, Camacdén, Mashacali, Puri, and Botocudo as “‘stocks’”’; Patashé and Malali as “‘isolated’’; Caingang he places under Ge; and Coropé with Puri. Apparently all the Macro-Ge groups were termed “‘Tapuya’”’ (en- emies) by the Tupi, and this name was formerly used as a synonym for Ge. They inhabit the infertile regions of eastern Brazil, off the rivers. From the coast they were displaced by the Tupi. Somato- logical and cultural evidence suggests that they were among the most autochthonous of all South American natives; as such it is unlikely that any connections will be found with other great families or phyla. As languages of people of low culture they have been neglected more than their scientific importance warrants, and the morphologies are not well known. They are said to be phonetically difficult, and harsh. Except for the Ge Proper and the Caingang, most of them were close to the Brazilian coast; most of them are now extinct, and the others practically so. GE The Ge group, as herein recognized, consists only of the Ge Proper, one section of the Ge family as previously constituted, which latter is herein termed Macro-Ge (q. v.). That is, a number of other groups, Mashacali, Camacdn, etc., formerly considered as constituting the Ge — Vol. 6] LANGUAGES—MASON 289 family, have been removed from it and, together with Ge proper, con- sidered as constituent parts of the phylum Macro-Ge. The Ge occupy a solid large area in eastern Brazil. They were displaced in some regions by the Tupi, and the language is losing ground to modern Brazilian Portuguese. They were formerly known as “Tapuya,”’ the Tupi word for “enemy.” Cran or Gueren, meaning ‘ancients’ or “natives,” is another synonym; their self-name is Nac-nanuc. Many of the names of groups end in zhe (Ge, Ze). The classification of the Ge languages here presented is taken pri- marily from Lowie and Métraux (Handbook, vols. 1 and 3). These, however, are probably based principally on present political associa- tion, cultural resemblance, and geographical propinquity rather than on linguistic grounds. A thorough independent study has not been possible, but a brief comparison of vocabularies of a few of the better- known Ge languages suggests that a truer picture of linguistic relation- ships may cut across the proposed divisions. This short study indicates: A. A rather closely connected group consisting of: Apinayé, Northern and South- ern Cayapé6, Carahé, Gradahé, aiid Mecubengocré. . Ushikring and Suyé are slightly more variant. Ramcocamecran and Aponegicran probably should also be placed in this group. (All the above languages are in the Northwestern division.) . More distantly affiliated: Shavanté, Sherenté, Tazé and Crenyé. . Possibly affiliated: Acroa and Jetcé. Of uncertain affiliation: Mehin, Purecamecrén, Piokobzé, Capiecrén, Crao, Shicriabé. (Crenyé shows some resemblance to Crao and to Mecubengocré; Capiecraén to Northern Cayap6.) aap a GE I. Northwest A. Timbira 1. West Timbira a. Apinayé! (Apinages) 2. East Timbira (Hott) a. Northern a. Gurupy b. Creyé (Crenye) of Bacabal c. Nucoecamecran of Bacabal b. Southern a. Canela:? Apanyecra (Aponegicran), Kencateye (Kencatazé), Ramcocamecra (Capiecran) Carateye Craho (Krao), Macamecran . Crepumcateye Crenyé (Crange) (of Cajuapara), (Tazé) 1 Traditionally the Apinayé are offshoots of the Krikati. Pericot suggests that they may be the same as the Aenaguig, but Lowie considers the latter independent. 2 Canela is the Brazilian name for the savage Timbira. Some other groups here listed are considered as Canela. Rivet (1924 a) gives as additional Canela bands: Temembt, Bucobt, or Mannoz6, Poncatgé. Some of these may be synonyms of others here listed. SSS) fe 290 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 GE—Continued I. Northwest—Continued A. Timbira—Continued 2. East Timbira (Hoti)—Continued b. Southern—Continued f. Cricati (Krikateye, Caracaty, Makraya) . Gavides . Nyurukwayé (Norocoage) Porekamekra (Purecamecran) Pucobye (Piokobzé) . Chacamekra (Sacamecran, Matteiro) . (Augutge) m. (Paicoge) n. (Mehin) ~ FO. 9. 58 B. Cayapé Ie Northern Cayapé (Corod, Carajd) . Carahé (Karahé) . Gorotire . Gradahé (Cradahé, Grada%) . Ira-Amaire Pau d’ Arco Purucarod (Purukaru) a. Curupite . Mecubengokre . Ushicring (Chicri, Byoré) Criiatire j. Cayamo 2. Southern Cayapé C. Suyd 1. Suydé (Tsuvd) monoee mae sel Set II. Central A. Akwé (Acua, Akwa) 1. Shacriaba (Chikriaba) 2. Shavante ? (Chavante, Crisca, Pusciti, Tapacud) 3. Sherenté (Cherente) B. Acrod 1. Acrod a. Northern b. Southern 2. Guegué (Gogué) III. Jeicé (Jaicé, Geicé) A. Jeicé 3 The name Shavanté (q. v.) is applied to a number of different groups. The Shavanté included here are those of the Rio dos Mortes. They must be distinguished from the Oti Chavanté and the Opayé Chavanté (q. v.), of other linguistic affinities. The Tupi Canoeiro, the Timbira Nyurukwaye and the Orajowmapré are also termed Chavanté. Other names applied to the Akwe-Chavanté are Criza (Curiza), Puziti,and Tapa- cud. Shavante and Sherente are essentially the same. BIBLIOGRAPHY Apinayé.—Castelnau, 1852, p. 270; Hurley, 1932; Leal, 1895, pp. 225-228; Martius, 1867, 2:147—149; Oliveira, C. E. de, 1930, pp. 99-104; Snethlage, E. H., 1931. Eastern Timbira,—Etienne, 1910 (Capiecran); Froes de Abreu, 1931 (Rameco- Vol. 6] LANGUAGES—MASON. 291 camecran); Kissenberth, 1911 (Kencatazé); Martius, 1867, 1: 525 (Purucamecran), 2:149-151 (Aponezicran); Nimuendaji, 1914 d (Crengez, Tazé), 1915 (Crenzé, Mehin, Ramcocamecran, Tazé), 1946 (Timbira) ; Pohl, 1832-37 (Timbira) ; Pompeu Sobrinho, 1930, 1931 (Canela, Merrime, Mehin); Sampaio, T., 1912 (Mecume- cran, Crao); Snethlage, E. H., 1931 (Ramcocamecran, Crao, Piocobzé). Cayapé6.—Castelnau, 1852, pp. 273-274 (Carahé); Coudreau, H., 1897 b, pp. 271-290; Ehrenreich, 1888, 1894-95 (Cayapé, Usicring, Gradahé); Kissenberth, 1911, pp. 53-54 (Mecubengocré); Krause, 1911, pp. 461-469; Kupfer, 1870, pp. 254-255; Maria, 1914; Martius, 1867, 2:134-135, 151-152 (Carahé); Nimuendaji, 1931-32 pp. 552-567; Pohl, 1832-37, 2:447; Saint Hilaire, 1830-51, 2:108—109; Sala, R. P. A. M., 1920; Socrates, 1892, pp. 95-96. Suy4.—Steinen, 1886, pp. 357-360. Central Group.—Castelnau, 1852, pp. 262-268 (Shavanté, Sherenté); Ehren- reich, 1894-95 (Acud, Shavanté, Sherenté); Eschwege, 1830, pp. 95-96 (Shicriabd) ; Martius, 1867, 2:139-146 (Sherenté, Chicriabé, Acroa mirim); Nimuendajt, 1929 a (Serenté); Oliveira, J. F. de, 1913 a, 1913 b (Sherenté); Pohl, 1832-37, 2:33 (Chavanté); Saint Hilaire, 1830-51, 2:289-290 (Shicriabé); Socrates, 1892, pp. 87-96 (Sherenté); Vianna, 1928 (Acuen). Geico.— Martius, 1867, 2:1438. ‘“‘Tapuya.’’—Barbosa da Faria, 1925; Ehrenreich, 1894-95; Koenigswald, 1908 a; Schuller, 1913 c. CAINGANG Caingang, also sometimes called Guayand, Coroado, Bugre, Shocleng, Tupi, Botocudo, etc., is still considered a member of Ge Proper by Nimuendaji, Lowie, and Métraux. Loukotka also placed it under Ge until his 1935 classification when he gave it independent status. A perusal of the lexical data indicates that it is at least as different from Ge Proper as most of the other components of Macro-Ge. Henry (1935) is also of the opinion that Caingang should be separated from Ge. Caingang seems to show even less lexical resemblances to Ge than do Mashacali, Camacdn, and Puri (Coroado). In spite of the large vocabularies available very few words show any resemblance to words of similar meaning in any of the other families. Nevertheless, as in the comparisons of all of these languages, there are a fair number of possible connections, and a small number of certain, close, or even identical resemblances, generally in common basic words that would not be likely to be borrowed. In spite of the apparent great lexical differences it is probable that Caingang is distantly related to all these languages. Though the differences are not great, either qualitatively or quantitatively, Caingang seems to show slightly greater resemblance to Puri. Loukotka considers the family as showing Ge intrusions, and most of the constituent languages to show Arawak or Camacdn vestiges. Bugre is an opprobrious term; Botocudo and Coroado are descrip- tive, and the Caingang groups to whom they are applied must be distinguished from the other Macro-Ge groups of these names. Few 79471150 —20 292 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Caingang languages seem to be extinct. They occupy an interior region coterminous with the Ge Proper, not, like the other Macro-Ge divisions, a coastal region. The subdivisions are uncertain and disputed. Loukotka divides them into 10 languages: four dialects of Caingdn, Kadurukré, Kamé, Wayana, Ivitorokai, Ingain, and Aweicoma. The classification of Métraux, herein adopted, is probably based on political and regional groups rather than on linguistic variations, but, nevertheless, gives the impression of greater reliability. According to him, Cayurucré (Kadurukré) is a moiety; the Wayand (Guayand) were the ancestors of the present Caingang; Ivitorocai and Ingain are synonyms of Taven and Tain, and Aweicoma a synonym of the more usual term Shocleng. There were apparently two groups of Wayand, one speaking Tupi- Guarani, the other ancestors of the Caingang. The best linguistic data are found in Father Mansueto Barcatta Valfloriana, 1918 a, 1920. CaINGANG I. Caingang A. Sao Paulo (Coroado) ! 1. Nyacfatettez B. Parand C. Rio Grande do Sul II. Shocleng (Socré, Chocré, Xocren, Bugre, Botocudo,? Aweicoma, Cauuba, Caahans, Caagua, Caaigua *) III. Taven A. Tain B. Ingain (Wayana, Guayané *) 1. Pattie (Basa) 2. Chowa 3. Chowaca C. Ivitorocai D. Gualacho (Coronado *) 1. Gualachi 2. Chiki 3. Cabelludo IV. Dorin (Bands: Jahuateie, Venharo) (Moieties: Cayurucré, Votoro, Camé) Possibly Caingang: Aricapu, Yabuti.® 1 Distinguish from other Macro-Ge Coroado (Puri). 2 Distinguish from other Macro-Ge Botocudo. 3 Distinguish from Chiriguano (Tupi-Guarant) Caaigua. 4 Distinguish from Guarand Guayani. (See Métraux, Handbook, vol. 1, p. 446.) 5 See Lévi-Strauss (Handbook, vol. 3, p. 372). BIBLIOGRAPHY Caingang.—Adam, 1902; Ambrosetti, 1894, 1895 a, pp. 354-87; Anonymous, 1852; Baldus, 1935, pp. 194-201; Barcatta de Valfloriana, 1918 a, 1918 b, 1920; Vol. 6] LANGUAGES—MASON 293 Borba, 1882, 1908, pp. 95-114; Chagas-Lima, 1842; Freitas, 1910; Henry, 1935, 1948; Hensel, 1869; Ihering, 1895, pp. 117-118, 1904; Martius, 1867, 2:212-214; Moreira-Pinto, 1894, p. 580; Paula-Souza, 1918; Pindorama, 1937; Saint Hilaire, 1830-51, 1: 456-457; Taunay, 1888, 1913; Teschauer, 1914, pp. 29-30, 1927, pp. 49-51; Vocabulario Comporado, 1892; Vogt, 1904. Shocleng Group.—Gensch, 1908; Ihering, 1907, p. 232; Paula, 1924, pp. 131-134. Guayana.— Borba, 1908, pp. 138-139; Ihering, 1904; Lista, 1883, pp. 112-113; Martinez, B. T., 1904. Aricapai, Yabuti.—Snethlage, E. H., 1931. CAMACAN, MASHACAL{, AND PUR{ (COROADO) On these three groups, formerly considered as members of the great Ge family, Loukotka has published linguistic sketches (1931-32 a, 1931-32 b, 1937), including vocabularies compiled from all known sources, and lexical comparisons with Ge and other neighboring lan- guages. His deductions are that all are independent from each other and from Ge, but with Ge “intrusions.”’ All three with their several language divisions are supposed to be extinct, though a few members may still live with other groups in some of the missions. Presumably, therefore, no new linguistic data will be found, and their relationships must be determined on the basis of the material at hand, compiled by Loukotka. Unfortunately, no grammatical studies are known, and the basic data consist of vocabularies of varying size, mainly of rather ancient date and all uncritically recorded. The few phrases afford very little morphological information. The phonetics of the three groups are similar in general type, and the few morphological deductions made by Loukotka show no great difference; on these grounds the three might be closely related. Lexi- cally, however, they are very different. The compiled vocabularies are large enough to afford sufficient data for tentatively conclusive results, nearly 900 words for the Coroado group, about 350 for Mash- acalt and Camacdn. Using very uncritical methods of comparison and noting every case of stems showing the slightest resemblance, many of which will doubtless be thrown out when a critical linguistic study is made, Loukotka finds the following proportion of possible stems connected with Ge and Caingang combined: Coroado, 10.7 percent; Mashacali, 12.6 percent; Camacdn, 17.2 percent. My reworking of Loukotka’s data, eliminating the most improbable of his correspondences, gave the following results: Camacdn showed most resemblance to Ge with 37 probable corre- spondences, 7 of them close; next to Mashacalié with 18 probable corre- spondences, 7 close; and next to Caingang with 25 probable corre- spondences, 2 close. There were 12 probable correspondences to Botocudo, 4 close ones. The correspondences with Jaté, Patashé, and 294 SOUTH AMERICAN INDIANS [B. A. B. Bull. 148 Opayé are ignored on account of the very slight amount of data on these languages. In spite of the large Coroado vocabulary, the largest of all, the correspondences are very few, only 7, with 2 of them close, less than the resemblance to Jaté, with 10 probable correspondences. Camacdn obviously stands in much closer relationship to Jaté than to Coroado. The closest resemblance of Mashacali, on the other hand, is about equally to Coroado, with 23 probable correspondences, 9 of them close, and to Ge with 26 probable correspondences, 8 of them close. Next follows Camacdn, with 18 probable correspondences, 7 of them close, and then Caingang with 20 probable correspondences, 4 of them close. Coroado has its closest resemblance to Ge, with 35 probable corre- spondences, 17 of them close; with Caingang, with 30 correspondences, 9 of them close; next with Mashacalét with 23 probable correspond- ences, 9 of them close; and last with Botocudo, with 13 probable corre- spondences, 3 of them close. The slight resemblance to Camacdn, a significant point, is noted above. As may be deduced from the above, Ge shows about equal resem- blance to Coroado and to Camacdn, the former showing 35 possible correspondences, 17 of them close; Camacdn, 37 possible correspond- ences, 7 close (but with a much smaller vocabulary to compare). Next follows Mashacali, with 26 possible correspondences, 8 of them close; and then Caingang, with 14 possible correspondences, 6 of them close. CAMACAN The Camacdn languages are all extinct, all the data being now on record. Loukotka (1931-32 b) has published a monograph on them, giving them independent rank. In this he is followed by Métraux and Nimuendajii (Handbook, vol. 1, p. 547). Rivet (1924 a), W. Schmidt (1926), and earlier authorities considered the group a com- ponent of Ge. It is here classified as a component of Macro-Ge (q. v.). It shows more and closer lexical resemblances to Ge Proper, Mashacali, Caingang, and Jaté than can be explained on grounds of borrowing. Though probably related to Puri also, the lexical re- semblances are surprisingly slight. The resemblance is about equal to all the Ge Proper groups, except to Suya and Jeicéd. The Camacdn are not an Acroa horde, as Martius thought. There is general agreement as to the languages composing the family. As regards the closer relationships of these languages, there is less agreement. The classification here accepted is based primarily on Métraux and Nimuendaji. Vol. 6] LANGUAGES—MASON 295 CaMaAckn I. Camacén (Kamakén) A. Mongoyé } B. Monshocé (Ezeshio) II. Cutashé (Kotoxz6) A. Catethoy (Katathoy) ? III. Menidn (Manyd) IV. Masacaré 3 1 Schuller’s identification (1930 a) of ZJaté or Fulnio (q. v.) with Mongoyé is certainly based on insufficient evidence. The vocabularies of Etienne and Guimaraes are said to be very incorrect plagiarisms of Wied. 2 No linguistic material on Catathoy is extant. 2? Loukotka (1935) differentiates Masacarda from the others as a language mixed with Ge, though he had earlier termed it merely a slightly variant form of Camacdn Proper. Bibliography.— Etienne, 1909 (Mongoyé); Guimaraes, J. J. da S., 1854 (Mon- goy6); Ignace, 1912; Loukotka (with complete bibliography), 1931-32 b; Martius, 1867, 2:153-54, 156-158 (Cotoxo), 155 (Meniens), 144-145 (Masacaréd); Métraux, 1930; Moreira-Pinto, 1894, 1:387; Wied-Neuwied, 1820-21, 2: 325-330. MASHACAL{ Mashacali, an extinct language, was first separated from the old inclusive Ge by Loukotka, who published the standard monograph upon it (1931-32 a);it is now accepted as an independent family by Métraux and Nimuendaji (Handbook, vol. 1, p. 541). However, it has obvious resemblances with Ge and has, therefore, herein been considered one of the members of the Macro-Ge phylum. The avail- able data are slight, old, and poor. The resemblance is about equal, and not great, to Puri and Ge Proper, slightly less to Camacdn and Caingang. In the Ge group, Mashacali seems to show the greatest connection with Cayapé, the least with Northern Ge. Six “languages” are placed by all authorities in this family, all given equal rank, and no further subdivisions proposed. With regard to three languages, Malali, Potashé, and Coropé, placed by some in the Mashacali group, there is great difference of opinion. These three are treated separately. Caposhé (Koposé) Cumanashé (Kumanar6) . Macuni (Makoni) . Mashacali (Mazakari) . Monoshé (Monozx6) . Panyame (Pafiame) Anrwne Bibliography.—Loukotka, 1931-32 a (containing full bibliography); Martius; 1867, 2:169 (Mashacali), 170-172 (Capoxé6, Cumanaché, Panhdéme), 173-176 (Macuni); Saint-Hilaire, 1830-51, 1:47 (Maconi), 213 (Mashacali), 428-429 (Monoshé); Wied-Neuwied, 1820-21, 2:319, 323-325 (Mashacali, Macont). PURI (COROADO) For this extinct group or family the name Puri is preferable to Coroa- do, to avoid confusion with Caingang and Bororo groups of the same 296 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 name, which means ‘crowned’ or “tonsured.”” The group was formerly considered a part of Ge, and is herein considered one of the components of Macro-Ge. It was separated from Ge by Loukotka, the author of the principal monograph upon it (Coroado) (1937); Nimuen- dajtii accepts it as independent (Puri), as does Métraux (Handbook, Vol. 1/p. 523). The languages are probably extinct but a number of vocabularies are extant, and the lexical data, compiled by Loukotka, amount to some 900 words (including Corepd). Lexically, Puri-Coroado shows the closest relationship with Ge Proper, closest with Cayapo, least with Suyd, about equal with Northern and Eastern Ge, little with Jeicé. Resemblances with the Caingang and Mashacali groups are a little less and about equal. The lack of resemblance to Camacdn is significant. A century ago, the Coroado remembered when they formed a single group with the Puri; the differences between them must, therefore, be in the degree of dialects, that of their component bands even less. Puri and Coroado are the only certain members of the group. Other proposed members are Coropé and Waitacd (Guaitaka, Goyataca). These are treated separately herein, for reasons there given. Métraux herein (Handbook, vol. 1, p. 523) considers Coropé related. Puri (Coroapo) I. Coroado . Maritong . Cobanipake . Tamprun . Sasaricon VanPr It. Purt . Sabonan . Wambori . Shamishuna Bibliography.—Ehrenreich, 1886; Eschwege, 1818, pp. 165-171, 1830, pp. 233-243; Loukotka, 1937 (full bibliography) ; Martius, 1867, 2:194-207; Noronha Torrezao, 1889; Reye, 1884, pp. 99-101; Saint-Hilaire, 1830-51, 1:46—-47; Wied- Neuwied, 1820-21. QS PATASHO The classification of Patashé is most uncertain. The older clas- sifications of Rivet (1924 a) and W. Schmidt (1926) placed it in the Mashacali group of Ge. Loukotka (1935) separated it and gave it independent rank on an equal footing with Mashacali; in this he is followed by Métraux and Nimuendajt. Métraux and Nimuendajt say (Handbook, vol. 1, p. 54) that Nimuendaji found a close rela- tionship between his Patashé and Mashacali vocabularies, but that Wied-Neuwied’s Patashé and Saint-Hilaire’s Mashacali vocabularies are very different. Vol. 6] LANGUAGES—MASON 297 An independent reworking of Loukotka’s published comparative material left the present author doubtful of the Macro-Ge relationship of Patasho. A brief comparison of Wied-Neuwied’s Patashé vocab- ulary (1820-21), however, showed a marked resemblance to Mashacali, and considerable to Coropé (q. v.), but little to Ge proper. Some 20 of the Patashé words show apparent connections with Mashacali, and more than half of these are very close, and mainly in words not likely to have been borrowed. The inclusion of Patashé in the Mashacali group or family thus seems to be strongly indicated. However, it is quite possible that the vocabularies showing this resemblance are somehow faulty in ascription, and the example of Nimuendajti and Loukotka have been followed herein in leaving Patashé apart as a separate member of Macro-Ge. Patash6 may not be entirely extinct. No grammar or linguistic study is known. The standard vocabulary by Wied consists of only 90 words, but Loukotka possesses an unpublished study, and ap- parently Nimuendajt also had unpublished material. Bibliography.—Ehrenreich, 1891, 1894-95; Loukotka, 1939 ¢ (full bibliogra- phy); Martius, 1867, 2:172-173; Métraux, 1930 b; Wied-Neuwied, 1820-21, 2:320-321. MALAL{ In view of the great disagreement regarding the affinities of the extinct Malali it should be regarded for the present as an independent member of the Macro-Ge group as do Métraux and Nimuendajt herein (Handbook, vol. 1, p. 542). Nimuendaju (map and index) puts it among the isolated languages. W. Schmidt (1926) makes it the sole member of the coastal division of his South Group of Ge, an opinion with which Loukotka records his disagreement. Loukotka (1931-82 a, 1935, 1939 a) and Rivet (1924 a) place it with Mashacali, though the former does so with a little hesitation, as a language mixed with Coroado. A hasty comparison of the available Malali data suggests that its closest lexical resemblances are with Patashé and Macuni (Mashacalt). Its resemblances to Puri-Coroado, Camacdn (mainly to Manyd or Menien), Ge Proper, and Caingang are much less, and those to Botocudo Opayé, and Jaté are very slight. The available lexical material is a little over 100 words; no textual material or grammatical sketch are known. Loukotka (1931-32 a) gives a critique of the value of the three extant vocabularies. Bibliography.—Loukotka, 1931-32 a; Martius, 1867, 2: 207-208; Saint-Hilaire, 1830-51, 1:428-429; Wied-Neuwied, 1820-21, 2:321-323. 298 SOUTH AMERICAN INDIANS [B. A. E. Bull. 143 COROPO The classification of the extinct Coropé language is uncertain and in disagreement; it is, therefore, treated separately and considered an independent member of Macro-Ge. Loukotka (1937) calls it the most interesting of all the languages in the old Ge group; he believes that it contains a large number of words borrowed from unidentified non-Ge languages. Unfortunately, it is extinct without any known textual material. A comparison by the present author of the two known small vocab- ularies appears to indicate a fair number of stems showing relationship to the languages of the Mashacali group, especially to Maconi, a little less to Caposhé, Cumanashé, and Monoshé, about the same to Malali and Patashé (q. v.), a little less to the Puri-Coroado languages. Considerable resemblance was also seen to Ge Proper, Caingang, and Botocudo, but little to Camacdén, Opayé, and Iaté. The authorities disagree greatly as to whether Coropé should be classified with the Purt-Coroado group (Nimuendajui, map; Loukotka, 1935, 1937, 1939 a; W. Schmidt, 1926) or with the Mashacali (Rivet, 1924 a). In view of this disagreement, Métraux’s statement (Hand- book, vol. 1, p. 523) that Coropé is ‘‘closely’”’ related linguistically (as well as culturally) with Coroado and Puri can hardly be accepted. Bibliography.—Eschwege, 1818, pp. 165-171; Loukotka, 1937; Martius, 1867, 2: 167-169. BOTOCUDO The name ‘‘Botocudo”’ signifies wearers of large lip-plugs and as such has been applied to several groups of different linguistic affinities which must be carefully distinguished. One Botocudo group, the Ivaparé, Are, Sheta, or Notobotocudo, is Tupi. The best-known Botocudo, however, are Macro-Ge peoples. Here two groups must be differentiated. One, of the State of Santa Catarina, is the Caingang (q. v.). The larger and better-known group, of Minas Gerais and Espirito Santo, has a language formerly considered Ge. It is, how- ever, quite different from Ge Proper, and has been accorded inde- pendent status by Loukotka (1935, 1939 a) and Nimuendaji (map) and accepted by Métraux (Handbook vol. 1, p. 532). It is here con- sidered as a member of the Macro-Ge phylum. It might be better to allow Botocudo independence. No study of it has been presented, and the morphology is unknown. The vocab- ulary shows a small number of words related to other Macro-Ge languages (but relatively few), and some probably due to borrowing. The greater number of resemblances are with Coroado, next with Camacdn. The Macro-Ge affinities in the data available are greater Vol. 6] LANGUAGES—MASON 299 than those of Opayé, Iaté, and Patashé, but this may be due to greater amount of data. The constitution of the Botocudo group, since the latter is not well known, is uncertain. The divisions are probably mainly political or geographic. At least four of the languages marked as Botocudo by Nimuendajtii (map), Anket, Nacnyanuk, Pimenteira, and Yiporok, are considered independent by various authors herein. Pimenteira (q. v.) is rather distant from the main Botocudo group, and is considered Cariban by some authorities. Botocupo (AIMBORE, Borun) Arané (Aranya) Crecmun Chonougn (Crenak) Gueren Gutucrac: Minya-yirugn (Minhagirun) Nachehe (Nakrehe) (Yiporok [Giporok]: Poicd [Poyishd, Pozitzd]) (Anket ?) (Nacnyanuk ?) Bibliography.—Adelung and Vater, 1806-17 (Engerekmung); Almeida, 1846; Anonymous, 1852 (Pozitra); Castelnau, 1850-59, pp. 249-259; Ehrenreich, 1887, 1896; Etienne, 1909 (Borun); Froes de Abreu, 1929 (Crenaque); Ignace, 1909; Jomard, 1846, 1847; Marliére, 1825 a, 1825 b (Pajaurum, Krakmun, Naknanuk) ; Martius, 1867, 2: 177-194 (Encreckmung, Crecmun, Djiopouroca); Reye, 1884; Renault, 1904; Rudolph, 1909; Saint-Hilaire, 1830-51, 1: 194-199; Silveira, A. A., 1921, pp. 529-543 (Pozitza); Simoes da Silva, 1924 (Crenak); Tranea, 1882; Tschudi, 1866-69, 2: 288; Wied-Neuwied, 1820-21, 2: 305-314. (Undifferenti- ated items are all of the Crecmun group.) SHAVANTE (CHAVANTE, SAVANTE) Four groups of Southern Brazil of very different linguistic affinities are known to the Brazilian natives by the name Chavanté. They must be carefully distinguished. Three of them, the Oti, Opayé, and Cucurd (q. v.), form small independent (provisionally) families; the fourth, the Akwé (q. v.), is a Ge language. Bibliography.—Chamberlain, 1910 a, 1913 c; Ihering, 1907; Martius, 1867, 2:1385-139. oTti The extinct Ott (Chavanté, Shavanté, Eochavanté) are one of the four groups, all of different linguistic affinities, known as Chavanté; they must be distinguished. Now extinct, the small group was named Hochavanté by Von Ihering. The language has been accepted as constituting an independent family by all authorities; Rivet (1924 a) terms the family Savanté. No suggestions as to larger affiliations have been made by anyone. Bibliography.—Borba, 1908, pp. 73-76; Ihering, 1912; Vocabulario Compa- rado .. ., 1892, 300 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 OPAYE Until recently accepted as one of the Ge languages, Opayé or Opayé Shavanté was separated from it and considered an independent family by Loukotka (1935, 1939 a). Nimuendaji (map) also calls it isolated. Though probably not extinct the data on it are scant. No gram- matical sketch and no linguistic study are known. ‘The lexical ma- terial is limited to vocabularies collected by Nimuendajt and published first by Von Ihering (1912) and later, enlarged, by Nimuendajt (1932 a). Each consists of less than 300 words. There are very few resemblances, even distant, with Ge, Camacdn, Mashacali or Coroado, and almost all words are very different. The Ge and Camacdn “intrusions” noted by Loukotka (1935, 1939 a) are not evident, and Opayé should be considered unclassified until future careful studies may prove otherwise. In his vocabulary, Nimuendajti gives a few words from a variant dialect Vaccaria, which Loukotka terms Guachi (Guaci) of Vaccaria. Bibliography.—lIhering, 1912; Nimuendaju, 1932 a. cucuRA The sole evidence for this ‘family,’ and apparently for the ex- istence of the tribe, seems to be a vocabulary of 31 words gathered by the Czech explorer Frié in 1901 and published by Loukotka (1931 b). These natives of the Rio Verde of Mato Grosso are one of a number of groups known to the Brazilians as Shavanté (q. v.). The Shavante- Cucurd are apparently mentioned by no other writer and do not appear in Nimuendajti’s map and index. The vocabulary seems to have no resemblance to any of the surrounding languages with which Loukotka compares it, Opayé, Oti, Akwé, and Tupi-Guarani, but might show affinity with some more distant stock. A very few words are apparently borrowed from Tupi-Guarani. At any rate such a small vocabulary, taken through an interpreter, can hardly be accepted as definitely establishing a new linguistic family. The language is now presumed to be extinct. Bibliography.—Loukotka, 1931 b; Nimuendajd, 1932 b. GUAITACAN Goyatacd (Guaitakd, Waitacd, etc.) was adopted as the name of a stock or family by Chamberlain (1913 a), and as a substock of “Tapuya’” (Ge) by Brinton (1891 a); the latter included under it the Mashacali languages, Patashé6 and Coropéd. W. Schmidt (1926) accepted it for the name of his subgroup that included the Purt- Coroado languages, and Rivet (1924 a) included it in that group. As Guaitacdé became extinct before a word of it was recorded (see Vol. 6] LANGUAGES—MASON 301 Handbook, vol. 1, p. 521; Métraux, 1929 b), it cannot be regarded as anything but an unclassified language, as Nimuendajt places it. There is no reason for classing it with Puri-Coroado or with any other group. It very likely, however, was a Macro-Ge language. Four subdivisions are known. GuaiTach 1. Mopi 2. Yacorito 3. Wasu 4. Miri Bibliography.—Ehrenreich, 1905; Koenigswald, 1908 b; Métraux, 1929 b; Steinen, 1886. SMALL LANGUAGES OF THE PERNAMBUCO REGION (Fulni6, Natit, Pancararu, Shocé, Shucurti, Tushé, Carapaté, Payact, Teremembé, Tarairiu or Ochucayana) Along and to the northeast of the San Francisco River in the States of Alagoas, Sergipe, Pernambuco, and Bahia are, or were, a number of small tribes the languages of which seem to be sufficiently variant from themselves and from others with which they have been compared to be classified by Nimuendajtii (map) and accepted by Lowie (Hand- book, vol. 1, p. 553) as isolated or independent. All are so small, unimportant, or newly identified that none of them is mentioned by Rivet (1924 a) or by earlier compilers, and only Fulnié (Laté) is listed by Loukotka (1935, 1939 a), and rates especial mention. On all but the last the lexical data seem to be very slight and difficult of access; most of them seem to be in unpublished notes and observations of Estevao de Oliveira and Nimuendajti, whose opinion as to the isolated status must therefore be accepted for the present. Fulni6.—Fulniéd (Fornié, Carnijé, Iaté) is the native name; the Brazilians of Aguas Bellas call them Carnijé6. Loukotka (1935, 1939 a) terms the family Jaté. There are no subdivisions. Loukotka sees Camacdn intrusions, and this is borne out by a superficial comparison of the data published by him; Fulnié seems to show closer resemblance to Camacdén than to any other of the Macro-Ge languages, but not enough to be itself placed in this group for the present. Schuller (1930 a) improperly identified the language with Mongoyé. Pancarart.—Pancararti. (Pankari, Pancarté) has sometimes been classified as a Cariri language but is better considered as isolated in agreement with the opinions of Lowie and Nimuendajti. (See Hand- book, vol. 1, p. 561.) Shocé.—Shocé (Soké, Chocd) must not be confused with the Isth- mian Chocé. Shucurti—Shucurt (Sukurt) is divided by Nimuendaji (map) into 302 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 two groups, those of Cimbres and those of Palmeira dos Indios. The latter appear to be known only by the notes of Oliveira. Teremembé.—Though more important historically than most of the above groups, nothing is known of the language of the Teremembé (Métraux, Handbook, vol. 1, p. 573). Tarairiu or Ochucayana.—Though generally classified as a Ge language, the available linguistic data do not support the affinity, and Lowie (Handbook, vol. 1, p. 563) thinks that it may be considered a distinct stock as proposed by Pompeu Sobrinho (1939)—probably too radical a decision, Nimuendajt’s preference (map) to leave it unclassi- fied is better. Schuller (1930 a) also called it “ein stamm”’; Khren- reich (1894) believed it affiliated with Ge and especially with Patashé. Loukotka (1935, 1939 a) calls it Carib, mixed with Ge, which is doubt- ful, in view of its distance from any other Carib group. Nati, Tusha, Carapato, Payaci.—Little is known of these languages. Nimuendajti leaves them unclassified; other authorities ignore them. Lowie (Handbook, vol. 1, p. 553) speaks of ‘‘six unrelated linguistic families within the area”: Fulnid, Shucurt, Pancarari, Nati, Shocod, and Tushd. Thus to accord them familial status is certainly not justified by the few data on them. Bibliography.—See bibliographies in Handbook, vol. 1, pp. 556, 561, 566, 571, 574. Most of the more recent works (Branner, 1887; Melo, 1927, 1929; Pompeu Sobrinho, 1935, 1939; Schuller, 1930 a) refer to the Fulnié. Pinto, 1938, treats of the Pancarizi; Schuller, 1913 ¢, of the Tarairiu. SOUTHERNMOST LANGUAGES ATAGUITAN Ataguita is here for the first time proposed as a hybrid term for the hypothetical Atacama-Diaguita linguistic group. It is unproved, and no definite proof of the relationship has been offered, but the con- nection has been accepted by several authorities. First suggested by Schuller (1908), W. Schmidt (1926) proposed a Cunza-Diaguita Group, and Jij6n y Caamafio (1941-43) adopts it as an Atacamefio-Diaguita phylum. ATACAMA Synonyms: Atacamefio, Cunza, Kunza, Likananta, Likanantai, Lipe, Ulipe. Though a few individuals may still speak the old Cunza language little is known of it. A modern study and grammar is urgently needed, though even a thorough study of the grammar of San Roman (1890) might link it to one of the larger linguistic families. Most authorities from Chamberlain (1911 b) down have accorded Atacama an independent position. Loukotka (1935) sees vestiges of Arawak in it. Von Tschudi (1866-69) suggested that it is a descendant of Vol. 6] LANGUAGES—-MASON 303 Calchaqui-Diaguita, and W. Schmidt (1926), accepting the arguments of Schuller (1908), proposes a Cunza-Diaguita group, uniting Atacama and Calchaqui. (See Handbook, vol. 2, pp. 599, 605, 606.) Bibliography.—Boman, 1908; Brand, 1941 ¢; Chamberlain, 1911 b, pp. 465-467; Darapsky, 1889; Echeverria y Reyes, 1890, 1912; Maglio, 1890; Moore, 1878; San Romdn, 1890; Schuller, 1908; Tschudi, 1866-69; Vaisse, Hoyos, and Eche- verria y Reyes, 1895. OMAWACA (OMAHUACA) The affinities of the extinct Omahuaca (Omawaka, Omaguaca, etc.) and Humahuaca are, and probably always will be, uncertain. It is one of the four South American languages that Loukotka (1935) declined to classify. Rivet (1924 a) places it with Quechua, Jijén y Caamafio (1941-43) with Atacameno-Diaguita. The Omahuaca are said to have been a mixture of Diaguita and Aymara, and spoke Quechua at the time of their extinction as a tribal entity. (See Handbook, vol. 2, p. 619.) DIAGUITA OR CALCHAQUI Since not one word of the extinct Diaguita or of its related languages has been positively identified, its status depending on early statements and proper-name etymologies, its independent position, relationship with other ‘‘families’” and with its probably component languages will probably never be conclusively determined, unless a copy of the lost Barcena grammar is found. Diaguita (Diaguite, Diagit) is the term most frequently used for the group, but Calchaqui(an) was the earlier term employed by Chamberlain (1912 a, 1912 b) and his followers, and Brinton (1891 a) preferred Catamarefa. The language of the Calchaqui-Diaguita was known as Cacan(a) or Kakan(a). It was replaced by Quechua in the 17th century. The Calchaqui were but one tribe or nation of the group; other affiliated languages as given in the table were probably of the status of dialects. The Lule enter to complicate the problem even more. This name was probably applied to several different groups in this general region—or else to a group speaking several different languages. The Lule of Padre Barcena seem to have been Diaguita, to be distinguished from the Lule of Machoni, which is Vilela. (See Lule-Vilela, Vilela- Chulupi, ete.) The relationship of the extinct Sanavirén and Comechingén is also in dispute. Most authorities consider these as forming the inde- pendent Sanaviron(an) family. Krickeberg (1922) and W. Schmidt (1926) place them under Diaguita. Loukotka (1935) puts Kakana (Calchaqui), Sanavirona, and Vilela together in his Vilela family. (See Handbook, vol. 2, pp. 657, 661-663.) 304 SOUTH AMERICAN INDIANS [B. A. E. Bull. 143 Brinton finally accepted the suggestion that Diaguita had affinities with Quechua. Relationships with the Atacama or Atacamefio family (q. v.) were suggested by Schuller (1908) and accepted by W. Schmidt (1926), who lists a Cunza-Diaguita Group, Kunza being the name of the principal Atacama language. The Diaguita ‘‘dialects”’ he lists as Kaka(na), Tonokote (placed by others in Lule, Vilela and Mataco), Zanavirona (though he also makes a Sanavirén family), and Indamu (generally placed with Sanavirén). All of the above conflicting opinions seem to be based on the most inferential evidence, from which every seeker after knowledge may take his choice. Diaguita subgroups:—Abaucan, Amaycha, Anchapa, Andalgald, Anguinahao, Calchaqui, Casminchango, Cope, Colalao, Famatina, Hualfina, Paquilin, Quilme, Taft, Tocpo, Tucumdén, Upingascha, and Yocabil. Possibly also: Acalian, Catamarca, and Tamano. Bibliography.—Barcena, MS., Boman, 1908; Cabrera, P., 1927, 1931; Canals Frau, 1943 a, 1943 b; Chamberlain, 1912 a, 1912 b; Kersten, 1905; Lafone-Quevedo, 1898, 1919 a, 1927; Lizondo Borda, 1938; Schuller, 1908, 1919-20 b, pp. 572-573; Serrano, 1936 b. CHARRUA, KERAND{, CHANA, ETC. Synonyms: TSarrua, Carrua, Chand, Giienoa. Charrua has been accepted as an independent family from the time of Hervas y Panduro (1800), but suggestions as to affiliation with all neighboring groups—Arawak, Ge, Guaicurt, Guarani, and Puelche— have been made, as well as its connection with Querandi, for which latter various connections have also been proposed (vide infra). Brinton (1898), D’Orbigny (1839), M. S. Bertoni (1916), Outes (1913 b), Serrano (1936 a, 1936 c), Schuller (1906), and others have entered into this argument, as well as the recent classifiers such as Rivet (1924 a), W. Schmidt (1926), and Loukotka (1935). One of the most recent writers, S. Perea y Alonso (1942), considers all the Chané-Chana languages, including Charrua, as Arawak. Even the present authors herein, Lothrop, Serrano, Cooper, and Métraux, do not agree. Ser- rano (Handbook, vol. 1, p. 192) considers it related to Caingang. Nor is there any agreement as to name; most authorities use variants of Charrua, but Serrano herein insists that the generic name should be Gienoa. Most of the arguments are based on historical evidence and infer- ence, since all of the languages have long been extinct with little re- corded data; no more than 7 words of Charrua were known. Recently, however, some 70 more words have been found and published (Gémez Hardo, 1937), but never scientifically compared with other languages. The opinions of present contributors may be cited as the most Vol. 6] LANGUAGES—MASON 305 modern. Serrano is certain that Charrua is a dialect of Chand and related to the Caingang of Rio Grande do Sul (cf. Macro-Ge). Mét- raux doubts the Ge affiliation, and suggests that a comparison with Puelche and other Patagonian and Chaco languages might prove fruitful; Lothrop wisely makes no suggestions as to wider affiliations. Nimuendajti’s decision to leave Charrua unclassified is doubtless the wisest one. Charrua subtribes are said to be Guayantiran, Palomar, and Negueguian. In the La Plata region were many other languages, now long extinct, on which the data are very deficient, with consequent great differences of opinion regarding their affiliations. All these had best be left unclassified. Most of them have been traditionally considered Charruan. Prominent among these is the Querandi (Kerandi), which has variously been considered of Guaicuri, Het (Tehuelche), Arauca- nian, and Guarani affiliations. Cooper (Handbook, vol. 1, p. 137) says there is good argument for considering it Puelche. Giienoa and Bohané may relate closely to Charrua, possibly also Caracand. Other of the more important groups are Minuané, Yard, Colastiné, Corondé, Timbi, Mbegud, and Carcarand. Of lesser importance are Kiloazd, Cayasté (Chaguayd), and Macurendé (Mocoreté). (See Lothrop, Handbook, vol. 1, pp. 177-190.) Bibliography.—Brinton, 1898 a; Chamberlain, 1911 b, pp. 469-471; Gémez Hardo, 1937; Kersten, 1905; Lothrop, 1932; Martinez, B. T., 1919; Orbigny, 1839; Outes, 1913 b; Perea y Alonso, 1938 b, 1942; Rivet, 1930 a; Schuller, 1906, 1917; Serrano, 19386 a, 1936 ¢; Vignati, 1931 d. CHANA Synonyms: T$andé, Tschané, Cand. One of the important groups in this area, undifferentiated in locale, is the Chand. Nimuendajt accepts the Chand as a linguistic entity, leaving all the other before-mentioned languages as unclassified. Affiliated with the Chand seem to be the Chand-Mbegud, Chand- Timbui, and the Yard. Perea y Alonso (1942) relates these Chand to the Chané (q. v.) of southern Brazil and apparently believes almost all the above-mentioned groups, including Charrua and all the natives of the Banda Oriental of Uruguay, to be Arawak. For geographic reasons this is open to doubt, pending further exposition. Chand is a descriptive term and as such applied to a number of dis- tinct groups of different linguistic affiliations that are liable to be con- fused. Itissaid to be a Tupi word, probably meaning “my relations.” It seems to have been applied to certain Tupi, Guarani, and Chiri- guano groups. It also seems to be a synonym for the Layand, a southern Arawak group (according to W. Schmidt, 1926; Nimuendaji considers the Layana to be Guaicuri., q. v.). 306 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Larrafiaga (1924 b) says the language was guttural, an amateurish characterization applied to many Indian languages. Larrafiaga’s vocabulary and grammar was published by Lafone-Quevedo and Torres. (See also Chané.) Bibliography.—Brinton, 1898 a; Cardus, 1886; Kersten, 1905; Lafone-Quevedo, 1897 a, 1922; Larrafiaga, 1924 b; Lothrop, 1932; Orbigny, 1839; Outes, 1913 b; Serrano, 1936 a, 1936 c. ALLENTIAC OR HUARPEAN Huarpe might be a better term than Allentiac for this linguistic group, and is preferred by some modern writers, but the latter name is probably too well established to make a change advisable. It has been accepted as an independent family or stock by all authorities since Brinton (1891 a) and Chamberlain (1913 a), generally under the name Allentiac or variations thereof. Huarpe (Guarpe) is a synonym of Allentiac. ‘There is general agreement that the Millcayac language was rather closely related. The languages became extinct in the 18th century. However, grammars of both Allentiac and Millcayac by Padre Valdivia are known, though the first editions are extremely rare. Though the group will probably eventually be tied up with some of the neighboring languages, and probably affiliated with some one of the larger phyla, few suggestions as regards such relationships have been made, and none accompanied by good evidence. Some early statements suggest a relationship with Pwelche, and Brinton (1891 a) placed Huarpe, Puelche, and Araucanian in his Aucanian linguistic stock. Canals Frau (1944) presents extensive evidence and argument that the Comechingén (q. v.) were related to the Huarpe, and terms the linguistic group Huarpe-Comechingén. He considers the group to consist of the following languages: (1) Allentiac or Huarpe of San Juan; (2) Milleayac or Huarpe of Mendocino; (3) Puntano Huarpe; (4) Puelche of Cuyo; (5) Ancient Pehuenche; (6) Southern Comechingén, language: Camiare; (7) North- ern Comechingén, language: Henia; (8) possibly Olongasta (Indians of Southern Rioja). (See Canals Frau, Handbook, vol. 1, p. 169.) As subgroups or dialects Pericot y Garcia (1936) names Zoquillam, Tunuyam, Chiquillan, Morcoyam, Diamantino (Oyco), Mentuayn, Chom, Titiyam, Otoyam, Ultuyam, and Cucyam. Bibliography.—Cabrera, P., 1928-29; Canals Frau, 1941, 1942, 1943 a, 1943 b, 1944; Chamberlain, 1912 b; La Grasserie, 1900; Marquez Miranda, 1943, 1944; Medina, J. T., 1918; Mitre, 1894; Schuller, 1913 a, 1913 d.; Valdivia, 1607 a, 1607 b. Vol. 6] LANGUAGES—MASON 307 SANAVIRON AND COMECHINGONAN SANAVIRON There are few linguistic data on which to classify the extinct Sanavirén and its affiliated languages, and the opinions are, therefore, very variant. Most authorities, such as Chamberlain (1913 a), Rivet (1924 a), and Nimuendaji (map and index), class it as an inde- pendent family. Krickeberg (1922) places it under Diaguita; W. Schmidt (1926) equivocates by establishing a Sanavirén family, but also placing Zanavirona in his Cunza-Diaguita group. Loukotka (1935) groups Sanavirona, Kakana (Calchaqui-Diaguita), and Vilela in his Vilela family. Jijén y Caamajfio (1941-43) also places Sanavirén in his Vilela-Lule phylum. Sanavirén is omitted from the accompanying linguistic map, the occupied area being allotted to Comechingén. Bibliography.—Chamberlain, 1910 a, p. 198. COMECHINGON So little is known of the extinct Comechingén that its affiliation may never be determined, and there is no present consensus. It has been connected with three families, also all extinct. Most authorities place it with Sanavirén; Krickeberg (1922) considers it related to Diaguita. The most recent writer, Canals Frau, (Handbook, vol. 1, p. 169; also 1944) links it with Huarpe (Allentiac). It is one of the few languages that Loukotka (1935) wisely refuses to attempt to classify. There seem to have been five subgroups or dialects. Michilingiie apparently belonged to the same group. IJndaméd or Indamu is gen- erally associated with Comechingén, but W. Schmidt (1926) puts it with Zanavirona in the Cunza-Diaguita group, not with Comechingén under Sanavirén. I. Comechingén A. Comechingén 1. Main 2. Tuya 3. Mundema 4. Cama 5. Umba B. Michilingwe C. Indama Bibliography.—Canals Frau, 1944. ARAUCANIAN The Araucanian (Araukan, Aucanian, Aucan) languages occupied a moderately large solid area in northern Chile and adjacent Argentina; their modern range is considerably reduced though the language is 7947115021 308 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 still vigorous, with speakers said to number upward of fifty thousand. Their relationship with Puelche, Het, and Tehuelche or Chon (q. v.), as well as with other ‘‘families” to the north of these, is likely but un- proved. Considerable confusion is caused by the fact that names of certain groups in almost all of these families end in ‘‘che,”’ and others ending in ‘‘het’’ are also thus divided. No one seems to have at- tempted to subdivide the family on a linguistic basis, or to have presented concise data on which this could be done. The linguistic divisions probably coincide with the political and geographical ones, but many groups, especially the extinct ones, are of doubtful rela- tionship, even as to the Araucanian family. The classification here given is based primarily on Brand (1941 c¢),so far as that goes. Few of the other authorities agree with him or among themselves as regards the minor groupings. The living groups are said to be of the order of dialects, all mutually intelligible. The linguistic affinities of the Pewenche (Pehuenche-Puelche) and the Huilliche Serrano are question- able. (See Cooper, Handbook, vol. 1, pp. 128, 132, vol. 2, pp. 688-696; Canals Frau, Handbook, vol. 2, pp. 761-766.) The Araucanian languages are said to be pleasant and harmonious. ARAUCANIAN I. North A. Picunche B. Mapuche 1. Pewenche a. Rankel(che) 2. Moluche II. South A. Wiliche (Huilliche) 1. Wiliche a. Serrano b. Pichi-Wiliche 2. Manzanero Veliche (Chilote) . Chikiyami (Cuncho) . Leuvuche III. East . Taluhet (Taluche) 3 . Divihet (Diviche) } 1Possibly member of separate Het family (q. v.). we Yow BIBLIOGRAPHY Adelung and Vater, 1806-17; Amberga, 1914; Ameghino, C., 1913; Anonymous, 1876; Augusta, 1903, 1916, 1922; Barbara, 1879; Brand, 1941 a, 1941 c; Cafias Pinochet, 1902, 1911 a; Darapsky, 1888; Echeverria y Reyes, 1900; Englert, 1934; Falkner, 1899; Febres, 1765, 1864, 1882; Figueroa, 1903; Groeber, 1926; Gusinde, 1936; Gutiérrez, 1871; Huaiquillaf, 1941; La Grasserie, 1898; Latcham, 1942; Lenz, 1895-97, 1904-05; Lillo, 1928; Loukotka, 1929-30, pp. 75-83; Mi- lanesio, Domingo, 1918; Mitre, 1894, 1909, pp. 311-338; Moesbach, 1929-31, 1936, Vol. 6] LANGUAGES—MASON 309 Outes, 1914; Rodriguez, Z., 1875; Santa Cruz, 1923 b; Schuller, 1908; Sigifredo, 1942-45; Speck, 1924; Valdivia, 1887, 1897; Zeballos, 1922 b. CHONO The Chonoan “‘family”’ of the Chilean coast, recognized by Chamber- lain (1913 a) and his followers, is no longer accepted. Only three words seem to be known, and its independence was presumed on grounds of early statements. Affiliations have been suggested with all the neighboring groups, Araucanian, Tehuelche, and Alacaluf. Rivet (1924 a) placed it with the last. It had best be left unclassified. It must be distinguished from the Chon (Tson) or Tehuelche, but doubtless these names are of common origin. (See Handbook, vol. 1, pp. 48-49). Bibliography.— Chamberlain, 1911 a; Cooper, 1917 a, 1917 b; Ferrario, 1939. PUELCHEAN Synonyms: PueltSe, Puelée, Kunnu, Gennaken, Pampa. The group has been recognized since early days but its constitution is greatly under discussion. Brinton (1891 a) grouped Puelche, Araucanian, and some other languages in his Aucanian stock; Cham- berlain (1911 a, 1913 a) and all subsequent authorities have accepted a Puelche(an) family. They are often referred to as northern Tehuelche, or merely Tehuelche, but belong to a separate family from the true or southern Tehuelche; the latter is the older name for the southern family but leads to confusion with the Puelche, so the modern name Chon (q. v.) is preferable for the former. All authorities recognize but one language in the family, Puelche, unless Chechehet is related; this is now often placed in a family of its own, Het (q. v.). Ten dialects are said to have been spoken, but today only two, eastern and western, are reported. Relationships have been suggested with Guaicurt, Araucanian, Het, Chon, and Charrua, none of which would be in the least surprising. The old source, Valdivia (1607), says that Puelche differs very slightly from Milleayac (Allentiac), but he may have been referring to another group of Puelche. Bibliography.—Adelung and Vater, 1806-17; Brinton, 1892 d; Chamberlain, 1911 b; Harrington, T., 1925; Milanesio, Domenico, 1898; Outes, 1928 a; Outes and Bruch, 1910; Valdivia, 1607 a. HET (CHECHEHET) Although only 15 words and some place names seem to bé known, Lehmann-Nitsche argued that the language of the Chechehet (TSetsehet, Ceéehet), formerly considered as a Puelchean (q. v.) language, is radically different from the latter and entitled to be considered an 310 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 independent family. He called the family ‘‘Het”; Rivet (1924 a) adopts the same name. Loukotka (1935) and Nimuendaji prefer their orthographic variants of Chechehet; the former accepts it as a family, the latter as “‘isolated.”” Together with the Chechehet go a group who lived with the Araucanian Divihet (Diuihet) and were known by that name only. The pertinent linguistic data are based on Falkner (1774), Hervds y Panduro (1800), and Dobrizhoffer (1784). The language became extinct about the close of the 18th century. The solution of the Het question is an historical, not a linguistic, one. (See also Handbook, vol. 1, p. 134.) Bibliography.—Dobrizhoffer, 1784; Falkner, 1774; Hervdés y Panduro, 1800; Lehmann-Nitsche, 1918 a, 1922, 1925 b, 1930 a. CHONAN OR TEWELCHE (TEHUELCHE) AND ONA The Chon or Tehuelche (Ton, Tschon, Con, Tsonekan, Tehuelchean) has been considered independent since earliest classifications, and no suggestions of larger relationships have been made except for those of Rivet (1925 a, 1925 b, 1925 c, 1926 a, 1926 b, 1927 b, 1927 c) whose revolutionary belief in a connection with Australian languages has been accorded ex-cathedra condemnation by all North American anthropologists, probably without sufficient scientific consideration. The term Tehuelche was often used in a geographical rather than a linguistic sense, and the northernmost Tehuelche, the Kiini, seem to have spoken a Puelche tongue. The three languages of Tehuelche proper were almost unintelligible, but now are less so (Cooper, Hand- book, vol. 1, p. 130). The two divisions of the Ona could understand each other only with difficulty; the dialects differ slightly (Cooper, Handbook, vol. 1, p. 108). However, Tehuelche and Ona are rather closely related. The various classifications differ but slightly. The affiliations of the Paya are uncertain. Ona was long considered as forming a separate family from Tehuel- che. Though the names are probably connected in origin, the Chon must be differentiated from the Chono (q. v.) of the Chilean coast. CHON I. Chon (Tehuelche) A. Tehuelche 1. Ta’utishn (Tewesh) 2. Northern: Payniken (Pa’dnktin’k) a. Poya 3. Southern: Inaken (Ao’nitiktin’k) B. Ona} 1. Haush (Manekenkn) 2. Shelknam a. Northern b. Southern 1 Brinton (1891 a, p. 331) gives Huemul and Peschere (Ire) as other divisions of the Ona. Vol. 6] LANGUAGES—MASON 311 BIBLIOGRAPHY Adelung and Vater, 1806-17; Ameghino, C., 1913; Beauvoir and Zeballos, 1915; Brinton, 1892 c, 1892 d; Chamberlain, 1911 a, 1911 b; Cooper, 1917 a, 1917 b; Ferrario, 1942; Furlong, C. W., 1917 b; Hunziker, 1910, 1928; La Grasserie, 1906; Lahille, 1928, 1929, 1934; Lehmann-Nitsche, 1912 a, 1912 b, 1913, 1923; Lista, 1896; Lothrop, 1928; Lowie, 1933; Miller, Frederick, 1882; Outes, 1913 a, 1926, 1928 a, 1928 b, 1928 c; Rivet, 1925 a, 1925 b, 1925 c, 1926 a, 1926 b, 1927 b, 1927 ¢; Schmid, 1860, 1912 a, 1912 b; Spegazzini, 1888; Tonelli, 1926; Zeballos, 1915. YAHGANAN The independence of the Yahgan (Yagan, Ydmana, Yahganan) family or stock has never been doubted. Except from this point of view it is unimportant; it is practically extinct; probably not more than 20 Yahgan survive. The tongue is said to be markedly euphonic, soft, melodious, agreeable, with a rich vocabulary. There is only one language, with five mutually intelligible dialects, of which the Central and Western are said to be most alike. I. Yahgan A. Eastern B. Central-Western 1. Central 2. Western C. Southern D. Southwestern BIBLIOGRAPHY Adam, 1885; Brand, 1941 c; Bridges, 1884-85, 1933; Brinton, 1892 c; Chamber- lain, 1911 a; Cooper, 1917 a, 1917 b;,Darapsky, 1889; Denuce, 1910; Garbe, 1883; Gusinde, 1928 b, 1934; Haberl, 1928; Hestermann, 1927 b, 1929; Knudsen Larrain, 1945; Koppers, 1926, 1927, 1928; Lothrop, 1928; Miller, Frederich, 1882; Outes, 1926-27 a, 1926-27 b; Platzmann, 1882; Spegazzini ,1888; Wolfe, 1924. ALACALUFAN The Alacaluf (Alikuluf, Alukulup, and similar orthographical vari- ants), the southernmost language of South America, has been recog- nized since earliest times as constituting an independent family. No relationships with any other group have been suggested, except Rivet’s (see Chon) belief in their connection with Australian languages. The language is said to be harsh, with explosives and gutturals, though not so strong as in Ona. Three “dialects” are reported, but 10 or more groups, presumably each with its dialect, are named; how they group in subdivisions is unknown. The northernmost, Chono (q. v.), is of uncertain affiliation. Most of the languages or dialects are ex- tinct. The same may almost be said of the group; estimates of their number vary from 250 to none. a Bed SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 The following groups or dialects are reported: Caucawe (Kaukahue, etc.) Adwipliin Enoo or Peshera (PeSera) Alikulip, Alakaluf, ete. Lecheyel (LetSeyel) Calen Yekinawe (Yequinahuere, etc.) Taijatof Chono (TSono) (q. v.), Caraica (Karaika), and Poya may also belong. Bibliography.—Borgatello, 1924, 1928 a, 1928b; Brand, 1941 c¢; Brinton, 1892 ec; Chamberlain, 1911 a; Cooper, 1917 a, 1917 b; Ferrario, 1939; Gusinde, 1927 b; Hestermann, 1927 c; Lehmann-Nitsche, 1918 b, Lothrop, 1928; 1921; Skottsberg, 1913, 1915; Spegazzini, 1888. BIBLIOGRAPHY The following bibliography is an attempt to cite all the known sources for South American native languages, without qualification as to value. Very few of 'them were consulted in the preparation of the preceding monograph. Many of them were taken from Dr. Paul Rivet’s bibliographies published annually in the Journal de la Société des Américanistes de Paris; many others from Jijén y Caamajfio’s EH] Ecuador Interandino y Occidental, vol. 3, 1943. Abreu, 1895, 1914, 1938 a, 1938 b; Acosta Ortegén, 1938; Acufia, 1641; Adam, 1878, 1879, 1885, 1890 a, 1890 b, 1891, 1892, 1893, 1896, 1897 a, 1897 b; 1897-98, 1899, 1902, 1905, 1906; Adam and Henry, 1880; Adam and Leclerc, 1877, 1880; Adelung and Vater, 1806-17; Agiiero, 1929; Aguilar, 1939; Aguirre, 1898; Ahl- brinck and Vinken, 1923-24; Albis, 1855, 1860-61, 1934; Albuquerque, 1916; Alemany, 1906 a, 1906 b; Alemany y Bolufer, 1929 a, 1929 b; Almeida, H., 1846; Almeida, W., 1931; Altieri, 1939 a, 1939 b; Alvarado, 1919 a, 1919 b, 1921; Amaral, 1920; Amberga, 1914; Ambrosetti, 1894, 1895 a, 1895 b; Ameghino, C., 1913; Amorim, 1928; Anchorena, 1874; Anonymous, 1852, 1876, 1879, 1882, 1897, 1902, 1905, 1908, 1909, 1914, 1917, 1918 a, 1918 b, 1918-19, 1919 a, 1919 b, 1919 ¢, 1919 d, 1919 e, 1921, 1924, 1925, 1926-27, 1927 b, 1928 a, 1928 b, 1928 c, 1929, 1930 a, 1930 b, 1930 c, 1930 d, 1931, 1932, 1933 a, 1933 b, 1933 c, 1933 d, 1934, 1936, 1937, 1939, 1941; Arcaya, 1918; Archivo Nacional. . . n. d.; Arcila Robledo, 1940; Armellada, 1936; Armellada and Matallana, 1942; Armentia, 1888, 1898, 1902, 1903, 1904; Astete, 1936-37, 1937 a, 1937 b; Augusta, 1903, 1916, 1922; Ayllon, 1926; Ayrosa, 1933, 1934 a, 1934 b, 1934 c, 1934 d, 1934 e, 1934 f, 1934 g, 1935 a, 1935 b, 1935 c, 1935 d, 1935 e, 1935 f, 1935 g, 1935 h, 1935 i, 1935 j, 1935 k, 1936 a, 1936 b, 1936 c, 1936 d, 1936 e, 1936 f, 1937 a, 1937 b, 1937 c, 1938, 1939, 1941, 1948; Aza, 1928, 1924 a, 1924 b, 1924 c, 1927, 1928, 1930 a, 1930 b, 1930-32, 1931, 1933 a, 1933 b, 1935, 1937; Azpilcueta, 1938. Bacarreza, 1910; Bach, 1916; Bachiller y Morales, 1883; Balbi, 1826; Baldus, 1927, 1931, 1932, 1935, 1937; Bancroft, 1874-76; Baptista, 1931-32; Barbara, 1879; Barbosa da Faria, 1925; Barbosa Rodrigues, 1885, 1892-94; Barcatta de Valfloriana, 1918 a, 1918 b, 1920; Barcena, 1590, 1893; Barranca, J. 8.,1914, 1915-20, 1922; Barranca, S., 1868, 1876; Barrett, 1925; Barros Arafia, 1893; Basadre, J., 1939; Basadre, M., 1884; Bastian, 1876, 1878-89; Beauvoir and Zeballos, 1915; Belaieff, 1931-34, 1937, 1940; Bell, C. N., 1862; Berckenhagen, 1894, 1905, 1906; Berendt, 1874 a, 1874 b; Berengueras, 1934; Bernal P., 1919; Bernard de Nantes, 1709; Berrfos, 1919 a, 1919 b; Bertoni, A. de W., 1924; Bertoni, G. T., 1924, 1926, 1926-27, 1927, 1939; Bertoni, M. S., 1914, 1916, 1920, 1921, 1922, 1927 b, 1932; Bertonio, 1879 a (1603), 1879 b (1612); Beuchat and Rivet, 1907, 1908, 1909, 1909-10, 1910, 1911; Bibolotti, 1917; Bingham, 1922, 1930; Boas, 1911, 1922, 1933; Boggiani, 1894, 1895, 1896, 1900, 1929; Bolinder, 1917, Vol. 6] LANGUAGES—MASON ale 1924, 1925; Bollaert, 1860; Boman, 1908; Booy, 1929; Borba, 1882, 1904, 1908; Borgatello, 1924, 1928 a, 1928 b; Botelho de Magalhaes, 1946; Bottignoli, n. d. (1938, 1940), 1926; Branco, 1935, 1937; Brand, 1941 a, 1941 c; Brandéo de Amorim, 1928; Branner, 1887; Breton, 1665, 1666, 1669, 1877; Brett, 1868; Bricefio-Iragorry, 1929; Bridges, 1884-1885, 1933; Brinton, 1869 a, 1869 b, 1871, 1884, 1885 a, 1885 b, 1885 c, 1886 a, 1886 b, 1886 c, 1887, 1889, 1890, 1891 a, 1891 b, 1892 a, 1892 b, 1892 c, 1892 d, 1894 a, 1894 b, 1895 a, 1895 b, 1896, 1897, 1898 a, 1898 b, 1898 c; Briining, E., 1904, 1913; Briining, H. H., 1928; Buchwald, 1908 a, 1908 b, 1915, 1919, 1920, 1921, 1922, 1924. Caballero, 1933; Cabarera, P., 1924, 1927, 1928-29, 1931; Caillet, 1930-33; Calandrelli, 1896; Caldas, 1899, 1903; Camara Cascudo, 1934; Campana, 1902; Campanella, 1938-39; Canals Frau, 1941, 1942, 1943 a, 1943 b, 1944; Cajfias Pinochet, 1902, 1911 a; Candelier, 1893; Capistrano de Abreu, 1938 a, 1938 b; Cardona Puig, 1945; Cardus, 1886, 1916; Carrasco, 1901; Carrera, 1939 (1644); Carrocera, 1935; Carvalho, 1929, 1931; Castellvi, 1934 a, 1934 b, 1934 c, 1938, 1939, 1940; Castelnau, 1850-59; Castillo, 1906; Castillo y Orozco, 1871; Castro, A., 1985; Castro, E. de, 19387; Cavada, 1920, 1921; Celedén, 1878, 1886, 1892 a, 1892b; Céspedes Marin, 1923; Chaffanjon, 1889; Chagas-Lima, 1842; Chamber- lain, [1903], 1907, 1910 a, 1910 b, 1911 a, 1911 b, 1912 a, 1912 b, 1912 c, 1913 a, 1913 b, 1913 c; Chandless, 1866, 1869; Chantre y Herrera, 1901; Christensen, 1917; Christian, 1924, 1932; Chuqiwanqa, 1928; Church, 1898; Cieza de Leén, 1881 (1918) ; Cimitile, n. d.; Cipriani, C. A., 1902; Clarke, 1937; Colbacchini, n. d., 1925; Colbacchini and Albisetti, 1942; Coleti, 1771; Collins, F., 1879; Colman, 1917, 1921, 1929 a, 1929 b, 1932, 1936; Cominges, 1892; Constancio, 1939; Conzemius, 1923, 1927-28, 1929 a, 1929 b, 1930 a, 1930 b, 19382; Cook, O. F., 1916; Cooper, 1917 a, 1917 b; Coqueiro, 1935 a, 1935 b; Cordeiro, 1875; Cordero Palacios, 1924; Coryn, 1922; Cosio, 1916, 1919, 1923, 1924; Cotheal, 1848; Coudreau, H., 1886-87, 1892, 1897 a, 1897 b, 1897 c; Coudreau, O., 1901, 1903 a, 1903 b; Couto de Magal- hades, J. V., 1876; Couto de Magalhaes, J., 1902; Créqui-Montfort and Rivet, 1912, 1913 a, 1913 b, 1913 c, 1913 d, 1913 e, 1913 f, 1914, 1914-19, 1916-17, 1917-20, 1921, 1921-23, 1925-27; Crévaux, Sagot, and Adam, 1882; Cueva, 1893; Cullen, 1851 a, 1851 b (1853), 1866, 1868, 1875; Cuneo-Vidal, 1913, 1914, 1915a, 1915 b, 1915 c, 1916, 1918 a, 1918 b, 1920 a, 1920 b, 1922, 1930 a, 1930 b. Dangel, 1930, 1931 a, 1931 b; Darapsky, 1888, 1889; Deletang, 1926-31; Del- gado, A, and Vacas Galindo, 1929; Delgado, E., 1896-97; Denison, 1913; Denuce, 1910; Diaz Romero, 1918; Diaz, L., et Diaz, hijo, 1939; Dijour, 1931-32; Dobriz- hoffer, 1784 (1822); Dominguez, M., 1912; Domville-Fife, 1924; Donay, 1890; Dornas Filho, 1938; Dorsey, 1898; Drumond, 1944, 1946; Ducci, 1904, 1905, 1911-12; Durand, 1915-18, 1921; Duroni, 1924 (1928), 1928; Dyott, 1926. Kcheverria y Reyes, 1890, 1900, 1912; Eckart, 1890; Ehrenreich, 1886, 1887, 1888, 1891, 1894-95, 1896, 1897 b, 1905; Englert, 1934; Eraso Guerrero, 1944; Ernst, 1885, 1885 a, 1887 b, 1891, 1895; Eschwege, 1818, 1830; Espinosa, L., 1935; Espinoza, A., 1938; Etienne, 1909, 1910; Ewerton Quadros, 1892. F. M., 1907; Fabo, 1911; Falkner, 1774, 1899; Farabee, 1918 b, 1922, 1924; Farfan, 1939, 1941-42, 1942, 1943, 1944; Febres, 1765, 1864, 1882; Febres Cordero, 1921, 1942; Feijoo Reyna, 1924; Fejos, 1943; Fellechner, Miiller, and Hesse, 1845; Fernandes, J. A., 1892; Ferndndez, L., 1884; Ferndndez Guardia and Ferrez, 1892; Ferndndez de San José and Bartolomé, 1895; Fernandez y Gonzdlez, 1893; Ferrario, 1927, 1933, 1934, 1937, 1938, 1939, 1942; Ferreira-Franga, 1859; Ferreira-Penna, 1881; Fiebrig-Gertz, 1927, 1932; Figueira, 1878; Figueredo, 1701; Figueroa, 1903; Flornoy, 1938; Fonseca, A., 1914 (1920); Fonseca, J. S. da, 1880- 81; Forero, 1934, 1939; Franco, 1882; Franco Inojosa, 1937; Freitas, 1910; Frié, 314 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 ms.; Frié and Radin, 1906; Friederici, 1926, 1930; Fritz, 1922; Froes de Abreu, 1929, 1931; Fuhrmann, 1922; Furlong, C. W., 1917 b. Gabb, 1875, 1886; Gagini, 1917; Galindo, 1834; Galvfo, 1934; Gancedo, 1922; Gandia, 1931; Garbe, 1883; Garcia, J. A., 1917; Garcia, R., 1927; Garcilaso de la Vega, 1723 (1869); Garraux, 1898; Garro, 1939, 1942; Gasso, 1908, 1910-14; Gatschet, 1900; Gensch, 1908; Gez, 1915; Ghinassi, 1938, 1939; Giacone, 1939; Giannecchini, 1896; Gili, 1902; Gilij, 1780-84; Gillin, 1936, 1940; de Goeje, 1906, 1908, 1910, 1924, 1928 a, 1928 b, 1929, 1929-30, 1930, 1930-31, 1931-32, 1932, 1932-33, 1934, 1935, 1989; Gémez Hardo, 1937; Goncalves Dias, 1855; Gonzalez Holguin, 1607, 1608, 1901; Gonzdlez Sudrez, 1904; Gorrochotegui, 1918; Greiffenstein, 1878; Grigdrieff, 1935; Grijalva, 1921; Groeber, 1926; Groeteken, 1907; Grubb, K. G., 1927; Grubb, W. B., 1911 (1913); Grupe y Thode, 1890; Giiegiience, The . . . 1883; Guerrero y Sosa, 1932; Guimardes, C. J. da S., 1865; Guimaraes, G., 1930; Guimaraes, J. J. da S., 1854; Guimaraes Daupias, 1922; Gumilla, 1745; Gusinde, 1926 c, 1927 b, 1928 b, 1934, 1936; Gutiérrez, 1871; Guzmdén, 1920 a, 1920 b. Haberl, 1928; Hanke, W., 1938; Hardenburg, 1910, 1912; Harrington, J. P., 1925, 1943, 1944, 1945; Harrington, J. P., and Valcdrcel, 1941; Harrington, T., 1925; Haseman, 1912; Hassel, 1902; Heath, E., 1883; Heath, G. R., 1913, 1927; Henderson, A., 1846, 1847, 1872; Henderson, G., n. d., 1811; Henry, 1935, 1936, 1939, 1948; Hensel, 1869; Hermann, 1908; Hernandez, E., and Pinart, 1897; Herrera, 1916, 1919, 1923, 1933, 1983-34, 1939 a, 1939 b, 1941 a, 1941 b; Hervds y Panduro, 1800-05; Herzog, G., 1941; Herzog, W., 1886; Hestermann, 1910, 1914, 1914-19, 1925, 1927 a, 1927 b, 1927 c, 1929, 1938; Hilhouse, 1832, 1834; Hocquart, 1916; Hoehne, 1915; Hoeller, 1932 a, 1932 b; Hoijer, 1941; Holmer, 1946, 1947; Holter, 1877; Horta Barbosa, 1922; Huaiquillaf, 1941; Hiibner, 1907; Humboldt, 1822-27; Hunt, 1913 a, 1913 b, 1915 a, 1915 b, 1917, 1937, 1940; Hunziker, 1910, 1928 a, 1928 b; Huonder, 1902; Hurley, 1931 a, 1931 b, 1932; Hurtado, G. O., 1924, 1937. Ignace, 1909, 1912; Igualada and Castellvi, 1940; Ihering, 1895, 1904, 1907, 1912; Imbelloni, 1928 b, 1936; Im Thurn, 1883; Isaacs, 1884; Iturry Nufiez, 1939; Izaguirre, 1927-29. Jahn, 1914, 1927; J&uregui-Rosquellas, 1937; Jij6n y Caamafio, 1919, 1921, 1923-24, 1938, 1939, 1941-43; Jiménez de la Espada, 1881-97; Jiménez Borja, 1937; Jiménez Moreno, 1936; Jiménez Seminario, 1924; Johnson, F., 1940; Jomard, 1846, 1847; Jorge, 1924; José de Marfa, 1918-19. Karsten, 1919, 1920 a, 1920 b, 1921-22, 1922, 1923, 1932, 1935; Katzer, 1901; Keller, Fr., 1874; Kersten, 1905; Kimmich, 1917-18; Kinder, 1936; Kissenberth, 1911, 1916; Knoche, 1931; Knudsen Larrain, 1945; Koch-Griinberg, 1902 a, 1902 b, 1902 c, 1903, 1906 a, 1906 b, 1906 c, 1908, 1909-10, 1910 a, 1910 b, 1911, 1912-16, 1913, 1914-19, 1915, 1922, 1928, 1932; Koch-Griinberg and Hiibner, 1908; von Koenigswald, 1908 a, 1908 b; Kok, 1921-22; Koppelmann, 1929; Koppers, 1926, 1927, 1928; Koslowsky, 1895; Krause, 1911, 1936; Krickeberg, 1922; Krug, 1925; Kruse, 1930; Kunike, 1916, 1919; Kupfer, 1870; Kysela, 1931. Lafone-Quevedo, 1872 a, 1892 b, 1892-93, 1893 a, 1893 b, 1894, 1895 a, 1895 b, 1895 c, 1896 a, 1896 b, 1896 c, 1896-97, 1897 a, 1897 b, 1898, 1905, 1910 (1912), 1912 a, 1912 b, 1917, 1919 a, 1919 b, 1922, 1927; La Grasserie, 1890, 1892, 1894, 1898, 1900, 1904, 1906; Lahille, 1928, 1929, 1934; Lange, 1914; Larco Hoyle, 1938, 1939, 1944; Lares, 1918; Larrafiaga, 1924 a, 1924 b; Latcham, 1910, 1939 b, 1942; Latham, 1851; Leal, 1895; Leeris, 1592; Lehmann, H., 1945; Lehmann, W., 1910 a, 1910 b, 1914, 1915, 1920; Lehmann-Nitsche, 1905, 1910-11, 1912 a, 1912 b, 1918, 1918 a, 1918 b, 1921, 1922, 1923, 1924, 1925 a, 1925 b, 1926, 1930 a, 1930 b, 1936-37; Lemos, 1920; Lemos Barbosa, 1937; Lengerke, 1878 a, 1878 b; Lenz, Vol. 6] LANGUAGES—MASON 315 1895-97, 1904-05, 1904-10; Leédn, A. M., 1922, 1927, 1928-29, 1929, 1929-31, 1930 a, 1930 b, 1932-33, 1939, 1939-40; Léry, 1599; Lillo, 1928; Lista, 1883, 1896; Lizondo Borda, 1927, 1928, 1938; Llosa, 1906; Loayza, 1930; Lopes, 1925, 1934; Lorenz, 1939; Lothrop, 1928, 1932; Loukotka, 1929, 1929-30, 1930, 1931 a, 1931 b, 1931-32 a, 1931-32 b, 1933, 1935, 1937, 1938, 1939 a, 1939 b, 1939 c; Lowie, 1933; Lozano, 1941; Lozano-Machuca, 1885; Ludewig, 1858; Lugo, 1619. Macedo y Pastor, 1931-35, 1936, 1939; Machado d’Oliveira, 1936; Machoni de Cerdefia, 1732 (1877), 1894; Magalhaes, 1919; Magalli, 1890, 1891; Magio, 1880; Maglio, 1890; Maldonado de Matos, 1918 a, 1918 b; Mamiani, 1877; Marbdn, 1894; Marcano, 1890 a, 1890 b, 1891; Marcoy, 1867 (1866, 1869), 1875; Maria, 1914; Markham, 1864 a, 1864 b, 1871, 1894-95 (1910); Marliére, 1825 a, 1825 b, ms.; Marqués, 1903, 1931; Marquez Miranda, 1943, 1944; Martinez, B. T., 1904, 1917, 1919; Martinez, T. A., 1916; Martinez Orozco, 1938; Martius, 1867; Mason, J. A., 1940, 1945; Massei, 1895; Mata, 1923; Matto de Turner, 1926; Mayntzhusen, 1917, 1919-20; McGee, 1903; McKinney, Medina, and Pefiaranda, 1930; Medina, J. T., 1918, 1930 a, 1930 b; Medina, M., 1919-20, 1920-21; Meillet and Cohen, 1924; Melo, 1927, 1929; Membrejfio, 1897; Mercante, 1924; Merizalde del Carmen, 1921; Métraux, 1929, 1930, 1936 a, 1936 b, 1940, 1942 a, 1942 b; Middendorf, 1890, 1891, 1892; Milanesio, Domenico, 1898, 1917; Milanesio, Domingo, 1918; Missao Salesiana, 1908; Mitre, 1894, 1909-10; Mochi, 1902-03; Moesbach, 1929-31, 1936; Mollien, 1824 (1924); Monoyer, 1905; Montesinos, 1920; Montolieu, 1882, 1895; Moore, 1878; Morales, 1929; Mordini, 1935; Moreira- Pinto, 1894; Moreno-Mora, 1922, 1934; Morinigo, 1931, 1935; Mosquera, 1866; Mossi, 1916; Mostajo, 1923; Miiller, Franz, 1934-35; Miiller, Frederich, 1882; Murdock, 1936; Murrieta, 1936. Narvaez, 1944; Navarro, 1903; Nimuendajd, 1914 a, 1914 b, 1914 c, 1914 d, 1915, 1923-24, 1924, 1925, 1926, 1928-29, 1929 a, 1929 b, 1931, 1931-82, 1932 a, 1932 b, 1946; Nimuendajii and Valle Bentes, 1923; Nino, 1917; Nordenskidld, 1905, 1910 a, 1910 b, 1911 a, 1911 b (1923), 1912, 1915 a, 1915 b, 1922, 1924 b, 1928, 1928-30, 1929 b, 1932, 19388; Noronha Torrezéo, 1889; Nusser-Asport, 1890, 1897. Oefner, 1942; Oiticica, 1933, 1934; Olea, 1928; Oliveira, C. E. de, 1930; Oliveira, J. F. de, 1913 a, 1913 b; Oliveres, 1934-35; Ollanta, 1878; Oramas, 1909, 1913 a, 1913 b, 1914, 1916, 1918 a, 1918 b, 1920; Orbigny, 1839, 1896; Ore, 1607; Orejuela, 1934; Ortiz, 1937, 1938, 1938-39 a, 1938-39 b, 1940 a, 1940 b, 1940 c, 1941 a, 1941 b, 1942, 1948, 1944, 1946; Orton, 1871; Osculati, 1854; Ossa V., 1938; Osuna, 1921, 1923 a, 1923 b, 1924 a, 1924 b, 1924 c, 1925, 1926 a, 1926 b; Otero, 1938-39; Outes, 1913 a, 1913 b, 1914, 1926, 1926-27 a, 1926-27 b, 1928 a, 1928 b, 1928 c, 1928 d; Outes and Bruch, 1910. Pablo del Santisimo Sacramento, 1933; Pacheco Zegarra, 1878; Palavecino, 1926, 1928, 1930, 1931-33; Panconcelli-Calzia, 1921; Pape, 1935; Paris, 1924; Patrén, 1918; Patr6é6n and Romero, 1923; Paula, 1924; Paula Martins, 1941, 1945 a, 1945 b; Paula-Souza, 1918; Pauly, 1928; Payer, 1906; Paz y Nifio, 1936-37; Paz Soldadn, 1877; Peccorini, 1910; Pelleschi, 1881, 1896; Penard, A. P., 1928-29; Penard, T. E., 1926-27, 1927-28; Penard, T. E., and Penard, A. P., 1926-27; Pefiaranda and Medina, 1923; Perea y Alonso, 1937, 1938 a, 1938 b, 1942; Pérez, 1935; Pérez Guerrero, 1934; Pericot y Garcia, 1936; Perret, 1933; Petrullo, 1939; Pfaff, 1890; Philipson, 1946 a, 1946 b; Pierini, 1908; Pinart, 1887, 1890 a, 1890 b, 1892 b, 1897, 1900; Pinart and Carranza, (eds.), 1890; Pindorama, 1937; Pinell, 1928; Pinto, 1938; Pittier de Fabrega, 1898, 1903, 1907; Pittier de Fabrega and Gagini, 1892; Platzmann, 1882, 1888, 1896, 1901; Pohl, 1832-37; Poindexter, 1930; Polak, 1894; Polo, 1901 (1910); Pombo, 1931; Pompeu Sobrinho, 1919, 1928, 1930, 1931, 1933, 1934, 1935, 1936, 1939; Porter, 1911; Portnoy, 1936; Posnansky, 1915, 316 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 1918, 1931, 1934; Preuss, 1919-27, 1921-23, 1925; Prieto, 1885; Prince, C., 1905; Prince, J. D., 1912, 1913 a, 1913 b; Pulgar Vidal, 1937; Puydt, 1868. Quandt, 1807, 1900. Radwan, 1929; Raimondi, 1862; Raimundo, 1934; Rat, 1897-98; Recalde, 1924, 1937 a, 1937 b; Reich, 1903; Reich und Stegelmann, 1903; Reichel-Dolmatoff, 1944, 1945 a, 1945 b; Reinburg, 1921; Remedi, 1896, 1904; Renault, 1904; Restivo, 1892; Restrepo Canal, 1936; Reye, 1884; Rice, 1930, 1931, 19384; Richter, 1928; Rimbach, 1897; Ripalda, 1923; Rivas, 1944; Rivet, 1905, 1907-08, 1910 a, 1910 b, 1911 a, 1911 b, 1912 a, 1912 b, 1916, 1920 a, 1920 b, 1920 c, 1921 a, 1921 b, 1924 a, 1924 b, 1924 c, 1924 d, 1925 a, 1925 b, 1925 c, 1925 e, 1926 a, 1926 b, 1927 a, 1927 b, 1927 c, 1928, 1929 a, 1929 b, 1930 a, 1930 b, 1934, 1941, 1943, 1943-44; Rivet, Kok, and Tastevin, 1924-25; Rivet and Oppenheim, 1943; Rivet and Reinburg, 1921; Rivet and Tastevin, 1919-24, 1920, 1921, 1924, 1927-29, 1931, 1938-40; Roberts, F. J., and Symes, 1936; Robledo, 1922; Robuchon, 1907; Rocha, 1905; Rochefort, 1658 (1667); Rochereau, 1926-27, 1932; Rodriguez, M. C., 1921; Rodriguez, Z., 1875; Rojas, 1942; Romano and Cattunar, 1916; Romero, C, A., 1931; Romero y Cordero, 1930; Rondén, 1910, 1915; Roquette-Pinto, 1912, 1917, 1935; Rosell, 1916; Rosen, 1904; Rosenblat, 1936; Roth, 1924; Rdéthlisberger, 1883-84; Rowe and Escobar, 1943; Rozo M., 1938; Rudolph, 1909; Ruiz Blanco, 1888 a, 1888 b, 1892; Ruiz de Montoya, 1876 a, 1876 b, 1876 c, 1876 d, 1892; Ruiz Palazuela, 1927; Rydén, 1941. Saavedra, 1931; Sdenz, 1876; Saffray, 1872; Sagot, 1882; Saint-Hilaire, 1830-33, 1847-48, 1851; Sala, R. P. A. M., 1920; Sala, G., 1897, 1905-06; Salas, 1918, 1919, 1924; Salathé, 1931-32; Saldias, 1912; Sampaio, T., 1890, 1912; SAnchez Labrador, 1896, 1910-17; San Rom4n, 1890; Santa Cruz, 1913, 1921, 1922, 1923 a, 1923 b Santa Teresa, 1924; Santo Thomas, 1891; Santos, N. C. dos, 1934 a, 1934 b, Safiudo, 1923; Sapir, 1925; Sapper, 1897, 1901, 1905; Schermair, 1934; Scherzer, 1855; Schmid, 1860, 1912 a, 1912 b; Schmidt, M., 1902, 1903, 1905, 1912, 1914 a, 1914 b, 1917, 1929 a, 1986 a, 1936 b, 1987 a, 1937 b, 1987 c, 19388, 1941, 1942; Schmidt, W., 1905, 1925, 1926; Schomburgk, 1847-48, 1849; Schuller, 1906, 1907 a, 1907 b, 1908, 1911 a, 1911 b, 1912 a, 1912 b, 1912 c, 1912 d, 1913 a, 1913 b, 1913 c, 1913 d, 1916, 1917, 1917-18, 1919-20 a, 1919-20 b, 1921, 1925, 1928, 1930 a, 1930 b, 1930 c, 1933, 1936; Schultz, T., 1816 (1850); Seemann, 1853; Seler, 1885, 1902 a, 1902 b; Selva, 1922; Senna, 1932; Serrano, 1936 a, 1936 b, 1936 c, 1941; Sigifredo, 1942-45; Silveira, A. A., 1921; Silveira, E. da, 1935; Silveira, G., 1938; SimGes da Silva, 1924; Simon, 1887; Simpson, 1940 c; Simson, 1879 a, 1879 b, 1886, 1899; Skinner, 1920; Skottsberg, 1918, 1915; Snethlage, E., 1910, 1982; Snethlage, E. H., 1931, 1932 a, 1932 b, 1936; Socrates, 1892; Solari, 1928; Solis, 1928, 1928; Soliz Rodriguez, 1926; Souza, A., 1916 a, 1916 b, 1920; Souza, C., 1875; Speck, 1924; Spegazzini, 1888; Spix and Martius, 1823-31; Squier, 1852, 1858, 1858; Stahel, 1944; Steere, 1903; Steffen, 1928; Stegelmann, 1903; Steinen, 1886, 1892, 1894, 1895, 1901, 1904, 1906, 1912; Steinthal, 1888; Stiglich, 1908; Stirling, 1938; Stoll, 1884; Storni, 1944; Stout, 1947; Stradelli, 1910, 1929; Strémer, 1932; Studart, 1926; Sudrez, 1930; Sudrez de Cepeda, 1923; Swadesh, 1939; Sympson, 1926. Tagliavini, 1928; Talbet, 1926; Talbot, 1931-32; Tapia, D., 1888; Tascén, 1934; Tasso Yatahy, 1918; Tastevin, 1908, 1919, 1923 a, 1923 b, 1923 c, 1923 d, 1924, 1928 a, 1928 b; Taunay, 1868, 1888; Tauste, 1888; Tavera-Acosta, 1907, 1921-22, 1930; Tavolini, 1856; Taylor, 1946; Tebboth, 1948; Techo, 1673; Tello, 1913 b, 1923, 1931; Teran, 1917; Teschauer 1914, 1921, 1927; Tessmann, 1928, 1929, 19380; Teza, 1868; Thiel, 1882, 1886; Thomas, C., and Swanton, 1911; Thouar, 1891; Tocantins, 1877; Tola Mendoza, 1939; Tonelli, 1926, 1927, 1928; Torres, A. M., 1931; Torres Rubio, 1603, 1619, 1754; Touchaus, 1908; Tranga, 1882; Triana, Vol. 6] LANGUAGES—MASON 317 1907; Trombetti, 1925, 1928; Tschudi, 1853, 1866-69, 1884, 1891; Tulcdn, 1934 a, 1934 b. Uhle, 1890, 1896, 1919, 1931; Up de Graff, 1923; Uribe, 1883; Uribe Angel, 1885; Uricoechea, 1854, 1871, 1877; Urteaga, 1895. Vacas Galindo, 1895, 1903 a, 1903 b; Vaisse, Hoyos, and Echeverria y Reyes, 1895; Valedrcel, 1933; Valdivia, 1607 a (1894, 1940), 1607 b (1918), 1887 (1606), 1897; Valenzuela, 1918-19; Valle Cabral, 1880; Vallejo, 1910; Vara Cadillo, 1931, 1937; Vasquez, 1902; Vazquez, 1921-24; Veigl, 1785 a, 1785 b; Velasco, 1840; Velazco Aragén, 1923; Velazquez, 1916; Vellard, 1934-35, 1937; Vellard and Osuna, 1934; Verneau and Rivet, 1912-22; Vianna, 1928; Vignati, 1931 d; Villareal, 1921; Villavicencio, 1858; Vifiaza, 1892; Vocabulario comparado . . ., 1892; Voegelin, 1941; Vogt, P. F., 1902-03, 1904. Wallace, 1853; Wassén, 1933, 1934 a, 1934 b, 1935, 1937, 1938; Wechsler, 1917; Wegner, 1934 a, 1934 b, 1934 c; Whiffen, 1913, 1915; White, 1884 a, 1884 b; Wied-Neuwied, 1820-21; Wiener, 1880; Wilczynski, 1888; Williams, J., 1924, 1928-29, 1932; Wissler, 1917 (1922, 1938); Wolfe, 1924. Yangues, 1888; Young, T., 1842. Zeballos, 1915, 1922 a, 1922 b; Zeledén, 1918; Zidek, 1894. we ir hi ats f ‘ bi bey Find ao ita Hin | ~ , b , ‘ tia? 4 . ‘ 1 ‘ 7 { t \ ' ' +) i - a : > e if 5 i) ; my \ oy r ‘ 4 sf } a” ‘ J Dhy PART 4. GEOGRAPHY AND PLANT AND ANIMAL RESOURCES GEOGRAPHY OF SOUTH AMERICA By Cart O. SAvER RELIEF In major design the two continents of the New World have certain striking resemblances. (For South America, see maps 7 and 8.) Both are triangular, broad at the north, ending in an acute tip at the south. Both have their great mountains at the west, fringing the Pacific Ocean; and hence this ocean receives only a minor part of the continental drainage. Both have wide, ancient highlands at the east, largely reduced by long-continued erosion to hill lands and rolling uplands. In both cases, between eastern uplands and western moun- tains, lie great plains, now or in the past subject to heavy aggradation from the waste of the elevated lands both to the east and west. A broad resemblance may be pointed out between the continental posi- tion of the Mississippi Basin and that of the La Plata, perhaps even between the Great Lakes-St. Lawrence and the Amazon Basins. Thus far, an almost identical schematic diagram could be drawn for the two continents, suggesting basically similar dynamics of crustal modeling. THE EASTERN HIGHLANDS From the mouth of the Orinoco River almost to the estuary of the Rio de la Plata an ancient land mass fronts the Atlantic Ocean, inter- rupted only by the trough of the lower Amazon River. The Amazon is no ordinary river valley but a great, and partly ancient, structural depression, which has long separated the old land mass of Brazil to the southeast from that of Guiana on the north. Together Brazil and Guiana form the core of the South American continent, a stiff fundament of ancient igneous and metamorphic rocks, partially covered by sedimentary beds, mostly of Paleozoic age. The parallel to the North American Laurentian Shield, the Old Appalachians, and the Ohio Basin cover of Paleozoic sedementary rocks is apparent. The Guiana Highlands.—These form the land surface between the Amazon-Yapura and the Orinoco Rivers. It is to be noted that they 319 320 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 occupy all but a narrow coastal stretch of the colonies that bear their name and also include most of Brazil north of the Amazon River, as well as the southern half of Venezuela and a large part of eastern Co- lombia. In length they are about equal to the North American Ap- palachian area (Alabama to Gaspé Peninsula) ; in width they surpass it considerably. The greater part of this surface is underlain by igneous and meta- morphic rocks, presumably ancient granites and gneisses, with in- folded bodies of schists and quartzites, said usually to lie in east-west strips. The quartzites commonly stand out as ridge makers. The watershed between the lower Amazon River, on the one hand, and the Essequibo River and streams of the northeast, on the other, constitutes one of the highest and most dissected parts of the ancient rock mass, culminating in the Sierra of Acarai, around 5,000 feet (1,500 m.). Between the Amazon (Rio Negro) and Orinoco Rivers the watersheds are in part indistinct. Yet about the Orinoco River, in southern Vene- zuela, lie numerous detached mountain chains, supposedly of eleva- tions to 6,000 feet (2,000 m.) and more and formed of granitic and eneissic rocks. Mostly the granites and metamorphic rocks constitute a rolling upland cut by many streams into an intricate and in part irregular pattern of hills and valleys. The rocks are deeply weathered and the residual soils, often brick red, are too greatly leached to be fertile. Beds of sedimentary rock, especially sandstones and conglomerates, once overlay much if not all of the Guiana region, but have been largely stripped off by erosion. There remain three major areas of sedimen- tary cap: (1) The southern margin has fringing slopes down to the Amazon, of a large variety of beds. (2) In the middle of the highland lies a great block of sandstones and conglomerates, culminating around 8,000 feet (2,600 m.) in the Sierra of Roraima and westward in the Sierra of Pacaraima. These beds form sheer cliffs enclosing table- lands, deeply trenched by streams. (38) In the far west, almost against the foot of the Andes, the mesas of Pardaos and of Iguaje separate the Llanos of the Orinoco River from the alluvial lowlands of the Amazon River. None of these uplands of sedimentary rocks is suitable for agriculture. Throughout the Guiana Highlands, it appears that agricultural advantage is limited to the valley bottoms and that these are not of remarkable fertility. The Brazilian Highland.—The Brazilian Highland, or upland as it might better be named, is traced in outline as to its interior limits by stream courses, not to be thought of as marking an abrupt change from plateau to lowland. However, the northwestern limit is quite well indicated by the lower Amazon River, west to its junction with the Madeira River, then by following the latter upstream to its junc- Vol. 6] GEOGRAPHY—SAUER BPA tion with the Beni River. Here is the most interior end of the Brazilian uplands. Continuing thence, the boundary swings sharply south- eastward, along the eastern side of the lower Mamoré River, and following the entire length of the Guaporé River to the interlacing of its headwaters with those of the Paraguay River. The rest of the land boundary is parallel to, and at varying distances to the east of, the course of the Paraguay-Parand, south to lat. 30° S. Here the bounding escarpment runs directly eastward as the Serra Geral, across the State of Rio Grande do Sul to the sea in the vicinity of Porto Alegre. The Highland fills not only the greater part of the political frame of Brazil, but also includes the Misiones Territory of the Argentine, the southeastern part of Paraguay, and may even be considered as entering the easternmost tip of Bolivia. This, the greatest geologic block of South America, resembles the Guiana Highlands in materials and surface, with the difference that in the Brazilian Highland, land surfaces derived from sedimentary rocks (sandstones and conglomerates) predominate over those formed by the degradation of granites, gneisses, schists, and metamorphic rocks. The ancient crystalline and metamorphic fundament of Brazil is widely exposed in three areas: (1) The eastern border of the High- land against the Atlantic; (2) the greater part of the State of Goias, where the drainage of the upper Tocantins River has uncovered the basement rocks over a large area; and (3) the northwest flank, sloping down to the Amazon and Madeira Valleys. A lesser fourth region might be added, about the headwaters of the Paraguay and Guaporé Rivers, below the scarp and Plateau of Mato Grosso, and situated in the State of Mato Grosso and in eastern Bolivia. Of these areas, all but the Goids one are marginal to the Highland; all appear to be un- distinguished as to relief, and are sometimes referred to as somewhat dissected peneplains. The soils are reddish, and where derived from schistose rock are likely to be tough and compact and of low fertility. The eastern area of the basement rocks may be divided into three subregions: The ‘“‘shoulder’”’ of Brazil, comprising the six small States of the northeast, behind Cape Sao Roque; the interior basin of the Sao Francisco; and the Serra do Mar from Bahia to Santa Catarina. The physical characteristics of these three subregions are as follows: (1) The northeastern part of Brazil is a plateau or peneplain of crystalline rocks attaining, behind Recife, maximum heights of about 3,000 feet (1,000 m.) in the highlands of Borborema. Here and there a sharp ridge of quartzite rises, as in the Serra da Itabaiana, straddling the lower Séo Francisco River. The descent from highland to coast is in general gradual, and the coast area itself has low slopes developed on various sedimentary rocks. 322 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 (2) The basin of the Sao Francisco, in the western part of the State of Bahia, is considered a peneplain, over which the river wanders at a very low gradient. Only upon reaching the State of Pernambuco, on its swing to the east, does its course accelerate. Here at the western boundary of the State of Alagoas, are the granite canyon and falls of Paulo Affonso. A short distance below, the river is at grade, from Piranhas to the sea. (3) The Serra do Mar, of granites and gneisses, forms an increasing- ly impressive background to the coast as one goes south from Bahia. A short distance behind Rio de Janeiro it attains a height of 7,000 feet (2,200 m.), and it continues as an unbroken front range to its termination in Rio Grande do Sul. Excepting the area from Santos to Rio de Janeiro, fan slopes and even coastal terraces form a narrow but important corridor between serra and sea; here mountain streams deposit their alluvium in flood plains and deltas. Behind the Serra do Mar at the city of Rio de Janeiro, the Parahyba River (of the south) drains a structural basin, which is parallel to the coast and runs the entire length of the State of Rio de Janeiro. Inland from the Serra do Mar, and stretching north from So Paulo into the great bend of the Sao Francisco River, is a highland of most complicated structure and relief. This is known in the north (Bahia) as the Chapada Diamantina, in the center (Minas Gerais) as the Serra do Espinhago, and at the south (behind the cities of Rio de Janeiro and Séo Paulo) as the Serra da Mantiqueira. Ancient folding and faulting has thrown a great series of rocks, ranging from old crystalline and metamorphic members of the Brazilian Shield to Paleozoic sedi- mentary rocks, into parallel strips, trending on the whole parallel to the coast line. The result of its erosion is a relief somewhat reminiscent of our folded Appalachians, parallel valleys in the weaker rocks, master streams that cut directly across the structural lines in gorges, with cliff-fronted, smooth-topped interfluves on resistant rocks. A be- wildering variety of highly metamorphosed sandstones (quartzites and itacolumites) provides the principal resistant, ridge-forming beds of this area. At the south and southwest, that is, in the States of Santa Catarina, Parané, and parts of Sao Paulo, the surface is formed by beds of basalt and red sandstone, of relatively late age, spread out over the older rocks of the plateau. This area is sometimes called the Plateau of Parané. The volcanic materials in general weather into deep, red, productive soils, which also contribute to the fertility of alluvial soils derived therefrom. Where basalts cap, or are interbedded with shales and weak sandstones, high cliffs develop, as along the edge of the Plateau and along the stream courses (e. g., the valley of the Uruguay River). a eee! (ozg *d 008g) OS—TTLPGL “ur 000'% “tH 000'T ‘i 00S “ul 002 ‘UI 000'F PUB “Ut O00'S SUMOYS OIE SINOJUOD BATSSBDONIS SUIMOT[OJ OU], “‘go11oUTy YIog jo deur jorpoy—'L IVIN eae ite Va bad ea oh Fee Beng Pi a ed vr ee 4 . os as. oy 1 ae fs: ; v | fe a | Fakes aa ae Pe , Abi rant ; a4 rh ; oe *" r Pyar i | are , . 7 . i] ; . | ae ae “ar Q00‘F PUB “UI 000'E “ar 90g “Ul 00% suMOYs eB $1N0}U0D SAISSEDONS SUIMOTIOJ OYT, “BOeUTY YRNOG jo deur jorey—'2 av (oze "4 oo8q) O9—TTLFBL “ar 000'% “™ 000'T 40 BULLETIN 148, VOL. 6, MAP 8 30 BUREAU OF AMERICAN ETHNOLOGY BULLETIN 143, VOL. 6, MAP 8 40 | ANCIENT BRAZILIAN-GUIANA HIGHLANDS A. Exposed crystalline basement complex (Pre-Cambrian igneous and meta-sediments) B. Folded and faulted indurated sedimentary covering C. Exposed surface of old lava flows (Triassic) — Approximate extent of Parana lava field AddLtLiL Major cuestas " AAA Major scarp of Brazilian Hightand (Serra do Mar) . ee i), ANCIENT PATAGONIAN PLATEAU A. Exposed crystalline igneous and meta-sediments B. Soft sedimentary covering (Tertiary) Uo Il] FOLDED AND FAULTED MOUNTAINS OF ANDEAN AND ANTILLEAN OROGENIES A. High mountain areas of folded and faulted sediments Areas of volcanic extrusives ai]. Major volcanoes C. Intermontane basin plains 1. Undissected surfaces 40+ i | a, Playas 2. Dissected surfaces GEOGNOSTIC MAP OF SOUTH AMERICA / Folded and faulted ancient crystalline rock (mainly Pacific coast ranges) E, Sediments covering area of ancient folded und faulted crystalline rock os BNNHER F. Basin and range structure in faulted ancient igneous rock (Pampean Sierras); solid lines indicate trend of principal ranges; dotted lines indicate margin of interior basins T 1V. LOWLAND PLAINS A. Undissected coastal and upland plains B. Slightly dissected plains surfaces C. Plain surfaces subject to widespread and frequent inundation BULLETIN 143, VOL. 6, MAP 9 i ee te ie? i. ao (pe ee: 4 Cy en a) yi r40 \ A TROPICAL RAIN CLIMATES DRY CLIMATES c MESOTHERMAL CLIMATES E POLAR CLIMATES A Tropical Rain Forest Climates am —[ | Topical «Monsoon Rain Forest Climates Tropical Savanna Climates, winter dry season Tropical Savanna Climates, summer dry season Hot Steppe Climates Cold Steppe Climates Deserts with winter cold n= foggy 8Whn Rag Warm Desert, fouay Hot Summer Mesothermal avanna,winterdry season Cwb— | Cool Summer Mesothermai Ly * Savanna,winter dryseason Mediterranean” Clim: summer dry season Humid Mesothermal Climate, hot summers o- 3 WL. Humid Mesothermal Climate, cool summers (Mainly Tundra) Cle — Humid Mesothermal Cli mate, very cool summers ZA Drought Ares of NE. Brazil A Areas subject to severe drought and flood ca- lamities, according to F.W. Freise \ Cool Summer Mesothermal \ ate \ BULLETIN 143, VOL. 6, MAP 9 CLIMATES OF (Letter symbols are according to the Képpen system) SOUTH AMERICA / 2 ; / 110 = 100 90 i ee Jr Ll i im a wah * weer yt eaiagit e-- tA pas : estamif fee104 2 ‘ zafemdo ennewae tenaqorT hiielie — wh ee woresr vb 1oithiw y SITAMLD \ t f Pee ajemnilD comeve? laargorl ’ ; , aie DA <_-s . pa * Pa nOoeese veo Wren kee re . rater} enrearc tor ats ) sausi2 bled | ~— #28 res Yao wiwe ZITAMLID Sten ee | 1 6 mage nit sae oan ee ming ete a - : | . / : 7 Ay 4 hay ; > : ; é My = | , Km } iB 4 ' 4 : a gay . : rn ‘ si 7 - (A vee: ee vee + ——— _ ae ~ ee; a *6s3e ee ¥ 2 ) 2A ee Bristle rat (Proechimys sp.), 200 to 300 gm., eastern Brazil. Top, Dourado (Salminus sp.), Rio 8 kg., (Bottom, right, courtesy Joao Moojen de Oliveira.) PLATE 41,—Rodents and fish. Top, left: right: Agouti (Dasyprocta sp.), 2 to 3 kg., eastern Brazil. Bottom, left: Parana, Brazil. Bottom, right: Plains viscacha (Lagostomus maximus), 5 to National Zoological Park, Washington; others, courtesy PLATE 42.—Mammals, reptile, and bird. Top, left: Nine-banded armadillo (Dasypus novemcinctos), 2 to 3 kg., Andpolis, southern Goias, Brazil. Top, right: Teju lizard (Tupinambis teqguixim), Andpolis, south- ern Goias, Brazil. Length, 1 m., Wt.1.5 kg. Center, left: Maned fox (Chrysocyon brachyurus), Anaépolis, southern Goias, Brazil. Center, right: “‘Kinkajou,”’ or jupara (Potos flavus), 3 to 5 kg., eastern Brazil. Bottom, left: Porcupine (Coendou prehensiliz), 3 to 5 kg., Anapolis, southern Goias, Brazil. Bottom, right: Tinamou (Crypturellus undulatus), Anapolis, southern Goias, Brazil. (Top, left and right, and bottom, right, courtesy R. M. Gilmore: center, left, and bottom, left, courtesy Clark Yeager; center, right, courtesy Joao Moojen de Oliveira.) Puate 43.—Llama and alpacas. Top, right: Llama (Lama glama) ; left: Alpaca (Lama pacos). Bottom: Alpaca. (Courtesy New York Zoological Society.) Puate 44.—Vicufia and guanaco. Top: Female vicufia (Lama, or Vicugna vicugna) in Zoological Park, Sucre, Bolivia. Bottom: Guanaco (Lama quanicoe). (Top, courtesy R. M. Gilmore; bottom, courtesy New York Zoological Society.) PLATE 45.—Birds. Top, left: Trumpeter (Psophia leucoptera), Riberalta, northeast Bolivia. Top, right: Young rhea (Rhea americana), Yacuiba, southeast Bolivia. Center: Adult rhea, tame, in Indian village, near Trinidad, eastern Bolivia. Bottom: Domesticated Muscovy ducks (Cairina moschata), Trujillo, Pera. (Courtesy R. M. Gilmore.) PLATE 46.—Birds. Top: Flightless steamer duck (Tachyeres pteneres), Hermite Island, southern Chile. Center, left: Condor (Vultur gryphus). Center, right: Razor-billed curassow (Mitu mitu), Madidi River, northern Bolivia. Bottom: Chachalaca, or guan (Penelope superciliaris), eastern Brazil. (Top, courtesy Rollo H. Beck and American Museum of Natural History; center, left, courtesy National Zoological Park, Washington; center, right, courtesy Sulo Sihvonen; bottom, courtesy Jodo Moojen de Oliveira.) tiie. PuatE 47.—Turtle, fishes, and snake. Top, left: Common tortoise (Testudo tabu- lata), 15 kg., Andpolis, southern Goias, Brazil. Top, right: Piranha (Serrasal- mus sp.?) and piranha wound on man’s foot, [Ilha de Marajé, Brazil. Bottom, left: Anaconda (Hunectes murinus), Amazonas, Brazil. Bottom, right: Piraruct (Arapaima gigas), Paraguay River (?), Brazil. (Top, left and right, courtesy R. M. Gilmore.) Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 435 (Rusconi, 1930). In the early Recent and late Pleistocene, animals, claimed to be indistinguishable on cranial characters from llamas, ranged in the nonglaciated intermontane valleys of the Andes (Tarija and Ulloma, Bolivia) and the eastern lowlands of Argentina and Patagonia (Lopez Aranguren, 1930; Cabrera, A., 1931). Description (pl. 43).—Size: Averaging slightly larger than the guanaco; total length, 175-225 cm.; height at shoulder, 90-125 cm.; weight, 75-125 kg.; smaller breeds along Peruvian Coast in Inca times, and at present in Ecuador. (See p. 437.) Color: Variable; generally black, brown, or white, or mixtures of these; some of “guanaco” color (light reddish brown dorsally and laterally, white ventrally and medially, and smoky gray brown on head); some spotted and even “‘maltese.”” Pelage: Heavy and long on body, slightly shorter on neck, very short on lower legs and head (abrupt change in length from body hair), almost bare on groin and axillae (sometimes depilated on rump and withers, because of cargo abrasion ?); composed of heavy, long, matty, slightly greasy wool hairs, 10-20 cm. if sheared annually, as long as 50 cm. (or 20 in.) if not sheared; wool hairs easily distinguished from overhairs (beard or guard hairs) scattered through the wool, which sometimes, in case of young animals, are longer than wool hairs and form a ridge on back of neck; no complete annual molt, but hairs continually growing and shedding like mane and tail of horse, though some individuals shed more hair in summer; at shearing time a tuft, “hualiaui,” is often left on throat as protection against exposure (Stroock, 1937 a). Juvenile pelage, up to 4 or 6 months, of a fine, silky, short hair (3 to 10 cm., or 1 to 4 in.) without apparent dimorphism into wool and overhair. Feet: Large for group; fore larger than hind; each hoof with a small ridge along top, giving it a clawlike aspect. Metatarsal glands: Pair on each metatarsum (between hock and toes, on hind shank); clearly visible (uncovered by hair) and variable in shape and size, but generally oval and outer larger than inner. Tail: About 25 cm. (10 in.), heavily haired, and characteristically carried in an arc up, out, and down. Lars: About 15 em. (6 in.), sparsely haired, and generally turned in at tips or slightly scalloped subterminally on inner edge to give turned-in appearance. Callouses: Absent except when produced on chest and legs by much kneeling in hard corrals. Skull: “Normal,” like guanaco, though probably even more variable, especially in rostral length; large; facial portion in front of orbits usually longer by 10-15 mm. than cranial portion behind anteorbital border to rear of condyle; facial pits large; teeth large, P 3-4/4 and heel of M/3 relatively large, but endostyle of M/3 usually small; choanae variable, but usually narrow V, open V, or nar- row U (rarely an open U); frontal profile slightly concave (greatly sunken in an Inca breed) ; lower incisors spatulate, rooted, imbricate (overlapping laterally), and with enamel on lingual and labial faces; symphysis of mandible long and narrow (as 2:1); submental foramina slightly anterior to menton (rear of symphy- sis). Reproduction and growth: Rut in summer; gestation 11 months (Mac- cagno, 1932, p. 45); one or rarely two young born the following summer; precocious, can rise and run in few hours; wean in 4 to 8 months; mature in 2 to 3 years; live 30 to 35 years (Romero, E. C., 1927), but in foreign zoos, 12 years normally, and as long as 20 (Flower, 1931, p. 223). Rutting probably not so sharply defined as in guanaco, and longevity probably longer as a usual result of domestication. Hybridization: Can be crossed with guanaco, alpaca, and vicuiia, but only com- mon hybrid is llama X alpaca, producing a huarizo or mixti, which is fertile; natural hybridization with guanaco probably rare, though producing fertile off- spring (MacDonagh, 1940), and with vicufia more infrequent if at all, and off- spring probably sterile. For discussion of huarizo and mixti, see page 445. 794711—50——29 436 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Intelligence: Low. Disposition: Generally calm, stolid, and ‘dignified’? unless annoyed, then characteristically spray-spitting (‘“‘cough spitting’’); often obstinate and sometimes panicky; curiosity pronounced; playful and affectionate when young. Habits: Greatly subordinated to man’s activities and control; naturally herbivorous, gregarious, polygamous, with male dominance and dependence of female and young on male leadership; diurnal and nocturnal, and preferring high altitude and treeless, arid country; voiding dung and urine in communal spots (source of ‘“‘taquia’’ fuel), and apparently often stimulated by sight of prior voider. Diseases: Liver flukes, and especially mange, which often sweeps through herds in epidemics causing many deaths and more morbidity; probably due to the llama itch mite (Sarcoptes scabiet aucheniae) and not to the llama mallophagan louse (Trichodectes breviceps). The belief that llamas have syphilis and trans- mitted it to the Inca herders through contact (zoophily ?), and hence to the Spaniards, who then brought it to Europe, is without foundation; no llama has been found to harbor Treponema (syn. Spirocheta) pallidum. Voice: Intermediate between a bleat and a moan (Cabrera and Yepes, 1940, p. 261); generally silent. Special physiology: Haemoglobin with high affinity for oxygen, red blood cells high (12.11 million per cu. mm.; Dill, 1938, pp. 127, 131), and probably a high efficiency in converting harsh and dry vegetation into carbohydrates for energy and metabolic water. Breeds: Several with vague characters; more in pre-Colum- bian times than today, one with five front toes (p. 487). Remarks.—The llama apparently intergrades in all its morphology with the guanaco, though no one individual may show all intergrada- tions, and they can be interbred freely, producing fertile offspring (MacDonagh, 1940). This supports the conclusion that the llama was derived from an ancient stock of wild guanaco. The much more restricted range of the llama (than that of the guanaco) can be ex- plained on cultural rather than biologic grounds, as can be explained the failure of the llama to become established where it has been introduced into foreign countries (United States, Australia, Argentina, etc.). It breeds in North American zoologic parks. Even the evidence that the Chimac and the Chincha of Coastal Peri had Highland breeding grounds for their llamas (Estruch, 1943, p. 118) can be explained by the absence of artificial (and natural) pastures and cultivated forage crops along the coast. However, herds of llamas existed on the coast of Pert in pre- Columbian times, as is shown by the many archeological remains and the numerous representations on pottery. (See Breeds, p. 437.) Maccagno (1932, p. 43) stated, without giving reasons for the custom, that llamas were brought from the Highlands to the Coast carrying human cadavers and were buried with their burdens; and that the Coastal climate was (ultimately ?) fatal to llamas. Pocock (1923), Lopez Aranguren (1930), and especially Cabrera (1931) contended that the llama was a distinct species, and the latter two authors identified fossils from Pleistocene deposits of southern South America as those of the llama. If these contentions are true, the llama can be considered a distinct species which became progressively a relict Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 437 in the postglacial period, confined to the high mountains and weak in competition, and was saved from extinction by the fortunate com- bination of its zoologic status (?) and geographic range contiguous in time and space with a human culture which was well developed enough to domesticate the species. All the facts, however, do not fit any theory. Origin.— On known distributional and archeologic evidence, the orig- inal domestication of the llama was made in or on the margins of the Central Highland area of southern Peri, Bolivia, North Chile, or Northwest Argentina. Latcham (1936 a, p. 611) localized domesti- cation in the Highland around Lake Titicaca, and by the Atacamefio before this tribe was driven south to North Chile by the bearers of Tiahuanaco culture. Breeds.—There seem to be several vague breeds of llamas today: (1) Common llama; (2) large burden-bearing Ilama (?) of the Lake Titicaca region, and (3) small llama (?) of Riobamba, Ecuador. Definite data on these breeds are absent. E. C. Romero (1927) stated that there are two breeds in Northwest Argentina, ‘‘brachy- morphics” and “‘dolichomorphics,”’ but he intimated later in the same paper that the latter are llama-alpaca hybrids (although one would suppose that alpaca blood would give a brachy effect). In pre-Columbian times, in the Highlands and on the Coast of Pert, there also existed several breeds, some as indefinite as those today: (1) Small llama on the Coast (definitely not an alpaca; speci- mens seen from Pachacamac, south of Lima); (2) normal-sized llama with sunken forehead from the Coast (specimens from Pachacamac) ; (3) large burden-bearing llama (?) of the Highlands, utilized especially by the Jnca army; and (4) aberrant llama with five front toes from Chancay, Central Coastal Pert. One complete skeleton and two baked-clay figurines of the five-toed llama were examined in the Museo Nacional de Antropologia, Lima, Pert, through the courtesy of Dr. Tello, the director, who has kindly given permission to describe the animal here. The extra front toes were the 1st, 2d, and 3d; the 3d and 2d arose from pro- gressively shorter additional articulation facets on the medial side of an otherwise normal metacarpal (front metapodial, or cannon-bone); and the Ist toe was represented by a short, closely appressed ‘“‘dew-claw”’ situated above the 2d on the inside of the leg, and seen on the figurines only. The color of the figurines was black above and white below; the color of the skin fragments found with the skeleton was, according to Dr. Tello, dark brown above and white below. This breed was undoubtedly a rare mutation, possibly saved for religious purposes or mere curiosity. Strangely, this mutation does not recapitulate phylogeny which is clearly shown, by the long Nearctic Tertiary fossil record, to have involved a loss of digits 1, 2, and 5, leaving 3 and 4 existing today in normal lamoids. The “‘chilihueque” of North Chile was probably the common breed 438 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 of llama, and was employed as a water carrier (Cabrera and Yepes, 1940, p. 263). Such southern range indicates diffusion by the Inca. Marelli (1931, p. 54) listed ‘‘chilihueque” under alpaca. Aboriginal use.—In aboriginal times in the middle Andean region, the llama was very important ethnozoologically. It was used as a beast of burden, as food, sacrifice in religious ceremonies, and as a source of fuel (dung, or taqufa), medicine (bezoar stones), wool, and hide. The animal thereby constituted a prime source of wealth, both for the individual and for the state (Inca). The success of the Inca conquests undoubtedly was partially dependent upon the llama as a burden-bearing and emergency food animal. The llama was never ridden, milked, or used for wheeled drayage. (The wheel was unknown in pre-Columbian America.) The llama can be mounted, but it is probably too small for continued riding. No animal was milked by American aborigines, and even after the Span- iards introduced milking with cows, no extension of the custom was made to llamas (nor to alpacas). There are hints that the llama may have been used for plowing or travois drayage, though such is doubtful. Romero (1927, p. 63) denied an assertion of ‘Teodoro Bry”’ on plowing, and Bennett’s record (Handbook, vol. 2, p. 607) of “toggles for lama harnesses’? may indicate use for burden bearing rather than some kind of draft. : Today, in high Peri and Bolivia, the llama is still very common, and though it is generally inferior to, and has been replaced partially by the burro and mule as a pack animal and the sheep and alpaca as a wool producer, it has remained numerous and important, probably because of (1) its combination as a burden-bearing, wool, meat, and dung-fuel animal, (2) its tolerance to altitudes probably higher than mules, burros, or horses can endure, and (3) the conservatism of the Indian. Maccagno (1932, p. 50) stated that the llama will disappear if not improved in quality and husbanded with care. Cargo transportation.—Today llamas are used mostly for transpor- tation of cargo, and probably were likewise used in pre-Columbian times. They carry agricultural produce from Highland farms to cen- ters of population and return with other goods, principally salt or coca. Sometimes large caravans make long trips. In Colonial times, 350,000 llamas were used for transporting metals alone, mostly silver, from remote mines (Maccagno, 19382, p. 49). The cargo appears to weigh from 10 to 50 kg., or 22 to 130 lbs. (45 to 80 kg., or 100 to 175 lbs., Maccagno, 1932, p. 28; but later he stated, p. 48, 25 to 60 kg., or 55 to 132 Ibs.), depending on the size and strength of the individual llama and the exigencies of the moment. Overloading is said to cause the animal to lie down and refuse to rise until relieved of some weight. Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 439 The daily travel is from 10 to 30 km., or 6 to 18 miles; the pace is slow, and the group is as often bunched as in file. An entire journey may be several hundred km. Maccagno (1932, pp. 28, 48) said that llamas travel 6 to 7 leagues (30 to 35 km. ?), can march 20 days steadily, and go 5 days without food (and water ?). The degree of thirst tolerance is not known, but it is believed to be high. Romero (1927, p. 44) stated as a proved fact that llamas can pass (work ?) 3 or 4 days without drinking. Water probably is derived, during forced abstinence, from carbohydrates and subcutaneous fat, not from a supply in the so-called “‘water-cells” of the rumen. There is usually one herder for every 15 to 20 animals. No frame packsaddles are used, but only a blanket, and the load is tied with plain rope without a cinch girth of canvas. The heavy wool on the back is not sheared but allowed to grow long, presumably for easing the burden. Llamas can find food even in the most desolate-appearing country, and grazing is done on the march; or at night when they are often turned loose. Stabling is done in stone corrals, or by securing a group in a circle with a neck rope, heads inward and kneeling. Also, it is stated that a llama will not pass a single-strand rope fence (Maccagno, 1932, p. 49). Wool and weaving.—The wool of a llama is strong, though greasy and relatively coarse and with a considerable amount of stiff guard hairs. It is usually plucked from the llamas by hand while on the march, and spun on hand spindles by the walking women soon after- ward. In some places, flocks seem to be sheared regularly every 2 years (to get a greater length of fiber than with annual shearing), beginning at the second year (Cabrera and Yepes, 1940, p. 263). Yield from one animal is 1,800 to 3,500 gm. (4 to 8 lbs.) and at the end of the second year the fibers reach a length of 30 cm., or 12 inches (Maccagno, 1932, p. 28). Part of the required labor of indentured Aymara and Quechua Indians on haciendas is the weaving of blankets and rugs on large wooden looms from llama wool supplied by the Indians. In Jnca times llama wool was used extensively by the com- mon people for fabrics; the better wool from alpaca and vicufia was reserved for the upper classes (Maccagno, 1932, p. 47). Meat.—Llama meat is said to be tasty, especially when from a young animal, but apparently it is little utilized today; perhaps sheep are too plentiful. In Inca times, probably the principal source of meat was that of llama, and it was often dried to form “charqui.” Maccagno (1982, p. 47) stated that fresh entrails and blood were relished, and that herds were driven to the Coast for slaughter and were taken with the armies to supply meat; also that the Spaniards 440 SOUTH AMERICAN INDIANS [B. A. B. Bull. 148 relished the brain, and slaughtered thousands from special herds which were kept for this purpose (op. cit., p. 45). Hides.—The use of llama hides is not indicated except for sandals in Jnea times (op. cit., p. 48); but it seems that the hide may have been used more extensively, even though there is no indication that the tanning art was developed. Dung fuel (taquia).—Llama dung is a common fuel on the treeless altiplano of Bolivia and South Pert. The bolus is small, about 2 to 3 cm. (% to 1 in.) in diameter, like that of a sheep, but it is deposited in communal voiding places and is thus gathered easily. When the llamas are on the march or billeted for the night, the act of defecation of one animal seems to be stimulated by that of another. The alpaca, guanaco, and vicufia have the same habit. This use of llama dung for fuel may be a relatively recent custom and may be correlated with the deforestation and general vegetational depletion of the Highland area in the past several millennia. Today the resinous tola bush (Compositae) also is used in some places for fuel, and has become scarce or extinct in easily accessible areas. Husbandry.—-Today llama husbandry is entirely in the hands of the Indians, both Aymara and Quechua, and the technique is usually simple. Many Indians have their own herds, which are kept in corrals or allowed to graze under the eye of a watchful child. Breeding is indiscriminate, without selection, and many alpaca-llama hybrids seem to be present. Males and females are used for burden bearing, and castration is not practiced to any extent. However, some herds are better managed (probably those of the larger haciendas and the property of the haciendado), and Maccagno stated (1932, pp. 49-50) that herds are divided into ‘‘extremas” (8 to 24 months), ‘‘ancutas” (2 to 3 years), “‘puntas de anachos”’ (breeding males over 3 years), “‘puntas’’ of burden-bearing llamas (males over 3 years), “puntas” of females with young, and “puntas” of females without young. For each 100 females, 2 to 10 males are designated. The females breed at 3 years of age. Castration is not generally practiced, though it was commoner in Colonial times for burden- bearing males. Tschudi (1844-46, p. 259) stated that castration was practiced by the Indians to obtain a savory meat and long wool. In Inca times llama husbandry was a complex and honored pro- fession. Maccagno (1932, quoting earlier sources) stated that the herds belonged to the Dynasty (the Inca Emperor or his family), the Sun, the Temples, and the Huacas (burial grounds): On certain occasions, particularly after a successful [military] campaign, gifts were made among the Kuracas who received 1,000, 500, 100, 50, 20, and 10 llamas each [according to rank], and some Indians were given a pair. The mayordomos (“l’ama kamayox’’) of the Dynasty of Sun herds (‘“‘kapax Vamakuna,”’ or ‘‘kapax l’ama’’) were generally men of distinction, and sometimes Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 441 princes of royal blood. They had under their orders many shepherds (‘‘l’ama nitsix’’) who directly attended the herds. By means of quipus they counted exactly the number of llamas in each herd, and each herd had its corresponding color on the quipu. [Maccagno, 19382, p. 44.] He also stated that the division of Znca herds was as follows: Young, 4 to 12 months (“ufiakuna’’); young, 1 to 2 [3?] years (“‘malta una”); breeding males over 3 years, according to colors (“apukura’’); cargo males, according to colors (“‘wacaywa”’); breeding females, by color (“tsina’”’); and sterile adults (females only ?), by color (‘‘komi’’); also, that castration was not known. Religious ceremony.—The llama was very important in Jnca reli- gious ceremony. The animals for sacrifice were taken from the Sun or Huaca (‘‘Waka’’) herds, according to the object of the sacrifice. . . . [The kind] most appreciated by the Peruvian Incas were the pure black llamas. . . . Contrary, the Kollas [Aymara- speaking Indians of Lake Titicaca] appreciated more the white. . . . At the prin- cipal temple, Korikautsa, in Cuzco, every morning one white llama was sacrificed; it was sheared to facilitate stabbing with a stone or copper knife in the heart. Each month at least 100 were sacrificed, and at large festivals, 1,000 or more. [Maccagno, 1932, pp. 44, 46, quoting earlier unnamed sources.] Garcilaso de la Vega (Markham’s ed., 1869, 2: 376-380) said that at the Yntip Raymi festival in “June,” pure black llamas were sacrificed with other llamas, lizards, toads, serpents, foxes, tigers [jaguars], lions [pumas], and many birds. He also stated that such use of animals prevented widespread human sacrifice (ibid., p. 131). Certain star constellations were sacred to llamas: Lira for the male, and Cygnus for the female and its young (Cabrera and Yepes, 1940, p. 262). Remains of llamas, as well as clay figurines, are common in the Inca burial mounds. Evidently the llamas were buried to supply food for the deceased. Maccagno (1932, p. 47) stated that the small llama figurines generally had a round hole in the back, some 5 to 10 mm. (0.15 to 0.30 in.) deep, that the wool was indicated by markings and the tail shown in a vertical or a horizontal position, and that some silver images have been found with a gummy matrix in the eye sockets, which once must have contained precious stones. Medicine.—The bezoar (gastric calculus) of the llama was prized, at least in Colonial times, as a specific remedy for certain ailments. Lama pacos: Alpaca Distribution.—At present the alpaca is found in the domesticated state only and is restricted to southern Pert, adjacent northern Bolivia, and extreme North Argentina. Formerly, in historic times, the alpaca ranged as far south as Catamarca, Argentina. In early Recent and late Pleistocene times (?), it is alleged to have occurred AAD SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 on the Humid Pampas of Argentina near Buenos Aires (map 13). (See Lopez Aranguren, 1930; Cabrera, 1931.) Description (pl. 48).— Size: Smaller than llama; total length 150 to 175 cm. (58 to 68 in.); height at shoulder 80 to 90 cm. (31 to 35 in.). Color: Generally pure brown or black, but in some pure white; rarely multicolored. Pelage: Long and heavy over body, neck and top of head, chest, belly, and upper legs; short but thick on face (often, however, obscured by the long forehead hairs) and on lower legs (obscuring metatarsal glands) ; almost bare on axillae and groin; length about 40 to 50 cm. on body (obscuring tail), and sometimes hanging to ground (suri breed); composed mostly of long, fine wool hairs which are slightly curly and greasy, and mat in heavy clumps. The wool hairs are not shed en masse annually, but individually at different times (like human hair, horse’s mane and tail), and the fibers are said to reach a length of 75 cm. when allowed to grow over many years (probably suri breed only). The wool of the common huacaya breed has a density of 90 to 100 fibers per sq. mm., and the yield is 2 to 3 kg. per animal; in the suri breed, density is from 140 to 170 fibers per sq. mm., and the yield is 8 to 5 kg. per animal (Maccagno, 1932, p. 27). Wool of the alpaca, strangely, is more closely allied to true hair (more medulla and scales) than is true wool of sheep, and more so than llama and vicufiia wool (Bowman, 1908, pp. 232-234). Tail: Short 20 to 25 cm. (8 to 10 in.) and obscured by heavy long body and tail hairs; said not to be carried out and then down in are as in llama (Romero, E. C., 1927). Feet: Small; fore larger than hind. Metatarsal glands: Present on hind shank as on other three species, but completely concealed by hair (as in vicufia); size about asin llama. Hars: Longish, 12 to 15 cm., and broadly lanceolate (not scalloped). Hyes: Large. Mammae: 4 functional (1 nonfunctional). Skull: Medium; in size between llama-guanaco and vicufia; and with features of both (appears to be a combination of small skull of vicufia with large teeth of llama); facial portion short, about same length as cranial portion; facial pits small or absent, teeth large, P 3-4/4 relatively large; lower incisors with enamel on labial and lingual faces and rooted, though some become very large, and these evidently have grown persistently from pulps which re- mained open longer than usual, and have not been worn properly; choanae generally open U; symphysis of mandible generally long and narrow; submental foramen 5 to 15 mm. anterior to root of symphysis (menton). Reproduction: Like llama. Longevity: To 15 or 17 years (av. 12; Flower, S. S., 1931, p. 223), probably longer in native land and out of zoos, though individuals are considered old at 7 years, and then usually killed for food (Maccagno, 1932, p. 35). Hybridi- zation: Possible with any of other three species of lamoids, but common with llama only, producing huarizo and mixti, and secondarily with vicufia, producing paco-vicufia (or vicufia-pacos) which are doubtfully always fertile. Intelligence and disposition: Like llama; and though some authors say that they are less playful and tamable when young (Romero, 1927); others state the opposite. Habits: At present (in domestication), like llama, except for alleged preference for wet and marshy ground of higher altitudes (Romero, 1927; Maccagno, 1932, p. 37), and for alleged greater dependence upon companionship (Tschudi, 1844-46, p. 26). The preference for marshy ground is said to be a function of the “‘soft”’ feet, and to result in small size of the animal because it eats the “‘poor and rachitic”’ vegetation in these places. Both assumptions are applied to the vicufia also, and both appear to be questionable. This close dependence upon water is said to be responsible for great losses during prolonged dry spells, and may be an ecologic barrier which has restricted the range of the animal. Tschudi (1844, p. 261) stated that the alpaca will not suffer a lone existence and must have the Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 443 Map 13.—Distribution of the alpaca. (Horizontal hachure, present range; diagonal hachure, Pleistocene.) (According to Lopez Ananguren, 1931, and Cabrera, 1932.) 444 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 companionship of other alpacas or llamas. Dzseases: Probably like llama; also afflicted with a specific acute chronic fever of diplococcic or streptococcic origin, which often is called ‘‘syphilis’’ (Preston, 1939, pp. 28-29), perhaps giving rise to the story of the derivation of syphilis from llama and alpaca; however, no spirochetes have been found in alpaca lesions. Breeds: Two; common, or ‘huacaya,”’ and “‘suri’’; both breed true. Special physiology: See below. Remarks.—The alpaca is the most restricted in range, and perhaps the most specialized of the four lamoids. It appears to be on the decline in numbers. In its physiology, it is probably the best adapted to high altitude; the normal pulse is high, 54 to 100 per minute; respiration rate high, 20 to 40 per minute; blood count very high, 20 million red blood corpuscles per cu. mm. (6 to 8 million for other domesticated animals and man at high altitudes, though sheep have been recorded with 10,530,000 (Dill, 1938, p. 131)); blood coagu- lates rapidly (Preston, 1939). These data indicate that most alpacas may have difficulty in adapting themselves to lower elevations, though they live long and breed in North American zoos. Their apparent dependence upon water has been mentioned above. The alpaca is considered here as a prebable distinct species. It is known in the domesticated state only, though some feral herds were said to exist by Tschudi (1844-46, p. 261). As the usual alpaca skull is distinctive enough to accept easily the identification by Lopez Aranguren (1930) and A. Cabrera (1931) of the species in the Pleisto- cene of Argentina, the alpaca may be considered a woolly Pleistocene relict, which became progressively more restricted in range and weaker in competition, losmg its adaptive plasticity, until, shortly before extinction, some were integrated into human culture, which for- tuitously was at a proper high level in the same region. J. A. Leén (1939) believed that the alpaca was domesticated later than the llama, when a higher civilization could use the wool; but he did not commit himself definitely as to the biological origin of the alpaca. Bowman (1908, p. 332) believed, after a study of hairs of llama, alpaca, and vicufia, that the alpaca was probably a “mere variety of the llama’ (despite his assertion that the wool of the alpaca was more hairlike than that of the other two tested). The vicufialike characters (concealed metatarsal glands, short snout, open-U choanae, and especially the long and continuously growing incisors from incipiently open rooted incisors) suggest hybrid origin for the alpaca from llama-vicufia crosses. But this is not necessarily true. The alpaca cannot be crossed easily with the vicufia, and the offspring are not unequivocally fertile, and when able to breed inter se, the hybrids soon revert to either parent. The vicufalike characters are quite constant, and may be validly specific. Origin.—The history and present distribution of-the alpaca indicate Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 445 that the Highland altiplano area around Lake Titicaca was the hearth of domestication. Breeds.—Two breeds of alpaca are known today and both are of pre-Columbian origin: (1) common or “huacaya,”’ and (2) ‘‘suri.”’ Both breed true. The suri is distinguished by its longer and finer wool. The alpaca is involved in the two common hybrids among lamoids: (1) Alpaca X llama = “‘huarizo’”’ (female alpaca X male llama = “mixti’’); (2) vicufia (male) < alpaca (female) = paco-vicufia. The huarizo is common but apparently does not breed true (Romero, E. C., 1927, pp. 25-26); in time, hybrid crosses inter se revert to parent types. Many apparent hybrids between lamas and alpacas, or at least llamas with alpacoid characters, are to be seen today in Bolivia, and it would seem that the two forms interbreed readily, or that genetic variation in the llama often tends toward the alpaca. Maccagno (1932, p. 57) stated that huarizos were common because the llama and alpaca interbred freely without help from man. How- ever, E. C. Romero (1927, pp. 25, 27) indicated that the hybrids showed no fixed characters, and had to be perpetuated by parental crosses. The breed presents no advantage, except to combine a cargo animal with one of good wool production. The paco-vicufia has been well known since the Conquest, and Maccagno (1932, p. 54) stated that they have been produced from time immemorial without intervention of man, but the statement would appear to require some proof. These hybrids today are deliberate results of attempts to establish a breed with wool of high vicufia quality and of high alpaca quantity. After the first shearing, the wool of paco-vicufia loses its fine quality, and the guard hairs increase; also, herds of up to 300 have been produced in the last half century, but the owners declared that they had no commercial value because the suri alpaca breed was superior (ibid., p. 56). Such attempts at hybridization are still in progress, though none has been signally successful because of the alleged infertility of some of the hybrids, or because of the return of the hybrid crosses to parent types, together with the lack of desired wool production. Famous is the first recorded attempt to establish a paco-vicufia breed. A priest, Cabrera, in Macusani, Puno, South Highland Pert, labored with patience and perseverance for 21 years (1821-42) and obtained 20 tame paco-vicufias. For this ‘‘new richness,’’ Perti rewarded him; but the hybrids became extinct, either because offspring crosses to the third generation reverted to alpaca or vicufia types (Maccagno, 1932, p. 54) or because all or most of the hybrids were sterile (Alvina, 1872; and Arinibar, 1906; quoted by E. C. Romero, 1927, pp. 26-27). Even earlier than Cabrera’s attempt, was that of the Spanish at 446 SOUTH AMERICAN INDIANS [B. A. B. Bull, 143 San Lucar de Barrameda, Andalusia, Spain. Here M. Bouy de Saint Vincent, a naturalist with the French invasion army under Marshall Soult during Napoleon’s campaign on the peninsula, saw three “‘alpa- vigonias’”’ with long heavy fleeces. These hybrids the Spanish thought would replace their declining flocks of Merino sheep, but unfortunately they left no descendants, being apparently sterile (B[rowne], 1858, pp. 69-70). Subsequent to Cabrera’s hybridization of alpacas and vicufias was the recorded attempt of Faustino Belon, Puno, Peri (Maduejfo, 1912, p- 18). Belon obtained fertile hybrids which ‘‘degenerated” after the third generation to one or the other of the parent stocks. The method of alpaca-vicufia hybridization is interesting. A young male vicufia is captured and given to a lactating female alpaca. The skin of a baby alpaca is used to disguise the vicufia during acquaint- ance. The vicufia, when 1 year old and sexually mature, is given a harem of female alpacas (Maccagno, 1932, pp. 55-56). Aboriginal uses.— Wool, meat, religious ceremony, dung fuel, hide, and medicine. The alpaca is not used for transportation of cargo. Wool and weaving.—The fine quality (luster, strength, and resist- ance to fading and dyes) and length of alpaca wool give it superior value. The Jnca reserved its use for the higher classes and for the royal family, and with it developed their magnificent textile art, which is perhaps unequaled for fineness and design by that of any other aboriginal group in the world. Even today, alpaca wool amounts to about half of the total wool production of Pert (including that of sheep), and about 65 percent of the total value (Tappy, 1944, p. 50); it forms about 95 percent of the lamoid wool exports. There are now five grades of wool besides the “alpaca primeira” and “‘alpaca suri’; only 3 to 4 percent of all wool needs rejection, and the fleece is much cleaner than sheep’s wool (ibid., p. 49). Each alpaca is sheared every 2 years, beginning when it is 2 years of age. Growth amounts in 2 years to about 30 cm. (12 in.) in the common breed and 60 cm. (24 in.) in the suri; the maximum yield is obtained at 5 years of age. Shearing is done, generally, with broken glass or a knife, rarely with clippers (Maccagno, 1932), in the rainy season (January to March), when the day temperature is high and sheared animals do not chill, and rain can wash the unclipped wool on the backs of the animals. Aboriginal shearing must have required special implements, perhaps obsidian knives. Meat.—Alpaca meat is eaten, especially in the form of dried “charqui,” or “cecina”; about 18 kg. (40 lbs.) is obtained from a male, and 9.5 kg. (21 Ibs.) from a female (Maccagno, 1932, p. 41). Dung fuel.—Alpaca dung is extensively utilized as fuel (the entire Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 447 range of the alpaca is treeless). The bolus is like that of the lama and is similarly deposited in communal spots. Hides.—The uses of hides are similar to those of llama hides. Husbandry.—Alpaca husbandry today is more complex generally than that of the llama. Herds are segregated according to age, sex, and use: Young, 3 to 8 months, both sexes; ‘‘extrema’’ males, 8 to 18 months; “‘extrema’”’ females; “‘ancuta’’ males, 18 to 24 months; “ancuta’’ females; females with young to 3 or 4 months; pregnant females; sterile and old females; reproductive males; and castrated males (‘‘capones,”’ for wool and meat only). The best breeding males are 3 to 7 years old; generally 50 percent of the males are used for breeding; others are castrated. Old females, over 7 years, are often kept with the capones which are destined to supply meat for annual June ‘‘Beneficios.”” Breeding is aided often by man (through manip- ulation, as with stallions). (See Maccagno, 1932, pp. 34-36.) Religious ceremony.—The alpaca had religious significance almost equal to that of the llema, and many were sacrificed to the Sun Gods. The black ones were reserved for Viracocha, the son of the Sun, and the white were reserved for the Sun God himself. The Collas tribe (Aymara, of Lake Titicaca), before their conquest by Lloque Yopanque, third Inca [Emperor] of Pert, about A. D. 1171, worshiped the white alpaca as their principal presiding deity and offered up sacrifices to him. They believed that their divinity had come among them in this form and, as a mark of favor to them, procreated more abundantly in their territory than elsewhere. [Tappy, 1944, p. 47.] Medicine.—Bezoar stones. were utilized, especially in Colonial times. Pets.—Baby alpacas are kept often as pets by children, perhaps more so than llamas. Lama glama guanicoe, or L. guanicoe: Guanaco Distribution.—The present range of the guanaco extends from Central Pert south through the Andes and coastal ranges of southern Peri and North Chile (where it reaches the Coast) to mid-Chile, thence east and south through the Eastern Andes and Argentine Pampas, from a rough line between Mendoza and Bahia Blanca, to the Patagonian steppes, Tierra del Fuego, and across the waters of Beagle Canal to Navarin Island (map 14). Formerly, in prehistoric, early Recent, and late Pleistocene times, the guanaco ranged also far northeast over parts of Paraguay and extreme northern Argentina, end perhaps as far northwest as southern Colombia, where a Paramo de Guanaco exists. The alleged llama bones from early archeological sites in Ecuador may be those of guanaco. Cieza de Leén (Mark- ham’s ed., 1864, p. 207) recorded it in early Colonial times from Loja, southern Ecuador. Description (pl. 44).—Size: Similar to llama, but slimmer; total length around [B. A. HB. Bull. 143 SOUTH AMERICAN INDIANS 448 (Horizontal hachure, present range; vertical hachure, prehistoric; diagonal, Pleistocene.) Map 14.—Distribution of the guanaco. Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 449 2 m.; height 90 to 110 cm. (35 to 43 in.); weight about 75 to 100 kg. (165 to 220 Ibs). Color: Rich tawny brown or faded brown dorsally and laterally; white ventrally and medially (including neck); smoky gray brown on head, with whitish orbital rings, chin, lips, and ear edges; albinism occurring, melanism absent. Pelage: Soft, silky, wool underhair, 30 to 40 mm. (0.90 to 1.20 in.), over body, neck, and chest, with varying amount of long but fine overhair (beard or guard hair), 60 to 90 mm. (1.2 to 2.7 in.), which often is ridgelike along back of neck; chest and upper belly with longer white hairs, 100 to 120 mm.; short hairs on lower legs and head; almost bare on axillae and groin. Winter specimens from Patagonia and Navarin Island have a profusion of long, rich brown overhairs (to 150 mm., or 4% in.) on body and neck, which almost obscure the woolly under- hair. Probably a complete annual molt is experienced in summer, The juvenile pelage of ‘‘chulengos”’ up to 4 to 5 months is composed of soft, silky wool hairs without dimorphism, and is highly prized. Tail: Short, about 25 cm., heavily haired with wool and hair, and carried out and down in an are, like llama. Feet: Large, and like llama. Metatarsal glands: Pair on each hind shank; clearly visible, ovoid, and large. No medial callosites on ‘‘knees’’ (wrists) or sternum. Ears: About 15 em.; rimmed with short white hairs; gently scalloped terminally on anterior edge. Skull: Large; rostrum long (premaxillae to antorbital border greater by 10 to 15 mm. than antorbital border to posterior face of condyle); choanae narrow or open V-shaped, rarely U-shaped; facial pits large; molar teeth large; P 3-4/4 large; heel and endostyle of M/3 usually large; mandibular symphysis long (about twice width); lower incisors imbricate, rooted, spatulate, with enamel labially and lingually; submental foramina just anterior to menton; condyle relatively high above angular process of mandible. The skulls of guana- cos are very difficult to distinguish from those of the llama; in the guanaco the mandibular condyle is often higher above the angular process, and the endostyle of M/3 more often larger; the frontals may be slightly less depressed also. A combination of skin and skull, however, usually can be identified as of one or the other species, and the two are always distinguished by natives. However, overlap occurs frequently enough to justify considering the two as conspecific. Reproduction: Rut in summer; gestation 11 months; one or rarely two young, very precocious, weaning in 4 to 6 months, and maturing in several years; longev- ity perhaps 30 years or so in wild state (Simpson, 1934, p. 191). Hybridization: Fertile crossing possible with llama, alpaca, and vicufia, though such hybrids doubtful in nature, especially those of guanaco-vicufia, and these probably not always fertile. Intelligence: Low. Disposition: Curious, calm, and easily tame- able when young, though adults probably less manageable, and captive males in rut dangerous. Habits: Herbivorous, gregarious, polygamous with male dominance and dependence of others on his leadership; found in arid, treeless country and deserts from 17,000 to 18,000 feet (5,200 to 5,600 m.) down to sea level, and in the forests of Navarin Island, and the southern Chaco (formerly). The natural social organization consists of herds of 4 to 10 females and young with a male leader, herds of bachelors up to 3 years or so, and often solitary males; some- times there are large composite herds of hundreds of individuals after calving season and before rutting. The male leader is also the sentinel; when he is shot the females are as likely to remain as to flee, but on the death by shooting of a female all immediately escape in a group. Severe fighting with teeth and hoofs takes place between males in rutting season for conquest or retention of harems. Deposition of dung is made in communal places which are conspicuous sights in guanaco country; also, common wallowing and trampling grounds are used, perhaps to take advantage of a dusty spot to remove parasites; common dying places, however, are now discredited. Bezoar stones (gastric calculi) common 450 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 and once highly prized for medicinal purposes. Voice: A distinctive tremulous neigh or querulous bray; a ‘‘yammer’’ (Simpson, G. G., 1934, p. 192). Remarks.—The guanaco is the most widely ranging and most adaptive of the four lamoid forms. It successfully meets environ- mental conditions from sea level to 17,000 feet (5,200 m.) altitude, and from barren hardpan deserts to forests, but all these have some degree of aridity. It is a good swimmer. Ethnozoology of the guanaco.—The guanaco is a wild form slightly distinct from the llama and probably the progenitor. Many Pleisto- cene and early Recent fossils have been identified as belonging to the guanaco. For countless centuries, young individuals have been tamed by aborigines for pets, perhaps to be killed later for food and perhaps consciously tamed for this purpose. The conclusions are that the guanaco may represent the wild representatives and ancestral stock of the domesticated llama. Today there are a few semidomesticated herds in Argentina, but they are somewhat refractory to domestication and it is doubtful that the present members ever will be fully domesti- cated so that they can be raised easily in semiconfinement and utilized as is the llama, which already serves the purpose. Hybridization with the llama confuses this situation. Though man utilized the guanaco over its entire range, the dry Pampas of Argentina and Patagonia were the areas of greatest num- bers of individuals and of the most complete cultural utilization. The guanaco was the most conspicuous and one of the most typical animals of these plains, as was the bison on the Great Plains of North America. With the rhea, perdiz, fox, mara, viscacha, cavy, tuco-tuco, and armadillo, the guanaco constituted part of a plains fauna that was relatively completely and easily utilized by man. Today the guanaco has been extirpated over much of its northern and southern Argentine range, and elsewhere it is much reduced in numbers. To the Tehuelche, Puelche, northeastern Araucanians, Huarpe, and Querandi, the guanaco supplied meat for food; fur and hide for clothes and shelter; bezoar stones for medicine; sinew for sewing; pets for pleasure (and food ?); a stimulus for myths and many verbalizations to account for age, sex, color, etc.; and an object of time-consuming hunting activity for the men and accessory duties for the women. Like the bison-hunting Indians of North America, the guanaco- hunting Pampa Indians were savage fighters, easily shifting their hunting techniques to warfare and fiercely resisting the inroads of the Whites. : The Inca obtained many guanacos in their periodic hunting drives for vicufia. (See p. 454.) The older females and many of the males were killed for meat, wool, and hide; the younger females were sheared Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 451 and released. The meat was given to the common people, as was the wool, which was considered coarse and of poor quality (probably be- cause of the quantity of guard hairs). The aborigines hunted guanacos by drives and ambuscades with bow and arrow, spear, and bolas, sometimes with the aid of dogs. Later, when the European horse was acquired from the Spaniards, the encircling “rodeo” with bolas prevailed. Finally, with the advent of firearms and the disappearance of the Indian, the hunting degenerated to a slaughter of the newborn to 3- or 4-months-old chulengos by White hunters for the skin only. Lama vicugna or Vicugna vicugna: Vicuiia Distribution.—At present the vicufia is distributed from North or Central Peri south through the High Andes and coast range moun- tains through Bolivia to adjacent North Chile and Argentina (map 15). In the late Pleistocene and early Recent, it ranged as far east as the Humid Pampas of Buenos Aires (Lopez Aranguren, 1930; Cabrera 1931). Cieza de Leén (Markham’s ed., 1864, p. 207) recorded the vicufia from Loja, southern Ecuador, in the 16th century. Description (pl. 44).— Size: Smaller and more slender than the guanaco, though neck equally as long (or longer ?); total length, around 175 cm. (70 in.); height at shoulder, 70 to 90 em. (28 to 35 in.) ; weight, approximately 35 to 60 kg. (77 to 132 Ibs.). Color: Rich ‘‘vicufia’”’ brown dorsally and laterally and on entire neck and head; white ventrally and medially and on chin and under eyes and on flanks, showing conspicuously on brisket, underside, and especially tip of tail; otherwise very uniform brown all over; small circle of black stiff hairs around eyes. Pelage: In summer, short and curly wool underhair on body and neck (30 to 40 mm., or 1.9 to 1.5 in.), with rare scattered guard hairs almost concealed in the wool; short but thick hair on legs and face, gradually shortening in length from body hair (this change in hair length is abrupt in llama and alpaca) ; almost bare on axillae and groin; long hair-fringe on brisket (100 to probably 300 mm., or 4 to 12 in.) on both males and females (perhaps longer in some males and certainly longer on both sexes in winter) and linked between bare axillae to chest and belly patch of long hairs (100 mm., or 4 in.). Single annual molt of entire pelage in summer. Wool hairs very fine, silky, though short. (See below.) Tail: 25 to 27 cm.; woolly hair, brown dorsally and laterally with white ventrally and term- inally (conspicuous white tip). Feet: Small; fore and hind subequal; ridge on nails not pronounced. Metatarsal glands: Present on hind shanks, though completely concealed by heavy, short hairs; more lanceolate-shaped than oval, and smallish. Ears: Shortish (13 em.), and lanceolate (without scallop). Skull: Very distinct in rootless, parallel-sided, ever-growing lower incisors with enamel on labial side only—characters which tend to support subgeneric (or generic rank)—otherwise, skull small; facial portion (from premaxillae to antorbital border) always slightly shorter than cranial portion (antorbital border to rear of occipital condyle) ; facial pits absent or very small; teeth small, especially P 3-4/4 and posterior lobe of M/3; choanae open and U-shaped; palate very narrow at diastema; vomer ending in a long posterior spine; incisive foramina not extending anterior of alveolus of I 3/; manibular symphysis short and broad with submental foramina slightly posterior to menton. Reproduction: Rut in late summer, gestation 794711—50 30 452 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Map 15.—Distribution of the vicufia. (Horizontal hachure, present range; diagonal hachure, Pleistocene.) Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 453 10 months, mature in a year (Cabrera and Yepes, 1940, p. 268). Hybridization: Possible with guanaco, llama, or alpaca, but rarely (or never ?) in the wild with any of these, and in captivity with alpaca only, to any extent; hybrids probably not always fertile, and if so, always reverting to parent stock with continued interbreeding. Intelligence: Low. Disposition: Curious, calm, and docile, affec- tionate and playful, especially when young; but generally reverting to wildness later in life; rutting males are vicious. Habits. Social in small herds; gregarious, herbivorous, polygamous with male dominance, and when male leader is killed the harem females remain; said to prefer marshy or other soft ground (like alpaca), but often seen in dry barren places. Diseases: Probably mange in captivity, but in wild state conditions not known. Voice: Peculiar neigh (‘‘chillido ou relincho peculiar’; Ridoutt, 1942); perhaps more of a whistle than a neigh, because it can be distinguished from the cry of any other animal (Tschudi, 1844-46, p. 218). Special physiology: Like llama (p. 4386); red blood cells 14.9 million per cu. mm. (Dill, 1938, p. 1381). Remarks: The vicufia is a species restricted to the arid high altiplano, or hard pampas of the adjacent Coastal mountains. Ethnozoology of the vicufia.—The vicufia is a good example of an animal that has been subjected unsuccessfully to many and carefully premeditated attempts at domestication, in this case because of the high value of the wool. Attempts of the Jnca and their predecessors to domesticate the vicufia are not known, though it may be assumed that they also tried and were unsuccessful and that they then adopted the method of periodic round-ups (drives) and capture of wild stock for shearing. A notable attempt at domestication was made by the priest Cabrera in southern Peri in 1821-42, who consciously or unconsciously aimed eventually at alpaca-vicufia hybridization; but neither his domesticated vicufia nor paco-vicufia hybrids became established. Other recorded attempts have failed also. (See p. 445.) Today several “‘criaderos”’ exist in Pert and Bolivia, but they have amounted to little more than semiconfined protection for the vicufias. Simon Bolivar, in 1825 in Pert, issued a governmental decree offering a monetary reward for domestication based on “experience [which] shows every day the facility with which they are domesticated”’ (Maccagno, 1932, p. 58)! The vicufia is prized for its silky, fine, richly colored wool. The Inca reserved the wool for the royal family and a few honored high officials. The value of vicufia wool lies in its fineness (0.00043 in. in diameter and 2,500 fibers to the inch; sheep, 0.0008 in.), soft and silky texture without predominating medulla, good surface cohesion of overlapping cortical scales, tendency to curl, elasticity, fine and uni- form pigmentation, beautiful rich natural color, resistance to dyes and fading, and luster (Stroock, 1937 b). About 500 gm. can be obtained from a single animal (Maccagno, 1932, p. 28); but 250 gm. seems to be the average individual yield; and of this, half is made up of worthless stiff guard hairs (Stroock, 1937 b). The Inca practiced true conservation of the vicuiia; they prohibited 4.54. SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 indiscriminate hunting, and harvested a crop of wool, meat, and hides periodically from the wild herds, allowing sufficient stock to remain for breeding and increase. The periodic vicufia hunts were called ““chacus,’”’ and were held every 3 to 5 years in certain provinces, some of which were divided into special sections pertinent to this purpose. On royal order, 20,000 to 30,000 Indians were assembled to build an immense corral 2 to 3 km. in diameter, with the entrance on one side about 120 m. wide, and posts of a man’s height connected with ropes for enclosing the animals and entangling them in attempted escape. Colored rags were hung from the ropes, and these fluttered in the wind and frightened the vicufias from trying to escape. Bolas were also placed at certain intervals along the fence. Then the Indians formed a semicircle and “‘beat’’ an area sometimes as much as 20 to 30 km. in diameter. All the thousands of vicufia, guanacos, and deer, and many other animals, such as bears, pumas, and foxes were driven into the corral. The bears, pumas, and foxes were killed as predatory vermin; but some were saved for pets or “torture animals.” The deer, guanaco, and vicufia were caught by hand or with bolas. All the guanacos and vicufias were shorn of wool; the young and repro- ducing females, together with enough young males for breeding pur- poses, were released; the old females and other males were killed for their meat and hides. Deer were treated similarly except that they were not shorn. The wool of the vicufia went to the Virgins of the Sun for weaving into garments for royal use; the guanaco wool went to the Indians, as did the meat of all and was dried into ‘‘charqui” for preservation. Counts of all species and their destiny were kept on quipus (knotted counting strings). Garcilaso de la Vega (Mark- ham’s ed., 1869, 2: 115-119) gave a detailed account of this “chacu,”’ and stated that the hunts were held every 3 years, because this interval was sufficient to reestablish normal numbers of game (20,000 to 40,000 head per chacu) without allowing them to become a nuisance, and also to allow the wool of the shorn vicufias to reach a maximum length; he did not mention the corral. (See also Cabrera and Yepes, 1940, p. 268.) Bezoar stones from the ruminants were a byproduct of the chacu and were greatly prized for medicinal purposes. THE DOMESTICATED CAVY, ‘‘GUINEA PIG’’ (CAVIA PORCELLUS) Cavies (sensu stricto), small, tailless, hystricomorph rodents of Neotropica, constitute the supergeneric group Cavia, or subfamily Caviinae, of the family Caviidae. The name ‘‘cavy” is often applied also to the related but much larger Patagonian hare, or mara (Doli- chotis). Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 455 Classification Supergeneric group Cavia, or subfamily Caviinae. Cavies (sensu stricto). Genus Kerodon. ‘‘Mocu’’; northeastern Brazil. Genus Microcavia. ‘‘Cui’’ (syn. Caviella; includes Nanocavia and Montica- via); Patagonia-Chilea cavies. Genus Galea. ‘Cui,’ “pampahuanca’’; Highlands of Bolivia to Pampas of Argentina; also northeastern Brazil (an isolated species). Genus Cavia. “Cui,” “pred,” ‘cori’ (C. aperea, C. “tschudii,” ete.); Guiana- Brazilia; includes the domesticated cavy (Cavia porcellus). We are concerned here primarily with Cavia; secondarily with Galea. The above classification of the cavies is modified slightly from that of Kraglievich (1930 a), who kept Monticavia separate from Microcavia, but included Nanocavia in Microcavia ‘‘because they are so much alike.’ As the generic character of Microcavia rests primarily on the postnatal replacement of DP 4/4, and as a specimen of ‘‘Monticavia”’ niata (USNM-172942, from Mount Sajama, Bolivia) also shows post- natal development of DP 4/4, these three groups are included in the same genus. Caviella (Osgood, 1915) is a synonym of Microcama (sensu lato). All cavids except Microcavia have prenatal replace- ment of DP 4/4 by P 4/4. (See also Osgood, 1915; Thomas, O., 1917; and Ellerman, 1940-41, 1: 240-247.) All of many specimens ex- amined of wild Cavia appear to be conspecific, C. aperea. Cavia porcellus (syn. C. cobaya): Domesticated cavy Distribution.—In late pre-Columbian times the domesticated cavy was found throughout the Andes from Central Chile north to Central America (?) and probably to the Antilles, where they were recorded by the Spaniards shortly after the Discovery (map 16). Description.— Size: Total length, about 300 mm. (12 inches); tail absent externally (but present under the skin and about 25 mm.); weight, 750 to 1,250 gm. (2 to 3 pounds). Color: Monochrome and polychrome; generally white, brown, or black, or of mixtures or intermediate shades; original color probably “agouti” (gray-brown-black grizzled or “ticked’’). Pelage: Short and silky generally; long, fine, and curly in one breed (Angora). Jars: Short, closely appressed to head and slightly ‘‘scalloped” posteriorly in typical hystricoid form. Legs: Short. Feet: Fore, short with 4 toes; hind, long with 3 toes. Skull and skeleton: Relatively large, heavy, and rugose; large antorbital foramen (as in all hystricomorphs); lacrymal not broadly interrupting zygomatic extension of maxillary (also other Cavia); incisor teeth unpigmented (all Cavia and Micro- cavia); rostrum relatively broad and flat; post-palatal margin smooth (rarely with spine as in other Cavia); naso-frontal suture truncate (rather than W- shaped); fronto-parietal suture convex posteriorly (rather than straight) ; para- occipital process relatively short and blunt; auditory bullae small with relatively long anterior projection and meatus; pterygoids smallish; no transparent areas in frontal region (Ubisch and Mello, 1940, p. 397); coronoid-condylar distance on mandible short, and angular process short; inferior surface of atlas vertebra singly perforated (like Kerodon; doubly perforated in Galea and Microcavia) ; neural spine of axis vertebra large and massive; and acromion of scapula wide. 456 SOUTH AMERICAN INDIANS [B. A. H. Bull, 143 Map 16.—Distribution of wild Cavia (horizontal hachure) and Galea (vertical hachure). Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 457 (For illustrations of some of these characters, see Detlefsen, 1914, pls.) Repro- duction: Breeds continually throughout the year, generally 4 times; gestation 60 to 65 days; young at birth 2, 3, or 4 (rarely 1, 5, or 6). Mammae: 2 (1 pair) inguinal. Growth: Wean at 2 months; mature sexually in 2 to 3 months. Lon- gevity: 5 to 8 years? Intelligence: Low. Hybridization with wild forms: (1) With Cavia “aperea” (=C. aperea pamparum?) of northern or Northwest Argen- tina, produced fertile offspring, both inter se and in backcross (Nehring, 1894); (2) With Cavia “‘rufescens’”? (=C. aperea fulgida?) of Campinas, Sao Paulo, southeastern Brazil, yielded fertile female hybrids, but male hybrids sterile down to and including many of the % wild hybrids (Detlefsen, 1914; Ubisch and Mello, 1940); (3) With Cavia ‘‘cutleri”” (=C. aperea pallidior?) from Arequipa, southern Pert, produced fully fertile hybrids of both sexes, either inter se or in backcross (Castle, 1916); and (4) with ‘‘wild Brazilian ancestor,” no interbreeding (Haeckel, quoted by Cumberland, 1886[?], p. 27). Habits: Sedentary; gregarious without social organization aside from usual male dominance over several (up to a dozen) females, accompanied by bitter fights between males; herbivorous and frugiv- orous; poor burrower. Diseases: None aboriginally (?); but with the advent of bubonic plague around 1900, the cavy has proved highly susceptible, and recently thousands have been destroyed systematically in Ecuador and Pert by public health officials amid scenes of grief, and antagonism from the Indian owners. Economic value.—Primarily for meat; secondarily as agent in medical mysticism, in religious ceremony, and also as pets (recently as a laboratory animal). Remarks.—The cavy, or “cui” (the names ‘‘guinea pig” and ““cobaya” are patent absurdities), the llama, and alpaca are the only South American mammals that have been domesticated. Cavies were encountered along the Andes from Ecuador (and Columbia ?) to Central Chile by the early Spaniards, among the Arawak of His- paniola and Cuba by Columbus (?), and in Yucatén by Cortes (?). Today cavies may be seen around most Indian dwellings throughout this area, and from 5 to 15 usually may be encountered in a house, where they remain voluntarily and scurry around the furniture and dark corners, squeaking plaintively when disturbed (hence the name “cui,” ‘“kwee”’). The native seems greatly attached to his cuis, but the animal apparently does not occupy a very important place in his economy save as an occasional source of meat prepared with hot spices, and occasionally as an adjunct in medical diagnosis through magical means and in medical treatment (the warm viscera and quiver- ing body of a freshly killed cavy are laid over the abdomen of a patient who is suffering intestinal pain). The cavies receive little care and have no special cages; their only food is table scraps and some greens and fruit. The Inca are said to have sacrificed 1,000 cavies to Frost, Air, Water, and Sun in ceremonies during the month of August (Rowe, Handbook, vol. 2, p. 310). The homeland of the domesticated cavy was thought at first to be Brazil, probably because of the accurate description of the animal in 458 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Marcgrave’s report (1648) of the animals of the Pernambuco region; and the stem form was considered to be the Brazilian C. aperea by the French authors. (See Cumberland, 1886 [?], pp. 26-27.) Rengger (1830, pp. 274-278) postulated Paraguay as the homeland, with the local population of C. aperea as the stem form, and this statement persisted for many years. Darwin (1876, 2:135, footnote) noted that the wild aperea of La Plata had a louse of a genus different from that infesting the domesticated cavy, and concluded that the latter had not been derived from the former, and that any alleged inability to interbreed was not a result of domestication. Nehring, in a series of papers (1888, 1889, 1891, 1893, and 1894), claimed that the home- land of C. porcellus was Pert, and that the stem form was C. cutleri Bennett (1835). However, O. Thomas (1917, p. 156) showed that Bennett’s C. cutlert was without exact locality on the coast of Peri, and that it was a melanistic individual which was identical with C. porcellus in morphology. Tschudi (1844-46, p. 195) wrote about a large agouti-colored cavy from Ica, southwestern Pert, under the name of C. cutileri (sic), and Fitzinger (1867) redescribed this form as C. tschudii. Castle (1916) obtained three individuals of a similar large wild cavy from Ica, interbred them freely with laboratory C. porcellus, and said that they were probably feral C. porcellus. However, the bare possibility exists that this Ica ‘‘Cavia cuttleri-tschudvu-porcellus’”’ may be a true wild cavy, and that it represents the wild stock of the domesticated form. Castle (1916, pp. 5-6) found seven mendelizing unit-character variations among the cavies in the Indian houses around Arequipa, as follows: Agouti, black, yellow, albino, red, smooth, and rosette. Aside from agouti-colored individuals, all were spotted; he found no self-colored ones. The Division of Mammals, United States National Museum, Washington, D. C., recently received four cavy skulls from pre- Columbian archeological sites in Peri (three from Ancén, one from Pachacamac near Lima), and some specimens of wild cavies from Arequipa and from near Ica, South Coastal Peri. The four archeo- logical specimens are presumed to be of domesticated stock because the three Ancén skulls had quantities of adhering white and/or pure brown hairs. These specimens show some intergradation in the morphologic characters which usually separate wild Cava from the domesticated stock. The frontoparietal suture of three of five wild skulls and of three of four archeological skulls is convex posteriorly (the domesti- cated character) ; in the others it is almost straight. The nasofrontal suture of all is flat M-snaped (the wild character; in the domesticated condition it is almost straight). A palatal spine (wild character) is Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 459 present in two skulls of the wild specimens (absent in two; one skull broken and not observable) ; this spine is present on one of the archeo- logical skulls. The broad rostrum of the domesticated stock seems to be a heterogonic character of old age. The tail vertebrae number six in all four skeletons of domesticated stock in which the tail is present (with four sacral vertebrae), six in all three wild Coastal Peruvian skeletons available (also with four sacral vertebrae); eight tail vertebrae existed in two specimens of wild Brazilian stock exam- ined in the field (Anapolis, Goias), counted without regard to sacral number. It is highly probable that the domesticated cavy can be linked to wild South American stock through archeological specimens from Pert, and through the Peruvian Coastal wild populations. It is apparent that a profitable study could be made of the morphologic changes concomitant upon domestication. Domestication here seems to have been made from certain populations of a widely ranging wild species, and to have resulted in increased size, selection of genetic color strains, heterogonic skull differences, and genetic loss of wild- ness (wild Cavia: C. aperea from Brazil and Colombia were incorri- gibly wild and refractory to captivity). Unfortunately, to date there is lack of knowledge of morphology of skull and skeletal characters of all the several homogeneous domesticated strains of cavy. Some characters which are attributed to the entire domesticated stock may be characters only of a single or of several strains. Genetics may be the science to elucidate the taxonomy and other problems of this domestication. The conflicting results in producing fertile offspring when crossing porcellus with wild populations may be a consequence of the actual specific difference of some of the wild stocks in question, though here all are considered on morphologic evidence to be provisionally con- specific (C. aperea). The various results may be caused by marked physiological differentiation in geographic races. Cases are known where subspecies on the opposite ends of an intergrading “Rassen- kreis’” are morphologically and physiologically as distinct as two species. The Central Andean region represents the locus of the highest populations of the domesticated cavy, and this condition probably has existed for several millenia. However, remains of domesticated cavies are very scarce in Peruvian archeology. Of many hundreds of pictorial pots seen in the museums of Chiclin and Lima, none unmis- takably portrayed cavies; and only two undoubted mummies of C. porcellus (from Nazca) and one pot of doubtful cavy form (origin unknown) were seen. (See also Schmidt, M., 1929 b, pls. on pp. 216, 561.) None was at the Museo Chiclin in northern Peri. On the 460 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 other hand, remains of llamas are multitudinous, and many other animals are represented on pottery drawings or in the shape of the pots themselves. (See p.357.) Hence, the rarity of the domesticated cavy in Peruvian archeology is a mystery and a paradox, and may be considered weighty evidence against the theory that the cavy was domesticated in the Peruvian area. However, the morphologic and genetic evidence points to the Andean Highland area, or its immediate environs, as the home of the domesticated cavy. Negative evidence for other areas supports this thesis. Today the genus Cava is found wild in the Highland area as far south as Cuzco and La Raya Pass (separating the basins of Cuzco and Puno). From Puno southward, the wild cavy is another genus, Galea, which is small, grayish, with four mammae, and is easily cap- tured and kept alive in captivity (generally difficult for wild Cavia). To any observer in this Highland region, it appears strange that Galea has not been domesticated either prior to and instead of Cavia porcellus, or subsequently, upon stimulation of such a cultural trait involving a closely related animal. THE MUSCOVY DUCK (CAIRINA MOSCHATA) This duck is the only domesticated bird from South America and it is found both wild and domesticated (the turkey is a southern Mexican domesticate). Distribution.—In the wild state the Muscovy duck ranges over Guiana-Brazilia and Central America (map 17). As a domesticated bird, it occurs over most of Neotropica, including the Antilles, where it was present when Columbus landed, though it does not occur wild on the islands. It is very numerous in Peri and Paraguay. Description (pl. 45).—Size: Large; total length (tip bill to tip tail feathers), males 70 to 80 cm. (28 to 32 in.); weight, 4 to 5 kg. (8 to 10 lbs.); females about half this size. Color: Males and females alike (except for caruncles on bill in males); wild populations dark brown or blackish all over, except for striking white upper and lower wing-coverts and axillars, and pinkish caruncles in males; greenish sheen on shoulders and on wings; domesticated individuals, as given above, or with various amounts of white, to almost pure white. Carwuncles: Warty excrescences between bill and eye, present and pinkish on males, absent on females. Habitat: In the wild, along jungle streams; apparently not common on savannas. Habits: In wild state, generally solitary; sometimes flocking on large bodies of water in dry winter season; in domestication, solitary or in small family groups; shy, almost voiceless. Reproduction: In wild state, nests in trees; in domestication, anywhere; polygamous, with much fighting between males in breeding season. Musk: None. Hybridization: In domesticated state, breeds with any other duck, but hybrids are sterile; probably wild and domesticated populations interbreed at any opportunity. Sterility of hybrid Muscovies has been questioned. Remarks.—The domesticated Muscovy duck seems to be conspecific with its wild populations; variation in morphology and color overlap. Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 461 Map 17.—Distribution of the wild Muscovy duck. (After Phillips, 1922, vol. 1, map 2.) 462 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 It was introduced into Europe in the middle 16th century, and soon afterward spread over Africa and Oceania, but it was never popular in Europe or North America. It became feral in parts of Europe and southeastern Asia, and Pallas in 1831 and Keyserling and Blasius in 1840 claimed that it was a native wild species. The name “Muscovy” probably comes from Pallas’ account of the duck in his treatise on Russian (Muscovite) animals; or it may come from the erroneous claim of Buffon that the bird has a musk gland, or from the alleged derivation of the species from the Mosquito coast of Central America or from the Muisca Indians of Nicaragua. (See Phillips, 1922, 1: 57-67, from which the above account is taken.) The Muscovy duck was and is used primarily, and generally exclusively, for meat. However, Ignacio de Armas (1888, p. 132, quoting Garcilaso de la Vega) noted that the pre-Columbian Incas dried the meat and produced an aromatic powder (presumably from the alleged natural musk of the animal). Pizarro is said to have received duck powder to perfume himself. Such uses are not easily explained as the bird is not considered to have musk. The origin of domestication is not known, but evidence indicates that it was in the area of high cultures in the Central Andean region, probably Peri. Representations of the Muscovy duck are common on pre-Columbian pottery in Pert. SUMMARY South America, Central America, and the Antilles, make up the Neotropical zoogeographic region of the world. This region is noted for its rich fauna, composed of a great number of unique or endemic forms, and invaders (some now distinct) from North America since Pliocene times. The fauna is divisible into primary endemics of pre-Pliocene Neotropical origin (when South America was a conti- nental island), secondary endemics from changed invader stock since Pliocene contact with North America, and unmodified invaders. Man is one of the latter, and long has been an important faunal element. At present there are many diverse species, but the Pleistocene and early Recent saw the extinction of many others, particularly large forms. Man perhaps hastened the extinction of some. This late Tertiary extinction, for one reason or another, reduced the plains faunas particularly, though less so in the southern temperate parts of Neotropica than in the central and northern tropical parts. Neotropica is divisible into four subregions: (1) Guiana-Brazilia, (2) Central America, (3) Antillea, and (4) Patagonia-Chilea. The first three constitute at least 75 percent of the whole region, and are tropical. The rich ethnozoology of these tropical subregions may be Vol. 6] FAUNA AND ETHNOZOOLOGY—GILMORE 463 separated into Continental Tropical, and Antillea Tropical on the basis of the different faunas. Patagonia-Chilea is south temperate, and is characterized by the very large number of its primary endemic forms, which include the guanaco-llama-alpaca-vicufia-viscacha-mara- rhea-perdiz fauna of the plains and mountains. This fauna, with and without domestication, was thoroughly utilized in many special ways by man. The domesticated llama probably was derived from similar wild guanacos. The alpaca probably was domesticated from a now extinct wild ancestor, Domestication of the llama and alpaca un- doubtedly was accomplished somewhere in the contiguous altiplanos of Pert, Bolivia, Chile, or Argentina. The other domesticated mammal of South America is the cavy (guinea pig); it probably was domesticated in the Central Andean region also, from the wild stock which exists there today. The remaining domesticated animal, a bird, the Muscovy duck, has conspecific wild populations, and probably was domesticated, as were the others, in the Central Andean region. ACKNOWLEDGMENTS AND SOURCES My first acknowledgment and greatest debt are to Dr. Julian Steward, editor of this Handbook series, who served as the inspiration for this paper. To the following members of the United States National Museum, I owe thanks for reading sections of the finished manuscript: Dr. Remington Kellogg, mammals; Dr. Herbert Fried- mann, birds; Dr. Doris Cochran, reptiles; Dr. Leonard Schultz, fishes; and Dr. E. A. Chapin, insects. To Dr. H. E. Anthony, Curator of Mammals of the American Museum of Natural History, New York, I am grateful for permission to study the lamoid specimens in that museum, in conjunction with those from the United States National Museum. Much of the information presented here comes from many authors, all of whom I hope have been credited, though some statements may have been derived from the literature so long ago that I have forgotten the source now. Apparently Latcham (1922) was the only author to have dealt extensively with South American ethnozoology. A great deal of the factual material and inferences, however, are my own, and have been gained from 5 years’ observation and field work in South America (principally Brazil, Colombia, Bolivia, and Pert), and primarily concerning mammals, other vertebrates, and insects of medical importance, in the order of importance as named. BIBLIOGRAPHY Allen, G. M. (see Tozzer and Allen, 1910), 1911, 1920, 1939, 1942, 1943; Allen, J. A., 1876, 1904; Allen, W. R. (see Eigenmann and Allen, 1942); Anthony, 464. SOUTH AMERICAN INDIANS [B. A. H. Bull. 143 1925-26; Ashley-Montagu, 1929; Ballou, 1897; Barbour, 1914; Barden, 1941; Bates, H. W., 1892; Bates, M., 1944; Beddard, 1902; Bennett, 1835; Bier, 1932; Bird, 1938 b; Bowman, 1908; Bridges, 1846; B[rowne], 1858; Cabrera, A., 1929, 1930, 1931, 1932, 1934, 1945; Cabrera and Yepes, 1940; Castle, 1916; Cieza de Leén, 1864; Coates, 1939, 1947; Coates, Cox, and Granath, 1937; Coker, 1919, 1923; Cook, O. F., 1925; Cooke, 1895; Couto de Magalhades, A., 1931; Cox (see Coates, Cox, and Granath, 1937), 1938; Crawshay, 1907; Cumberland, 1886[?]; Cutright, 1940; Darlington, 1938; Darwin, 1876; De Sola, 1932; Detlefsen, 1914; Dill, 1938; Ditmars, 1920, 1928; Ditmars and Greenhall, 1935; Dunn, 1931, 1944; Eaton, 1912, 1918, 1916; Eigenmann and Allen, 1942; Ellerman with Hayman and Holt, 1940-41; Elliot Smith, 1915; Estruch, 1943; Ewing, 1927, 1934; Fitzinger, 1867; Flower, 8. S., 1931; Garcilaso de la Vega, 1869; Garman, 1875; Gilmore, 1947; Goeldi, 1906; Goldman (see Kellogg and Goldman, 1944; Young and Gold- man, 1944); Granath (see Coates, Cox, and Granath, 1937) ; Greenhall (see Ditmars and Greenhall, 1935); Gudger, 1919, 1923, 1925, 1927, 1930, 1943; Harper, 1945; Hayman (see Ellerman with Hayman and Holt, 1940-41); Hertig, 1942; Holt (see Ellerman with Hayman and Holt, 1940-41); Hooton, 1942; Housse, 1940; Howes, 1930; Hrdliéka et al., 1912; Hudson, 1892, 1920; Huysmans, 1945; Ihering, R. v., 1934, 1940; Ilgin, 1941; Ignacio de Armas, 1888; Jull, 1927, 1930; Kellogg and Goldman, 1944; Killip and Smith, 1931; Kraglievich, 1930 a, 1930 b; Krieg, 1925, 1929, 1939; Krieger, 1929; Lane, 1943; Latcham, 1922, 1936 a; Leén, J. A., 1939; Ley, 1941; Lopez Aranguren, 1930; Lotsy and Kuiper, 1923-24; Lydekker, 1896; Maccagno, 1932; MacDonagh, 1940; Madueno, 1912; Maregrav de Liebstad, 1648; Marelli, 1931; Matthew, 1939; Mello (see Ubisch and Mello, 1940); Mello Leitao, 1937; Métraux, 1942 a, 1944 a; Michaelson, 1936; Miller, 1912, 1916, 1927, 1929; Miller and Krieger, 1929; Montandon, 1929 c; Moodie, 1923; Morais Rego, 1944; Murphy, 1936; Myers, 1938; Nehring, 1888, 1889, 1891, 1893, 1894; Nordenskidld, 1922, 1930; Osgood, 1915, 1934; Oviedo y Valdés, 1851; Peterson, 1904; Phillips, 1922; Pocock, 1923; Preston, 1939; Rengger, 18380; Ridoutt, 1942; Romero, E. C., 1925, 1927; Rusconi, 1930, 1933; Santesson, 1935 a, 1935 b; Santos, E., 1942; Schmidt, M., 1929 b; Schmitt, 1931; Schwartz, 1939; Sclater, W. L., and Sclater, P. L., 1899; Scott, 1937; Sellards, 1940; Silva Coutinho, 1906; Simpson, G. G., 1934, 1940 a, 1940 b, 1948, 1945; Smith, A. C. (see Killip and Smith, 1931); Smith (see Elliot Smith) ; Speck, 1935; Strong, 1934, 1942; Stroock, 1937 a, 1937 b; Swales (see Wetmore and Swales, 1931); Tappy, 1944; Tate, 1932; Thomas, O., 1910, 1917; Townsend, 1910; Tozzer and Allen, G. M., 1910; Tchernavin, 1944; Tschudi, 1844-46; Ubisch and Mello, 1940; Verissimo, 1895; Vogt, 1942; Wassén, 1934 c, 1934 d,; 19385; Wetmore, 1926, 1929, 1981; Wetmore and Swales, 1931; Yepes (see Cabrera and Yepes, 1940); Yacovleff, 1985 b; Young, S. P., and Goldman, 1944. THE USE OF WILD PLANTS IN TROPICAL SOUTH AMERICA By Criaupe Lévi-Strauss INTRODUCTION It is not always easy to distinguish between wild and cultivated plants in South America, and there are many intermediate stages between the utilization of plants in their wild state and their true cultivation. Karl von den Steinen (1894) gives several examples of these transitional stages in Central Brazil: among the tribes of the upper Xingi River (Handbook, vol. 3, p. 321), he saw paths lined with piqui trees, which generally grow wild, and with mangaba and uruct trees that had been transplanted near the settlements and arti- ficially irrigated. On the other hand, actual cultivation was very rudimentary. One native tried to plant discarded matches; others blew on tobacco plants to insure their growing. The Tupi-Cawahib of the upper Madeira River gather the seeds of an unidentified wild grass that grows in the forest, and in order to facilitate the harvest they tie together several stems before they are ripe, so that the seeds of several plants fall on the same spot and pile up in small heaps. The tribes of the Pimenta Bueno River leave on their clearings some palm trees, in the bark of which edible grubs develop. These are the first steps toward cultivation. W. EK. Roth (1924, p. 214) writes: It must be borne in mind that in the clearing of the forest the Indian will usually save from destruction any economic palms or edible fruit trees. Dance says that kushi ants will not have their nests near a cunaparu (Phyllanthus sp.) plant, the milky juice of which is acrid and insufferably irritant, and it is for this reason that many fields contain two or three of these plants. In the Tropical Forests, gathering as well as cultivation may be highly developed, for the utilization of wild plants often entails refined exploitative techniques that require far more than the mere collection of wild foods. Few people, for example, have made a staple of a food as highly poisonous as manioc. The great skill shown in utilizing the vegetable environment is also shown in the various uses made of the same plant. For example, manicoba (Manihot dichotoma, M. glazi- ov, M. heptaphylla, M. piauhyensis, M. violacea) is a source of poison; of rubber (borracha do Ceardé, de Jequié, de Manicoba); and of food, its grated roots being consumed as flour after the poison has been extracted and its oily seeds being eaten (Pio Corréa, 1909). Protium 465 466 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 heptaphyllum provides a balsam, a rosin for glazing pottery, and a drink, the last prepared from its fruits. The preparation of several wild foods requires various complicated processes, such as those for preparing bitter manioc and green-heart seeds (Nectandra rodioet). Roth (1924, p. 218) describes the latter: The seeds are grated and put in fresh water, and a matter precipitates similar in appearance to starch. It is repeatedly washed to lessen its bitterness, which is never lost entirely. It is then mixed with rotten wood, pounded previously and sifted, and those who have it in their power mix a little cassava flour with it. In tropical South America, the general cultural levels are determined historically rather than by the local plant resources, for no fundamen- tal culture traits appear to depend directly on the botanical environ- ment. Pine nuts in southern Brazil and Brazil nuts in northern Brazil are two important foods not found elsewhere, yet no special aspect of the culture of the tribes exploiting them can be directly related to their exploitation; conversely, no special traits are found in the areas lacking these nuts. Fibers from palm trees (Astrocaryum sp.) and from a bromeliad (Bromelia sp.) are used indifferently in the same area, though palm fibers are more commonly used in the north and Bromelia in the south, and differences in materials and techniques between these two areas are insignificant. Nordenskiédld (1924 a) is responsible for the notion that wild plants ‘‘set their stamp on the culture of the Indians.’”’ The example which he gives is unconvincing. He writes: | Thus, in 1909 I came across a couple of Guarayt Indians on the Rio Parapeti. They had long portable baskets woven out of paripinnate palm leaves. As we entered the Parapeti territory the baskets became worn out, but as there were no paripinnate palms in this part, they could not make new ones. If, for any reason, the Guarayt tribe were forced to migrate from their present region to the Parapeti region they would have to change the type of their portable baskets. [Nordenskidld, 1919, p. 15.) The statement draws its importance from a former statement by Nordenskidld that ‘fanshaped leaves are of little use, while paripin- nate are so useful’ (ibid., p. 4). As a matter of fact, the Guiana Indians used both fan-shaped and paripinnate palm leaves, and both kinds have about equal value in basketry, so that the presence or absence of one or the other is of little consequence. The difficulty of the Guarayt% mentioned was culturally, not environmentally, caused. The facts, indeed, point in a quite different direction. Many vegetal species in South America have a widespread distribution, and the same vegetal environment surrounds tribes far distant from each other. For purely cultural reasons, these tribes make a very different use of their environment. The distribution of Ficus, Bombaz, Bertholletia, and Cariniana does not explain the presence or absence of bark cloth; the two great centers of bark cloth, i. e., Vol. 6] USE OF WILD PLANTS—LEVI-STRAUSS 467 the upper Amazon. and northeastern Bolivia, are cultural—not geo- graphic—centers. The failure of Central Brazil to develop this industry was not because of the lack of convenient material; the Bororo, for example, make bark cloth, although only for the perineal band of women’s dress. The striking fact is that, far from depending wholly on the natural environment, South American Indians throughout the tropical area show exceptional ability to discover substitutes wherever a vegetal species is lacking. For example, Pardal quotes the substitution of the decoction of the bark of pariah (Simaruba, Simaba, Picrasma) for uructi (Biza orellana) in the southern part of the tropical area where it is difficult to grow uruci. The principle of body ointment and adornment is preserved; the plant varieties used for this purpose differ. The same is true of the balsams: in the south, Copaifera langsdorfit replaces Copaifera multijuga of the Amazon; and when the Leguminosae listed in pharmacopoeia as yielding benzoin are lacking, they are replaced by either Myrocarpus or liquidambar (Pardal, 1937, pp. 104-105). For the varnishes, Protiwm hepta- phyllum served in the north, Bulnesia sarmienti in the south; for stimulants, guarandé in the north, maté in the south; for weapons, arrow shafts are made either of taquara (Chusquea sp.) or of Gynerium sagittatum, according to the lack of one or the other in a definite region. The Chané, who lack even the latter, have replaced it with Arundo donaz (Nordenskiéld, 1920). It is also difficult to agree with another statement by Nordenskiéld (1919, p. 4), who says, ‘‘that the abundance of wild truits, as well as the intensive dryness during part of the year and the flood during another part, account for agriculture being so underdeveloped in the Chaco.”” Nowhere in South America has the abundance of wild resources impeded agriculture. On the contrary, the various inde- pendent places of origin of agriculture postulated by Vavilov (1926, and after Sauer, 1937) all have many kinds of wild foods, and in South America incipient farming and developed exploitation of wild resources tend to be associated rather than mutually exclusive. That the abun- dance of wild foods does not preclude farming is shown in the case of Guiana: One Indian (Akawai) will clear and, with his wife, plant 2 or 3 acres in as many weeks, and 7 or § acres will supply them with a year’s food, so that 10 or 12 weeks in the year is absolutely all that is required for actual labor, and the rest of the time remains for pleasure, hunting, and fishing. [Roth, W. E. 1924, p.214.] In the mind of the South American Indian, the principal geo- graphical distinction is that between the savanna and the forest. The first is unsuitable for cultivation as well as for gathering and collecting wild foods; both animal and vegetal life on it are sparse. 79471150 31 468 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 The forest offers abundant wild plants and game, and its moist soil is fertile. The stupidity of the deer which in a myth tries to culti- vate manioc in the savanna filled the Bacairi with mirth, according to Steinen (1894, p. 488). Cooper (1942 a, 1942 b) has suggested that the tropical area of South America could be divided into two subareas according to the level of cultural achievement: “. . . the Orinoco-Amazonian farmers and the scattered tribes subsisting by a purely collecting economy or else with a rudimentary or recently acquired horticulture’ (Cooper, 1942, b p. 147). The same author suggests that, considering the fairly close correlation between the cultural groups and the natural areas, the first group should be called Silval and the second Marginal, the latter subdivided into a Savannal and an Intrasilval subgroup. Irrespective of the usefulness of such a classification for practical purposes, it is necessary to keep in mind that farming always accompanies, and is never a substitute for, the exploitation of wild resources. The Silval area is not only an area of farming but is one with abundant wild vegetal food and industrial plants. Moreover, few tribes subsist solely by a collecting economy, and they are distributed at random in such varied places and in such geographic environments (the forests of Paraguay and the Guajira Peninsula, for instance) that their lack of farming seems to depend much more on the cultural history of each separate region than on geographical factors. Finally, there is no reason to consider that the rudimentary agriculture of the great majority of the savanna tribes was recently acquired. These remarks lead to the following conclusion: The characteristics both of farming and of the exploitation of wild plants in South America show that their place of origin was the Tropical Forest or the banks of the northern streams which are naturally bare and remain un- covered by water during most of the year (Roth, W. E., 1924, p. 214; see also Sauer, this volume, pp. 331-344). This Silval culture, based altogether on farming and on the exploitation of wild resources, which requires as much skill as farming, is the only genuine culture of tropical South America. (See also Handbook, vol. 3, pp. 883-886, on the two culture types in the Tropical Forests.) All South American tribes clung to the forest whenever they were forced to change their habitat. This was true in the case of the Tupi during their long and widespread migrations. Petrullo (1932) noticed that the inhabitated area of the Xingt River begins only at the points where the gallery forest becomes a true rain forest spreading inland. The savanna, where manioc does not grow, is always avoided and probably was occupied only by tribes driven into it by stronger populations. In the savanna the horticultural pattern was partially retained by turning to the best possible account the strips of gallery Vol. 6] USE OF WILD PLANTS—LEVI-STRAUSS 469 forest along the streams. It was sometimes improved, as shown by Nimuendajii’s discovery among the eastern Ge of a cultivated Cissus not reported elsewhere. Farming was abandoned in favor of hunting (Bororo) or of collecting and gathering wild foods, or of both. But there is little doubt that all nonhorticultural South American tribes were formerly farmers. The well-known text by Karl von den Steinen (1894) about the behavior of the Bororo in the presence of the gardens opened by the Brazilians is of little weight when compared to the fact that these very Indians were acquainted with an elaborate harvest ritual. Farming might have been forgotten among some Bororo as a result of the abundance of game alorg the marshes, but agriculture was not new to them. Utilization of wild foods exists in the tropical area on two different levels: a basis level, in which it coexists with farming and is centered in or around the forest; and a subsidiary level, which is one of col- lecting brought about by compvu!sive adaptation to the savanna and which often remains partial and is always secondary. THE PALMS Several species of palms played an outstanding part in native cultures. Thus, for instance, Gumilla (1791, 1: 145) remarks that the muriche palm (Mauritia flexuosa) was the mainstay of the Warrau economic life. From it these Indians obtained wood for their pile dwellings, fiber for their clothes, ornaments, hammocks, and fishing tackle, starch for making bread, sap for their wine, the fruits for a sort of punch, and leaves for their baskets. They also extracted large edible larvae from its decayed trunk. The pupunha, or peach palm (Guilielma gasipaes), is a palm long cultivated by the Indians, though it still grows wild. The edible fruit of the cultivated tree lacks the thick shell characteristic of the wild varieties. Palms are semicultivated, for wild palm trees are often spared on a clearing and tended together with the cultivated plants. About 20 genera of palms were widely used, being exploited for the following purposes: Edible fruits.—Several genera yield nuts which are edible after the shell of the fruit has been broken. Most important in the native diet are the uaguassi (baguassu, babassu), or pindoba nut (Orbignya speciosa), which is rich in oil, and the nuts of the genera Acrocomia, Astrocaryum, Attalea, Catoblastus, Cocos, Copernicia, and Maximiliana, which have different food values. With other species, it is not the nut but the fleshy substance sur- rounding it which is consumed. Both the nut and the flesh are eaten of the mucaja or bacaiuva (Acrocomia), but only the flesh is eaten 470 SOUTH AMERICAN INDIANS [B. A. B. Bull. 148 of the caranai (Mauritia horrida) and the buriti (Mauritia vinifera) in central and western Brazil, and of the miriti or ite (Mauritia flexuosa) in Amazonas and Guiana. This fruit is all-important in the diet of many tribes, because of the many vitamins contained in the mush prepared with its orange-yellow pulp. Thevet (1878) describes the uricuri or buri da praia (Diplothemium maritimum), a small tree with edible fruits relatively abundant between Rio de Janeiro and Cabo Frio. The fruits of several palm tree genera are used only to prepare beverages or mushes. The most important are the assai (Huterpe oleracea, E. precatoria), the manicol (Huterpe edulis), the bacaba or turu (Oenocarpus distichus, O. bacaba), the lu (Oenocarpus sp.), the patua or pataua (Oenocarpus patua), the aeta (Mauritia flexuosa), the kokerit or anajé (Marimiliana regia), the awarra or jawari (Astrocaryum tucumoides), and the maraj& (Bactris minor). In most cases the ripe palm fruit is soaked in lukewarm water—boiling water would harden instead of softening them—and then the pulp is sepa- rated from the shell or kernel and made into a thick, oily, fragrant drink, which has a high nutrient value. These drinks may be con- sumed immediately or after standing a night, which gives them a slightly sour taste. Sometimes manioc flour is added to them. Palm wine.—The sap of the Mauritia vinifera is drunk fresh or slightly fermented. It is collected in a trough-shaped cavity dug in the trunk of a felled tree (Warrau). The coroxo wine is made from the fruits of the Acrocomia aculeata. (See also Handbook, vol. 1, p. 418.) Palm cabbage or palmito.—The “‘palmito,” i. e., the terminal shoot of several species of palm, is one of the few fresh vegetables in native diet. It is eaten raw, broiled, and sometimes boiled. The palmito of almost all the palm species can be consumed, but some have a bitter taste, as for instance the Acrocomia. The Brazilian Indians show a marked preference for the palmitos of the Euterpe, Cocos, and of several species of Iriartea. In the Chaco, the Indians consume the palmitos of the caranday (Copernicia cerifera). Starch.—The Warrau extract starch from Mauritia in the following manner: When an ite tree begins to fructify it is cut down, a large slice is cut off one side, and the stringy substance of the interior is cut into shreds, the remainder of the trunk serving as a trough, in which it is triturated with water, by which is dis- engaged a considerable quantity of starch. The fibrous particles are then ex- tracted, and the sediment, or aru, formed into molds like bricks. This is spread out on stones or iron plates over the fire, and makes a very nutritive but at the same time unmasticable bread. [Roth, 1924, pp. 215-216; cf. Gumilla, 1791, 1: 149.] This starchy food is known under the name of sagu in northern and Vol. 6] USE OF WILD PLANTS—LEVI-STRAUSS 471 eastern Brazil (Pio Corréa, 1909). The Guayaki extract a starchy flour from the pindo palm (Cocos romanzoffiana) (see Handbook, vol. 1, p. 486; also Vellard, 1939, p. 84). Oil.— Oil can be extracted from several palm fruits by crushing and boiling them. It may be used in cooking, for lighting purposes, or in medicine; but more often the Indians mixed it with uructi or some other pigment to smear on their bodies. The palm species which produce oil are: Orbignya speciosa, Astrocaryum tucuma, Astrocuryum tucumoides, Attalea speciosa, Mazximiliana regia, and Oenocarpus (O. bacaba and patua). Salt.—The ashes of the fibers and of the fruits of some palm trees, such as jara (Leopoldinia major), and of the leaves of some other species, such as Mauritia flexuosa, are boiled and the decoction is allowed to evaporate in order to obtain a brownish powder which is used as salt. Staden (1928, pt. 2, ch. 11) saw and describes the whole process among the ancient Tupinamba: From the ashes of a palm trunk, they make a solution which they boil until the salt is separated. “Tt tasted like salt and was grey in colour.” House thatching.—Palm leaves are the most common plant ma- terials for thatching the roofs and frames of native huts. The method of thatching depends upon the nature of the leaves. If the fronds are paripinnate, such as those of the anajé, the leaflets are made to fall limp and loose by tearing loose the “‘eye,”’ 1. e., the internal articu- lation of the leaves with the midrib. The palms are attached hori- zontally to the purlines, overlapping like tiles. For fan-shaped leaves, the techniques are more elaborate. The ancient Tupinamba parched the leaves of the pindoba over a fire and then plaited them before thatching their huts. Among the Guiana Indians, palm leaves preferred for thatching are: the truli or bussi (Manicaria saccifera), caranai (Mauritia horrida), buriti or ite (Mauritia vinifera, M. flexuosa, or M. armata), dallibana (Geonoma baculifera), ubim and several Geonoma (G. elegans, G. paniculata, G. pohliana, G. schottiana), anaja or kokerit (Maximiliana regia), manicol (Huterpe edulis), turu or bacaba (Oenocarpus bacaba), assai (Euterpe oleracea), etc. (Roth, W. E., 1924, pp. 265-270). Basketry.—Max Schmidt (1905) classifies twilled basketry in two classes depending on whether fan-shaped or paripinnate leaves are used. He believes that many decorative motifs in the art of these Indians come from the basketry patterns that are inevitably produced by using fan-shaped buriti palms. Paripinnate leaves, such as those of Mazximiliana regia, Orbignya speciosa, Orbignya phalerata (cusi of the Chaco), two species of Astro- caryum (respectively, awarra and akko-yuro in the Guianas, tucim and tucumé4 in eastern Brazil, murumuru or Astrocaryum murumuru 472 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 in the Amazon, and several species of Desmoncus (kamwarri or jacitara) are particularly suitable for making fans, mats, and temporary carry- ing baskets. With the fan-shaped leaves of the buriti or ite palm, the Indians of eastern Brazil weave fire fans, containers, trays, and rectangular baskets which are characterized by geometrical patterns (lozenges, ete.). Twine, cords, and strings——The young unopened leaves of the Mauritia flecuosa reduced to the cortical strips and soaked in water for several days are made into cords which have many uses in the Amazon Basin. The fibers of several Atialea, mainly those of the piacaba (Attalea funifera) or chiquichiqui (Leopoldinia piagaba) pro- vide material for thick or small ropes. The name ‘‘tuctim”’ is given to several species of palm trees, mainly Asérocaryum or Bactris, particularly Bactris setosa, which give excellent fibers for strings and ropes used for making hammocks, nets, etc. These species are as important to the Indians of the Amazonian or Orinoco Basins as the Bromeliaceae known as caraguaté are to the Indians of the Chaco. The name “‘jupati’” is given to plants of the genus Raphia. Wooden objects.—Posts, fences, and palisades are often made of the wood of the catizal or paxiuba (Socratea exorrhiza and Socratea durissima). The natives of eastern Bolivia and the upper Amazon make their bows of the hard, black wood of chonta palm (Guilielma insignis). Clubs and spears are often carved of the same wood. The stem of the paxiuba (Socratea exorrhiza) serves to encase the Arundinaria tube of the blowgun. The gigantic trumpets of the Uaupés River Indians are made of sections of paxiuba palm (Socratea exorrhiza) wrapped with long strips of iebaru (Hperua grandiflora). Finally, the vegetal wax of the carandai or carnauba palm tree (Copernicia cerifera) must be mentioned here. This species is especially important in northeastern Brazil. Nordenskidld (1929 a, fig. 1) has published a Chacobo manioc grater from Bolivia. It consists of a section of the trunk of a thorny palm tree. Often, to make a grater, thorns are imbedded in rows in a wooden plank. Beads and ornaments.—The black polished shells of the small nuts of the Astrocaryum are practically everywhere carved into beads, earrings, and other types of ornaments. The wood of other palm trees is occasionally used for miscellaneous purposes. These are pati (Orcus sp. and Cocos botryphora), buri and buri-assu (Diplothemium caudescens, D. campestre), aracuri (Cocos coronata), curua or acuri or auacuri (motacu in the Chaco), several species of Attalea (A. speciosa, A. phalerata, A. princeps, A. spectabilis), buritirana (Mauritia aculeata) etc. Vol. 6] USE OF WILD PLANTS—LEVI-STRAUSS 473 TIMBER WOOD The number of species of trees used by the Indians in their industries is so large that a complete list would fill a volume. Here are enu- merated the names of species most frequently mentioned in the old literature dealing with the culture of the Brazilian Indians. Several kinds of aroeira are used: aroeira branca (Lythraea moleoides, L. brasiliensis), aroeira molle (Schinus molle, which also yields the so-called American mastic), and aroeira vermelha (Schinus terebin- thifolius). ‘“Cedro”’ comes from several families of plants: imbuia (Bignonia sp.); cabreuva (Myrocarpus sp.); acareuba (Calophyllum brasiliense) ; conduru, a red wood (Brosimum conduru); ubiraeta or iron wood (Caesalpinia ferrea); barauna (Melanoxylon brauna); jurema (Pithe- cellobium tortum and Mimosa verrucosa); ivory white or ‘pau marfim” (Balfourodendron riedelianum); red guarabu (Peltogyne confertiflora) ; black caviuna or jacaranda (Dalbergia nigra); ‘“‘Vinhatico,” a yellow reddish wood (Plathymenia reticulata); and piquihi (Caryocar barbi- nerve). The genera Tecoma and Couralia provide various reddish and blackish woods. Jatahi and jatoba are trees of the genus Hymnenaea; macarandiba is Lucuma procera; guapeveira is a species of Chrysophyllum; andira or “pau de morcego” is Andira rosea or A. fraxinifolia; jequitiba is Couratari brasiliensis; sucupira is Bow- dichia wirgioides and Pterodon pubescens; arariba or araruva, a striped wood, is Centrolobium robustum; urucurana is Hieronymia oblonga and a species of Alchornia. In addition, several palms, especially Orbignya, Astrocaryum, Guilielma, and Iriartea, are used for hut frames, weapons, fences, ete. Some woods are traditionally used for making specific objects. Clubs and macanas are generally carved of the hard wood of various Leguminosae, especially purpleheart (Copazfera pubiflora, Caesalpinia sp., and Myrocarpus sp.), snakewood (Brosimum aubletir), and amara (Schwartzia tomentosa). The Tupinamba used ibiratinga (Funifera sp., of the family Thymelaeaceae) to make the staves of their spears. The Guiana indians made their best paddles of the fluted projections of the yaruru or paddle wood (Aspidosperma excelsum) ; the Tupinamba of Genipa americana or of uaca (Ecclinusa ramiflora). The light woods or ‘‘gameleiras” include a great many species of Ceiba (e. g., copaubucu, Ceiba erianthos) and Ficus, as well as ubira- gara (“barriguda,” or “‘barrigudo” tree, Cavanillesia arborea, and several other Bombacaceae), umbaubeira (Cecropia adenopus), apeiba (Apeiba sp.), and paraparaiba (Cecropia and Triplaris). These light woods are used mainly for making ear or lip plugs (Suya, Botocudo, etc.), cylindrical containers for feathers and orna- 474 SOUTH AMERICAN INDIANS [B. A. E. Bull. 143 ments (Bororo, etc.), rafts or jangadas (on eastern Brazilian coast, Apeiba sp. or apei is used), and canoes (Cavanillesia arborea and Ceiba pentandra). Canoes.—In Guiana, canoes and corials were made out of the following trees: Siruaballi (Nectandra spp.), tenyari or mara (Cedrela odorata) ; purpleheart (Copaifera pubiflora), kabukalli (Goupia glabra), itenalli (Vochysia tetraphylla), silk-cotton tree (Ceiba pentandra), crab-wood (Carapa guianensis), incense tree (Protium guianense), Dimorphandra mora, and several species not yet identified. In northern Brazil canoes were dug out of the trunks of Cedrela odorata and Ceiba pentandra. The Indians of central Brazil make their canoes from the bark of jatoba (Hymenaea courbaril). The same bark was probably used by the Tupinamba. The Tupi dug canoes out of a Bombacaceae or of Ficus doliaria. Iriartea ventricosa is used for the same purpose. Bows.—In the Guianas bows are made from at least half a dozen different timbers. Those which have been identified are the purple- heart (Copaifera pubiflora), burakura, burukuru, burokoro, leopard- wood, or snakewood (Brosimum aubletii), and Lecythis ollaria. In Brazil the most common bow wood is Tecoma conspicua, called for that reason pau d’arco. Perfumed woods.—Beads of necklaces are often carved from fragrant woods. These are: carunje (species of Ocotea and Nectan- dra), vanilla (Vanilla sp.), cinnamon wood or anhaybataa (Pseudo- caryophyllus sericeus, Cinnamodendron azxillare, and Capsicodendron pimenteira), rosewood or jacaranda (Dalbergia nigra), and pau santo (species of Bulnesia and Zollernia). Certain other woods have an unpleasant odor: ubirarema or “‘canella merda” (Nectandra myriantha), pau d’alho (Gallesia scoro- dodendrum), and several vines which smell like garlic (Lundia longa, Clytostoma noterophilum, Segueira floribunda, Adenocalymna allia- ceum, etc.). FIBERS Fibers used by tropical Indians in their industries come mainly from palm trees (see p. 472). Fibers are also extracted from several Bromeliaceae, mainly Bromelia fastuosa and B. serra, which are known as caraguatd, gravata (Tupi), chaguar (Quechua), pita, kuraua, etc. In the Guianas and in many regions of Brazil, the Indians utilize the fibers of both palm trees (tuctim) and of bromeliads, though the first give thinner and better strings. In the Chaco and southern Brazil, the Indians employ almost exclusively fibers of Bromeliaceae. In Colombia, Ecuador, and Pert, the Indians obtain the fibers for their ropes and textiles from the agave. Vol. 6] USE OF WILD PLANTS—LEVI-STRAUSS 475 The Brazilian Indians use vines and creepers of many species for ropes, cables, or strings. Those called cipo belong to many families and genera. Cipo-embé is the adventitious root of a plant of Philo- dendron. The timbé group includes thinner varieties of vines (Ser- jania and Paullinia) which may be twined or plaited. Timborana (Lonchocarpus sp. and Malpighiaceae sp.) serve the same purposes. In the northwestern parts of South America, the name bejuco is given to Vitis tiliifolia, Trichostigma octandrum, Entada gigas, ete. Roth (1924, p. 118) lists nibbi or sippi, mamuri, and muna (Carludo- vica sp.) as bushropes which in their natural condition may be used as twines. Paina (Chorisia speciosa), embiriti (Bombax munguba), and goay- aimbira (Cecropia concolor) are trees which yield a fibrous substance used by the Indians. The cottonlike wads which the Indians attached to the buts of blowgun darts come from the fruits of Bombax globosum and EHriodendron samauma. Bark cloth.—Bark cloth may be obtained from several species of trees. In eastern’ Bolivia and on the upper Amazon the Indians use bark of trees of the genus Ficus, which are known in Bolivia as bibosi. From the Guaporé to the Orinoco River cloth is made from the bark of species of Bertholletia and Cariniana; in the northern parts of the continent and in the West Indies, from cabuya (Fucraea gigantea), majagua (Hibiscus tiliaceus), memiso (Muntingia calabura), and man- barakrak (Lecythis ollaria); and in eastern Brazil, the bark of embiriti (Bombax munguba). MISCELLANEOUS According to Von den Steinen (1894), the Xingi River Indians cultivated a wild grass which they used as razor blades. The fruit of a Bignoniaceae (‘‘pente de macaco,” Pithecoctenvum echinatum) was used as a comb by the Tupi and other tribes. The Tapirapé use the fruits of a grass (‘‘capim flecha,”’ Streptogyne crinita) as tweezers for plucking the hair (Baldus, quoted by Hoehne, 1937, p. 115). As the two more important fruits used as containers were cultivated (Crescentia cujete, cuia, gitira, and Lagenaria siceraria), only the shell of Lecythis blanchetiana (one of the numerous sapucaia nuts) and the hollowed-out seeds of several palm trees (Astrocaryum) and the staunch flour containers made from the leaves of Heliconia and Cala- thea may be mentioned here. The leaves of Heliconia bihar and of several Geonoma were also used for roof and wall thatching. The calabashes of the upper Rio Negro are lacquered with a decoc- tion of carayuru—Bignoniaceae and cassava leaves sprinkled with human urine (Roth, W. E., 1924, p. 302). In addition to numerous palm nuts (see The Palms, p. 469), a great many nuts (genera Bertholletia and Lecythis) and seeds (“olho de 476 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 cabra”’ or comedoi: Ormosia nitida, Omphalea diandra, Myrozylon toluiferum, and others) are used as beads in necklaces, tassels, and the like. Several hard-shelled seed pods, including Thevetia peruviana and Juglans sp., were made into rattles for accompanying songs and dances. Also used for rattles in Guiana were many nuts and seeds known only by their local names: Kawa and cerewu or cerehu seeds and caruna and ahouai nuts. The hollow cylinders used to strike the ground as a dance accompaniment are made of trumpet wood (Cecro- pia) or bamboo (Gadua). The drums of the Orinoco are hollowed from the trunk of several trees: silverballi (Nectandra); karuhoho (Arawak), simaruba (Warrau) or muratatau (Carib); omu (Warrau); and some- times of Mauritia flecuosa (Roth, W. E., 1924, pp. 464-466). Bundles of palm leaves (Mazximiliana maripa) enter into the construction of the Oyana drums dug in the ground and struck with the feet (ibid., pp. 468-469). GUMS AND RESINS Most of the gums known to the Indians are obtained from the Leguminosae. They are generally used as drugs, but may also be employed in the native industries. Lighting substances.—All kinds of rubber burn quickly with a bright flame. In addition to its other industrial uses, rubber is collected in lumps of coagulated latex and carefully kept to light fires. The lump is drilled with the fire drill, which produces a highly in- flammable dust. The rosin of the locust tree (jatoba, jatahi, simiri, algarroba—names which in different regions are given to the same or different species: Hymenaea courbaril, Cassia blancheti, and others) when lighted or thrown in a fire gives a brilliant light. Guaconax (Amyris maritima) and almecega (Hedwigia balsamifera and Protium heptaphyllum, the latter ‘the haiowa of the Arawak, shipu or sibu of the Warrau, and sipipio of the Carib’? (Roth, W. E., 1924, p. 80)) serve the same purpose. Plastic substances.—On the upper Guaporé River, the translucent rosin of Hymenaea courbaril and probably of many other trees is made into nose and lip plugs by means of wooden molds. The Guarani used the rosin of a tree called abati timbaby in the same way. Several rosins are applied as a glaze to the whole or to parts of newly fired ceramics: in southern Brazil, Paraguay, and northern Argentina the rosin of “pau santo,” ‘‘palo santo” (Bulnesia sarmienti); and in the north, that of icica (Protium brasiliense, P. heptophyllum, P. guianense, P. aracouchili, P. carana) and of Hymenaea courbaril. The Nambi- cuara make their pots waterproof by washing them when still hot with a decoction of the rosinous bark of a Mimosa. The same thing is said of the Warrau (Roth, W. E., 1924, p. 133). Vol. 6] USE OF WILD PLANTS—LEVI-STRAUSS 477 Gums and glues.—The rosins already mentioned and especially that of balata or turara (Mimusops globosa) and of ‘‘pau breo,’”’ manni, ohori, or manil (Moronobea coccinea) are widely used for fastening points and feathers to arrows, for waxing threads, for calking canoes, and for attaching stone chips or thorns on cassava graters, blades to knives, and axes to their handles. Balsams.—The rosin of the Hymenaea courbaril is chewed for pains in the stomach and flatulence, and is burned for fumigations in case of colds and headaches. The balsam of acouchi (Protiwm sp.), umiri (Humiria floribunda), and wallaba (Eperua sp.) are said to heal wounds. ‘The same virtues are ascribed by Indians and Whites to Coparfera multyuga, C. officinalis, and C. langsdorfii, which are known as copayba, cabima, curucay, curaki, purukai, mawna, mararen, and maran. Corohiba or cabureiba (Myrozylon toluiferum, the tolu balsam of Colombia and Venezuela) is popular with the Indians of central and southern Brazil. According to Soares de Souza (1851), the ancient Tupinamba collected it with pads of cotton, which they later squeezed. Other balsams used by the tropical Indians are: cabreuva (Myro- carpus frondosus, M. fastigatus), obira (Apocynaceae), imbauba or ambay (Cecropia adenopus), corneiba (Schinus terebinthifolius or Luthraea brasiliensis), and gayac (Guaiacum officinale). The Guayba, Tunebo, and Chiricoa used a rosin called mara (Protwum sp.) for hunting deer, which are said to be attracted by its odor. OILS AND UNGUENTS To make oil of crab wood (caraba or andiroba, Carapa guianensis) “the Roucouyenne of Cayenne . . . preserve the seed for a year by burying it in the ground and making veritable silos of it . . . The Oyampi of the Oyapock River boil the seeds, expose them for several weeks to the air in a scooped-out tree trunk, crush them with their feet, and finally let them drip on an inclined palm leaf” (W. E. Roth after Crévaux, 1924, p. 85). The oil is used to anoint the hair and skin and to prepare paint. From makeima bark (Mespilodaphne pretiosa, Roth, W. E., 1924, p. 86) the Macushi extract an ethereal oil for use against diarrhea and dysentery. Most species of Lecy- thidaceae, some of Hugenia and Virola, and Bertholletia excelsa (the Brazil nut) have fruits which, when boiled and crushed, yield an oil or a vegetable tallow which may be used as a food, as an unguent, or for lighting purposes. Palm oils are described under The Palms (p. 471). PIGMENTS AND DYES The most important pigment in all tropical South America is cer- tainly the uructi or roucou (called achiote in México, mantur in A78 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Quechua, bija in the West Indies, and, in other dialects, arnotta, faroa, kuseve, shiraballi, mubosimo, and majepa). JBiza orellana is cultivated by most tribes, although many of them merely transplant young wild plants to places near their huts. JBiza orellana is, how- ever, a widespread wild plant and thus belongs to the large group of plants which have not been modified through cultivation. The red pigment of the uructii comes from a thin skin covering the seeds. These are washed and mashed, and the pigment, which settles to the bottom of the container, is dried, mixed with animal or vegetal oil or gum, and made into balls or cakes. Uruct dye is used to color cotton thread and to paint weapons, ceramics, and implements, but it is employed especially to anoint the body and even the hair. Whether this widespread custom has a predominantly esthetic or hygienic function (protection of the body against heat and insects) is a much- debated question. In northern Argentina, where cultivation of Biza orellana is difficult, a decoction of pariah bark (Simaruba, Simaba, and Pricrasma) or “palo amargo,” is used instead. Bodily oint- ments are also made from guavira (Campomanesia), taperihud (Cassia), caburehi (Myrocarpus), and isipo kati (Aristolochia) (Pardal, 1937, pp. 99 ff.). Thevet (1878), Léry, Gomara, and others have de- scribed the unguent used against body parasites and made of hibou- couhu; Hoehne considers this plant a Myristica (Hoehne, 1937, p. 126). Another red paint used for body and pottery decoration and varying from orange to purple according to the technique of preparation is caraweru, barisa, barahisa, biauro, etc., which comes from the boiled or fermented leaves of Bignonia chica. It is kept in small straw baskets or in tubes. Unripe Genipa fruit of the jagua or xagua, launa or lana, tapuriba, tabuseba, etc. (Genipa americana) yields a juice which becomes black or dark blue when exposed to the air. From Argentina to the Guianas, it is used as a dye and as a paint for the body and for pottery and utensils. The tree is sometimes cultivated but also occurs wild. From arrisaura or karasaru berries the natives of the Guianas extract a clear blue used for body paint. The following dyes are used more for native handicrafts than for the body: ‘Pau brasil” of the old travelers or oroboutapn of the Tupinamba (Caesalpina echinata), used to dye feathers red; other red dyes are yzipo roots mentioned by Dobrizhoffer (1822), mespil or itarra (Bellucia aubletii) used in the Guianas to paint the paddles, the inside of calabashes, etc., maba bunakara (Coussapoa latvfolia) and buri-badda (Homalium ?), Maparakuni erythroxylum of the north- eastern Arawak, and kuruwatti (Renealmia exaltata) used as a remedy for ophthalmia, as a dye, and on the Pomeroon supposedly as the pigment for tattooing in the old days (Roth, W. E., 1924, pp. 90-91). Vol. 6] USE OF WILD PLANTS—LEVI-STRAUSS 479 Besides Genipa americana, a black dye is extracted from inga or shirada bark (Inga lateriflora). It is mostly used in basketry. Sey- eral yellow pigments come from tatajiba or tayuva (Chlorophora tinctoria), which has an edible fruit, and from an unidentified plant which Soares de Souza called caapiam (Hoehne, 1937, pp. 241-242). A blue, indigolike dye comes from Anil trepador (Vitis sicyoides) and from anil-assi (Hupatorium sp.). The old literature mentions several other vegetable dyes which are not yet identified. The sakuapéra of the Arawak and Warrau is Henriettea succosa (Roth, W. E., 1924, p. 90). SHAMPOOS Shampoos include the Brazilian“ arvore de sab4o’’; this is the Guiana and West Indian ‘‘I”’ of the Taino, identified by Roumain (1942, pp. 65-66) as Gouania lupuloides or G. polygama, certain roots and fruits of Sapindus divaricatus used by the Tupi, cjaru (Colletia spinosa) of Bolivia, Chile, Uruguay, and Argentina, and Sapindus saponaria of Venezuela, Brazil, and Argentina. RUBBER = According to W. E. Roth (1924, pp. 83-84), Sapium jenmani, or S. cladogyne, and some species of Hevea were probably the original sources of rubber. The Omagua made balls, rings, and syringes from the milky sap of a creeper which, from the structure of its fruits and flowers, must be ascribed to a genus of the Apocynaceae. The rubber balls of the Paresst and Nambicuara are made of the latex of mangabeira (Hancornia speciosa). The name ‘‘caucho”’ (rubber) is given to Castilla elastica and to Sapium eglandulosum. The rubber latex of soveira or sorveira (Cowma guianensis, C. macro- carpa, C. utilis) can be drunk either pure or diluted with water. The Indians coat their skin with the latex of several rubber trees in order to suffocate parasite worms lodged under the epidermis. FOODS Tubers.—-Wild food plants include roots, fruits, nuts, and shells. Few tubers have been identified, because the Indians generally col- lected them only in time of scarcity, when they were substitutes for manioc, Dioscorea, Zanthosoma, and other cultivated tubers. Among the wild tubers eaten is mandioquinha do campo (Zeyheria sp.). Green vegetables.—Green vegetables are scarce in native diet, being limited to manioc leaves, palm shoots (q. v.), a species of Cissus, which Nimuendaji found cultivated among the eastern Ge, and a very few others. Nuts and seeds.— Besides palm nuts (p. 469), which are consumed from the Chaco to the Guianas, there are two other nuts of great value. 480 SOUTH AMBRICAN INDIANS [B. A. B. Bull. 143 The Caingang and Guarani of southern Brazil depend for several months each year on the nuts of Araucaria angustifolia. The Tupt called these iba, the fruit “‘par excellence.” To the Araucanians they are just as important. In the Amazon Basin, sapucaia (Lecythis ol- laria or L. pisonis) and Brazil nuts, tocari (Bertholletia excelsa), are not negligible items of native diet. These nuts were a favorite food among many tribes of the Beni and the Madre de Dios Rivers. Other nuts with food value for the Indians are those of piqui (“Almendras del Brasil’’), sawari or chachapoya (Caryocar barbinerve, C. brasiliense, C. tuberculosum, C. amygdaliforme) comanda-iba (Sophora tomentosa), comanda-assu (Mucuna altissima), and jatoba or locust tree (Hymenaea courbaril). In the Guinas, cassava flour is often increased, mixed with, or even replaced by flour made of the following seeds. Mora (Dimorphandra mora), greenheart (Nectandra rodiaei), dakamballi (Vouacapoua americana), pario, and nuts of the sawari tree (Caryocar tuberculosum). W. E. Roth (1924, pp. 219, 230) quotes Schomburgk on the occurrence of wild maize (?) on the eastern foot of the Pacaraima Range. Not only seeds and nuts are occasionally added to cassava flour, but also soft wood. Wild rice (Oryza subulata) is abundant in Uruguay, Rio Grande do Sul, in the marshes of the upper Paraguay and of the Guaporé Rivers (Hoehne, 1937, pp. 33-39), and in the Orinoco Valley, but the Indians of the last region do not seem to have consumed it. According to Hoehne, Oryza sativa may be aboriginal in South America. Although peanuts (Arachis hypogaea, A. nambyquarae) were generally cultivated, ‘southern Brazil, and particularly Sao Paulo, Parana, and Mato Grosso, is the land of origin of the different peanuts. All known species still exist there in wild state . . .’ (Hoehne, 1937, p. 216). Fruits.—Some widely distributed fruits were used both cultivated and wild: caraguaté (Bromelia fastuosa) ; inga, shirada, or pacay (Inga vera, I. lateriflora, I. bahiensis, I. fevillei); maracuja (Passiflora quadrangularis, P. alata, P. edulis); and pineapple (Ananas sativus). The use of the following centers around the Chaco, southern Brazil, southern Bolivia, and northern Argentina: The important algarroba or aloja (Prosopis alba), principally used as a drink; guabiroba (Myrtus mucronata and Psidium multiflorum, P. corymbosum, P. cinereum, P. guazumaefolium); guavira (Campomanesia); tamarin, common in the Chiquitos region but lacking in Paraguay, according to Dobriz- hoffer; tusca (Acacia aroma); taruma (Vitex montevidensis); and chafiar (Gourliaea decorticans). A curious use of quebracho, wood yielding tannin, may be mentioned here: The Guarani burn pieces of the tree tayy, receive the smoke or soot arising from them into a clean dish, and by pouring hot water upon it, convert it into ink which Vol 6] USE OF WILD PLANTS—LEVI-STRAUSS 481 mixed with gum and sugar is by no means to be despised. [Dobrizhoffer,1822 1: 398.] The “fructa do lobo” (Solanum grandiflorum, yielding the alcaloid, grandiflorina) has a central and southern distribution in the savanna lands of central Brazil. Its delicious large peachlike fruit seems to be the object of a food prohibition in several regions. Some varieties may be toxic. Cashew (Anacardium occidentale) is generally cultivated, but another wild species, Anacardium giganteum, yields small fruits which the Indians collect at the foot of the tree after the monkeys have thrown them down. The Brazilian Indians consume the fruits of several other species of Anacardiaceae: umbti (Spondias tuberosa), hobo, jobo (Spondias monbim, S. dulcis, S. robe), caja-mirim, maropi or hog-plum (Spondias lutea), and acaju or acaja (Spondias monbim). The tuber- like roots of umbit are edible. Mangaba fruits (Hancornia speciosa) are so important to the savan- na tribes that when they are in season the Indians undertake large expeditions for the sole purpose of collecting them. Likewise of great importance are the fruits of several Psidiwum (P. turbiniflorum, P. guayava, P. variabile), and of several Myrtaceae, such as cambuy and jaboticaba (Mouriria pusa), both common trees in eastern Brazil. The following species yield fruits which are eaten occasionally by the Indians: Cambuca (Myrcia sp.), massaranduva or macarandiba (Lucuma procera), mucugé (Couma rigida), ubauba (Pourouma cecrop- iaefolia), ubacaba (Britoa triflora), murici (Byrsonima), canapu (Phy- salis pubescens), Cereus sp., Eugenia sp., Genipa macrianae and G. edulis, Malpighia sp., ‘banana do brejo”’ (Mostera deliciosa), ete. In the northern part of the continent and in the Antilles, fruits eaten include the following: Oiti coro (Couepia rufa), oiti da Bahia (Moquilea salzmannit), piquia (Macoubea guianensis), bacopary (Rhee- dia brasiliensis), icaco (Chrysobalanus icaco), bacury (also cultivated) (Platonia insignis), abio (Lucuma caimito and Pouteria caimito, which are different from the caimite of the West Indies, Chrysophyllum caimito), the mammee apple or ‘‘abrico do Para”? (Mammea americana, to be distinguished from the mamey of Cuba, Calocarpum mammo- sum), several species of Cowma, several Annonaceae (Annona muricata, A. reticulata, aratigu), and several species of cacao (Theobroma cacao, T. bicolor, T. grandiflorum, T. speciosum). The wild kakaui (Theo- broma sylvestre) and the ‘‘cacau selvagem”’ (Pachira insignis), which are eaten raw, are important food items for the tribes of the upper Madeira River (Lévi-Strauss, ms.). Some kinds of mushrooms were consumed on the Orinoco, and are also considered a delicacy by the Nambicuara. 482 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 DRINKS When no water is available, the Indians know how to quench their thirst with the sap of several vines and creepers. This sap resembles pure, clear water, and it can be gathered easily in a calabash. The best known vine is the waterwhithe (Vitis sp., Entada polystachya) and salisali (Lonchocarpus rufescens or Lonchocarpus nicou), a creeper also used for drugging fish. Its water is clear and fresh, but only its first flow can be drunk, because later it becomes white and milky, and is toxic (Crévaux, 1883, p. 278). In periods of drought the Arawak of Pomeroon obtained water from truli fruits (Manicaria saccifera). Water may also be obtained from the sheath bases of the leaves of several plants: some Tillandsia, the buriti palm (Mauritia fleruosa), and caraguaté (Bromelia sp.). A great many beverages, some of them fermented, are prepared from the fruits, seeds, and roots of wild plants. A popular liquor is made with the cultivated and wild pineapples (Ananas sativus). The ancient Zupinamba prepared a fermented beverage with cashews (Anacardium occidentale). In the Chaco, beer is made with chafiar (Gourliaea decorticans), mistol (Zuzyphus mistol), tusca, and algarroba pods (Prosopis), ete. Refreshing drinks are obtained from hitchia (Byrsonima spicata), hlawaraballi (Protium heptaphyllum), guavira (Campomanesia), and several species of Psidium sp. and Eugenia sp. CONDIMENTS The preparation of salt from the ashes of leaves or fibers of some palm trees has been described (see The Palms, p. 471). Another type of vegetable salt is obtained by boiling an aquatic plant, oulin, weya, weira, weyra, or huya (Mourera fluratilis). It is a dirty brown and inferior in quality. Roth, who describes its preparation among the Guiana Indians, considers it the same as the caruru salt mentioned by Coudreau (Roth, 1924, p. 223), and it is probably the same as the Trumai salt made from waterlilies (Quain, ms.; see Upper Xingt, Handbook, vol. 3, p. 326). Numerous tribes, for instance the Nam- bicuara, cannot bear the taste of salt, but tribes which do enjoy it use native salt, bitter as it is, in large quantities. There is, in fact, a strong contrast in the like and dislike of various tribes for ‘‘hot”’ foods. Peppers (Capsicum) are usually cultivated, but wild peppers (Capsicum rabenii, C. baccatum) occur on the Brazil coast (Soares de Souza, quoted by Hoehne, 1937, p. 218). The coastal Tupi used a “long pepper which is crushed together with salt, pinches of which are swallowed after each mouthful’; this may be Piper longum or the bitter grass jambi or nhamby (Eryngium foetidum), which was Vol. 6] USE OF WILD PLANTS—LEVI-STRAUSS 483 also known to be used as a condiment (Hoehne, 1937, pp. 157-158, 252). Jambi has also been identified as Ageratum conyzovdes. The Nambicuara flavor ‘‘tonka beans,’”’ known in Brazil as cumaru (Dipteryx odorata), by crushing handfuls of them with grasshoppers. The Tupi-Cawahib add tocari (Brazil nut, Bertholletia excelsa) to maize when preparing maize beer (Lévi-Strauss, ms.). There are other condiments of animal or mineral origin. POISONS The composition of curare has been the object of many discussions. Sampaio (1916) lists the following plants which are used in the preparation of the curare of the Nambicuara and Paressi: Strychnos, Insianthus virgatus, Cassia rugosa, Dioscorea sp., and species of Apocynaceae, Marceraviaceae, and Sapindaceae. Vellard (1939), however, has proved that the Nambicuara curare consists only of the extract of a plant of the genus Strychnos. There is no doubt that this and other species of Strychnos (S. medeola, S. toxifera, S. cogens, S. crevauri) provide the active element in the preparation of the more elaborate poison of the Amazon Basin. Dance, quoted by W. E. Roth (1924, p. 151), makes mention of two other arrow poisons: the heauru-canali and the hurubuh, similar to the hog tannia. “Bresillet”’ or carasco (guao of the Taino) was perhaps used as an arrow poison, and Oviedo mentions it as a cosmetic for whitening the skin (Roumain, 1942, p. 29). It is doubtful if the poisonous manceniller (Hippomane mancinella) was ever used for arrow poison. A poisonous bamboo used in the Guianas as an arrow point is said to be Guadua latifolia (Roth, W. E., 1924, p. 151); Roth quotes Barrére that arrows were poisoned in Cayenne with the milk of the pougouly tree (Ficus venenata) and with several other ingredients. Among other poisons were Thevetia ahouai (eastern Brazil), T. peruviana (northern Brazil and West Indies), and 7. bicornuta (Mato Grosso), commonly known in Brazil as ‘‘Chapeu de Napoleao,” which Thevet (1878) stated were used for revenge in love affairs. The Nambicuara used the rosin of certain Bombacaceae as a magical poison, and there are many other unidentified native poisons, such as pakurti-neard, a cardiac poison of the Chocéd mentioned but not identified by Nordenskiéld (1930) and studied by Santesson (1929). In Surinam there was an especially poisonous arum called punkin, Arum venenatum surinamense (Roth, W. E., 1924, p. 564). Most poisons are kept secret by the natives. Several wild grasses are known as poisonous to animals. The so-called fishing poisons include a large number of plants 794711—50 32 484 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 the physiological action of which is not at all identical. Some are true poisons, some act only by suffocating the fish. Hoehne (1937, pp. 98-99) lists for Brazil three groups of fish poisons: guarana timbo (Dahlstedtia pinnata); many species of Tephrosia, in particular Tephrosia toricaria; and the group of the “timbo do cerrado” comprising Magonia pubescens, Indigofera les- pedezoides, and a species of Sapindaceae. The Tephrosia species are known under the name of tingui in most parts of Brazil and as yarro-conalli by the Macusht. The barbasco of eastern Bolivia and the upper Amazon is Serjyania perulacea. In Pert the name barbasco is also given to Tephrosia toxicaria. The ochoho of eastern Bolivia is Hura crepitans. In the Guianas, the followimg fish poisons were identified: Tephrosia, haiari, heri, nako (kumu or cube of Pert) (Lonchocarpus nicou, L. densiflorus, L. rufescens), quanami, gonami, kunalli, etc. (Clibadium asperum, C. surinamense). On the Demerara, cumapuru (Phyllanthus conami) leaves are bruised with leaves of kunami, a shrub, and the dried light pericarp of the arisauru (Derris pterocarpas), which give buoyancy to the mass, and cast into the river along with pellets of dough to tempt the fish and to paralyze and kill them (Roth, W. E., 1924, pp. 203-204); haiara-balli (Muwellera frutescans) is an Arawak poison. According to Gumilla, alligators were shot with arrows made of poison bamboo (Roth, W. E., 1924, pp. 202-207). There are several antidotes for poison, among them a decoction of Potalia amara leaves for cassava poisoning, the only one identified by W. E. Roth (1924, p. 711). (See Medicines, below.) MEDICINES Early travelers were surprised at the number of herbs known to the Indians and by the fact that the natives always used ‘‘simple’’ remedies, each employing only one plant at a time, whereas Europeans relied more on semimagical combinations of several herbs. Few primitive people have acquired as complete a knowledge of the physical and chemical properties of their botanical environment as the South American Indian. With the exception perhaps of the cinchona bark (Cinchona sp., several Rubiaceae, especially ‘‘cascarilla”’ of the Spaniards, Cinchona pubescens), there is no species used in modern pharmacopoeia which was not familiar to the natives in pre-Columbian days. Furthermore, it is probable that only a fraction of the herbs used by modern Indians are presently known and ex- ploited. The following list is only partial and fragmentary, and it is limited to the species mentioned most frequently in the literature. For most internal disorders, the natives administer an emetic followed by a purgative. The principal emetic is the root of the Vol. 6] USE OF WILD PLANTS—LEVI-STRAUSS 485 ipeca or poaia (Cephaelis ipecacuanha), but in certain regions of the Guianas, the Indians use the bark of the wallaba tree (Hperua sp.), a small creeper (Vandellia sp.) and tobacco juice (Roth, W. E., 1924, p. 704). Other emetics known to the Indians in central Brazil are Cissampelos glaberrima, Manettia ignita, the kaamara’ tai (Asclepias curassavica), and kaa’ chsa (Chiococca anguifuga). A great many purgatives are known to the Indians: several Caya- ponia (tayuya of the Twpi); cassia (taracu, Cassia occidentalis); several kinds of nuts, such as anda-uassu (Johannesia princeps) and ‘pinhio do Uruguay” (Jatropha curcas); jeticucu or “batata da purga”’ (Operculina convolvulus); several species of Ipomoea; sarsa- parilla (Herreria salsaparilha) and its numerous substitutes (Smilaz aspera). Against worms the Indians use gameleira branca, the latex of Ficus anthelmintica or F. glabrata, F. dolaria and of some other Moraceae, the seeds of Andira, etc. For gastric disturbances, the Brazilian Indians take paico (Chenopodium) and Dryopteris; the Guiana Indians, Jatropha gossypi- folium, Boerhaavia hirsuta, Chelonanthus alatus, Allamanda aubletii; and the marginal Indians of the Tropical Forest area, yerba del moro (Amaranthus sp.) or urutu (Alternathera repens). For healing wounds, the Brazilian Indians used besides balsams (q. v.) the crushed seeds of Mucuna altissima, tupixaba (Scoparia dulcis), hiboucouhu (Virola sp.), crushed seeds of the toxic fruits of Carapa guianensis, pounded and parched leaves of Piper jaborandi, and Pilocarpus pennatifolius. Astringent herbs of native pharmacopoeia are camara (Lantana camara), Polypodium crassifolium, Oxalis tuberosa, and O. angusti- olvum. To stop bleeding, the Indians used Arenaria lanuginosa, Oenothera rosea, Chrysophyllum glycyphloeum, and some mushrooms (Polyporus coccineus and Geaster saccatus). Among the drugs used to cure eye pains, a constant complaint of South American Indians, Roth mentions red-pepper juice, the leaves of mokumoku (Caladium arborescens), a decoction of Wansimai roots, and the purplish red juice of kuruwatii (Renealmia exaltata). eee Be ke Il Nazionale, Revista di Centro Italiano di Studi Americani. Rome; Turin. rete te eres © oN ke Beek at hs eee Inca, Revista de Estudios Antropolégicos, érgano del Museo de Arqueologia de la Universidad Mayor de San Marcos. Lima, Peru. rvclNGtes iS: ce Mee Saeed sh aa Indian Notes, Museum of the American Indian, Heye Foundation. New York, NX: Ind Notes Monogr: 2-2 2.522422 Indian Notes snd Monographs, Museum of the American Indian, Heye Founda- tion. New York, N. Y. Inst. Biol. Univ. Auté6. México- -_ --- Instituto de Biologia, Universidad Auté6- noma de México. Inst. Etnol. Univ. Nac. Tucum4n___. Instituto de Etnologfa de la Universidad Nacional. Tucuman, Argentina. Inst.Int. Anthrop. .....-..--3- 2222 Institut International d’Anthropologie. Paris, France. Inst. Panamer. Geogr. Hist. _-__- -_- Instituto Panamericano de Geografia e Historia. México, D. F. Sint. Archiy lthnogr.:22-- 2. 2-26: International Archiv fur Ethnographie. Leyden, Holland. Int. Journ. Amer. Ling. ___.________ International Journal of American Lin- guistics. New York; Baltimore. Jahresber. Geogr. Gesell. Bern - - - - -- Jahresbericht der Geographischen Gesell- schaft zu Bern. Bern, Switzerland. Jounr. Allied: Denti Soc... = 652. s2._- The Journal of the Allied Dental Societies. New York, N. Y. Journ. Amer. Dent. Assoc. --------- Journal of the American Dental Associa- tion. Chicago, IIl. Journ. Amer. Dietet. Assoc. ------~-- Journal of the American Dietetic Associa- tion. Chicago; Baltimore. Journ. Amer. Holklores: 2-52) 2522-2 = Journal of American Folklore. Boston, Mass.; New York, N. Y.; Lancaster, Pa. 554 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Journ. Amer. Med. Assoc. __-------- Journal of the American Medical Associa- tion. Journ. Amer: Stat. Assoc. -__....._- Journal of the American Statistical Asso- ciation. Boston, Mass. Journ: Anat, Physieli
... -22- 22... Revista Médica Peruana. Lima, Pert. Rev. Men. Sec. Soc. Geogr. Lisboa, Revista Mensual de Seccao da Sociedade Brazil. de Geographia de Lisboa, Brazil. Rio de Janeiro, Brazil. Rev. Men. Soc. Geogr. Rio de Janeiro. Revista Mensual da Sociedade de Geo- grafia do Rio de Janeiro. Rio de Janeiro, Brazil. Beemayalox., SOG) 22 kook ae ee Revista Mexicana de Sociologia, Instituto de Investigaciones Sociales de la Univer- sidad Nacional Auténoma. México, Dir, ev. is: Colombia. ol 2s221....2.2 Revista de Misiones. Bogot4, Colombia. Rev. Mus. Arch. Pub. Rio Grande do Revista do Museu e Archivo publico do Sul. Rio Grande do Sul. Porto Alegre. Rev. Mus. Hist. Nac. Chile________- Revista del Museo Histérico Nacional de Chile. Santiago de Chile. mee Mus. la, Plata. oo eo5 ok es Revista del Museo de La Plata. La Plata, Argentina. eve Nius. Nac. Lima-..2.2 2-2-5. Revista del Museo Nacional de Lima. Lima, Peri. Rev. Mus. Nac. Rio de Janeiro. __-_- Revista do Museu Nacional. Rio de Janeiro, Brazil. Remeron. PAM: ose oo ee Revista do Museu Paulista. Séo Paulo, Brazil. feewenberapiiaya 2-2 ook Revista Paraguaya. Asuncién, Paraguay. Rey. Philol. Hist. Rio de Janeiro____ Revista de Philologia e de Historia, Ar- chivo de Estudos sobre Philologia, Historia, Ethnographfa, folclore e critica literaria. Rio de Janeiro, Brazil. Me MEGNAVEN ooo ee ea ee Revista Popayaén. Popaydn, Colombia. 560 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143 Rev. Real Acad. Cienc. Exact. Fis. Revista de la Real Academia de Ciencias Nat. Madrid. Exactas, Ffsicas y Naturales de Madrid. Rev: Rio dé la ‘Platac 22 sess: 5 Revista del Rio de la Plata, Periéddico Mensual de Historia y Literatura de América. Buenos Aires, Argentina. Rev. Serv. Patr. Hist. Art. Nac. ____ Revista do Servigo do Patrimonio His- torico e Artistico Nacional, Ministério da Educagéo e Saude. Rio de Janeiro, Brazil. Rev. Soc. ‘‘Amigos Arqueol.” _______ Revista de la Sociedad ‘‘Amigos de la Arqueologia.’’? Montevideo, Uruguay. Rev. Soc. Argent. Cienc. Nat. -___-_- Revista de Sociedad Argentina de Ciencas Naturales. Buenos Aires, Argentina. Rev. Soc. Cient. Paraguay__________ Revista de la Sociedad Cientffica del Paraguay. Asuncién, Paraguay. Rev. Sud-Amer. Cienc. Med. Farm. _ Revista Sud-Americana de Ciencias Medi- cas y Farmaceuticas. Buenos Aires, Argentina. Rev. /Sudamer, Morfol. 22). 22- Revista Sudamericana de Morfologfa. Concepcién, Chile. Rey. Trim: Inst. Cearg-_..-..._-244 Revista Trimensal do Instituto do Ceard. Fortaleza, Ceard, Brazil. Rev. Trim. Inst. Hist. Geogr. Brasil._ Revista Trimensal do Instituto Histérico e Geographico Brasileiro (or do Brasil). Rio de Janeiro, Brasil. Rev. Trim. Inst. Hist. Geogr. Revista Trimensal do Instituto Histérico, Ethnogr. Brasil. Geogrdphico e Ethnographico do Brasil. Rio de Janeiro, Brazil. Rev. Univ. Arequipal.i22 2 =