CULTURAL AND NATURAL AREAS OF NATIVENORTH AMRCA By A. L. KRQEBER UNIVIMIHTY OIF CM IOItNU~ PUBLICATIONS, IN "Am RcAP ARCHAEOLO.GY AND ETHNrotoGY Volume 38, pp. xii-+ 1-242, 28 maps UNIVER'SITY OF CALIFORNIA PRESS BERKEEY, CALIFORNIA 1,939' UNIVERSITY OF CALIFORNIA PUBLICATIONS IN AMERICAN ARCHAEOLOGY AND ETHNOLOGY VOLUME XXXVIII EDITORS A. L. KROEBER R. H. LOWIE R. L. OLSON UNIVERSITY OF CALIFORNIA PRESS BERKELEY, CALIFORNIA 1939 CULTURAL AND NATURAL AREAS OF NATIVE NORTH AMERICA BY A. L. KROEBER UNIVESITY Or CAuroRNU PUBLICATIONS IN A31MICAN AOHAzOLOGY AD ETHNOLOGY EDIToRs: A. L. KROErn, 1. H. LOwI, R. L. OLSON Volume 38, pp. xii + 1-242, 28 maps Transmitted October 9, 1936 Issued December 12, 1939 Price: cloth, $3.50; paper, $3.00 UNIVRSITY OF CAIPORNIA PRESS BERKEY, CALIJoRNiA CAMBRIDGE UNIVERSITY PRSS LONDON, ENGLAND PRINTED IN THE UNITED STATES OF AMERIOA PREFACE ALTHIOUGH the manuscript of this work was completed in 1931, publication was delayed by the depression, both at the University of California and at the Bureau of American Ethnology, which latter, generously, for a time hoped to be able to print it. The years that have elapsed have brought sub- stantial additions to knowledge, and references to the more important of them will be found in footnotes and special sup- plementary passages inserted in 1936 and (a few of them) in 1939. The main body of the text.stands as written in 1931. Alterations that might have been made in parts of several maps have not been undertaken, on account of cost; but the newer data, if deviant, have been cited. A. L. K. Berkeley, California, May, 1939. CONTENTS BFXfON PAGE I. Objectives ................................................................ 1 II. History of concepts ....................................................... 3 Environment in anthropology ..................................... i ..... 3 Culture areas, climaxes, and boundaries ................................. 4 Relation of natural to cultural areas .................................... 6 III. Tribal areas .............................................................. 8 List of works used for the tribal map (map 1) ........................... 9 List of works appearing since preparation of the tribal map ............... 11 Tribal synonyms ........................................................ 11 Pronunciation of tribal names ........................................... 12 IV. Vegetation areas .......................................................... 13 Authorities used for the vegetation maps (maps 2-5) ...................... 14 Concordance areas ..................................................... 15 V. Culture areas: Arctic Coast .............................................. 20 Sources of Eskimo culture .............................................. 20 Ecological phases .......... ............................ 22 Cultural classification and history ...................................... 24 Summary .............................................. 27 VI. Culture areas: Northwest Coast ........................................... 28 VII. Culture areas: Southwest .............................................. 32 1- 2. Pueblo subculture type ......................................... 34 3-10. Sonora-Gila-Yuma subculture type .............................. 38 History .............................................. 45 VIII. Culture areas: Intermediate and Intermountain areas ....................... 49 1. Great Basin .............................................. 49 2. California .............................................. 53 3. Columbia-Fraser Plateau ......................................... 55 IX.FCulture areas: East and North ............................................ 60 Eastern areas .............................................. 60 1. Southeast .............................................. 61 Muskogian and Creek .......................................... 64 2. South Florida .............................................. 67 3. South Texas: Northwest Gulf Coast ............................... 70 4. Red River area .............................................. 74 5. Plains areas .............................................. 76 Southern Plains .............................................. 79 Northern Plains .............................................. 80 6. Prairie areas ........................... ................... 84 Summary of tribal history in the plains-prairies ...................... 86 7. Wisconsin or Wild Rice area ....................................... 88 8. Ohio Valley .............................................. 89 9. Lower Great Lakes .............................................. 91 10-12. Atlantic Coast areas ........................................... 92 13. Appalachian Suummit .............................................. 95 Northern areas .............................................. 95 14. Northern Great Lakes ............................................. 96 15. Eastern Subarctic ...................................... 97 16. Western Subarctic .................. ..................... 98 Relations of Eastern and Northern areas ................................ 101 Eastern archaeological areas .........................................- . 101 [vi] viii Contents SECTION PAGE X. Culture areas: Mexico and Central America ................................ 109 Isthmus ................................................ 109 1. Atlantic Nicaragua-Honduras ..................................... 110 2. Pacific Nicaragua ................................................ 110 3. Salvador ................................................ 110 4. Upland Guatemala ................................................ 110 5. Yucatan Peninsula ................................................ 111 Relations of Maya upland and lowland ............................... 111 6. Oaxaca-Tehuantepec ............................................... 115 7. Guerrero .................................................. 116 8. Vera Cruz .................................................. 116 9. Southeastern Central Mesa ........................................ 116 10. Michoacin .................................................. 117 11, 12. Jalisco Highland and Jalisco Coast ............................. 117 13. Northeastern Central Mesa: Guanajuato-Quer6taro ................. 119 North Mexican areas .................................................. 119 Documentary sources ................................... ....... 119 Archaeology .................................................. 122 Language .................................................. 124 Areas .................................................. 125 14. South Sinaloa: Aztatlan-Culiacan .................................. 126 15. Sierra del Nayarit: Southern Sierra Madre ......................... 127 16. Central Sierra Madre ............................................. 128 17. North Mexican Interior Plateau ................................... 128 18. Tamaulipas: Northwest Gulf Coast ................................ 130 XI. Population .................................................. 131 Discussion of Mooney's figures .......................................... 132 Population and density by tribes ........................................ 134 Population and density by areas ........................................ 141 Coast land and farm land ............................................... 143 The agricultural East .................................................. 146 Comparison with Mexico ............................................. 150 The Southwest .................................................. 151 California .................................................. 153 The Northwest Coast .................................................. 155 Eskimo .................................................. 156 Mexico and Central America ............................................ 157 The hemisphere .................................................. 164 Shore-line population density ........................................... 166 Population size of language groups ...................................... 172 Appendix: later data .................................................. 177 XII. Physiographic areas .................................................. 182 United States .................................................. 182 A. Laurentian Highland (Pre-Cambrian Shield) ........................ 182 B. Atlantic Coast Plain .............................................. 183 C. Appalachian Highlands ............................................ 183 D. Interior Plains .................................................. 185 E. Interior Highlands ................................................ 187 F. Rocky Mountain System .................... 187 G. Intermontane Plateaus .................... 188 H. Pacific Mountain System .................... 191 Costent8 ix SECTION PAGE XII. Physiographic areas (Continued) Canada ....................................... 193 A. Canadian or Pre-Cambrian Shield .................................. 194 B. Atlantic Coast Plain ........................................ 194 C. Appalachian Highlands ....................................... 194 D. Interior Plains ........................................ 195 E. Interior Highlands ....................................... 195 F. Rocky Mountain System ....................................... 195 G. Intermontane Plateaus ....................................... 196 H. Pacific Mountain System ....................................... 196 Mexico ....................................... 197 Physiography and culture ....................................... 201 XIII. Relations of environmental and cultural factors ............................ 205 Physiography .................................... 206 Natural vegetation .................................... 206 Climate .................................... 207 Water ..................................... 213 Drainage .................................... 215 XIV. Agriculture .................................... 218 XV. Cultural intensity and climax .................................... 222 Index .................................... 229 TABLES 1. Concordance key of vegetation areas .................................... 15 2. Regional variants of Eskimo economic culture ................................. 23 3. Florida vegetation types ................................... 70 4. Algonkin dialect groups and culture areas ...................................... 98 5. Holmes and Wissler archaeological areas ................................ 107 6. Classification of the Mayan languages ................................ 114 7. Tribal populations, territorial extent, and densities north of Mexico ............ 134 8. Population densities of principal areas of culture .............................. 142 9. Population densities by major areas .......................................... 143 10. Grand population divisions ................................................... 144 11. Mexico-Central America: areas, population, densities ........................... 158 12. Estimated increase of population in Mexico ................................... 163 13. Population 'of the Americas, 1650 A.D ............................. 165 14. Population of the Western Hemisphere, 1500 A.D ............................... 166 15. Shore line and population ..................................... 170 16. Population by linguistic stocks ..................................... 174 17. Principal correspondences of physiographic provinces with cultural and ethnic areas ..................................... 202 18. Summary list of North American culture areas ............................ in pocket Content8 MAPS PAGE la. Native tribes of North America ......................... in pocket lb. Native tribes of North America (key for use with other maps) . .......... in pocket 2. Vegetation areas (Harshberger) ............ ............................. in pocket 3. Vegetation areas (Shelford) .............................................. in pocket 4. Vegetation areas (Shantz-Zon, Lynch) ................................... in pocket 5. Vegetation areas (Shreve, Malte, Kellogg, Sanders) ...................... in pocket 6. Native cultural areas of North America .................................. in pocket 7. Physiographic areas of North America ................................... in pocket 8. Creosote bush and sagebrush .................................................. 34 9. Open and wooded country in the United States ............................... 58 10. Culture areas of Mason and Wissler compared ................................ 61 11. Muskogian dialect groups ..................................................... 66 12. Range of southeastern deciduous tree species ................................. 68 13. Precipitation-evaporation ratio in the United States .......................... 72 14. Evergreen broad-leaved and microphyllous trees in the United States ....... . 73 15. Mound areas of the eastern United States .................................... 102 16. Pottery types of the eastern United States ................................... 104 17. Subsistence, ethnic, and political groups of native Mexico ..................... 121 18. Native population densities by cultural areas ........................... facing 134 19. Population densities in native California ..................................... 154 20. Density of rural population in modern Mexico ................................ 162 21. Coast population densities in the native period ............................... 168 22. Natural regions of Meidco (McBride) ........................................ 198 23. Natural regions of Mexico (Sanders) ............. ............................ 199 24. Dry climates of the United States ........................................... 208 25. Seasonal distribution of precipitation in the western United States ............ 209 26. Dates of killing frosts in the Southwest ...................................... 210 27. Growing seasons in the United States-120-day and 100-day ................... 212 28. Intensities or "levels" of native cultures by areas ...................... facing 222 x BIBLIOGRAPHICAL ABBREVIATIONS USED AA American Anthropologist AAA-M American Anthropological Association-Memoirs AMNH-AP American Museum of Natural History-Anthropological Papers AMNH-B American Museum of Natural History-Bulletins AMNH-H American Museum of Natural History-Handbook BAAS British Association for the Advancement of Science BAE-B Bureau of American Ethnology-Bulletins BAE-R Bureau of American Ethnology-(Annual) Reports CED Culture Element Distributions, a series of monographs begun in UC-PAAE and now continued in UC-AR CNAE Contributions to North American Ethnology ICA International Congress of Americanists (Comptes rendus, Proceedings) JAFL Journal of American Folk-Lore JRAI Journal of the Royal Anthropological Institute MAIHF-C Museum of the American Indian, Heye Foundation-Contri- butions MAIHF-IN Museum of the American Indian, Heye Foundation-Indian Notes PM-P Peabody Museum (of Harvard University)-Papers SI-AR Smithsonian Institution-Annual Reports SI-MC Smithsonian Institution-Miscellaneous Collections UC-AR University of California-Anthropological Records UC-IA University of California-Ibero-Americana UC-PAAE University of California-Publications in American Archaeology and Ethnology UC-PG University of California-Publications in Geography USNM-R United States National Museum-Reports UW-PA University of Washington-Publications in Anthropology YU-PA Yale University-Publications in Anthropology ZE Zeitschrift fur Ethnologie [ xi ] CULTURAL AND NATUR AREAS OF NATIVE NORTH AMERICA BY A. L. KROEBER I. OBJECTIVES THIS STUDY has two objectives. It aims, first, to review the environmental rela- tions of the native cultures of North America. Its second purpose is to examine the historic relations of the culture areas, or geographical units of cultures. Three points are best stated explicitly at the outset, to prevent possible mis- conception. The first is that the present work in no sense represents a relapse toward the old environmentalism which believed it could find the causes of culture in environment. While it is true that cultures are rooted in nature, and can therefore never be completely understood except with reference to that piece of nature in which they occur, they are no more produced by that nature than a plant is produced or caused by the soil in which it is rooted. The immediate causes of cultural phenomena are other cultural phenomena. At any rate, no anthropologist can assume anything else as his specific working basis. But this does not prevent the recognition of relations between nature and culture, nor the importance of these relations to the full understanding of culture. The second point is to guard against the possible misconception that the determination of culture areas is here considered an end in itself. The concept of a culture area is a means to an end. The end may be the understanding of culture processes as such, or of the historic events of culture. The study of processes tends to be analytic, and therefore to disregard time and space relations except so far as they condition the particular phenomena whose processes are being examined. In proportion as the study advances and learns to deal more directly with cultural processes as such, the time and space relations become a sort of frame. They remain factors that for scientific pur- poses must be controlled, but this control becomes a limitation, almost an encumbrance. This type of study is akin to the dissecting technique of the laboratory, even though cultural anthropology has neither laboratory nor experiment. It is the method which has been carried farthest, in penetration and exactness, by Franz Boas. This method can use culture areas only to a limited extent, as a sort of preliminary; and its practitioners therefore esteem the concept as of only incidental utility. On the contrary, the historic approach, remaining concerned with events as they occur in nature, always stresses the time aspects of phenomena as part of its ultimate objective. Ethnology, particularly when concerned with peo- ples which, like the native ones of America, have left few or no documentary records, perforce has recourse to spatial classifications such as culture areas. In themselves these yield only a momentary and static organization of knowl- edge, whereas the purpose of history is genetic. In proportion as the recog- [1] University of California Pu'blications in An. Arch. and Ethn. nition of culture areas becomes an end in itself, it therefore defeats really historic understanding. The conception on which the present monograph is based is that space and time factors are sufficiently interrelated in culture history to make the culture area a valuable mechanism, rather than a distrac-_ tion, in the penetration of the time perspective of the growth of cultures so relatively undocumented as aire those of native America. The third point to be kept in mind is that the present study deals with cul- ture wholes, and not, except incidentally, with culture elements or "traits," nor with those associations of elements which are sometimes called "culture complexes" but which always constitute only a fraction of the entirety of any one culture. Culture wholes as a concept correspond in many ways to regional floras and faunas, which are accumulations of species but can also be viewed as summation entities. The term "culture area" is employed because usage has established it. It is an unfortunate designation in that it puts emphasis on the area, whereas it is usually the cultural content that is being primarily considered. We mean a regionally individualized type or specific growth of culture when we say "cul- ture area," much as a historian may use "the Eighteenth Century" as a short way of referring to the culture that was characteristic of eighteenth-century Europe. It would be well if there were a brief technical term for the naturally individualized growths of culture with which historical anthropology is more and more dealing. But it seems impossible to find an unambiguous term with- out coining it.1 Evidently the general thought of our day is not yet sufficiently concerned with such growths of culture to feel the need of a designation for them. 1 "Diaita" (Angl. diaeta) has been suggested to me by J. L. Myres as an etymologically adequate term to denote a culture whole or actually cohering culture mass, corresponding to the "biota" of biologists. It would be useful if adopted. 2 Kroeber: Cultural and Natural Areas of Native North Aimerica3 II. HISTORY OF CONCEPTS ENVIRONMENT iN ANTHROPOLOGY FOR A GENERATION American anthropologists have given less and less attention to environmental factors. In part this represents a healthy reaction against the older naive view that culture could be "explained" or derived from the environment. For the rest, it is the result of a sharpening of specific anthro- pological method and the consequent clearer perception of culture forms, pat- terns, and processes as such: the recognition of the importance of diffusion, for instance, and of the nature of the association of culture elements into "com- plexes." Most attention came to be paid, accordingly, to those parts of culture which readily show self-sufficient forms: ceremonial, social organization, art, mythology; somewhat less to technology and material culture; still less to economics and politics, and problems of subsistence. Much of the anthropology practiced in this country in the present century has been virtually a sociology of native American culture; strictly historic and geographic interests have re- ceded into the background, except where archaeological preoccupation kept them alive. We have had intensive studies of the internal social grouping of peoples of whom we did not know whether they constituted one or several national units; analyses of the patterns of maize- or acorn-utilization com- plexes, rather than consideration of whether such a complex provided a tenth, a half, or four-fifths of the subsistence of the various tribes who adhered to it; and so on. This diversion of attention to cultural forms was necessary and de- sirable; the attendant shift of interest away from historical and subsistence problems was probably inevitable. There is also often a readier productivity in work along the formal lines, especially among Indians on reservations. An old informant can sometimes give exact data on the sequence of details of a ritual that has been abandoned for forty years, but is vague about the propor- tion of acorns or salmon in his father's diet, or the months of each year spent by his group on the river or in the mountains. However, such facts are also of consequence in their relation to culture, since every culture is conditioned by its subsistence basis. The culminations of culture obviously rest on a certain degree of economic surplus, for instance. Such a surplus will not explain why the lines in a given art are curved instead of straight, or why a people derives the origin of mankind from below ground rather than from the sky. But it may help to explain why Haida art is esthetically richer than Kwakiutl, or Pueblo ritual more complex than Havasupai. And these are also legitimate problems; and strictly historical ones. We need not edge away from them be- cause they involve qualitative judgments or a concern with culture wholes. Anthropology does not have to be exclusively analytic in order to be valid. 3 4University of California Publications in An. Arch. and Ethn. CULTURE AREAs, CLIMAXES, AND BOUNDARIES The concept of the culture area has had a gradual, empirical, almost uncon- scious growth. It probably began, as Boas points out, with the classification of museum collections on natural geographical lines instead of evolutionistically schematic ones. By 1916, Sapir in his Time Perspective discussed culture areas as something in general use; in 1917, Wissler codified those of native Amer- ica,-on the basis, largely, of current usage. There have been no serious modi- fications or criticisms of his scheme. But it is significant that Wissler does not develop his interpretation of the growth of American culture through use of the culture areas which he defines. He follows agriculture, the textile arts, architecture, and so on, one by one through the two continents; and it is the summation of these findings, essentially, that yields his picture of hemispheric history. The culture-area classification remains a nearly static one, and apart. There has been another method of geographical attack: consideration of the distribution of single culture elements or limited complexes. This is the method pursued with such eminent success by Nordenskiold in South America. Noth- ing equally systematic has been attempted for North America. But on a more limited scale the method has been applied by the Danes to Eskimo culture, by Spier to the Havasupai and their neighbors, and by several students to mytho- logical material, although these latter have applied it without primarily histor- ical objective. Wissler has used the method abundantly in somewhat different form: for larger complexes, or for summary outlines, or in elaboration of the age-and-area principle. This method is analytic in the sense that it deals with detached parts of culture. But cultures occur in nature as wholes; and these wholes can never be entirely formulated through consideration of their ele- ments. The culture-area concept does attempt to deal with such culture wholes. Boas has attempted to limit the significance of culture areas by asserting that these areas do not coincide when they are formulated on the basis of dif- ferent parts of culture: technology, social organization, ritual, art, music, myth, etc. This view must be doubted as contrary to the overwhelming run of the facts, though no doubt occasionally true. An unusually rich development in almost all these lines is normally fbund coincident in highly specialized and distinctive cultures, such as those of the Pueblos or North Pacific Coast In- dians.1 Navaho altar paintings may be the most developed in the Southwest, but Navaho culture is after all close to that of the Pueblos and in many ways obviously dependent on it. That at some points the pupil departs from the master or surpasses him does not invalidate the reality of a school or tradition. In general, the experience of Old World history is to the same effect. As a matter of fact, the points in time and space at which historically known culture growths culminated usually show a virtual coincidence of florescence 1 Negative developments in relatively rich cultures are an apparent exception which really confirms the situation depicted, because absences tend to be due to strong positive develop- ments in allied directions: the shaman is lacking in Pueblo life because the priesthood is strong, Lower Colorado tribes use a minimum of ritual paraphernalia because of their ex- treme emphasis on dream experience, and so on. 4 Kroeber: Cultural and Natural Areas of Native North America in the several facets of culture: the peaks of empire, wealth, sculpture, drama, philosophy, science in fifth-century Athens, for instance. Augustan Rome is an- other classical example; so is sixteenth-century Spain. Among other scholars, Flinders Petrie has gone so far as to try to demonstrate a fixed order in which the respective peaks of each of these facets of culture are reached in any civilizational culmination.! This attempt must be regarded as somewhat forced into a scheme. But it does show clearly the correlation of the parts, their close relation or overlapping coincidence in time and space, whenever the culmi- nation is strong. There is no reason to believe that the course of events was materially different in native America. For the Maya and Pueblos we have archaeological justification that it was similar. The whole subject of cultural climax is evidently related to that of the cul- ture area. Since ethnologists normally deal with relatively timeless data they have been cautious and slow to approach problems of time climax. They have, however, evolved a spatial substitute: the culture center, or district of greatest cultural productivity and richness. This obviously is the regional expression of a culmination whose temporal manifestation is the climax. As so often, Wiss- ler has pioneered the way. He makes the point that the center is the integral thing about an area. The area may therefore be conceived and represented somewhat diagrammatically. Hence the straight lines and sharp angles on Wissler's culture-classification maps. No serious exception could be taken to these maps if the centers were decisively defined; but Wissler more often than not leaves them as indefinite as the areas. His Plains group comprises thirty- one tribes, of which eleven are the most typical; his Southeast centers among the Muskogians, Yuchi, and Cherokee, who occupied half of the total region. For the Mackenzie and Eastern Woodland areas, the localization of centers is attempted very half-heartedly. Wissler also makes but little more use of his culture centers than of his culture areas when he reconstructs the outline his- tory of the hemisphere. In short, it is clear that he has perceived the signifi- cance of focal points of growth, resulting in ciulminations definable in spatial and presumably temporal terms; but his working out of these has remained summary and indefinite. The weakest feature of any mapping of culture wholes is also the most con- spicuous: the boundaries.' Where the influences from two culture climaxes or foci meet in equal strength is where a line must be drawn, if boundaries are to be indicated at all. Yet it is just there that differences often are slight. Two peoples classed as in separate areas yet adjoining each other along the interarea boundary almost inevitably have much in common. It is probable that they normally have more traits in common with each other than with the peoples at the focal points of their respective areas. This is almost certain to be so where the distance from the foci is great and the boundary is not accen- tuated by any strong physical barrier or abrupt natural change. But the same holds true of the faunal and floral areas used by naturalists. In short, what 2Discussed further in the final section of the present work. 8 This is less true of complexes or associations than of wholes, and is not at all appli- cable to atomic culture elements which can be mapped in terms of presence or absence. 5 University of California Publications in Am. Arch. and Ethn. boundaries really show is not so much clefts occurring in nature, as relative extent and strength of influences emanating from foci. They represent some- thing comparable to political spheres of influence expressed by devices suitable for showing artificial political entities. It would be desirable, therefore, to con- struct cultural maps without boundary lines, on some system of shading or tint variation of color; but the mechanical difficulties are great. For the pres- ent, it seems necessary to use the old devices and leave it to the reader to translate what his eye sees into the dynamic aspects that are intended. This difficulty inheres in all attempts to express in static two-dimensional space terms, phenomena that have a sequential as well as a spatial aspect; a flow as well as a distribution. RRLATION OF NATURAL TO CuLTuRa ARAs We can accept Wissler's findings on the relation of culture areas to environ- ment.' He concludes that environment does not produce a culture, but stabilizes it. Because at many points the culture must be adapted to the environment, the latter tends to hold it fast. Cultures therefore incline to change slowly once they have fitted themselves to a setting, and to enter a new environment with more difficulty than to spread over the whole of the natural area in which their form was worked out. If they do enter a new type of territory, they are subject to change. Once fitted to an environment, they are likely to alter radically only through some factor profoundly affecting subsistence, such as the introduction of agriculture. Beyond these sound general principles, however, Wissler does not go very far. In his American Indian he enumerates some suggestive rough correspond- ences between altitude contours and linguistic or culture groups.' His later work, The Relation of Nature to Man in Aboriginal North America (1926), is concerned with the spatial distribution of culture traits and complexes. Na- ture in the sense of the varying physical and organic environment does not really enter into the argument, except in the last section of the last chapter, which points out, with a few examples, that ecological factors may be of im- portance, but does not pursue the subject to any intensive conclusions. Wissler's ten North American culture areas really rest on the six "food" areas which he reviews at the beginning of his book on the American Indian, although the relation of the two classifications is not wholly exact and does not become very explicit. These subsistence areas seem to refer primarily to the basis of culture, but of course involve environment also, especially its eco- logical aspects. Some years before, Otis T. Mason had dealt directly though summarily with the environment of cultures, in the Handbook of American Indians. His twelve "ethnic environments" are defined in both geographical and cultural terms; and the environments are largely faunal and floral, that is, ecological. This stimulating essay has attracted little attention, in spite of its obvious sound- 'The American Indian (1922 ed.), 372-374. The same, 368-369. 6 Kroeber: Cultural and Natural Areas of Native North America 7 ness of classification.' Mason's areas partly coincide with Wissler's, as the fol- lowing comparison shows: WISSaLn W1rBL3a MASON Food areas Culture areas Ethnic environmnents rEskimo .............. Arctic Caribou . .... Mackenzie ........................... Yukon-Mackenzie L(Northern part of Eastern Woodland) . . St. Lawrence-Lakes Bison ......... Plains ........ Plains Salmon ......... o........ North Pacific Coast .............. North Pacific Coast Pateau ............... ColumbiaFraser* Wild Seeds ...... Caliifornia . Interior Basin California-Oregon mastern Woodland (southern p)art). 5 Atlantic Slope Eastern Maize .......... { Wtssissippi Valleyt outheastern ....... Gulf Coast Intensive Agriculture. Southwestern. .. ... Pueblo "Nahua" ..... (Not dealt with) * Assigned to Wild Seed food area, largely to Plains culture area, by Wissler. t Divided by Wissder between the Plains and Eastern Woodland culture areas. Ratzel concludes the second volume of his Anthropogeographie with a world map in which native North America is divided into four areas7 corresponding rather closely to the primary culture areas laid down in the present work. They are, however, only briefly discussed.6 Ratzel clearly knew much ethnog- raphy, had thought about it, and possessed definite ethnographical insight. But in the modern view his work is deficient in not sufficiently separating pop- ulation and culture. Somatological, populational, and cultural aspects are only partly differentiated by him. Hence he evolved a clear concept of marginal peoples without advancing to that of marginal cultures, which Sophus Muller grasped concretely in dealing with the prehistory of Europe some years later. Ratzel, in short, remained primarily a geographer. But he did conceive of culb ture as more than an incidental epiphenomenon, and was far from being the crass environmentalist which Semple's misrepresentatively selective adapta- tion makes him out to be. Environmental factors have not been wholly neglected in monographic studies in the North American field; but treatments have either been intro- ductory, or, like Jenks's Wild Rice Gatherers, concerned with special mnifes- tations. There seem to be no general classifications besides those reviewed. In an earlier work, Influence of Environment upon Human Industries or Arts, SI-AR 1895:639-665, 1896, Mason recognizes eighteen "environments" or "culture areas" as he in- discriminately calls them (pp. 646, 651), in the western hemisphere: Arctic (Eskimo); Athapascan (Yukon-Mackenzie); Algonquin-Iroquois; Muskhogean; Plains of the Great West; North Pacific Coast; Columbia Drainage; Interior Basin; California-Oregon; Pueblo; Middle American; Antillean (including southern Florida and part of the northern coast of South America); South American Cordilleran (Colombia to Peru); Andean At- lantic Slope (Colombia to Bolivia); Eastern Brazilian (from Tocantins east); Central Brazilian (Matto Grosso, between Araguaya and the western boundary of Brazil); Argen- tinian-Patagonian; Fuegian. ' Hyperborean, Northwestern, Northeastern, Civilized Peoples of Middle America. 8 Anthropogeographie, 2:775-779, 1891. 7 8University of California Publications in Aim. Arch. and Ethn. III. TRIBAL AREAS NEARLY FIFTY YEARS AGO, Powell published his classification and map of In- dian linguistic families north of Mexico, and this has been reissued with minor corrections by the Bureau of American Ethnology. Thomas and Swanton fol- lowed with a similar map of Mexico and Central America. So far as major speech groups are concerned, the continent has accordingly long been plotted with considerable accuracy. Not so, a tribal map. There have been many sec- tional ones; but the first continental one was that of Wissler in 1917. This, however, gave no boundaries, and the apparent area attributed to any group was sometimes a function of the number of letters in its name rather than of its actual geographical holdings. The latter difficulty was partly remedied in a small map, based mainly on Wissler's, issued in 1919 by the University of California, in which a nuimber near the center of each tribal range corre- sponded to the name as given in a key list. There were also added some tribes not included by Wissler. However, no boundaries of tribal areas were shown, and the tribes represented were only those most frequently cited in recent ethnological literature.' Evidently, maps as loosely defined as these offer little opportunity for exact comparison of tribal and cultural areas with environmental ones. The only recourse was to compile a tribal boundary map; which herewith appears as map 1. It makes no pretense of original research or of finality. It has in- volved many judgments between differing delimitations. It follows at every point some one of the authorities listed, except where irreconcilable conflicts have had to be more or less arbitrarily compromised, and weight has then been given to natural features; for instance, watersheds rather than streams have generally been postulated as boundaries whenever a departure from the sources was forced.! The map does not, as it should in principle, represent conditions at one ab- solute date nor even at one relatively consistent historic moment, such as that of discovery. It attempts to indicate tribal territories approximately as they were constituted at the time of first occupation by Europeans. This time varied from the early sixteenth to the early nineteenth century in North America. It is this variation as well as conflict of authority that has forced the com- promises mentioned. A map dated for the period of discovery would be incom- 1 The Department of Anthropology of Yale University has recently (in 1938 7) pre- pared and manifolded, apparently in connection with its "cross-cultural" program, a tribal map showing boundaries as well as names. This, then, is the first map of the kind to be issued. The size is adequate-16 inches high; dralnage is not shown; nor the areas south of Tehuantepec. 2 The territorial relations of the Bannock and Shoshone, in which Mooney has mainly been followed, are almost certainly wrong. The Lemhi are Shoshone, not Bannock in speech. I suspect either that the "Shoshone" and Wind River Shoshone held a fringe of territory along the Rockies and Bitterroots which ineluded the Lemhi; or that the Bannock had all the upper Snake, virtually cutting the Lemhi off from the "Shoshone" and the "Shoshone" from the Western Shoshone, the "Shoshone" and Wind River Shoshone being one people. In addition to other inconsistencies, the relation of the ethnic distribution on the map to the drainage seems unlikely to be true; but I do not know how to make correction. This and related problems are clearing up, owing to recent field studies by Steward and others. See the supplemental bibliography in this section (p. 11). 8 Kroeber: Cultural and Natural Areas of Native North America plete in many areas, or filled with doubtfully identifiable names. Besides, this plan would be subject to much the same variability of time represented as the plan actually followed. Most of the maps used as sources because they show boundaries refer to the period of occupation rather than to that of discovery. The situation is not wholly fortunate; but the method followed seems the most feasible and useful. Below are given the sources used in the preparation of the map, a list of some of the more important synonyms not appearing on the map, and memo- randa on pronunciation. Tribal names abbreviated on the map appear in full in its margin. LIST OF WORKS USED FOR THE TRIBAL MAP (MAP 1) BEAuCHAMP, W. M. 1900. Aboriginal Occupation of New York. Bull. N. Y. State Mus., no. 32. BoAs, FRANZ 1888. The Central Eskimo. BAE-R 6. 1889, 1890. Fifth and Sixth Reports of the Committee ... [on] the North-Western Tribes of ... Canada. BAAS. BoAs, FRANZ, and HAEBEELIN, HERMAN 1927. Sound Shifts in Salishan Dialects. Internat. Jour. American Linguistics, 4:117-136. BoAs, FRANZ, ed.: (HArBERLIN, T=T, ROBERTS) 1928. Coiled Basketry in British Columbia and Surrounding Region. BAE-R 41:119- 484. (Same map as preceding, colored.) BOYLE, DAVID, ed. 1906. Annual Archaeological Report [for Ontario], 1905. Toronto. (Boas, Chamberlain, Hill-Tout, Morice, and others.) CADzow, D. A. 1925. Habitat of Loucheux Bands. MAIHF-IN 2:172-177. COOPER, J. M. 1928. Northern Algonkian Scrying and Scapulimancy. P. W. Schmidt Festschrift, 205- 217. DAL, Wm. H. 1877. Tribes of the Extreme Northwest. CNAE 1. DORSEY, J. 0. 1890. Gentile System of the Siletz Tribes. JAFL 3:227-237. GIBBS, GEORGE 1877. Tribes of Western Washington and Northwestern Oregon. CNAE 1. HILL-TOuT, C. 1907. Report on the Ethnography of the South-Eastern Tribes of Vancouver Island. JRAI 37:306-374. HODGE, F. W., ed. 1907-1910. Handbook of American Indians. BAE-B 30 (pts. 1 and 2). HRDLIdKA, ALrA 1903. The Region of the Ancient Chichimecs. AA 5:385-440. 1904. Notes on the Indians of Sonora. AA 6:51-89. JENKS, A. E. 1900. The Wild Rice Gatherers of the Upper Lakes. BAE-R 19 (pt. 2). KIDDER A. V. 1924. An Introduction to the Study of Southwestern Archaeology. New Haven. KROEBER, A. L. 1907. Shoshonean Dialects of California. UC-PAAE 4:65-166. 1925. Handbook of the Indians of California. BAEB 78. 9 University of California Publioations in Am. Arch. and Ethn. LEzMANN, WALTEz 1920. Zentral-Amerika. I. Teil: Die Sprachen. Berlin. (North American part of map also in Seler, Gesammelte Abhandlungen, 4, Berlin, 1923.) MATHISSEN, Tmn1 L 1927. Arehaeology of the Central Eskimos. Copenhagen. MICHELSON, TRUMAN 1912. Preliminary Report on the Linguistic Classification of Algonquian Tribes. BAE-B 28. MOONEY, JAMES 1894. Siouan -Tribes of the East. BAE-B 22. 1896. The Ghost Dance Religion. BAE-R 14 (pt. 2). 1898. Calendar History of the Kiowa Indiana. BAE-R 17. 1900. Myths of the Cherokee. BAE-R 19 (pt. 1). 1907. The Cheyenne Indians. AAA-M 1 (pt. 6). MURDOCH, JOHN 1892. Ethnological Results of the Point Barrow Expedition. BAE-R 9. NELSON, E. W. 1899. The Eskimo about Bering Strait. BAE-R 18. NEwcoMBE, C. F. 1909. Guide to Anthropological Collection in the Provincial Museum. Victoria. (Map re- produced in reduction and without colors in P. E. Goddard, Indians of the North- west Coast, AMNH-H, 8er. 10, 1924.) ORozco Y BEnRA, M. 1864. Geografla de las Lenguas y Carta Etnografica de M6xico. Mexico. POWELL, J. W. 1891. Indian Linguistie Families North of Mexico. BAE-R 7. ROYCE, C. C. 1899. Indian Land Cessions in the United States. BAE-R 18. SAPPER, CARL 1897. Das nordliche Mittel-Amerika. Braunschweig. 1902. Mittelamerikanische Reisen und Studien. Braunschweig. SPECK, F. G. 1924. The Ethnic Position of the Southeastern Algonkian. AA 26:184-200. SPIER, LEsLIE 1927. Tribal Distribution in Southwestern Oregon. Oreg. Hist. Quart., vol. 28. STEENSBY, H. P. 1917. An Anthropogeographical Study of the Origin of Eskimo Culture. Meddelelser om Gronland, vol. 53. STOLL, OTTO 1884. Zur Ethnographie der Republik Guatemala. Zurich. SWANTON, J. R. 1904. The Development of the Clan System and of Secret Societies among the North- western Tribes. AA 6:477-485. 1911. Indian Tribes of the Lower Mississippi Valley. BAE-B 43. 1922. Early History of the Creek Indians and their Neighbors. BAE-B 73. TEIT, JAMES 1906. Notes on the Tahltan Indians. Boas Anniversary Volume, 337-349. 1928. The Middle Columbia Salish. U1W-PA 2:83-128. THOMAS, CyYRUS, and SWANTON, J. R. 1911. Indian Languages of Mexico and Central America. BAE-B 44. TuRNER, L. M. 1894. Ethnology of the Ungava District. BAE-R 11. WINCHELL, N. H. 1911. The Aborigines of Minnesota. Minn. Hist. Soc. 10 Kroeber: Cultural and Natural Areas of Native North America LIST OF WORKS APPEARING SINCE PREPARATION OF THE TRIBAL MAP BMAN, J. V. 1937. Tribal Distribution in Oregon. A A A-M 47. KELLY, ISABEL T. 1934. Southern Paiute Bands. AA 36:548-560. OSGOOD, C. 1934. Kutchin Tribal Distribution and Synonymy. AA 36:168-179. 1936. Contributions to the Ethnography of the Kutchin. YU-PA no. 14. 1936. The Distribution of the Northern Athapaskan Indians. YU-PA no. 7. RaY, VBNE F. 1936. Native Villages and Groupings of the Columbia Basin. Pacific Northwest Quart., vol. 27, no. 2. RAY, PARK, and others 1938. Tribal Distribution in Eastern Oregon; and: Tribal Distribution in the Great Basin. AA 40:384-415, 622-638. (V. F. Ray on Northeastern Oregon, 384-395; G. P. Murdock, Tenino, Molala, Oregon Paiute, 395-402; B. Blyth, Oregon Paiute, 402- 405; 0. C. Stewart, Northern Paiute, 405-407; J. Harris, Western Shoshoni, 407- 410; E. A. Hoebel, Eastern Shoshoni, 410-413; D. B. Shimkin, Wind River, 413-415; W. Z. Park, Paviotso, 622-626; E. E. Siskin, Washo, 626-627; A. M. Cooke, Northern Ute, 627-630; W. T. Mulloy, Central and Southern Nevada, 630- 632; M. K. Opler, Southern Ute, 632-633; I. T. Kelly, Southern Paiute, 633-634; M. L. Zigmond, Kawaiisu, 634-638.) SAUER, C. 1934. The Distribution of Aboriginal Tribes and Languages in Northwestern Mexico. UC-IA no. 5. SPR, L. 1936. Tribal Distribution in Washington. Gen. Ser. in Anthr., no. 3. STEWARD, J. H. 1937. Linguistic Distributions and Political Groups of the Great Basin Shoshoneans. AA 39:625-634. 1938. Basin-Plateau Aboriginal Sociopolitical Groups. BAE-B 120. 1939. Some Observations on Shoshonean Distributions. AA 41:261-265. STEWART) OMxR C. 1939. The Northern Paiute Bands. UC-AR 2:127-149. TRIBAL SYNONYMS Akansea, Arkansas = Quapaw Bungi = Plains Ojibwa (part) Cahita = Yaqui, Mayo, Tehueco, etc. Cajuenche = Kohuana Carrizo = Comecrudo Cayuse = Wailatpu Central Wintun - Wintun Chippewa = Ojibwa Chontal = Tequistlatec (or Mayan) Chuj = Mame (part) Etago-tine = Daho-tine Etchimin = Malecite Gros Ventre = Atsina Halkomelem = Cowichan and Lower Fraser Hareskin = Hare Hasinai = Caddo (part) Iglumiut = Tahagmiut Iroquois = Seneca, Cayuga, Onondaga, Oneida, Mohawk Irritila = Lagunero Jacaltee = Mame (part) Kawchodinne = Hare Kinipetu = Caribou Eskimo Koso = Panamint Laimon = Cochimi (part) Loucheux = Kutchin tribes Mangue = Chorotega Mascouten = Prairie Potawatomi Meskwakwi = Fo'x Middle Columbia Salish = Sinkiuse (and Wenatchi) Minitari = Hidatsa 11 University of California Publicatios in Am. Arch. and Ethn. TRIBAL SYNONYMS- (Continued) Mohave-Apache = Yavapai Mohegan = Pequot (part) Nahane = Tahltan, Taku-tine, Kaska, Abbato-tine, Etehao-tine, Daho-tine Nestucca = Siletz Niantic = Narraganset (part) Nishinan = Southern Maidu Northern Dieguefno = Western Dieguefno Northern Shoshone = Lemhi Northern Wintun = Wintu Ntlakyapamuk = Thompson Paipai = Akwa'ala Paviotso = Northern Paiute Peau de Libvre = Hare Pinto = Pakawa Pison = Janambre Quicama, Quiquima = Halyikwamai Ree = Arikara Sahaptin = Nez Perce Salish = Flathead Saulteaux = Ojibwa (western part) Seminole = late Creek offshoot Siciatl = Seshelt Sioux = Dakota Snake = Shoshone (and Bannock?) Songish = Lkungen Southern Dieguenlo = Eastern Diegueio Southern Wintun = Patwin Stlatliumq = Lillooet Susquehanna = Conestoga Takulli = Carrier Taratin = Abnaki Tlingeha-tine = Dogrib Tobacco Nation = Tionontati Tojolabal = Chafiabal Uspantec = Ixil (part) Warm Springs = Tenino, etc. Westo = Yuchi Wishosk = Wiyot Wyandot = Huron Xuala = Sara Yopi = Tlapanec Yukaliwa = Kiliwa PRONUNCIATION OF TRIBAL NAMES Vowels in tribal names have their approximate Continental values, consonants the English ones. In Latin America, Spanish orthography has been retained. The principal exceptions follow. a has the value of e: Ojibwa, Iowa, Salish, Waco, Nehalim, Chehalis ai, ay = e: Nottoway, Yanktonai, Kootenay au, aw = o (originally a or ae): Quapaw, Pawnee, Choetaw, Chickasaw, Shawnee, Mohawk, Siuslaw, Sauk, Nauset, Eufaula ee = i: Cree, Creek, Cherokee, Pawnee, Shawnee, Wateree, Pedee, Santee, Congaree, Sewee, Coree, Occaneechi, Oconee, Chattahoochee, Okfuskqe e silent: Seminole, Mobile, Nanticoke, Osage, Spokane, Sinkiuse eu = yu: Eufaula i = ai: Iowa, Kiowa, Siuslaw, Tenino oo = u: Tillamook, Chinook, Kootenay, Lillooet, Bela Coola, Kickapoo, Yazoo ou = u: Missouri ow = au: Powhatan, Cowlitz, Methow, Cowichan y = ai: Chipewyan ch = sh: Cheyenne, Chasta Costa x = sh: Mixe, and others in Spanish orthography x = ks: Comox Accented on first syllable: Navaho, Papago, Opata, Cahita; also, in English, Otomi, Zapotec, Totonac, etc. 12 Kroeber: Cultural and Natural Areas of Native North America IV. VEGETATION AREAS OF THE VARIOUS geographical and environmental classifications which might be compared with the native cultural classifications, those dealing with vege- tation perhaps prove on the whole the most useful. This is expectable, since culture, through houses and fire, enables even the most backward peoples to work out a residence adjustment in almost any climate or terrain, but does not make possible nearly so decisive a control, even through agriculture, of the general vegetation on which, directly or indirectly, most subsistence is based. Wissler has pointed out several ethnic correspondences to altitude, as al- ready mentioned; but on the one hand these are of language groups rather than of cultures, and on the other it seems doubtful whether it is the altitudes as such or their respective climates and plant covers that constitute the con- ditioning factors of the human grouping. Where Wissler has gone farther, as in his Tundra, Mesa, and Jungle division, the classification is too summary to be useful. The culture of his American Mesa, to consider just one example, reached its highest culmination among the Maya proper, whose older as well as newer seats were in the tropical forest. The strongest case for relation of climate and culture could expectably be made with a classification taking into consideration all or several important elements of climate, such as KEppen's, which is based on temperature, precipi- tation, and seasonal change. Unfortunately, no detailed classification of North American climates on Koppen's principles is yet available. The limited maps (nos. 13, 24-27) which have been compiled on this plan are briefly considered below in Section XIII, on "Relations of Environmental and Cultural Factors." Of classifications of the organic environment, the earliest to be developed for North America, the one most intensively mapped, and the one still most influential in geographical studies of faunas and floras is C. Hart Merriam's grouping into "life zones."' These zones are in theory empirical, but avowedly depend on temperature-not mean annual isotherms, but cumulative heat- as determinative of physiological activity in plants and reproductive activity in animals. They run, therefore, generally from east to west, with marked swings and convolutions where altitude or other temperature factors are in- volved. Theoretically, temperature seems too simple a determinant for culture; and a glance at Merriam's map of the United States suffices to show that the life zones have practically no correlation with recognized cultural areas. As a matter of fact, Merriam distinguishes an eastern and a western area, separated approximately by the hundredth meridian, within his life-zone scheme. These two areas obviously differ considerably in both average altitude and precipita- 'See "Authorities Used," below; also Nat. Geogr. Mag., 6:229-238, 1894. 2Normal mean daily heat above 00 C. (60 C. in theory) added up in degrees for the year. This is taken to give the northern limit of species and the life zones based upon them. The southward range of northern species is assumed to conform to the mean temperature of the six hottest consecutive weeks of summer. The life zones conform in general to the first of these two climatic factors, except on most of the Pacific coast of the United States, where cool summers are accompanied by a more northerly flora and fauna than the temperature suimmations determine elsewhere. 13 Universsity of California Publications in Am. Arch. and Ethn. tion. The fact, however, that the zones are run across them shows that the intent is to accord primacy to temperature. A number of areal classifications of North American natural vegetation have been attempted in the past ten to twenty years. The approach has been some- what variable. Harshberger's work, for instance, has been phytogeographic, and is characterized by long species lists. Shelford's is ecological and regional, with fauna considered as well as flora. Shantz and Zon attempt to define and map characteristic and prevalent plant covers: a few typical species rather than the totality represented are taken as determinants. Livingston and Shreve base their work on a classification similar to the last named, but use it for ob- jectives that are physiological and etiological. Nevertheless, the major findings of these and other authors are on the whole fairly concordant; and here, then, we would seem to have something detailed with which the classification of native cultures may profitably be compared. There are several reasons why plant areas should be of special importance in a consideration of culture variations. First of all, they necessarily reflect climate in its totality pretty well, besides accounting for soil influences. Sec- ondly, they underlie fauna, and therefore provide the whole subsistence set- ting of nonagricultural and nonmaritime peoples; while even agriculture must find itself limited by the conditions which express themselves in natural areas of plant cover. Thirdly, the vegetation areas are, like culture areas, strictly empirical, and not devised according to any preconceived scheme of the pri- macy of this or that factor. The plan here followed in the consideration of North American vegetation types is this: The principal areal classifications have been brought together on a series of maps (2-5), drawn to a scale uniform with that used in the tribal, cultural, and physiographic maps (1, 6, 7), and reproduced on transparent paper to allow of superimposition for comparison. In the consideration of cul- ture that follows, such reference as seems appropriate is made to the vegetation of each area. In Section XIII, on environmental factors, some of the more prominent correspondences between vegetation and culture are summarized. AUTHORITIJ S USED FOR THE VEGETATION MAPS (MAPS 2-5) DOMINION or CANADA 1930. Map Indicating Vegetation and Forest Cover, 100 Miles to 1 Inch. Department of the Interior, National Development Bureau, F. C. C. Lynch, Director, 1930. (Present map 4.) HARSHBERLE, J. W. 1911. Phytogeographical Survey of North America. (Engler and Drude, Die Vegetation der Erde, 13.) (Present map 2.) KELOa, R. S. 1910. The Forests of Alaska. U. S. Dept. Agr., Forest Service, Bull. no. 81. (Map from Professional Paper no. 45, U. S. Geol. Survey.) (Present map 5.) LIVINGSTON, B. E. and SmvE, F. 1921. The Distribution of Vegetation in the United States, as Related to Climatie Con- ditions. Carnegie Institution of Washington. (Present map 5.) MALTE, M. 0. 1922. The Flora of Canada. Canada Year Book [for] 1921, pp. 73-81. (Present map 5.) 14 Eroeber: Cultural and Natural Areas of Native North Ainerica MmRLMX C. H. 1898. Life Zones and Crop Zones. U. S. Dept. Agr., Biol. Surv., Bull. no. 10. (The map is reproduced in Livingston and Shreve.) SANDES, E. M. 1921. The Natural Regions of Mexico. Geogr. Rev., 11:212-226. (The map is reproduced in Shelford, fig. 13, p. 576.) (Present map 5.) SHANTZ, H. L., and ZON, RI 1924. The Natural Vegetation of the United States. U. S. Dept. Agr., Bur. Agr. Econ., Atlas of Am. Agr., Pt. I, The Physical Basis of Agr., Sec. E, Natural Vegetation. (Present map 4.) SHELFORD, V. E. 1926. Naturalist's Guide to the Americas. ('Prepared by the Committee on the Preserva- tion of Natural Conditions of the Ecological Society of America, with assistance from numerous organizations and individuals, assembled and edited by the chair- man, Victor E. Shelford." Mr. Shelford has been good enough to provide me with blueprints of the original full-size drawings from which his small maps, figs. 3, 4, and 5, had been engraved. These blueprints have served for the preparation of my map 3, which is therefore more accurate than it would have been if based on the published reductions. This courtesy is gratefully acknowledged.) (Present map 3.) SHR1IvE, F. 1917. A Map of the Vegetation of the United States. Geogr. Rev., 3:119-125. (The map is larger than that in Livingston and Shreve, which is credited to Shreve in that work; but otherwise they appear to be identical.) These sources aggregate four for the United States and Canada, three each for Mexico and Alaska, and two for Central America. In spite of some differ- ences of objective and method among the several authors, their findings agree nearly enough to make the compilation of a generalized map feasible with no very great difficulty. I have been tempted to make such a combination, but the work should properly be done by a botanist. The tinting or shading of the original maps has had to be omitted, and key numbers have been substituted. These numbers have been assigned according to a general scheme, so that the same number denotes the most nearly corre- sponding areas of the different authors. The authors' own terms have, how- ever, been retained for their areas. The concordance or uniformized key list of areal designations follows. CONCORDANCE AREAS TABLE 1 CONCORDANCE KEY OP VEGETATION AREAS REPRODUCED IN MAPS 2-5 Dom, Dominion of Canada (map 4); Ha, Harshberger (map 2); K, Kellogg (map 5); Shr, Shreve (map 5); Ma, Malte (map 5); San, Sanders (map 5); Shl, Shelford (map 3); Sha, Shantz and Zon (map 4). 1. TuNDRA 1. Tundra. Ha, Shl: 1, Tundra. Ma: 1, Arctic. K: 1, Tundra, and la, Area above Timber. Dom: 1, Treeless Plains and Mountains above Timber Line. 2-7. DESERT 2. Salt Desert. Sha: 2, G, Greasewood, Salt Desert Shrub. 3. Desert. Ha: part of 4a, Sonoran Desert. Shl: 3a, Desert, 3b, Extreme Desert. Shr: 3, California Microphyll Desert. Sha: part of 5, CB, Creosote Bush, Southern Desert Shrub. San: 3-4, Desert, including Alkaline Wastes. 15 University of California Publications in An. Arch. and Ethn. 4. Succulent Desert. Ha: 4a, Sonoran, and 5a, Chihuahuan Desert. Shl: 4, Succulent Desert. Shr: 4c, Arizona, 4d, Texas Succulent Desert. Sha: part of 5, CB, Creosote Bush, Southern Desert Shrub. San: 3-4, Desert; see also 17a. 5. Creosote Btush Desert. Ha: part of 4a, Sonoran Desert. Shr: 3, California Microphyll, and 4c and 4d, Arizona and Texas Succulent Deserts. Sha: 5 (=3-4), CB, Creosote Bush, Southern Desert Shrub. 6. Sagebrush-Juniper Semidesert. Ha: 6, Great Basin. Shl: 3a, Desert; 6a, Small-Tree Semi-Desert. Shr: 6, Great Basin Microphyll Desert and part of Western Xerophytie Ever- green Forest (remainder appearing on map 5 as 20x); 6a, Texas Semi-Desert. Sha: 6, SB, Sagebrush, Northern Desert Shrub, and part of J, Pifnon-Juniper, Southwestern Coniferous Woodland (remainder appearing on map 4 as 20x). Dom: 6-11-20; see 20. 7. Chaparral Semidesert. Shl: 7, Broad-leafed Evergreen Semi-Desert, Region of Winter Rains. Shr: 7, Pacific Semi-Desert. Sha: 7, C, Chaparral, Southwestern Broad-leaved Woodland. 8-13. GRASSLAND 8. Swamp Grass. Shl: 8, Grass Swamp. Shr: part of 8-26, Swamps and Marshes. Sha: 8, MG, Marsh Grassland. 9. Tall Grass. Ha: included in 9-10, Prairie-Great Plains. Shl: included in 9-10, Moist Grassland, or Temperate Steppe. Shr: included in 9-10, Grassland. Sha: 9, TG, Tall Grass, Prairie Grassland. Ma: 9, Second Prairie Steppe. 10. Short Grass. Ha, Shl, Shr: part of 9-10. Sha: 10, SG, Short Grass, Plains Grassland. Ma: 10, Third Prairie Steppe. Dom: Prairie, Short Grass. 11. Bunch Grass. Ha: 11, San Joaquin district. Sha: 11, BG, Bunch Grass, Pacific Grass- land. Ma: 11, Dry Belts (of British Columbia). Dom: 6-11-20; see 20. 12. Desert Grass. Shl: 12, Dry Grassland, or Semi-Desert Grassland (Bush Steppe). Shr: 12, Desert-Grassland Transition. Sha: 12, DG, Mesquite Grass, Desert Grassland. San: 12a, Short Grass. 13. Alpine Grass. Sha: 13, A, Alpine Meadow, Alpine Grassland. Shr: see 24. 1-17. PAnKLAND AND SAVANNA 14. Poplar Parkland. Shl: 14, Poplar Savanna. Ma: 14, First Prairie Steppe. Dom: Grove Belt (mostly poplar in prairie). 15. Oak Parkland. Ha: 15a, Transition Prairie-Forest, Oak Openings, and 15b, Texas Cross Timber and Coast Plain Belt, with Live Oaks and Prairies, and part of 29b, Ozark, and 29c, Edwards Plateau Forest. Shl: 15, Oak Savanna. Shr: 15, Grassland-Deciduous Forest Transition. San: 15-30, Deciduous Trees, chiefly Oak. 16. Moist Savanna. Ha: various. Shl: 16, Moist Savanna, not distinguished by symbol from 15. San: see 15-30. 17. Dry Savanna. Shl: 17, Arid Tree or Bush Savanna. Sha: 17, DS, Desert Savanna, Mesquite and Desert Grass Savanna. San: 17a, Scrub, chiefly Mesquite, Yucca, Agave, Cactus. 18-24. CONIFEROUS FOREST 18. Northern Coniferous Forest. Ha: 18, Subaretie (Hudsonian), Northern Coniferous Forest. Shl: 18, Northern, or Moist, Coniferous Forest. Shr: 18, Northern Mesophytic Ever- green Forest; (20w, below, is also ineluded by Shr in 18). Sha: 18, S, Spruce-Fir, Northern Coniferous Forest, and 18b, JP, Jack, Red, and White Pines, Northeastern Pine Forest. Ma: 18, Sub-Arctic. Dom: 18, Sub-Arctic Forest, and 18-20, Northwestern Coniferous Forest. K: 18, Timbered, 18a, Sparsely Timbered; (see also la, Above Timber; and 21). 19. Northeastern Coniferous Forest (with deciduous admixture). Ha: 19, St. Lawrenee- Great Lakes. Shl: see 25. Shr: part of 18. Sha: part of 18 (S), 18b (JP), and 25 (BM). 16 Kroeber: Cultural and Natural Areas of Native North America Ma: 19, Hardwood Forest. Dom: 19c, Eastern Coniferous Forest, and 19m, Mixed Forest, also Cleared Portions of Eastern Forest Belts." 20. Western Mountain Coniferous (Pine) Forest. Ha: 20, Rocky Mountain; 20e, Sierra Nevada, and 20f, San Bernardino; 20g, Santa Lucia area of California Coast Range; 20h, Western Sierra Madre. Shl: 20, Desert, or Mountain, Coniferous Forest. Shr: 20, Western Xerophytie Evergreen Forest (mostly merged in map 5 in areas 6, 3, 4c, 12, 9-10, 4d); 20w (of map 5) is treated by Shr as part of 18, Northern Mesophytic Evergreen Forest. Sha: 20a, P, Yellow Pine-Douglas Fir, 20b, LP, Lodgepole Pine, and 20c, SP, Yellow Pine-Sugar Pine, the three constituting the Yellow Pine-Douglas Fir area of Western Pine Forest; also 20d, WP, Western Larch-Western White Pine, part of Cedar-Hemlock or Northwestern Coniferous Forest; also 20x, Pinon-Juniper, Southwestern Coniferous Woodland (partly merged, in map 4, in areas 5, 6, 12, 10). San: 20h-23, Pine Forest. Ma: 20, Rocky Mountains, and 20d, Selkirk Mountains (see 21s). Dom: 6-11-20, Semi-open Coniferous Forest of South- ern Interior British Columbia (sagebrush, bunch grass, yellow pine, Douglas fir, according to elevation); 18-20, Northwestern Coniferous Forest (see 18). 21. Northwestern Coniferous Forest. Ha: 21a, Sitkan, and 21b, Columbian; and 21c, Mendocino area of California Coast Range district (= 20g and 21c). Shl: Northwestern Coniferous Forest. Shr: 21, Northwestern Hygrophytic Evergreen Forest. Sha: 21, DF, Pacific Douglas Fir, and 21c, R, Redwood, constituting Cedar-Hemlock or Northwestern Coniferous Forest (in which Sha also includes 20d, WP, here reckoned under 20). Ma: 21, Coast Mountains. Dom: 21, Western Coniferous Forest, Coastal, and 21s, Western Conif- erous Forest of Interior Wet Belts of British Columbia (= Ma: 20d). K: 21, Timbered (not distinguished from 18 by K; the broken line in map 5 has been added). 22. Southeastern Coniferous (Pine) Forest. Ha: 22, Atlantie-Gulf Coastal, with Pine Barren-Strand vegetation. Shl: 22, Southeastern Coniferous Forest; and 22a (=26b), Flat- woods. Shr: 22, Southeastern Mesophytic Evergreen Forest. Sha: 22, LLP, Longleaf- Loblolly-Slash Pines, Southeastern Pine Forest. 23. Arid Coniferous (Pine) Forest. Ha: 23a, Eastern Sierra Madre, and 23b, United Cordilleran. Shl: 23, Arid Coniferous Forest. San: 20h-23, Pine Forest. 24. Alpine Coniferous Forest. Shl: 24, Sub-Alpine Evergreen Forest, and 24a, High Moun- tain Forest. Shr: 24, Alpine Summits (see 13). 25. CoNnrERous-DEcmuous FoREsT 25. Northeastern Mixed Forest. (Ha: see 19). Shl: 25, Mixed Coniferous and Deciduous Forest. Shr: 25, Northeastern Evergreen-Deciduous Transition Forest. Sha: 25, BM, Birch- Beech-Maple-Hemlock, Northeastern Hardwoods. Ma: 25, Carolinean. Dom: 25, Southern Hardwood Forest (includes southern strip of Dom: 19m). Southeastern Mixed Forest. See 28, Piedmont Deciduous Forest. 26-32. DnIcmuous FORZST -26. Swamp Forest. Shl: 26, Cypress Swamp, or Tree Swamp, and (26b =) 22a, Flatwoods (pine forest interspersed with cypress swamp), and 26c, Magnolia Hammock (higher por- 8 Area 19m of my map 4 is represented by two differently colored areas on the Dominion map, "Mixed Forest" and "Cleared Portions of Eastern Forest Belts, Including the Hard- wood Forests of Southern Ontario and Southern Quebec." So the legend in the key. The legend on the map itself reads "Cleared Portions of Hardwood Forest." The color, however, is continued to the very mouth of the St. Lawrence, into New Brunswick and Nova Scotia, and in patches north into the Coniferous Forest as far as 490, all of which are well beyond the limits of any hardwood forest. The species listed in the key legend for "Cleared Por- tions" also are nearly the same as the species characterizing the "Mixed Forest." It is therefore evident that while the "Cleared Portions" represent elearing and not any one ex- clusive type of native vegetation, the great preponderance of the area was natively in "Mixed Forest"; and the whole of it has been so designated, except for the patches wholly within Coniferous Forest. A strip along the St. Lawrence and Lake Ontario was undoubtedly Hard- wood, connecting with the area designated as Hardwood on the north side of Lake Erie; but there is no way of demarking it from the major Mixed Forest portion of 19m, except by reference to the maps of Malte and other authorities. 17 University of Calfornia Publications in Am. Arch. and Ethn. tions of Tupelo low hammock). Shr: part of 8-26, Swamps and Marshes. Sha: 26, CT, Cypress-Tupelo-Red Gum, River Bottom Forest, and 26a, M, Mangrove, Subtropical Forest. .27. Appalachian Deciduous Forest. Ha: 27, Appalachian Mountain Deciduous Forest. Shl: included in 27-28-29, Temperate Deciduous Forest. Shr: included in 27-29, Deciduous Forest. Sha: 27, o0, Chestnut-Chestnut Oak-Yellow Poplar, part of Southern Hardwood Forest (= 27-28-29). 28. Piedmont Deciduous Forest (with coniferous admixture). Ha: 28, Piedmont. Shl: included in 27-28-29, Temperate Deciduous Forest. Shr: 28, Southeastern Evergreen- Deciduous Transition Forest. Sha: 28, OP, Oak-Pine, part of Southern Hardwood Forest (27-28-29). 29. Mississippi Valley Deciduous Forest. Ha: 29a, Lacustrine and Kentucky-Tennessee areas, and 29b, Ozark area, of Alleghanian-Ozark distriet, and part of 29c, Edwards Plateau Forest. Shl: included in 27-28-29, Temperate Deciduous Forest. See also 26c. Shr: included in 27-29, Deciduous Forest. Sha: 29, OH, Oak-Hickory, part of Southern Hardwood Forest (= 27-28-29). 30. Arid Deciduous Forest. Ha: 30c, Jaliscan. Shl: 30a, Arid Deciduous Forest, and 30b, Deciduous Thorn Forest. San: 15-30, Deciduous Trees, chiefly Oak. 31. Tropical Bain-forest Subclimax. Ha: 31c, Gulf Mexican. Shl: 31a, Montane or Cloud Forest, and 31b, Drier Tropical Rain Forest. San: 31d, Jungle. 32. Tropical Bain-forest Clima. Ha: 32e, Floridian and Insular areas of Bahaman re- gion; 32d, Antillean region; 32e, Guatemalan region, Central American province; 32f, Costa Rican region, South American province. SM: 32a, Luxuriant Tropical Rain Forest, and 32b, Tropical Rain Forest Climax. San: 32, Tropical Rain Forest. This concordance key together with maps 2-5 seems to go as far as is proper for a nonbotanist in blocking out the major vegetation areas on which ecologi- cal botanists are in substantial agreement, without attempting to decide upon the respective merits of their bases of classification or the relative accuracy of their areal limitations. At any rate, it provides something against which areal classifications of culture can be compared with reasonable approximation. The areal limits of the originals have been altered in maps 4 and 5 for cer- tain simplifications, which are here enunmerated. These simplifications have been enforced by the nonuse of color, without which many of the minute or irregularly narrow areas, especially of the Shantz-Zon atlas, cannot be repro- duced with effectiveness to the eye. Numerous long tongues of deciduous forest Oak-Hickory bottom lands (29) extending up the western afifuents of the Missippi, and of southeastern River-bottom Forest (26): omitted or shortened. Small areas or narrow fringes of Alpine Meadow (13), Tall Grass (9), Marsh Grass (8), Salt Desert or Greasewood (2): omitted. Small high-altitude patehes of Eastern Spruce-Fir (18) enclosed in areas of Northeastern Hardwood Forest (25) in the Appalachian ranges: omitted. Western Spruce-Fir has throughout been merged in the Douglas Fir (21) or Western Pine (21a, b, ?, d) areas in which it is enclosed or to which it is marginal. On both maps 4 and 5 the western Piflon-Juniper areas of Shreve (Western Xerophytic Evergreen Forest) and Shantz-Zon (Southwestern Coniferous Woodland) have been some- what summarily simplified. It is clear that this plant eover represents in the main a contour vegetation between the pine forests of higher altitudes and the desert shrub and grassland of lower levels. Particularly evident is its association with sagebrush, from which it rises islandlike or marginally as a funetion of increased altitude or slope, and therefore in numer- ous patches and irregular fringes. In both maps 4 and 5 the plan has therefore been fol- 18 Eroeber: Cultural and Natural Areas of Native North America lowed of converting Pifion-Juniper outright into Sagebrush (Great Basin Microphyll Desert, 6) wherever the original maps show the two in contact. Similarly, it has been merged with Creosote Bush (5), Desert Grass (12), and Short Grass (10) of Shantz-Zon; and the Cali- fornia Microphyll Desert (3), Arizona Succulent Desert (4c), Desert-Grassland Transition (12), Grassland (9-10), and Texas Succulent Desert (4d) of Shreve.' This leaves as Western Xerophytic Evergreen Forest (20) of Shreve only a compact area in southern Texas, and as Southwestern Coniferous Woodland (20x) of Shantz-Zon a fringe bordering the Yellow Pine (20a) mountain areas of northwestern and central Arizona and western New Mexico, Texas, and Colorado. Shreve makes no distinction between the eastern (St. Lawrence-Great Lakes) and western (Rocky Mountain) portions of his Northern Mesophytic Evergreen Forest. The former has been retained as 18 on map 5, but the latter redesignated as 20w. Shreve also does not distinguish between forested Swamps and grass Marshes. His areas of these have accordingly been variously designated in map 5 as 26 (swamp forest), 8 (marsh), or 8-26. At the points mentioned, therefore, recourse must be had to the original Shantz-Zon and Shreve maps where accuracy of detailed reference is desired. The simplifications introduced in maps 4 and 5 seem unavoidable if effective comparability is the end sought, and seem to do a minimum of violenee to the intent of the originals. Kellogg's Alaskan map shows the presence and density of timber, not the affiliations of the forest growth. His uniform "Timbered" area has therefore been divided, in map 5, be- tween Northern (18) and Northwestern (21) forest, as shown by the broken line. His "Sparsely Timbered" area has been designated as a variant of the Northern Forest, 18a. His areas "Above Timber" are designated as a variant of Tundra, la; "Glaciers and Snow- fields" are included in this. Shelford uses a single symbol for Tundra and for Paramos and High Mountain Forest, which are distinguished as 1 and 24a in map 3. In the reproduction of the DoTninion of Canada map, "Cleared Portions" have been mainly counted as Mixed Forest, as already discussed in a footnote to 19m. In the west, Treeless and Above Timber Line have both been designated by the same symbol, 1, Tundra, because of lack of distinetion in the original. I have also introduced some simplification of the end- less minor interdigitations of "Above Timber" with the various forests: 18; 18-20; 6-11-20; 21s; and 21. ' Where Pifnon-Juniper is adjacent to two of these vegetations it has been assigned to the one that is mentioned first in this paragraph. 19 University of California Publications in An. Arch. and Ethn. V. CULTURE AREAS: ARCTIC COAST THE NATIVE CULTURES and their areas will now be considered, points of differ- ence from the classifications in current usage being discussed as they arise. The chief characteristics of the present classification are the following: 1. Specific attention is given to geographical and ecological factors. 2. The cultures are treated as historical nonequivalents. 3. Centers or climaxes of culture are defined as sharply as possible. 4. Relations of subordination between and within cultures being sought and expressed, the number of basic areas is fewer, and of specifie ones greater, than it has been customary to recognize. The segregation of the eighty or so areas dealt with is into six groups, namely: A. Arctic Coast (A in map 6). D. Intermediate and Intermountain (I). B. Northwest Coast (NW). E. East and North (E). C. Southwest (SW). F. Mexico and Central America (M). With the partial exception of the fourth, each of these is believed to repre- sent a substantial unit of historical development, or of a prevailingly charac- teristic current of culture. Of course, these six units are also interrelated; and on the grounds of cul- tural primacy and prevailing historical priority Mexico ought to be considered firs.t. But incompleteness and lack of organization of data make analysis of this area the least satisfactory; so that the reverse order of procedure, from peripheral to central, is for the present almost enforced. The findings are embodied in map 6. ARCTIC COAST SOURCES OF ESKIMO CULTURE Eskimo culture is the most differentiated of lower-grade cultures in America. It therefore deserves to be considered as constituting a primary division. This conclusion is strengthened by the unchallenged separateness of Eskimo speech from any other American language, and the marked racial differentiation of the Eskimo from other American natives. Over its whole eastern extent the culture has mixed little with that of the Indians, on either side of the boundary. Traits have crossed, but the culture wholes have remained conspicuously dis- tinct. The culture has, however, numerous Asiatic relations; especially to the northeastern Palaeo-Asiatics, but traceable as far south as the Kamchadal or beyond and west to the Samoyed and perhaps Lapps. Its Magdalenian resem- blances, while easily exaggerated and difficult to evaluate, are almost certain to carry some historic significance. This, accordingly, seems the most non- American culture of the continent in its major specific origins. Such a con- clusion, however, does not contravene the possibility that the characterization of Eskimo culture as known to us was worked out in America. As to ecology, there has of late been a tendency to emphasize the importance of the tundra and the caribou as against the shore and the seal in Eskimo cul- 20 Kroeber: Cultural and Natural Areas of Native North America ture. In wider historical perspective this seems correct, with reindeer equated to caribou, and with reference to ultimate Eurasiatic origins. The use of coast and of sea mammals would then represent mainly the development of a later American, or Northeast Asiatic-American, phase of the culture. If so, the tundra-caribou form of Eskimo culture found about Chesterfield Inlet and the Back River would have to be interpreted as a secondary, local reconvergence to a much earlier or pre-Eskimo phase. Steensby's view' is the opposite one: he regards Eskimo culture as having originated inland in the Mackenzie drainage, in a caribou habitat probably centering about Great Slave Lake, and as having only later pushed to the sea, where the seal provided winter food, while caribou hunting was retained, wherever possible, as the chief means of summer subsistence. This maritime adaptation was worked out in the region of Coronation Gulf and the isthmuses of Boothia and Melville peninsulas; and there it has persisted in purest form. This argument of Steensby's can, however, be read backward, as Hatt has done, just as well as forward; and the following reasons seem to favor an in- terpretation the reverse of Steensby's: 1. The formation of the distinctive speech and physical type associated with so much of Eskimo culture is hard to account for in a particular part of a con- tinental interior which lies open, without geographical barriers or peculiari- ties. The selection of one portion of the Mackenzie drainage as the former home of Eskimo culture is arbitrary. If a caribou origin is to be hypothesized, the entire range of the animal from Alaska to Labrador might as well have been Eskimo. 2. The cultural similarities with Asia are underweighted by Steensby. These are undoubtedly strongest about the Bering Sea; but the fact that there has been recent influencing in this region does not mean that all influencing is re- cent. It is rather an argument, in the absence of anything specific to the con- trary, that the influences are ancient also. 3. Steensby's hypothesis makes the original sea-adapted culture persist in purest form at its original point of characterization, which is contrary to the age-and-area principle that persistences tend to occur at the peripheries. This principle, indeed, applies rather to traits or relatively small clusters of traits than to whole cultures. But while whole-culture types may appear with less purity toward their peripheries, this implies an intensity, complexity, and richness of characterization at the center which the Coronation-Melville area does not possess, being in fact more meager than the Alaska and Greenland peripheries. Its cultural quality is merely a certain "purity" of narrow spe- cialization along selected lines; which is most simply explained as a selection enforced by the extremity of high Arctic environment. 4. Mathiassen' has shown that the late prehistoric "Thule" form of Eskimo culture of the Coronation-Chesterfield-Melville area is closer to that of Alaska 'An Anthropogeographical Study of the Origin of Eskimo Culture (Meddelelser om Gron- land, vol. 53), 1917. 2Archaeology of the Central Eskimos, pts. 1 and 2, 1927, constituting vol. 4 of Report of the Fifth Thule Expedition [of] 1921-24. 21 University of California Publications in An. Arch. and Ethn. and Greenland than is the present Eskimo culture of the same region. Of 1.52 elements determined as characteristic of this Thule phase, he first eliminates 57 as common to all Eskimos except where the environment inhibits their use. Of the remaining 95, nearly half, or 47, reappear in recent Alaska and Green- land but are lacking among the Coronation-Chesterfield-Melville tribes, his Central Eskimo proper. Eighteen traits are confined to Thule and Alaska; only 3, all scraper forms, are exclusive property of Thule and Central. Out of 95 nonuniversal Thule elements, 71 reappear among the recent Eskimo from the mouth of the Mackenzie west ;' 58, among the Greenland Eskimo; 27, in Baffinland and Labrador; only 16 among the four most specialized Central Eskimo groups-Copper, Caribou, Netsilik, Iglulik-Aivilik. In short, a rela- tively uniform phase of Eskimo culture not long ago prevailed uninterrupt- edly from Alaska to Greenland, but was later modified, with a shift from whale to caribou or winter-seal dependence, in the very region in which Steensby supposes Eskimo culture to have been formed; whereas the western and east- ern ends of the Eskimo range preserved this old phase much more fully.' Even Baffinland and Labrador remained somewhat conservative; and here and there, especially on Southampton Island and Smith Sound, isolated communities re- tained much of the Thule culture relatively uninfluenced by the later Central Eskimo developments, even though local exigencies caused them to adopt modi- fied subsistence habits. ECOLOGICAL PHASES While Steensby's conclusion that Eskimo culture in the Coronation-Melville area developed out of a pre-Eskimo interior culture can therefore be rejected, his work is of the highest importance as an ethnogeographic study. He has for the first time outlined, for the whole of Eskimo territory, the importance of shore line, seasonal open water, drift and shore ice, driftwood or timber, and other natural features as they determine the presence or accessibility of vari- ous animal species and the habitual movements, occupations, and implement types of the Eskimo. What emerges from the total array of his succinctly an- alyzed data is not the primacy or priority of one particular economic adapta- tion, but a picture of the totality of Eskimo culture as a unit, modified by emphasis or reduction of its traits in direct response to local exigencies. Here seals are the important food, there whales, or walrus, or caribou, or birds, or salmon, while others are as good as unavailable. According to ice and water and season, seals are taken by maupok or waiting at the blowhole, utok or creeping, at cracks or the edge of the ice, from the kayak, or by. nets. Even this last method, which is so specially developed in Alaska as to look at first as if its spread were determined culturally instead of ecologically, was known in Greenland, Labrador, and the Central regions. Where continuous ice or snow fields are lacking, the sled of course goes out of use, both in southern 8This would not mean that an equal proportion of Alaskan elements would be found in the Thule culture, because Eskimo culture especially in southern Alaska has absorbed many elements presumably non-Eskimo in origin. Mathiassen, however, considers the Point Bar- row the most similar of all modern Eskimo cultures to the ancient Thule culture. 'Birket-Smith, as referred to below, accepts this change in the Central region, but con- strues it as confirmatory of views similar to Steensby's. 22 Kroeber: Cultural and Natural Areas of Native North America Greenland and southern Alaska; but it is employed to the limit of its utility. Caribou are eagerly hunted wherever they can be got. Whether for the most part they are surrounded, driven in fences, intercepted at passes, or kayaked in lakes depends on the opportunities afforded by the country; more often than not, in fact, two or more of these methods are used in support of one another. So with houses. Where, as on Coronation Gulf and in parts of Baffin- land, seals far from shore are the only dependable subsistence available during a considerable part of the year, and the Eskimo have therefore to live on the ice, the snow house may wholly displace that of stone or sod. In southern Greenland and on the Mackenzie, on the contrary, driftwood is abundant, good-sized timbered houses are built, and the snow house is lacking except as a travel shelter. On the rocky islets and headlands of Bering Strait, wood is again abundant and the houses stand on piles against the steep face of a slope. If whale hunting is productive, the umiak is well equipped and paddled; elsewhere, it is a freight boat, rowed by women; or where there are no whales and the short season of open sea is spent inland to get caribou, as on the shores of Coronation Gulf and on Boothia Peninsula, the umiak is absent. The list herewith shows the principal regional variants of Eskimo economic culture, some twenty-five in number. These are direct ecological adaptations from the basis of a cultural inventory that is or apparently was substan- tially uniform over the entire Eskimo range: skin boats, harpoon, bladder or inflated skin, spear thrower, three- or four-pronged bird spear, two-winged salmon spear, lamp, stone pot, house platform, type of clothing, ivory carving, kashim or social house, shamanism, type of myth or tale. TABLE 2 REGIONAL VARIANTS or ESKIMO EcoNouC COULTuR Northeast Greenland. Extinct. Southeast Greenland. Angmagsalik. Southwest Greenland. Subarctic culture, without sled, snow house, caribou, maupok or utok seal hunting; kayak hunting highly developed. Northwest Greenland. A rather generalized type of Eskimo adaptation. Smith Sound, Polar Eskimo. Loss of kayak, umiak, sled, salmon and reindeer taking, until renewed contacts with Baffinland about 1865; seal and walrus hunting; special dependence on birds. Baffinland. Seal hunting and winter dwelling on the ice, hence maupok and utok methods and snow house. North Labrador. Sealing from ice edge and kayak; reindeer important. South Labrador. Same but more subarctic. Southampton Island. Ancient (Thule) type of culture modified by a specialization on rein- deer hunting; no skin boats. Chesterfield Inlet and Back River: Kinipetu, Caribou Eskimo. Tundra habitat, with de- pendence almost wholly on caribou, secondarily musk ox; almost no use of coast or sea mammals. Melville Peninsula, including northwest Baffinland: Aivilik, Iglulik. Walrus, seals, rein- deer important; snow house replacing stone or sod house. Boothia Peninsula and King William Land: Netsilik. Seals by maupok and utok method, reindeer, no walrus or whales, no umiak, snow house for winter habitation. Coronation Gulf: Copper Eskimo. Much the same as last. 23 4University of California Publications in Am. Arch. and Ethn. Mackenzie River. Large and small whales in summer, seals in winter, salmon. Much wood, timber houses. Here begin the first traits of specific Western Eskimo culture on the super- subsistence level. Point Barrow. Whaling of primary importance; taking of seals especially by netting; reindeer hunting left to essentially inland groups. No snow house here or beyond. Kotzebue Sound, including neck of Seward Peninsula. Beal netting; taking of large whales important. Seward Peninsula, and Diomede and King islands. Whaling, walrus, seal netting, high development of umiak for voyaging; houses on piles. Northeast Siberia: Yuit. Generally similar to last. St. Lawrence Island. Similar especially to last. Norton Sound, especially south side. Similar to Kotzebue, but with more southern influ- ences, such as development of masks. Subarctic conditions begin here. Yukon-Kuskokwim deltas. Shallow shore waters; no whaling; little sealing; prime de- pendence on salmon, supplemented by other fish and birds; no reindeer. Masks, feasts, wood carving in full development. Bristol Bay. Little known. Aleut. An open-sea culture, with dependence on fish and kayak-hunted seals. Kadiak Island and opposite mainland. Temperate climate; salmon and other fish; high development of kayak. Social attitudes savor of Northwest Coast. Kenai Peninsula-Copper Biver. Similar. CULTURAL CLASSIFICATION AND HSTORY In contrast to this uniform array of culture elements varied only according to local needs, there is a series of traits, little connected with subsistence, which mark off the western from the central and eastern Eskimo. These include la- brets, masks, hats in place of hoods, coiled basketry or other weaving, pottery, grave monuments, mourning feasts or ceremonies, property distributions, war parties, perhaps clans or moieties. None of these extends beyond the Mac- kenzie, except for sporadic occurrences like occasional masks; many of them stop at or before Point Barrow and are therefore wholly Alaskan. In the main these traits seem to reflect the influence of the Northwest Coast tribes, espe- cially the Tlingit, or, in part, of the Athabascans influenced -by the Tlingit. Many may be ultimately Asiatic in origin; some, like pottery and coiled bas- ketry, may have drifted in from a long distance away. The primary division of Eskimo culture, then, apart from local adaptations comparable to those of shore and interior or valley and hill tribes in Califor- nia, is into a Central-Eastern and a Western or Alaska-Siberian form, the former being "pure" Eskimo, the latter Eskimo plus a Northwest American and Northeast Asiatic addition. It is a fair logical question whether the sequence implied in the word "addi- tion" could not be reversed, and Eskimo culture be construed as having de- veloped in its present richer Alaskan form in Alaska, the region of fullest con- tacts, and then diffused eastward, the rigor of the Coronation Gulf environ- ment ifitering out many of its supersubsistence elements, while necessity, and paucity of alien contacts, preserved the subsistence devices relatively unal- tered, except for a measure of modification among the Coronation-Melville groups. This view involves a further one, namely, that the contact of cultures in and about Alaska which resulted in the formation of Eskimo culture caused 24 Kroeber: Cultural and Natural Areas of Native North America not only absorptions from the contributing cultures, such as masks and labrets, but also new productions such as lamps and skin boats, and that on the spread of this culture eastward out of Alaska the absorptions were in gen- eral lost and the new specific products retained. While this seems theoretically improbable, it may well have happened to a considerable extent because of the definite utility of the new productions. Really, the two views are not incompatible. Influences from several seaboard cultures situated on subarctic or temperate shores mnay have met in the region of Alaska and produced an Eskimoid type of culture, which then in its east- ward spread through the high Arctic became strained out into "pure" Eskimo culture as we know it today, both because of the unusual but necessary con- centration in high latitudes on subsistence activities, and because of the spe- cialization of these with reference to sea mammalian life. At the same time the culture impingements in Alaska continued, leading to further absorptions and a general enrichment of the culture, but also to less homogeneity and uniqueness of cast. On this view, the shores of the vicinity of Alaska would have been both an ancient and a modern meeting ground of various cultural influences, pre-Eskimo, non-Eskimo, and Eskimo; and from the stock of sea- adapted culture there accumulated, the shore peoples eastward selected, not only once but more likely several times or continuously, such elements as they could use, besides of course modifying them. Alaska then would be the point of origin in the sense of point of crystallization-of Eskimo as contrasted with non-Eskimo culture as a whole, and at the same time the area where this culture remained most "mixed," least set apart by rigorous restriction to its own specializations.' This interpretation of the culture, incidentally, accords well with the situa- tion in racial type and speech, both of which are "purer," more characteristi- cally or undilutedly Eskimo, in the east than in the west, especially if the Aleut are included. The fundamental difficulty about deriving Eskimo culture from the north- ern interior of America is that it is hard to conceive of an inland culture origi- nating the many definite and accurate devices relating to the sea and sea life which constitute the most fundamental and distinctive aspects of Eskimo cul- ture. To take as an example Birket-Smith's "two main props of coastal life" in the far north, the blubber lamp and seal hunting at breathing holes,7 these both depend on and relate exclusively to sea mammals. The antecedents for the invention or development of these traits are much more nearly given in a sub- arctic sea-adapted culture than in a ruminant-hunting, wood-burning tundra or forest culture. The case is much like that of a people practicing a specialized agriculture, such as desert irrigation, under rigorously limiting natural con- " Boas, Die Resultate der Jesup-Expedition, ICA 16 (1908, Vienna) :3-18, 1910, inclines to the view, on folkloristic grounds, that there once existed a connection between the peoples of the Sea of Okhotsk and of British Columbia, which later was more or less interrupted by the arrival of the Eskimo about Bering Strait. If for "arrival of the Eskimo" we sub- stitute "development" or "crystallization of Eskimo culture," Boas's opinion is not incom- patible with that advanced here. ',Boas, AMNH-B 15:369, 1:907; Hrdlicka, BAE-R 46:364, 1930. 7AA 32:623, 1930. 25 6University of California Publications in An. Arch. and Ethn. ditions. All we have learned of the nature of culture processes in the last gen- eration would lead us to expect such an agriculture to be derived from a more generalized, less conditioned type of agriculture evolved elsewhere, rather than from a tour-de-force "invention by necessity" by a nonagricultural popu- lation finding itself in a habitat with insufficient wild food. My division of Eskimo culture into primary Western and Eastern types is therefore not only statically descriptive of recent conditions, but also likely to reflect a fundamental historic current. The Western form is at once older and more heterogeneous, the Eastern is strained out. Both are littoral, and have been such as far back as they may properly be designated Eskimo.8 Within the Western or Alaskan area, the Aleut evidently constitute a sub- area, whose validity is reenforced by the relative distinctiveness of Aleut speech and somatic type. Some of the specialists in the Eskimo field seem to regard the Aleut as an "Eskimoized" population; that is, an originally non- Eskimo group which took on something of Eskimo language and culture. It does not seem necessary to go quite so far in hypothesis as this. The Aleut may represent merely a specialization away from the other Eskimo. They live in a cul de sac, rather isolated from contacts; and their environment certainly is distinctive: oceanic islands, a damp, foggy, windy, raw climate. One could perhaps speak with more assurance of the place of Aleut culture if more were known of the Eskimo to the east of them. Whether these Eskimo of the stretch of coast east of the Aleutians, from the Alaska Peninsula to the Copper River, are to be classed rather with the Aleut, with the Alaska Eskimo generally, or as a distinctive subunit of these, it is difficult to decide without an intensive comparative study, and for this modern ethnographic data are not available. The subarctic environment per se of these Eskimo does not seem to have differentiated them much if any more than it has the southern Greenland Eskimo; they make kayaks, for in- stance, in an area of good growing timber. But on the cultural levels above those connected with subsistence they have been exposed to strong Indian in- fluences, as the Greenlanders have not. These influences, Tlingit in recent times, have presumably been strongest at the eastern border, about the Cop- per River. Also, the stretch from the Kenai Peninsula to the Copper River is sometimes reckoned as ecologically more nearly related to the habitat of the northwestern Tlingit than to the Bering Sea and Arctic coast of Alaska.9 The inland culture of the Chesterfield Inlet-Back River or Caribou Eskimo may probably best be regarded as primarily a specially marked instance of the ecological response variations discussed above. This group seems never wholly to have lost touch with the sea. They have merely gone one step farther than the inland minority of the Point Barrow division. These two groups are of interest as true tundra dwellers; but it is doubtful if they are very much more 8 Steensby's "Neo-Eskimo area of acculturation" differs from the Alaska Eskimo area as here defined. He makes its distinctive features recent, mainly derived from Asia, and local- izes it about Bering Strait, with Kotzebue and Norton sounds. My Western area takes in, with its variants, all the Eskimo-inhabited shores of Alaska, and is both ancient and modern, with the recent absorptions rather from American Indian than Asiatic sources. Compare below, Northwest Coast, Northern Maritime subarea, p. 29. 26 Kroeber: Cultural and Natural Areas of Native North America specialized away from "normal" Eskimo sea-mammal and shore life than are the Yukon and Kuskokwim salmon-eaters.10 The recent Eskimo may therefore be classified culturally as follows: la. Central-Eastern: From Coronation Gulf east. lb. Barren Ground: Caribou Eskimo. 2a. Western: Mackenzie, Alaska to Bristol Bay, Siberia. The Yukon-Kuskokwim Delta may prove to belong with 2c rather than here. 2b. Aleut. 2c. Pacific Coast: Alaska Peninsula to the Copper River.n SUMMARY The origin of Eskimo culture is unknown. Its ultimate affiliations seem Asiatic rather than American. The area of specifically Eskimo characterization may have been American or Asiatic-American; but it is unlikely to have lain east of Alaska, and it was coastal, with primary dependence on sea mammals and fish. This culture came to extend from Sibhria and Alaska to Greenland. After a time it became somewhat modified in the Central area, especially west of Hudson Bay, partly through the lure of caribou hunting, partly through im- poverishment due to arctic rigor. Meanwhile, too, perhaps even earlier, the Western Eskimo culture began to alter as a result of the fairly developed cul- tural contacts to which it continued to be exposed. The most important of these influences were much diminished north of Bering Strait, more so beyond Point Barrow, and practically terminated at the mouth of the Mackenzie, though a few of the older elements may have penetrated sporadically even as far as Greenland. Also, these Northwest Coast and Asiatic influences have con- tinued to recent times, possibly with increased force. Otherwise, Eskimo cul- ture has retained its stock relatively unaltered, except for a modification into about two dozen local phases, which are essentially ecological subsistence adap- tations with resultant reduction or emphasis of common culture traits. 10 K. Birket-Smith takes the opposite view in The Caribou Eskimo, Rept. Fifth Thule Exped., vol. 5, pts. 1 and 2, 1929 (esp. pt. 2, 212-233), and in a controversy with Mathias- sen, AA 32:591-607 and 608-624, 1930. He postulates an inland Proto-Eskimo stage, more or less represented today by the Caribou Eskimo, and only by them. This on pushing to the littoral became Palaeo-Eskimo culture, which in turn developed into Alaskan, Central- Thule, and Greenland phases of Neo-Eskimo. This was still later replaced in the Central region by the Eschato-Eskimo culture, which is closely allied to the Palaeo-Eskimo, and therefore represents a reversion due to renewed influences or advances by Eskimo who had remained inland with the Caribou group. See especially Car. Esk., fig. 5, p. 232; also ICA 23 (1928, New York) :470-475, 1930. The evidence on which his and Mathiassen's construals rest is too detailed to be gone into here. H H. B. Collins, Jr., Culture Migrations and Contacts in the Bering Sea Region, AA 39:375-384, 1937, reviews judiciously the recent archaeological and other data which at once illuminate and complicate Western Eskimo culture history. The Thule culture, he con- eludes, entered Alaska from the east, and late, contemporary with the Punuk phase (post- Old Bering Sea of St. Lawrence Island and post-Birnirk). It is not known archaeologically south of Cape Prince of Wales, and in the historic period it is well represented at Point Barrow. Collins also directs attention to the finding of Jenness that the greatest break within Eskimo speech comes between Norton Sound and the mouth of the Yukon. All this suggests that my primary classification above may have to be revised, the "Central-Eastern" Eskimo division extending westward beyond the Mackenzie to Bering Strait, the "Western" lying south thereof. The two grand divisions would then be Eskimo on the Aretic Ocean and Eskimo on the Pacific. 27 8University of California Publications in An. Arch. and Ethn. VI. CULTURE AREAS: NORTHWEST COAST TRE CULTURE of the Northwest or North Pacific Coast is that one of the more highly developed and differentiated cultures in America which has been least affected by influences from Middle (Nuclear) America. It has been reached to an unusual degree by influences from Asia. Some of these, slat or rod armor and hats, for instance, show distributions as far southwest as the higher civili- zational centers of eastern Asia. Many other resemblances are vaguer, or show interrupted distributions, but carry even farther, to Indonesia and Oceania: carving, masks, wealth emphasis. Similarities to the eastern Palaeo-Asiatics, however, may be due to cultural currents from America as much as into it. A third trend of the culture is the unusual degree to which its material, native and imported, has been worked over into its own patterns. The area is evidently one of unusual intensity of cultural activity. This intensity seems to have been still heightening at the time of discovery, and to have received a further temporary impetus from the first European contacts. This powerful repatterning has probably disguise?d the foreign origin of much Northwest Coast culture material. The historic source of material of this kind should prove discernible when intensive knowledge of the area is combined with a willingness to consider the probability of remote origins. The present indica- tions are that perhaps as much of the reworked material derives from Asiatic as from distant American centers. Recent conditions at the southern end, as weLl as the slender archaeological evidence available, suggest that the Northwest Coast culture was originally a river or river-mouth culture, later a beach culture, and only finally and in part a seagoing one. This means that the recent hinterland cultures of the Columbia-Fraser drainage (Plateau) and of the Intermountain Athabascans evidently provide approximate illustrations of an early stage of Northwest Coast culture. This situation is implicit in Wissler's basing of both the North- west Coast and the Plateau culture on a Salmon Area. Of course no mechanical subtraction of hinterland from coast culture suffices for a true estimate of the kind or amount of culture specialized on the coast, even apart from the variant conditioning of subsistence, because the hinterlands have secondarily absorbed culture material and forms from the coast as well as from the east. The ecological correspondence is remarkably close for the Northwest Coast. The vegetational-climatic area of the Northwestern Hygrophytic Coniferous Forest (maps 2-5) tallies almost absolutely with the cultural one. This forest is generally considered as extending into northern California. The culture ex- tends to Cape Mendocino and the lower Eel River, which lie about at the mid- dle of the Redwood belt (map 4). This Redwood strip may be viewed as a specialized southern extension of the northwestern forest; its denser and more characteristic part is its northern half, which belongs clearly to the Northwest culture. The areal types of the Northwest culture can be formulated only tentatively. While this is one of the more intensively studied regions of the continent, in- terest has been away from classificatory and developmental problems. 28 Eroeber: Cultural and Natural Arqa8 of Native North Ameriwa 1. Northern Maritime. Tlingit, Haida, Tsimshian; probably also the Haisla.1 Three sub- types can be distinguished. la. Northern Maritime Mainland. The Tlingit northwest of the Alexander Archipelago, on the coast backed by glaciated mountains. Resemblances to Athabascan inlanders seem fairly strong. This is also a separate ecological region. Osgood2 distinguishes a Southeastern and a Glacial Coast region in Alaska, separated approximately by the Lynn Canal. The present or Northern Tlingit subarea corresponds with the Glacial Coast region; but Osgood carries this farther west, to include the Kenai Peninsula. From the Copper River to Kenai the coast was Eskimo; and, as already stated, these Eskimo, the Ugalakmiut and Chuga- chigmiut, seem to deserve setting apart as a subtype. In any event, if the Glacial Coast region of Alaska is a valid natural area, it marks the meeting place of two deeply different cultures, Eskimo and Northwest Coast. Whether Eskimo or Tlingit are the later intruders is not known; but the Eskimo in this tract have taken over more obviously Tlingit traits than have the Tlingit adopted the Eskimo ones. It is of course possible that at an earlier period, when the Northwest culture was as yet less developed, the Eskimo influence was the more potent, but that the elements derived from it" have long since been worked over so as to seem native Northwestern. Very little is known about the phenomena of border contact between Tlingit and Eskimo; and an important study is indicated here if the two cultures have not yet disintegrated too completely.4 lb. Northern Maritime Archipelago. Southern Tlingit, Haida, Tsimshian proper. By gen- eral agreement these tribes represent the culmination of Northwest Coast culture during the nineteenth century. l. Northern Maritime Biver. Niska, Gitskyan, Haisla. A less intensive variant of the Northern subeulture, localized on rivers or inlets rather than on the sea. 2. Central Maritime. Central British Columbia coast, northern and western Vancouver Island, Cape Flattery. The peoples are Bella Coola, Heiltsuk, Kwakiutl proper, Nutka, Makah, Quileute, Quinault, perhaps Chehalis. According to Dr. Olson, whale hunting and secret societies extended to the Quinault. It is on the basis of these traits that the limit of the area has been drawn just north of Shoalwater Bay. This area is predominantly Waka- shan. The interior water boundary comes about Cape Mudge in latitude 500, which seems to mark also a climatic and minor vegetational change: to the south, the east side of Van- couver Island is relatively dry. Two subdivisions are recognizable in the Central Mari- time area: 2a. Northern Central Maritime. Kwakiutl, Heiltsuk, Bella Coola, with more developed art, ritual, and social organization, but mainly facing protected water. 2b. Southern Central Maritime. Nutka and seaward tribes of Washington, with whale hunting. 3. Gulf of Georgia. Southeastern Vancouver Island, mainland coast of southern British Columbia, north side of Olympic Peninsula. Wholly Salish and facing protected salt water; climate somewhat less humid than in the preceding. In terms of water, the specifying elements are the mouth of the Fraser, the Gulf of Georgia, and the straits of Georgia and Juan de Fuca. 'The northern mainland Kwakiutl have not been studied systematically and are difficult to place. The Haisla are tentatively assigned to area 1, and the Heiltsuk (Bellabella, Rivers Inlet) to area 2, on the basis of Boas's statement (AA 26:323-332, 1924) that the former have, and the latter have not, matrilinear exogamic clans. 2 "Alaska, in Shelford work cited, 141-146 1926. Harshberger has the Sitkan region (map 2, no. 21a) extend from northern Vancouver Island to beyond the Copper River, excluding the Kenai Peninsula. 8 Whale hunting, for instance, which in the historic period was practiced in the Northwest area only on Vancouver Island and about Cape Flattery. 'K . Birket-Smith and F. de Laguna, The Eyak Indians of the Copper River Delta, Alaska (Copenhagen, 1938), have described the remnant of a newly determined tribe which is non- Eskimo, non-Tlingit, and wholly distinct from the previously recognized Athabascans of the Copper River above the delta. The speech carries Athabascan suggestions, but if Atha- bascan it is greatly deviant; it may prove to be a fourth member of Na-Dene (Athabascan, Haida, Tlingit). 29 0University of California Publications in An. Arch. and Ethn. 4. Puget Sound. Salt but still water. Salish, plus probably the Chimakum. Groups with- out true secret societies. The Skagit probably belong to this group; the Lummi and Nutsak, also the Elallam, to the last." 5. Lower Columbia, with coast from Shoalwater Bay to Umpqua Mountains. Chinook, Chehalis, Tillamook, and Yaquina-Alsea-Siuslaw. 6. Willamette Valley. Interior. Kalapuya. 7. Lower Klamath. Northwestern California with Rogue and upper and middle Umpqua drainage in Oregon. Mainly Athabascan, but also including Kus, Yurok, Wiyot, Karok. Culmination on lower Klamath, among Yurok, Karok, Hupa.6 A subperipheral transition region is recognizable, extending in an are from the Shasta on the middle Klamath to the Wailaki and Sinkyone on the middle Eel, but is here reckoned as part of the California culture area. These areas are far from equivalent in cultural intensity and depth. The climax of the region seems long to have lain in its northern half. The four southern areas are distinctly subelimatic and culturally peripheral. During the last half of the nineteenth century, the climax must be credited to the Northern Maritime tribes, on account of their aggressiveness and the vigor of their art. Their culture was then in an expansive, acquisitive phase. Pre- viously, the climax was probably situated in the second or Wakashan group,' who worked out the Hamatsa cannibal ceremonies which the northerners later borrowed. Still earlier, the climax may have lain in the third area, about the mouth of the Fraser and the opposite shore of Vancouver Island. If the theory is correct that the Northwest culture as a whole originated on rivers and only slowly ventured on the open sea, this area would be the logical one for the first stages of its characterization. The Lower Columbia area may have experienced similar impulses, but these would have been checked by the debouching of its river on a straight, rugged coast, without sheltered salt waters to encourage the apprenticeship of transformation. Puget Sound is a backwash. It may have been an important area in early stages of the culture, but its very shelteredness from the sea destined it to relative lag as the oceanward development pro- ceeded. The Willamette Valley formed even more of a. pocket. It is the only interior culture in the Northwest region, and is probably best construed as an inland modification of a form of the primitive river phase. The fact that the valley contains enough prairie to cause it to be classified by some authorities as grassland (map 5) would have contributed to its cultural differentiation. 5H. Haeberlin and E. Gunther, The Indians of Puget Sound, UW-PA 4:1-84, 1930, print a map of Puget Sound tribes (p. 8) which shows a distribution somewhat different from that given in map 1 accompanying the present work. It is significant that several tribal territories (Skykomish, Snuqualmie, Muckleshoot) are shown entirely away from salt water, and others (Skagit, Nisqually) barely touching it. Another map has recently been issued by Spier in Tribal Distribution in Washington, Gen. Ser. in Anthr., no. 3, 1936. 6 The Tolowa are clearly subelimax as against the Yurok, and the Tututni apparently more so. With the Kus and Siuslaw, Lower Columbia elements begin to appear and are stronger among the Alsea and Tillamook. The Kus and Siuslaw thus cannot be split, as the text has it, but whether they both go rather with Lower Klamath or with Lower Columbia is less clear. These findings rest on field studies in 1934 and 1935 by Philip Drucker, The Tolowa and their Southwest Oregon Kin, UC-PAAE 36:221-300, 1937, and H. G. Barnett, CED VII-Oregon Coast, UC-AR 1, no. 3, 1937. 7Among the Heiltsuk Kwakiutl, to be exact, according to Boas, USNM-R 1895:661, 664, 1897. The evidence is native tradition, but confirmed by ceremonial names which are Kwakiutl. 30 Kroeber: Cultural and Natural Areas of Native North America It is the only tract in the Northwest area which is not continuously forested. The Northwest California subelimax has clearly been built up on a basis of river habitat. Its center lies on the only stream south of the Columbia to drain from the interior of the Sierra-Cascades mountain wall, and nearly at the meet- ing point of three forests, namely, the Northwest Coast Douglas Fir, the North- west Extension Redwood, and the California Pine (map 4). It is evident that the descriptive subdivision of the long north-south North- west area into seven to ten approximately transverse segments resolves itself, as soon as the relations of the segments are viewed with interest in environ- mental adaptation and historic development, into a classification into longi- tudinal belts, nearly but not quite parallel to the coast and expressive of degrees of utilization of water, from river to mouth to still salt water to ocean, with a subsidiary use of ocean replacing primary adaptation to inland salt water where this is not available. According to these degrees of water adap- tation, the areas group thus: 1, Willamette; 2, Klamath, Columbia, Puget Sound; 3, Gulf of Georgia; 4, Central Maritime, Northern River, Northern Mainland; 5, Northern Archipelago. Within each belt the more northerly sub- areas usually have the more intensive culture. Also, except in the most south- erly area, the center of intensity within each area seems to lie in its northern portion. The degree of development of such luxury aspects as art and society rituals is in agreement with this environmental-historical view. From both the northward centering and recent northward trend of the climax of the whole Northwest Coast, it is expectable that more refined analysis will confirm the conjecture that Asiatic influences perhaps were more potent than Nuclear (Middle) American ones in the specific shaping of Northwest Coast culture. If direct Oceanic influences have ever to be reckoned with, they may complicate the picture. 31 University of California Publications in Am. Arch. and Ethn. VII. CULTURE AREAS: SOUTHWEST THE DISTINCTNESS of the Southwest was recognized long before there was any thought of general areal classification. The name refers of course to posi- tion within the United States. Wissler, however, included northern Mexico, nearly to the Tropic of Cancer, in the area; and in this he was followed by me, in my modification of his hemispheric classification.1 As this inclusion has provoked no criticism, it may be assumed that dissent has not been lively. Ac- cording to this view, about half of the native Southwest lay in what is now Mexico. But this half is little known. Both axchaeological and ethnological studies have been extremely meager, and until recently the Spanish ethno- graphic documentation from the period of exploration and settlement had not been gone over systematically. Now at last there is available a digest and interpretation of the documentary data by Beals.2 This has been specially drawn upon for the consideration of Mexican areas, farther on in the present monograph. The Beals data were nec- essarily brought together primarily with reference to the situation in Mexico; just as the current data on the American part of the Southwest have been gathered as relating to the situation in the United States, especially to the Pueblos and their relations to the east, north, and west. The two sets of data thus by no means integrate fully; and it will require much fuller information, and its gradual digestion, before anything more than tentative classifications and attributions of the cultures south of the international boundary can be made. Along the Pacific coast, to be sure, a line of demarcation between the Southwestern and Central Mexican spheres of culture influence can be drawn with a certain degree of confidence, so as to include the Cahita in the South- west, the central Sinaloa peoples in Mexico.8 In the interior, however, it is much more dubious how groups like the Tarahumar and Concho should be construed as affiliating. The Tarahumar are here provisionally classified as in the South- west, the Concho in the Mexican sphere. The situation is considered further in the Mexican section, especially with reference to the linguistic relations that might be pertinent.' All in all, however, the question of the Mexican-South- western froutier must be left an essentially open one for the present. I have recently pointed out' that the known Southwest appears to comprise two related but consistently distinctive culture types: one characterized by the Pueblo culmination, and one which might be named the Sonora-Gila-Yuma. The common elements such as agriculture, cotton, pottery are obvious. The Pueblo culture shows masonry, clustered houses, stories; the kiva ceremonial chamber, altars and sand or meal paintings, masks and ancestor impersona- tion, priestly offices, elaborate ritual, much visual and verbal symbolism with 1Anthropology, fig. 34, 1923. 2 The Comparative Ethnology of Northern Mexico before 1750, UC-IA no. 2, 1932. ' C. 0. Sauer, Aztatlin, UC-IA no. 1, 1932. 'For instance, the Tarahumar and Concho affiliate linguistically with the Opata and Cahita, who are here reckoned in the Southwest; the Tepehuin with the Pima, also con- sidered Southwestern. See below, and UC-IA no. 8, 1934. 6UC-PAAE 23:375-398, 1928. 32 Kroeber: Cultural and Natural Areas of Native North Anerica special reference to colors, directions, fertility, and emergence; matrilinear descent; pacific inclinations; pottery with a whitish ground, polychrome or glazed painting, and texture decoration by corrugating. The Sonora-Gila- Yuma culture possesses adobe, wattled, or brush houses, village instead of town type of settlement; no kivas and few altars, little visibly expressed sym- bolism; simple rituals with few masks; shamans rather than priests; patri- linear institutions; warlikeness; a pottery reddish, monochrome or with one design color, uncorrugated; canal or river overflow irrigation.' As the vegetation maps show, the Pueblo area lies fundamentally within the sagebrush-juniper-pinfon association, with good-sized areas of short grass and desert grass, and pines in the mountains (maps 2-5, 8). The Sonora-Gila- Yuma area is prevailingly one of true desert, with the creosote bush selected by some authors as the characterizing plant (maps 4, 8), the succulence of the aridity-resisting vegetation of certain districts emphasized by others (maps 3, 5); and, except in the Sierra Madre, with an almost complete absence of forest growth. These two distinct plant covers are of course a function of alti- tude and climate. The Sonora-Yuma subarea averages much lower than the Pueblo; the heat equator passes through it; evaporation is as high as precipi- tation is low; and a number of included tracts are reckoned as extreme desert (map 3). The Pueblo region is high, cold in winter, and subdesert (map 3) a borderland between technical desert and steppe,-in fact, more largely the latter (map 24). The correspondence of environment and culture is close for these two subareas. Geographically, they lie roughly northeast and southwest toward each other. In Arizona, the Mogollon rim forms the boundary between the Colorado Plateaus and Basin-and-Range physiographic provinces, as well as between the two cultural subareas.7 Desert conditions extend southward through much of Sonora and Mexican California. Whether Chihuahua forms part of the same desert or a somewhat differentiated one, is not clear. A dif- ferentiation seems more likely on account of the greater altitude. It would apparently also fit the cultural situation better. On its other side, the Pueblo environment extends northwestward beyond the limits of Pueblo or Southwestern culture. The sagebrush-juniper associa- tion prevails over the Great Basin and beyond into the Snake portion of the Columbia drainage. Here, then, the correspondence of ecology and culture, at least in the recent distribution of the latter, breaks down. It holds sharply within the Southwest-at least its known portion; it does not hold beyond. The fact that the environment of one of the two Southwestern subareas runs far outside the cultural Southwest strengthens the probability that the two " Among archaeologists Hohokam has now come into general usage for the prehistoric phases of what is here called Sonora-Gila-Yuma culture. Sequences within Hohokam are now almost as well known as within Basket Maker-Pueblo, thanks especially to the work of Gila Pueblo as directed by H. S. Gladwin and published in the Medallion Papers since 1928. Kidder has recently proposed Anasazi as a counterpart term to replace Basket Maker- Pueblo (The Pottery of Pecos, 2:590; with Anna 0. Shepard). 7 The Sonora-Gila-Yuma subarea lies largely in the Basin-and-Range and allied Sonoran Desert and Sierra Madre provinees. The Pueblo subarea is physiographically more varied, extending over portions of the Colorado Plateaus, Rocky Mountains, Basin-and-Range, and Great Plains provinces. See map 7. 33 4University of California Pusblications in An. Arch. and Ethn. subcultures are fairly distinct, because it suggests that the history of one of them contains influences lacking in the other. It seems best first to delimit and subdivide the two areas in their recent manifestation, and then to consider their inferable history. Map. 8. Creosote Bush and Sagebrush; from Livingston and Shreve. The Sonora-Gila- Yuma area falls typically within the occurrence of the former; the ancient Pueblo area, in both, plus forest and grassland,-but in its present range is restricted to sagebrush or immediately adjacent vegetation. The sagebrush range, however, is far greater to the north than the widest Pueblo extension at any period. 1-2. PUEBLO SUBCULTURE TYPE 1. Pueblo: Tano, Keres, Zufni, Hopi. If Sapir's conjectures in regard to the ultimate linguistic affiliations of these groups are correct, half or more of them would be of Uto-Aztecan origin in the wider sense-Aztec-Tanoan. The true Pueblo culture is so distinctive, and so well known both ethnologically and archaeologically, that its detailed discussion here is uinnecessary. It forms 34 Kroeber: Cultural and Natural Areas of Native North America a very definite climax of established antiquity and of an intensity possibly equaled only at one or two other points north of Mexico. This climax culture appears to have reached its peak, at least in certain aspects, some centuries before Caucasian discovery, and its greatest areal extension several centuries earlier still. At no period of its history is there indication of its having in- fluenced surrounding or distant cultures at all strongly. It constituted a lo- calized and self-contained culmination. 2a. Inter-Pueblo: Navaho; and 2b. Circumn-Pueblo: Apache. The Navaho have accepted somewhat heavier Pueblo influencing than the Apache. Both these Athabascan groups made pottery and farmed only to a subsidiary degree. The cultures of both gradually became, as it were, parasitic on Caucasian cul- ture in their economic aspects, although in different ways: the Apache frankly predatorily, with the taking over chiefly of horses and weapons; the Navaho rather by theft and imitation, with rearing of flocks, weaving of wool, and working of silver. It is not known how much of these practices came to the Navaho through the Pueblos and how much directly from Caucasians. At any rate, their culture had essentially taken on its present-day aspect by the middle of the eighteenth century, possibly considerably earlier. It has also flourished, mainly along the lines then set, since the progressive American- ization of the Southwest, until today the Navaho constitute a definitely per- ceptible factor in the economic life of New Mexico and Arizona. They have multiplied, are still spreading territorially, and have worked out a unique and interesting subsistence system which is different from both the native and the Caucasian economies out of which it has been hybridized. In origin the Navaho and Apache are of course one people, as shown by their close dialectic relationship and by the Spanish habit of classing the Navaho as Apaches. The differentiation between them8 seems the result less of differ- ence in natural environment than of difference in cultural geography. The Navaho habitat lay between the Hopi and Rio Grande Pueblos, with Zufni on a third side. They were also fairly effectually shut off by Apache groups from direct involvement in the unsettled, war-embroiled life of the western edge of the Plains. Distrusted and feared though they might be by the Pueblos, especially after Spanish pacification, they were removed from the atmosphere of war as a prime occupation of life; took up the gainful arts of their Pueblo and Spanish neighbors; and laid the foundation of the special economic system which they still adhere to. Hand in hand with this went two other develop- ments: a greater receptiveness toward the material of Pueblo ritual, and an accelerating increase in numbers. The result of the latter factor was that whereas three or four hundred years ago the Navaho constituted a small and culturally scarcely distinguishable fraction of the Apache, they are now well set apart in customs from this parent body, and perhaps five times as numer- ous as all its other divisions combined. In terms of precise ethnological knowledge, the Apache are, with the pos- sible exception of the Ojibwa, the least-known surviving North American 8 This differentiation is similar in some ways to that of the Yaqui and Mayo, as discussed below. 35 6University of California Publications in Am. Arch. and Ethn. group among any of like areal extent and historic importance.9 Their numer- ous tribes or bands may be grouped according as they lived west or east of the Rio Grande. Roughly, the two divisions correspond to the modern official and reservation classification into White Mountain and San Carlos Apache and Mescalero and Jicarilla Apache. The westerners comprise the Tonto, Coyotero, Pinal, Arivaipa, Pinalefno, Chiricahua, Mogollon, Gilefno, and Mimbrefio.1' Some of these are probably subdivisions of others. Their total range was from the Tonto Basin in central Arizona to the Mimbres-Guzman Basin southwest of El Paso in Chihuahua. The beginning of their habitat formed the effective Spanish northern frontier in the eighteenth century, and thus largely determined the modern interna- tional boundary along western New Mexico and eastern Arizona. These West- ern Apache groups lived away from the plains and the dependable range of the bison, and were indubitable southwesterners.l1 The Eastern Apache,12 on the contrary, seem all to have depended consid- 9 This was true when written in 1931, but fortunately will not hold much longer, because of the intensive studies by Opler, especially on the Eastern Apache, by Grenville Goodwin on the Western, and by Gifford through an element survey of both divisions in 1935. The results should be available in print soon. Goodwin has published a valuable preliminary paper (AA 37:55-64, 1935). It appears that the Apache are excellent and willing inform- ants: the neglect has been by ethnologists. 10 This classification of Apache tribes follows primarily the 1796 account of Cordero cited in Orozeo y Berra, 368. Goodwin, in the paper cited in the preceding footnote, classifies the Western Apache into five tribal groups: White Mountain, Cibecue, San Carlos, Southern Tonto, Northern Tonto. These subdivide into bands-White Mountain: Eastern (much the largest territory of any) and Western; Cibecue: Carrizo, Cibecue, Canyon Creek; San Carlos: Arivaipa, San Carlos, Apache Peaks, Pinal; Southern Tonto: Mazatzal band and semibands 1 to 6; Northern Tonto: Fossil Creek, Bald Mountain, Oak Creek, Mormon Lake. The twenty-one territories are shown on a map. Their total range is small: about 110 miles by 65. Goodwin's and my Western Apache are, however, not the same. In default of knowledge, I have carried their eastern boundary to the Rio Grande. He defines them as Apaches within present Arizona during historic times except the Chiricahua, Warm Springs, and allied divisions, and the Mansos of Tucson. Only my first five divisions are therefore comprised in Goodwin's Western Apache: the Chiricahua, Mogollon, Gilefno, and Mimbrefno he excludes. He does not say whether the setting apart of his Western Apache rests on dialect, native sentiment, com- mon relations with the whites, or some extrinsic consideration. I hold no brief for the Rio Grande as a line of division: rivers rarely are frontiers in native America. But it would be surprising if the Apache of the upper Gila drainage really belonged ethnically with those beyond the Rio Grande; and Goodwin does not say that they did. Quite likely his Western Apache are simply those now on reservations in Arizona. This would account for his omitting from them the Chiricahua, who were placed with the Mescalero on a New Mexico reservation. 1 Goodwin, 61, 62, estimates farmed food at 20-25 per cent of the total Western Apache consumption, with the proportion of families farming varying from a majority of those in a band to none, the ratio in general diminishing from southeast to northwest. Of nine wild staples, he singles out mescal (agave) and acorns as most important; the others are sahuaro, mesquite, yucca, sunflower, tuna, pifion, juniper. 12 Gifford, as a result of his 1935 field survey, classifies the Eastern Apache (that is, those not called Western by Goodwin) into four larger divisions and a total of fourteen sub- divisions, as follows. (1) Chiricahua-Warm Springs: Chokalene and Chihene of the San Francisco and Alamosa rivers, upper Gila drainage in New Mexico (Mogollones ?); Shaia- hene or "westerners" of the Huachuca Mountains (Nogales-Bisbee area, Arizona); another division to the west of the last-named (these must be the Mansos of Tucson); Indedai of Sonora-Chihuahua. There is no mention of a Chiricahua division proper between the first two and last three and adjoining the "Western Apache" Pinalefno and Arivaipa on the southeast. All this division is well west of the Rio Grande. (2) Mescalero division: Kahoane, the most westerly group, apparently east of the Rio Grande; Ni'ahane, central, presumably about the Capitan Mountains and the Sierra Blanca; Huska'ane, or "plains people," to the 36 Kroeber: Cultural and Natural Areas of Native North America erably, and some of them perhaps primarily, on the bison hunt. They included the Jicarilla of the headwaters of the Rio Grande-sometimes considered a branch of the next; the Faraones between the Rio Grande and Pecos; the Mescalero along the Pecos; the Llaneros or "plainsmen" between that stream and the Colorado; and the Lipan southeast as far as to the Karankawa of the marismas or swamps of the Texas coast. Of these the Lipan, although true Apache in origin, formed an outpost, and are included below, on geographical grounds, though perhaps wrongly, in the South Texas culture area. The others all appear to have fronted the plains or to have lived on them until partly crowded back by the Comanche after 1700. They were thus part of the tribes within the old, prehorse, Plains culture; perhaps the principal southern plains tribes. The Kiowa-Apache apparently are a fragment that remained actually in the plains. The Jicarilla, somewhat isolated from all the others in their northerly habitat, became less predatory and effected a quasi relation with the Spaniards and northern Pueblos. The other tribes, or their remnants, have lately come to be known as the "Mescalero." How far the southwestern elements in recent Mescalero and Jicarilla culture predate or postdate the horse and the rolling back of the Eastern Apache by the Comanche, remains to be ascertained. It would seem that their nineteenth-century culture contains absorptions from the Plains culture of that period, probably in the main by way of the Comanche and Kiowa. But if the views set forth below on the de- velopment of historic Plains culture are true, these absorptions would prove little concerning relations before the horse. The Eastern Apache lived in territory which in the main seems to have been unoccupied by peoples of Pueblo culture, or only peripherally or sporadically utilized by them. The Western Apache habitat, to the contrary, contains pre- historic Pueblo ruin almost throughout. Several recognized ancient Pueblo areas, Upper Gila, Mimbres, Casas Grandes, lie wholly in historic Western Apache territory; and the westernmost extension of both groups was about the same: nearly to the Verde. It may therefore be assumed that when the east, in the Pecos Valley; Tuetenene, south of the Rio Grande below the mouth of the Pecos namely, in Coahuila, and said to be "half Lipan"; Zitachisene, of Azfil, toward Chihuahua City, perhaps belonging rather with Chiricahua than with Mescalero. (3) Jicarilla: Setide, "sand people," or Ollero, to the west; Gusgayi, "plains people," or Llanero, on the east (Opler, AA 38:202, 1936, calls them Saidinde and GuLgahen, and defines their range as on the upper Rio Grande, claiming north to the Arkansas and east to the Canadian). (4) Lipan: Tuensane, "big-water people," westerly; Chishene, "woodland people," easterly; perhaps also Tuetenene, mentioned above. Gifford's "Eastern" Apache totality, like Goodwin's "West- ern," apparently reflects modern reservation habitat. This in turn may rest on ethnic affilia- tions; but geographical probability is to the contrary. Until there is specifie evidence linking the Chiricahua with the Mescalero rather than with the White Mountain-Cibecue-Tonto in prereservation days, it seems most reasonable to consider all the Apache west of the Rio Grande, or at least in the Gila drainage, as an ethnic unit. Opler, Chiricahua Apache Social Organization (in F. Eggan, ed., Social Anthropology of North American Tribes, Univ. Chicago, 1937), p. 176, makes the Chiricahua closer, culturally and linguistically, to the Mescalero than to any other Apache group. H. Hoijer, The Southern Athapaskan Languages, AA 40:75-87, 1938, classifies as fol- lows. The Athabascan languages of the Southwest have a single origin within Athabascan, and have diverged: I, Western group, consisting of IA, Navaho, IB1, San Carlos (Goodwin's "Western" Apache), IB2, Chiricahua and Mescalero; II, Eastern group, consisting of IIA, Jicarilla, IIA2, Lipan, IIB, Kiowa Apache. Group II thus consists of Apaches on or front- ing the plains; I, of Apaches west of the Rio Grande, except that the Mescalero have rela- tively recently detached themselves from the Chiricahua to live east of the Rio Grande. 37 University of Calif ornia Pusblications in Am. Arch. and Ethn. Pueblos abandoned their southern territory after having held it through periods 2, 3, and in parts through early 4, the Western Apache were their principal if not sole heirs or dispossessors. Thus the Mimbren-o Apache seem to have ranged in Spanish times over most of the area of the extinct Mimbres and Casas Grandes forms of Pueblo culture. The farthest south of the Pueblos at the time of the discovery was in the valley of the Rio Grande about Socorro; and valleys, although important to the farming Pueblos, were not typical Apache habitats, which, apart from the open plains, are often definable in terms of mountain masses. The Eastern Apache habitat varied a great deal vegetationally. In terms of the Shantz-Zon classification (map 4), it included short grass, tall grass, desert grass, desert savanna, creosote desert shrub, with juniper-pinton and yellow pine along and in the mountains. Wherever agave was available, it is likely to have furnished a staple food much as among the Western Apache; or sotol in its place. 3-10. SONORA-GILA-YUMA SUBCULTURE TYPE I retain provisionally the term Sonora-Gila-Yuma for this moiety of South- western culture, although its extent from the Santa Barbara Archipelago to the Sierra Madre makes a broader as well as less cumbersome designation desirable. The area occupies the southwestern half of the Southwest, with prevailing Sonoran (Uto-Aztecan) and Yuman speech, as against the Pueblo languages and Athabascan in the northeastern half. 3. Fuerte-Yaqui Lowland. The Cahita- (Ka'ita-) speaking tribes: Yaqui, Mayo, Tehueco. The area is that of the deltas and lower valleys of the Yaqui, Mayo, Fuerte, and Sinaloa rivers. The early Spanish accounts make both language and customs change definitely, in a northward progress, at the Si- naloa (Petatlin) River. The archaeological remains indicate a marginal or sub-Mexican culture along the Sinaloa coast about as far north as the Moco- rito.' The archaeology of the northern rivers, probably including the Sinaloa, is much sparser and its types simpler. Cahita, like Pima, means "no" or "noth- ing" in the speech in which it occurs, and seems a desirable term to reestablish, because the ethnic group which it denotes appears to have formed also a dis- tinct cultural unit. The Cihita, though farmers in rich bottom lands,' were politically broken up into independent tribes. The open nature of their low- lands presumably contributed to this condition, as it did among the Yumans of the lower Colorado, in contrast with the isolating, canyonlike character of the Pueblo habitat in which permanent towns grew up. The modern Mayo and Yaqui appear to be two of an unknown number of Ahita tribes which pros- pered, grew, and absorbed remnants of less prosperous ones, until they alone retained their identity. They do not adequately represent the former native ethnic situation any more than the modern Navaho and Mescalero-San Carlos- 18Sauer, Aztatlun. 14 It is not clear how preponderantly they lived along the actual bottom lands. There is much unflooded, dry plain in their territory covered with monte or thorn-scrub forest; as well as isolated hills and small ranges. But the CAhita are obvious lowlanders as compared with Pima and Opata. 38 Kroeber: Ctultral and Natusral Areas of Native North America White Mountain Apache give a picture of the ethnic line-up of the Apache four centuries ago. Orozco y Berra's map shows a "shatter belt" of small tribes along the lower Fuerte and Sinaloa. The languages of these tribes, except for that of a body of introduced Pimas, are unknown, other than for statements that this one is similar to that, or distinct-which may mean dialectically-from another. Thomas and Swanton have reviewed the conflicting and inadequate evidence,u and have been followed in map 1 in the union of Tehueco, Zuaque, Cinaloa, Ahome, Guasave, etc., into the Tehueco group; as one of three main Cahita units. The several "tribes" may have been political entities, but all spoke Cfhita, and may not have been more distinct than the modern seven "na- ciones" or towns of the Yaqui, except in the accident of Spanish terminology. The early visitors speak of a single people from the Petatlin (modern Sina- loa) River to the Yaqui. The Nio and Zoe, who are on the southern margin, lowland and interior, of the C'ahita area, I have, also following Thomas and Swanton, left as separate groups. Here again we have only statements, not vocabularies, and it seems quite possible that they also represented only dialectic variants. The ultimate disposition of their relationships will probably depend on the decision yet to be made concerning the speech of Sinaloa south of the Cahita, where "Mexi- can" (Nahua) has usually been shown by the maps, but with reasons for dis- belief which are reviewed below in the discussion of the Sinaloa area. The Orozeo and Thomas-Swanton Tepahue area on the lower middle Mayo I have left so designated. The stretch immediately above, from San Bernardo on, is held today by the Huarejia, who speak a dialect about equally distinct from Tarahumar and Cihita,1 and who evidently correspond in name, though not so exactly in situation, to Orozco's Varohio or Varogio, who are also men- tioned as related to the Tarahumar." 4. Sonora. This term is used in the sense of the old province of Sonora, that is, the territory drained by the middle and upper courses of the Mayo, Yaqui, Sonora, Altar, and Gila rivers, and containing two ethnic groups, the Pima and the Opata. The Pima lived in the foothills, the Opata (O'pata) in moun- tain valleys to or nearly to the crest of the Sierra Madre. While both speak languages of the Sonoran division of Uto-Aztecan, these languages belong to quite different branches of Sonoran. Opata affiliates with Cfhita and Tara- humar, Pima with Tepehufn to the south.' The geographic dispersal of these 1BAE-B 44, esp. pp. 11-17. 16 Field record by myself at San Bernardo in 1930. See UC-IA no. 8:13, 19, 1934. 1TOrozco y Berra, 326. His map shows them in the Sierra Madre of Chihuahua in upper Fuerte drainage, which is an error. "I This whole area is given rather differently by Sauer in The Distribution of Aboriginal Tribes and Languages in Northwest Mexico, UC-IA no. 5, 1934, with map. As Cahita proper he recognizes Yaqui, Mayo, and Tehueco, Cinaloa, Zuaque on the Fuerte. On the Sinaloa the Ocoroni and Nio constituted small foreign enclaves. The Mocorito on the river of that name probably belonged with the Tahue of central Sinaloa. The coastal fishing tribes from just south of the mouth of the Mayo to include the mouth of the Culiaean he calls collectively Guasave: they included the Ahome of the Fuerte. These people could not farm their alkaline flats and sand dunes. The Spaniards distinguished them in speech from the C&hita; Sauer tends to throw Cahita, Guasave, and Tahue into one closer linguistie subdivision; for which, certainly so far as the Tahue are concerned, there seems to me no warrant (Uto-Aztecan 39 University of California Publications in Am. Arch. and Ethn. two branches gives rise to an interesting ethnohistorical problem, the fuller setting and import of whiche9 are mentioned below in the section on Mexican areas. 5. Northern Sierra Madre: The Tarahumar. The position of this group is uncertain, but chiefly as between the Sonora-Gila-Yuman and the Mexican group of cultures. The Pueblo form of Southwestern culture seems scarcely to be in question in their relations, since the Tarahumar territory lies mainly south of the known range of the Casas Grandes type of Pueblo remains. Judg- ment on affiliations is rendered difficult by the hybridization of all surviving native Mexican cultures with Spamnish culture, plus secondary local differen- tiations in retention and emphasis of elements. However, there is little reason to believe that the Tarahumar were markedly different from their speech kinsmen the Opata and Cihita. At any rate, what is known of them shows no striking excess of elements of Central Mexican culture. They are therefore provisionally classed as within the Southwest. The habitat in which they re- main is one of deep, hot clefts in a rugged, pine-clad cordilleran mass; but they formerly extended farther east into the lower, open Chihuahua plateau.' 6. Sonora Coast: The Serian tribes. These people are sharply marked off from their neighbors by being nonagricultural. This fact rests on an environ- mental limitation, their territory being almost rainless, and at the same time not reached regularly by flow in the rivers which descend into the coastal plain from the Sonoran highland. The next stream south, the Yaqui, does flow to the sea, and is occupied by the farming CAhita.' The question arises Languages of Mexico, UC-IA no. 8:15, 17, 1934). North of the Tahue he includes with them the Comanito of the upper branches of the Mocorito, the Zoe, and the Tubar of the Urique fork of the Fuerte. The Tepahue, Conicari, Macoyahui, and Baciroa, above the Mayo, he affiliates closely with the CAhita proper. They have at any rate absorbed into the modern Mayo; were probably not very different in speech; but, as inhabitants of streams flowing through hill country, were presumably distinct from the bottom-land Mayo culturally and nationally. Above these, he unites into another group the Varohio, Chinipa, Guasapar, and probably Temori in the canyon country of the Mayo and the Otero branch of the Fuerte; with the Chinipa culturally dominant. They were later displaced or assimilated by the Tarahumar. The Huite on the Fuerte between the Tubar and the Cinaloa are unplaced. In brief, the Cahita, on the regularly flooded bottom lands of the lower Yaqui, Mayo, and Fuerte, were the distinetive people of the area. On the Sinaloa, and above the CAhita on the three larger rivers, but below the high Sierra Madre, were a dozen or more territorially smaller nations on whose speech and affiliations we have various Spanish statements, but no specimens, and who have become extinct or submerged; with the exception of the Huarojio- Varohio, whose surviving language is of the CAhita-Tarahumar-Opata group of Sonoran Uto-Aztecan but neither CAhita nor Tarahumar. Finally, there were the coastal Ahome- Guasave, whose subsistence relation to the Cahita must have been much like that of the Seri to the Pima, though there is nothing to indicate that they were non-Sonoran. Sauer has been repeatedly on the ground, as Orozeo, Thomas, and Swanton have not. He has also adduced new documentary sources. However revolutionary his conclusions at times, they are therefore always entitled to most serious consideration. 19 Discussed in Sauer's and my papers in UC-IA just cited. See also the next footnote for Pima, Opata, and Cahita cultural relations. "O W. C. Bennett and R. M. Zingg have published an excellent modern monograph, The Tarahumara (Univ. Chicago, 1935), based on field residence. Their analysis of the culture makes it non-Pueblo, "Sonoran" or Northwest Mexican, built up on a Basket Maker-like foundation. Ceremonially its relations seem mostly with the South; otherwise, similarities are marked also with CAhita, Opata, and Pima. This is an important study, in detail and conclusions. Bennett and Zingg also (p. 392) modify Beals's and my culture grouping: the Opata are classed with the Cahita, not with the Pima. This accords with speech; but Zingg's manuscript trait lists will have to be published before the evidence can be judged. 21 There is also somewhat more rain in the lower Yaqui area. 40 Kroeber: Cultural and Natural Areas of Native North America whether the Seri group of tribes represent a populational remnant from pre- agricultural times, a former farming people which was pushed into the area and perforce gave up farming, or a nonfarming element that came in to oc- cupy the coastal desert which was worthless to the surrounding agricultural tribes. The last of these possibilities is favored by their situation on the nar- rowest part of the Gulf of California and by the fact that the peninsular tribes across the Gulf were also nonfarming. In terms of mere geography, therefore, a derivation of the Seri from peninsular California would be the simplest explanation of the gross facts. Actually, the evidence is not in hand to settle the question. I have discussed the pertinent available data elsewhere ;. and will only add here that there seems to be little to substantiate McGee's view of extreme uniqueness of the Seri. They certainly resembled the penin- sular Californians greatly in level of culture, and appear to show numerous specific resemblances in culture content to the Sonora and northwestern Ari- zona areas. This area, then, may or may not have to be classed ultimately with the Peninsular Californian one. 7. Northwest Arizona: Yavapai, Walapai, Havasupai. These three tribes are closely similar in speech, forming a distinct subgroup of the Yuman fam- ily, with closest affiliations, apparently, with the Akwa'ala-Paipai of northern peninsular California. The Walapai consist of seven subtribes or bands. The Yavapai, according to Gifford, comprise three divisions of at least near-tribal rank: western, southeastern, northeastern. These are again divided into local- ized bands, of which 2, 2, 6 respectively are enumerated. These Yavapai "bands" evidently correspond to the Walapai "subtribes." The Havasupai look like a Walapai band or subtribe which has acquired somewhat greater ethnic, cultural, and historic independence. All three tribes farmed where they could. This, however, they did sporad- ically and insignificantly, the Havasupai excepted. Even the Havasupai lived half the year out of the canyon in which they farmed, and their life during this winter half was scarcely distinguishable from that of the Walapai and Yavapai. The culture shows many resemblances to that of Peninsular Cali- fornia (including the Diegue'no) as well as to that of the Great Basin Sho- shoneans, especially the Southern Paiute across the great chasm of the Colorado. There are also a good many specific resemblances to the Seri. We have in this group, then, a culture related primarily to the nonfarming desert cultures of the region. Upon this basis there have been built superficial local differentiations: Havasupai semisystematic agriculture and use of a few masks adopted from the Hopi, for instance; matrilinear sibs which the South- eastern Yavapai share with the Apache; Mohave song cycles and mourning rites taken over in the American period by the Walapai. In each of these, the influence of the import remains local, and appears to be rather recent. Spier,' '2 The Seri, Southwest Museum Papers, no. 6, 1931. 2 Walapai Ethnography (Contrib. Lab. Anthr., 1), AAA-M 42, 1935. Gifford, The Southeastern Yavapai, UC-PAAE 29:177-252, 1932; Northeastern and Western Yavapai, UC-PAAE 34:247-354, 1936. 25AA 31:213-222, 1929. 41 2University of California Publications in Amn. Arch. and Ethn. even before the Walapai and Yavapai data became available, neatly analyzed Havasupai culture much along these lines, pointing out the essential smallness and overlay quality of the Pueblo ingredient, and aligning the culture pri- marily with that of the Great Basin. The resemblance of Northwest Arizona to Great Basin culture lies not only in considerable specific content, but especially in similar meagerness of defined patterns. In land form, the Northwest Arizona area is not a unit. The line between the Basin-and-Range and Colorado Plateaus areas strikes diagonally through Walapai and Yavapai territory. Almost coincident is the line that separates the vegetational areas of creosote bush and juniper-pifnon. However, the larger half of the habitat seems to lie in Basin-and-Range and creosote bush, and the smaller remainder lies mostly on the lower levels of the Plateau where the juniper struggles near its lower limits. The environmental fit of the fact that the area belongs in the Sonora-Gila-Yuma half of the cultural Southwest is therefore closer than the sharp lines on the map would indicate. 8. Lower Colorado River: The "river Yuman" tribes; in order upstream, the Cocopa, Halyikwamai, Kohuana, Yuma, Halchidhoma, Mohave; plus the Maricopa on the Gila. The first three belong to one dialect group of Yuman, the last four to another. The Maricopa have been on the Gila since before 1700. The Halyikwamai, Kohuana, and Halchidhoma took refuge with the Mari- copa during the nineteenth century and have lost their tribal identity among them.' The river culture is specialized from that of the Yuman tribes in the desert and mountains on both sides. It is characterized by consequential agriculture depending wholly on river bottom-land flooding, not at all on rains or artificial irrigation; by pottery which is a direct descendant of the prehistoric red-on- buff ware of the Middle Gila; by a lack of interest in many aspects of material culture and resulting degeneration, as in basketry; and by a religion which largely suppressed visible ritual and symbolism and substituted emphasis on song acquired by quasi-shamanistic dreaming, or pseudo dreaming, within a 2f Spier, Yuman Tribes of the Gila River (Univ. Chicago, 1933), has clarified the picture, especially for the river Yumans off the Colorado. His identifieation of a new tribe, the Kavelchadhom, brings the number of Yuman tribes and tribal remnants on the Gila up to five, instead of the Marieopa alone, as long assumed. These are: (1) Maricopa, between the Salt River and Gila Bend in the eighteenth century, and perhaps off the Colorado already in Alarc6n's time, 1540; (2) Kavelchadhom, perhaps a Halchidhoma subtribe and at any rate identical in speech; on the Gila from 30 to 50 miles below Gila Bend in the eighteenth century; joined the Maricopa between 1838 and 1852; (3) Halchidhoma, joined 1833-1838; (4) Kohuana, and (5) Halyikwamai, joined 1838-1839. After about 1800, and therefore when the four other tribes merged among them, the Maricopa were living above instead of below the mouth of the Salt. In short, at the opening of the historic record there were at least six Yuman tribes on the Colorado, two on the Gila (and these evidently recently from the larger stream). In 1840 there were three on the Colorado, five merged remnants on the Gila. Obviously, the Colorado was the breeding ground, from which the losers in war were expelled, following the Maricopa lead up the tributary. Speech classifieation, on the basis of my own vocabularies: Maricopa, Kavelchadhom, Halchidhoma are very close, and similar also to Yuma, somewhat less so to Mohave. Kohuana and Halyikwamai, however, are essen- tially Cocopa dialects, and Cocopa differs thoroughly from Maricopa-Yuma-Mohave, show- ing definite Akwa'ala-Dieguenio resemblances instead. 42 Eroeber: Cultural and Natural Areas of Native North America highly conventionalized mythological pattern. As Herzog has pointed out, the river Yuman music also follows a highly specialized style.' Certain specific traits are shared by the river Yumans and the Gila Pima. I have listed these elsewhere.' Some of the common traits are almost certainly the result of interchange within Arizona, and most may prove to be so; but others may extend through the various Pima groups of Sonora. This prob- lem, and the involved one of the relation of river Yuman to "Sonoran" (Pima- Opata) culture, depend for solution on fuller knowledge of the Pima in Mexico. On the whole, river Yuman culture gives the impression of being more specialized than Piman, though quite likely no fuller in content; and there- fore of being largely due to a development on the spot. The Shoshonean Chemehuevi have been considerably influenced by the Mo- have on the side of religion, but apparently without appreciable effect on their economic life. It is not clear whether or how far this influence antedates the Caucasian period. It may well be that the somewhat hazy distinction between the Chemehuevi and the other Southern Paiute rests essentially on this in- fluence; in other words, that the term Chemehuevi denotes those Southern Paiute who have been affected by the Mohave." 9. Peninsular California. This area comprises all the groups of the peninsula and somewhat beyond northward, namely, Pericui, Waicura and subdivisions, all the Cochimi, Akwa'ala" or Paipai, Kiliwa or Kilyuwa," Diegueiio, and 27 The Yuman Musical Style, JAFL 41:183-231, 1928. 28The Seri, Southwest Museum Papers, no. 6:44-47, 1931. Spier, Cultural Relations of the Gila River and Lower Colorado Tribes, YU-PA no. 3, 1936, gives a much longer list of traits. He affiliates river Yuman with Gila Pima and Arizona Papago culture, as against that of the Yumans and Athabascans of the "Ari- zona Plateau." This position seems sound for the United States; but it is incomplete through ignoring the long range of the Pima in Sonora, and the fact that the river Yumans and Cihita seem to have shared much more than flood bottom-land agriculture: for instance, simple technology, loose organization, meagerness of rituals, warlikeness, and unrest. Gifford, AA 38:679-682, 1936, takes issue with Spier concerning the closeness of river Yuman and Gila Piman culture. The difference of opinion seems to be one of taxonomic preference; they agree that the Maricopa relate culturally to the Colorado Yumans more than to the Gila Pima. Underlying Spier's alignment of the Arizona Plateau Yumans with the Apache and Basin Shoshoneans, as against the river Yumans and Pima, seems to be the consideration that the former do not and the latter do farm regularly; and underlying this, in turn, is of course the ecology of the two regions. The question is, Have we here two "cultures," or two facies extending through a series of cultures? Descriptively, Spier may be right; though then the Seri, Diegueflo, and Cochimi should presumably be included in his first group, the Cahita and others in the second. Historically it may be questioned whether the culture development was so simple that it can be resolved into two streams differ- ing essentially according as habitat forbade or allowed farming, important though this factor was. Isabel Kelly, AA 36:548-560, 1934, distinguishes fifteen Southern Paiute bands. Much the largest of these territorially is no. 14, the Las Vegas band, west of the Colorado from where this turns to flow south. From out of this band the "Chemehuevi" (band no. 15) pushed south to about 331/20 north latitude before 1850. "Chemehuevi" refers to the group called Chemehuevi by the Americans; the Mohave, and following them Spanish authors like Garc6s, call all Southern Paiute known to them Chemehuevi, at least as far northeast as the Moapa (no. 13) and Shivwits (no. 6). " Gifford and Lowie, UC-PAAE 23:339-352, 1928. Drucker has obtained an Akwa'ala and a Mexican Dieguefno element list, which will be published in the Culture Element Dis- tributions series in UC-AR. " Peveril Meigs, 3d, The Kiliwa Indians of Lower California, USC-IA no. 15, 1939. 43 University of California P?blications in Am. Arch. and Ethn. Kamia; possibly also the Seri, as discussed under Sonora Coast, and fragments of river Yuman tribes extruded into the desert. All, except in a measure the Kamia, were almost perforce nonagricultural; but the northern groups made simple buff-red pottery. Pitahaya and other cactus fruit, and locally agave, were the only abundant food supply, and that mainly seasonal. Alongshore, fish and mollusks must have been important. Subsistence through most of the desert peninsula was meager, and the population was compelled to remain scattered, even after mission reduction. In the north, from the San Pedro Mar- tir massif to the Cuyamacas, altitude and fog allowed some amelioration of food conditions; and the same holds in the extreme south, about Cape San Lucas, where the maps show a subhumid vegetation. The scant accounts of the Periciu at the southern tip, however, do not seem to differentiate them cul- turally much from the Waicura and Cochimi of the body of the peninsula. In the north, the level of the culture seems to have been raised more than the type was changed. Certain religious features of the Diegue-no, such as the Chungichnish Datura cult, which they share with the Shoshoneans of southern American California, are at least in part, and probably mainly, post-Cauca- sian imports.' Kamia agriculture and other river Yuman resemblances also look like rather recent additions to an eastern Diegue'no basis of culture.' 10. Southern California: Shoshoneans and Chumash south of the Tehachapi. The Dieguefno probably belong rather to the peninsula. The Southern Cali- fornia area is nonagricultural throughout, and ceramic only at its southeast- ern margin. The subsistence basis is Californian, many of the elements of culture Southwestern. Some of these, like the sand-painting altar, are of Pueblo rather than Sonora-Yuma type, and may be the result of ancient radiations from the former people across the territory of the latter. There is a definite climax in this area among coast and island Gabrielino and Chumash, whose culture was semimaritime, with seagoing plank canoes. Although this climax culture was likely to have been further developed locally once it had taken root on the Santa Barbara Islands, its spontaneous origin on the main- land coast and growth to the point where it could reach the islands are hard to understand on the basis of either a Californian or a Sonora-Yuman culture basis. There is therefore a possibility that its impetus came in part either from the Northwest Coast or from across the Pacific, to both of which regions there are sporadic but fairly specific parallels: harpoon, canoe, round shell fish- hooks, psychological cosmogony. The double-bladed paddle and spear thrower of the area might possibly be construed as taken over from Aleuts imported by Russian sea-otter hunters in the course of the Mission period; but the abun- dant archaeological evidence shows that this puzzling local climax culture as SSWaterman, Religious Practices of the Dieguefno Indians, UC-PAAE 8:271-358, 1910. These features are found chiefly among the Dieguefio of the coast and mountains, not of the desert side of the mountains. Im Gifford, The Kamia of Imperial Valley, BAE-B 97, 1931; esp. pp. 1-3, 83-86. Philip Drucker, who in 1935 visited the southern California tribes for an element survey, looks upon Diegueflo territory as extending east to the Yuma, and the "Kamia" as those families or lineages of the Desert Dieguefno who from time to time went to live among the Yuma. See Drucker, CED:V-Southern California, AR 1, no. 1, 1937. 44 Kroeber: Cultural and Natural Areas of Native North America a whole far antedates any Caucasian contacts. Of late, archaeological data have at last begun to throw a little light on part of its development.' HISTORY The prehistory of the Pueblo culture, attacked for a long time with little con- ception of historic problem and less of method, was finally synthesized by Kidder,"6 and accords well with the close ecological relationship of the Pueblo and Great Basin areas. Pueblo culture grew by continuous transitions out of a Basket Maker culture similar to that of the ancient and modern Basin, but incipiently agricultural though still potteryless. Most of the Basket Maker remains have to date been found in the northwestern part of the main range of historic Pueblo sites, toward its Colorado drainage and Great Basin side. Pueblo culture itself, in an early period, spread temporarily into the Basin cultural areae through parts of Utah and southern Nevada.' These relations are discussed again in connection with the Great Basin. Coincident with the recession from this northwesterly spread of Pueblo culture came a concen- tration into large towns and a flowering of the culture; and, more or less as- sociated with this, the Puebloid development of Casas Grandes in northern Chihuahua. Thereafter, Pueblo culture contracted in range; and it varied or specialized, rather than grew, in its forms and content. This process continued through the historic period, in which there also occurred an assimilation of Caucasian culture, most obvious in economics and technology, but far from negligible on the nonmaterial side. In summary, the sagebrush-juniper area did once harbor a relatively uni- form culture, but after this began to differentiate into cultures of Great Basin " Olson, Chumash Prehistory, UC-PAAE 28:1-21, 1930 (stratigraphic, two periods, plus a transition); David Rogers, Prehistoric Man of the Santa Barbara Coast (Santa Barbara, 1929) (three successive cultures). The documentation for Rogers' distinctive middle period is insufficient. Olson finds rude metates (plus some mortars) and charmstones characteristic of his earlier period; mortars, circular fishhooks, and perhaps perforated stones, of the later. The earliest deposits yet discovered on the Chumash islands are similar in type to the transitional rather than to the characteristic early remains of the mainland. Two recent papers by R. F. Heizer are important: on the spear thrower in American Antiquity, 4:137- 141, 1938, and on the plank canoe in Ethnological Studies (Goteborg), 7:193-227, 1938. 8Southwestern Archaeology, 1924. "Even to the Mohave Sink region of southern California, according to M. J. Rogers, San Diego Museum, Archaeology, 1:1-13, 1929. 87 It is doubtful how far the "Pueblo" culture north and west of the Colorado may not be Puebloid rather than true Pueblo. It contains genuine Pueblo traits, but lacks others, and possesses specific non-Pueblo features. The approach has been from the side of knowl- edge of the classical Pueblos, with a natural tendency to construe as Pueblo any culture which still showed definite Pueblo elements. Had the approach been from another side, it is conceivable that these northwestern cultures would have been described as non-Pueblo with a greater or less degree of Pueblo influencing. Noel Morss, in the Summary of his recent Ancient Culture of the Fremont River in [South Central] Utah, PM-P 12, no. 3, 1931, shows this rather clearly: Pueblo maize and pottery present; Pueblo masonry, kivas, cotton, turkey, plaited (twilled) basketry absent; non-Pueblo cists, moccasins, fur cloth, coiled basketry, snares, figurines, anthropomorphic pictographs well developed or abundant. On a broad view, does such a culture deserve to be called Pueblo I J. H. Steward, Archaeological Problems of the Northern Periphery of the Southwest, Mus. of Northern Ariz. Bull. no. 5, Flagstaff, 1933, makes such Southwest culture as en- tered Utah mainly Basket Maker 3-Pueblo 1. The Northern Periphery mainly or wholly lacked the grooved ax, turkey, cotton, sandals, and a whole series of Pueblo pottery forms and decoration techniques. Steward maps four areas (five with 1A, IB) of this Northern Pueblo Periphery in and about Utah. 45 6University of California Publications in An. Arch. and Ethn. and Pueblo type, the latter shrank back into a limited portion of the area and has been tenaciously on the defensive since. It is notorious that the Pueblos are nonpropagandist and that their exceptionally high culture has left little spe- cific impress upon others. The question arises, What was it that caused the differentiation of incipient Pueblo culture from the Basket Maker-Great Basin basis ? On the one hand, gradual development of pottery, masonry, and commuinity towns on the spot has been followed out in such detail by Morris, Prudden, and others, as to give a strong impression of a spontaneous, purely local growth.' On the other hand, there are a series of facts pointing to irradiations from the south. Maize, of flint variety, and squash, both almost surely of Mexican origin, appear in Basket Maker period 2, pottery in period 3; slab construction with masonry augmentation arises in Pueblo period 1; communal houses or small towns of masonry, in Pueblo period 2.' These successive appearances lend themselves to the interpretation of continuing or repeated influences from the south which gradually became effective in crystallizing what we know as Pueblo culture. A special injection is likely to have occurred at the beginning of Pueblo period 1, when a broad-headed population, which has persisted, began to replace the long-headed Basket Makers, whose head type continues among the recent Basin Shoshoneans. In any event, the explanation of a foreign southern origin of the stimulus or ferment of Pueblo culture also helps to explain the anomaly of two quite different culture types-Pueblo and Basin-within the same nat- ural area. Why, however, these have persisted side by side for at least a thousand and perhaps two thousand years without assimilation or without the replacing of one by the other has not been altogether clear. One is inclined to look for the cause as lying in something in the character of Pueblo culture itself, in those factors which early gave it its exceptionally nonexpansive, self-centered qual- ity. These factors in turn seem to be two: one cultural, the other natural. The cultural element is no doubt the relatively high degree to which Pueblo cul- ture even in early times already had its basis in farming subsistence. On account of the habitus of maize, this necessarily means an ultimate southern origin; though whether the importation was due more to diffusion of the art of agriculture or to populational movements, we cannot at present say. The natural factor is the limitation which climate puts upon maize growing. This is illustrated in maps 25 and 26, and discussed further in Section XIII, under "Climate." In essence, it appears, Pueblo agriculture, and therefore the Pueblo type of culture, were prevented from spreading westward either by downright aridity or, where there was enough rainfall, by the concentration of this into winter; northward, by decreasing temperature expressed specifically in too short a growing season for maize to mature between the last frosts of spring and the first of fall.'0 Where the Pueblos live today, they can depend on corn "Bibliography in Kidder, Southwestern Archaeology. "Kidder, 118-135; also Science, 66:489-491, 1927; Roberts, BAE-B 92:2-7, 1929, 100: 2-5, 1931, 111:2-27, 1932; Kidder, pt. 3, Discussion, pp. 589 ff. of Kidder and Shepard, The Pottery of Pecos, vol. 2, 1936. 40 Toward the east, a limiting climatic factor is not elear. 46 Kroeber: Cultural and Natural Areas of Native North America with reasonable safety though with little margin. Parts of the areas which they once occupied are also farmable for them; but others, like most of their former holdings in Utah and Nevada, must have afforded an extremely precar- ious subsistence at best. In short, sagebrush and juniper thrive about equally well in the Basin and in the Pueblo country; maize does not, even with the most careful nursing. Natural vegetation is not an index of the determining factors of a culture like that of the Pueblo. The Pueblo culture did push its southern-derived subsistence basis, which was integral to its nature, as far north as was possible; at times beyond the limits of success. In its basis, it was and remained definitely a marginal culture. The wonder is that upon this mar- ginal basis it succeeded in erecting so rich a social and religious superstructure of climax growth. Besides the early and rather meager flow into Utah and Nevada, and per- haps some sporadic efforts to penetrate the Plains, only one notable Pueblo expansion is yet authenticated: that which brought polychrome pottery into the Middle Gila region of the Sonora-Yuman area during the Great and Late periods (Pueblo periods 3 and 4). Here the Pueblo invasion found red-on-buff bichrome ceramics established, continued alongside them for a while, but retreated or died out again before the historic period, leaving red-on-buff somewhat altered but in possession of the field." In the local history of the Gila region, this Pueblo or Puebloid invasion was no doubt a momentous event. But its transience evidences the firmness with which the Sonora-Gila-Yuman area held its line against the Pueblo. Reciprocally, the eastern limit of red-on- buff ware about Solomonsville corresponds closely with the boundary of the creosote bush or succulent desert (maps 3, 4, 5). The Verde drainage, again, is mostly juniper, and its pottery, except near the mouth of the Verde, is Pueblo. Kidder suggests the prehistoric Casas Grandes River culture as a Pueblo proliferation in period 3 (or 4); but so little is known of this old north Chihua- huan culture-and nothing beyond it-that it might conceivably prove to be the result of the impingement on Pueblo culture of northward radiations of some Mexican development. Its pottery is well differentiated from all other local Pueblo styles; its architecture, from all but that of the Mimbres and Gila.' In the historic period, Pueblo contacts with the Plains were largely through the uppermost group on the Rio Grande, the Tiwa, where Taos shows much Plains influence. It is significant that the nearer Plains tribes-Comanche, Kiowa, Southern Arapaho, and Cheyeune-show very few Pueblo traits. The "E. F. Schmidt, Proc. Nat. Acad. Sci., 13:291-298, 1927; AMNH-AP 30:247-302, 1928; H. S. Gladwin, Southwest Museum Papers, no. 2, 1928; the same (no author given, privately printed for the Medallion, Pasadena, later, Gila Pueblo, Globe), 1-72, 1929, 135-161, no date; Kroeber, review of first and third, AA 31:513-516, 1929; F. M. Hawley, AA 32:522- 536, 1930; Sauer and Brand, UC-PG 3:415-448, 1930. Excavations at Snaketown, Medallion Papers, nos. 25, 26, 1937, by H. S. Gladwin, E. W. Haury, E. B. Sayles, N. Gladwin (with full bibliog.), summarizes knowledge of Hohokam culture and shows how much has been learned since the foregoing citations were written. " H. A. Carey, An Analysis of Northwestern Chihuahuan Culture, AA 33:325-374, 1931, points out Mexican resemblances, but aligns the culture primarily within the Southwest. See also D. D. Brand, The Distribution of Pottery Types in Northwest Mexico, AA 37:287- 305, 1935. 47 48 University of California Pu4blications in Am. Arch. and Ethn. Jicarilla and Mescalero Apache are Southwestern, by general estimation, but with a non-Pueblo basis of life-open plainsmen and buffalo hunters. This does not mean that they were Plains tribes in the nineteenth-century sense, but more likely that they were dwellers at the foot of the Rockies and southern ranges who roamed into the plains-members of a contingent of which a part later went into the making of the Plains tribes as we know them. The Kiowa- Apache would be a band that finally stayed in the plains. Somewhat similarly, the Kiowa, on the basis of their speech, apparently are a group that anciently broke away from the Tanoans of the Rio Grande-somewhat like the Co- manche from the Shoshone much later on. These movements illustrate the greater vigor of late Plains over that of Southwestern culture. Kroeber: -Cultural and Natural Area. of Native North Anrerica VIII. CULTURE AREAS: INTERMEDIATE AND INTERMOUNTAIN AREAS 1. GREAT BASIN CAirEoRNu has generally been reckoned a distinct area ever since American culture began to be classified geographically; but the Great Basin1 has been bandied about. It has frequently been included with the interior Columbia and Fraser drainages in a "Plateau area," the concept of which before long came to be unduly colored by the culture of the Fraser Salish, the only tribes then intensively monographed. Otis Mason recognized a separate Interior Basin. Wissler united the Basin with California into a Wild Seed area in his food-area classification. This is undoubtedly correct so far as subsistence is concerned, and was followed by myself when I constituted a California-Great Basin area of general culture.' In his culture-area classification, however, Wissler departs from this solid basis and dissolves the Basin away, assigning its territory to the adjacent Southwest, California, Plateau, and Plains, most largely to the last named. His schematic boundaries diminish the arbitrariness of this division, which would appear starkly on a map following physiographic or tribal features. No one seems ever to have doubted the close internal cul- tural unity of the Shoshonean Basin tribes. It is the meagerness of their culture on levels above that of mere subsistence which has made it difficult to specify their affinities. The union of the Basin with the Columbia-Fraser drainage into a Plateau area seems to rest on the recognition of a negative fact: the absence of nearly all the more intensive culture manifestations of the coast on one side and of the plains on the other. This, however, still leaves the Columbia-Fraser a hinterland to the Northwest Coast, the Basin to California. Also, food habits are built respectively about salmon taking and bulb digging and about seed gathering. The positive similarities of the Basin and Columbia-Fraser areas appear to be rather few. Their relationship is one of level or saturation stage rather than of specific content. Their union into a larger Plateau area there- fore leads to little opportunity for historic utilization. Wissler's inclusion of all the easterly Basin tribes in the Plains area has validity for the last century or so, but would misrepresent earlier conditions. It is true that, viewed against the Teton and Blackfoot, the recent Ute and Bannock cultures look like peripherally diminished Plains cultures. However, this interpretation ignores the recency of the Plains culture represented in our museum collections and in many modern monographs; and it also sees the Plains focus in the far western plains, where relations with the eastern Basin would be strongest. The view here developed is that the eastern Basin and I In the Great Basin there is here ineluded the part of the Colorado River drainage which lies outside the Southwest area. The plant eover is the same, though high mountain masses with pine forests are somewhat more extensive in the upper Colorado drainage than in the Basin proper. "Great Basin-Upper Colorado" would therefore be the more exactly descrip- tive term; but it is cumbersome and not wholly accurate, since the Little Colorado and San Juan affluents of the Colorado belong in the Pueblo Southwest area. 2UC-PAAE 17:151-169, 1920. See also Lowie, UC-PAAE 20:145-156, 1923. 49 University of California Publications in An. Arch. and Ethn. Rocky Mountains areas indeed had pre-Caucasian relations with the western Plains, but as influencing perhaps more than influenced. This point will be referred to more fully when the Plains culture is discussed. As for the relation to California, it is clear that the basic subsistence type of the Basin is similar, and that there are also close relationships in basketry and dwellings. It is to be noted, however, that climate and, in the main, vege- tation change sharply as soon as the Sierra Nevada is crossed; and in both these matters the Basin and most of the Southwest belong together, as all the maps show. Some presumption is therefore at once raised that the Basin be- longs with the Southwest in culture also. This connection has been disguised by the hitherto prevalent habit of thinking of the Southwest in terms of its specialized Pueblo phase. As a matter of fact there is a large amount of evi- dence pointing to close relations of Southwest and Basin. The first Basket Maker discoveries were recognized as showing Californian similarities. In the standardized Southwestern scheme of horizons, the hypothetical, pre- agricultural stage, Basket Maker 1, is formulated$ as a seed-gathering, basket- using culture of general Basin-like type. The Lovelock Cave of central Nevada, in the heart of the Great Basin, yields in its lowest stratum an atlatl culture which M. R. Harrington reckons as akin to Basket Maker.! The upper strata are on the whole more similar to recent California. Early Pueblo culture has been traced by Judd northward in western Utah to the Idaho line, and by Harrington westward across southern Nevada to the California boundary.' This means that before Pueblo culture attained its full specialization it ac- tually held a large part of the Basin. As specialization increased, territorial contraction took place, and tribes of Basin type of culture flowed back into the vacated area. Reciprocal relations must, however, have been fairly active. Spier's study of the Havasupai,' the first monograph on a non-Pueblo tribe in the general Pueblo range (except for the Navaho, who are Puebloized super- ficially), reveals a culture far more Basin than Pueblo in general habitus. The same is even clearer for the Walapai. In spite, then, of the striking differences between cultures like those of the modern Paiutes and Pueblos, their remote antecedents were closely similar if not substantially common, in a common environment mainly of sagebrush- juniper semidesert. Within the environment, the boundary between Basin and Pueblo culture has fluctuated, and that between Basin and sub-Pueblo has always remained ill defined. In the light of this, the relation of California to the Basin, which cannot be denied, is best viewed as resting on an early kinship of Californian and primi- tive Basin-Southwest cultures. In part, influences flowed from the latter into California, resulting in growths like that of Yokuts-Mono pottery.7 In part, sKidder, Science, 66:489-491, 1927. 4 UC-PAAE 25:1-183, 1929. Significant affinities must not be stretehed into an identifica- tion. The Loveloek culture is not classical Arizona-New Mexico Basket Maker eulture. 5Judd, bibliography 1917-1920 as cited in Kidder, Southwestern Archaeology; M. R. Harrington, AA, 29:262-277, 1927; MAIHF-IN 5:235-240, 1928 (map). 'AMNH-AP 29:81-392, 1928; also AA 31:213-222, 1929. 7UC-PAAE 23:382, 1928; Gayton, UO-PAAE 24:239-255, 1929. 50 Kroeber: Ctltural and Natural Areas of Native North America perhaps, reciprocal influences flowed from California into the Basin, as spe- cific Pueblo influences retracted there. In the main, however, the California and Basin cultures are alike because they have not risen very far above their early, closely related forms. Where there has been such rise or divergence, as in the Californian climax area, none of the secondary or specialized manifesta- tions-Kuksu cult, Pomo basketry-has crossed the Sierra Nevada, even in fragments. The Great Basin is a hinterland to California as the Columbia- Fraser drainage is to the Northwest Coast, in the sense that both have tended to preserve an early phase of culture which has advanced to specialization in the coastal areas. The Basin is not a hinterland to California in the full sense that Columbia-Fraser is to the Northwest, because it has not been influenced by the coastal culture to the same degree. The position of the Bannock and the Lemhi Shoshone is not clear. They live in Snake and therefore Columbia8 drainage, but in an area of sagebrush- juniper plant cover, except for pine in the higher Salmon River Mountains (map 4). They subsist to some degree on salmon, but their speech is that of the Great Basin. They are here tentatively classified contrary to physiography, and according to their ecological and linguistic relations, as constituting a Basin subarea. Another subarea is that of the Klamath-Modoc and Achomawi-Atsugewi, who live in Northwest Coast and Californian drainage, but seem largely Great Basin in culture. This classification of them is given a certain historic depth by the occurrence, in the Lovelock Cave deposits of central Nevada, of flexible twined basketry of modern Klamath-Achomawi type in the lowest or atlatl- bearing strata.9 The nineteenth century brought into the Klamath Lakes re- gion an importation of Columbia and Plains traits. These came from the north, by way of the Deschutes River, and represent an extension of Plains culture in its final exuberant horse phase. Achomawi territory is partly sagebrush- juniper, partly pine; Klamath, pine forest surrounding a characterizing area of marsh (map 4). Both territories lie high,10 at about 4000 feet elevation, and while they have nearly complete sea drainage, they are situated inland of the Sierra-Cascades axis, which here is somewhat broken down. Physiographi- cally, both territories are reckoned as in the Basin, that is, Basin-and-Range province (map 7); and climatically they are cool and still within humid limits (map 24). The Achomawi-Atsugewi subtribes segregate into an eastern and a western division, which C. Hart Merriam' and Kniffene' have shown to differ somewhat in culture, as well as in the plant cover of their habitats. The west- ernmost Achomawi group, the Madesi,"8 seem to belong culturally with their neighbors, the Wintu, who are clearly Californian. The Northeastern or Moun- 8,Boas, works cited in Tribal Map bibliography (p. 9 above), 1927, 1928, has only the Bannock in Snake drainage before 1800, all the Shoshone in this latitude being west of the continental watershed. This seems very doubtful. 9U-PAAE 25:26, 1929. 10 L. Spier, Klamath Ethnography, UO-PAAE 30, 1930. n Classification and Distribution of Pit River Indian Tribes, SI-MC 78, no. 3, 1926 (publ. 2874). Achomawi Geography, UC-PAAE 23:297-332, 1928. 18C. Hart Merriam, An-nik-a-del, 1928. 51 2University of Califormia Publications in Am. Arch. and Ethn. tain Maidu should perhaps be classed in the Klamath Lakes-Pit River area. Theirs is also a 4000-foot habitat, as compared with sea level to 3000 feet for the other Maidu. Culturally they agree at many points with the Achomawi, as in their basketry and lack of ritual organization. The Mountain Maidu as well as western Achomawi-Atsugewi can probably be included with about equal justice in the California and the Great Basin cultures. I reckon them here with the Great Basin in order to draw attention to their status, and to break down the tradition, to which I have myself contributed, that because they live in the state of California they are to be assumed as Californian cul- turally. Separated by the southerly Sierra Nevada are the Western and Eastern Mono, locally known as Mono and Paiute respectively. The former, at least about the Kings and Kaweah rivers, are culturally almost indistinguishable from the hill Yokuts, and therefore Californian." The latter, according to studies undertaken by J. H. Steward, promise to show a number of Califor- nian traits. Their habitat, however, and presumably their communications and outlook, are in the Basin; and they are here included in that area.' Farther south, the Chemehuevi, who essentially are only the westernmost bands of the true or Southern Paiute, have come under some influence of the Lower Colorado culture; as have the Paiute of the Virgin-Muddy drainage: song cycles, mourning, a little agriculture and pottery, though the last seems more likely to be a Pueblo inheritance.'6 The subsistence habits and manner of life, however, continue to be Basin Shoshonean.'7 On the east, tribes like the Ute and Shoshone are of Basin affiliations with a late Plains overlay, as discussed below. Even the Wind River Shoshone, across the divide in Missouri drainage, can best be included in Basin culture. Their nineteenth-century habitat was one of sagebrush. 1A. H. Gayton, UC-PAAE 24:239-255, 1929; 24:361-420, 1930; 28:57-82, 1930. The Mono of the north fork of the San Joaquin differ appreciably from the adjacent Yokuts and Miwok; see Gifford, UC-PAAE 31:15-65, 1932. ' Steward's study has now appeared: Ethnography of the Owens Valley Paiute, UC- PAAE 33:233-350, 1933. It shows these "Eastern Mono" (he declares the term a misnomer in spite of the fact that Mono Lake is east of the Sierra) to have a Basin rather than Cali- fornian culture. This is confirmed by element surveys by himself and H. E. Driver comparing the Owens Valley Northern Paiute with the tribes east and west respectively. Steward has also rendered a long-needed service in determining the territorial bands of the Shoshone, and 0. Stewart those of the Northern Paiute: see the supplemental bibliography in See. III, "Tribal Areas" (including briefer articles listed under "Ray, Park, and others"). These works, with Kelly's (see note 17, below), at last give a reasonably accurate picture of the many small groups that constitute the Great Basin Shoshoneans. "I Gifford, UC-PAAE 23:372, 1928. 17Kelly, AA 36:548-560, 1934, groups the Southern Paiute-Chemehuevi into fifteen terri- torial bands: San Juan (this band is the only one south of the Colorado), Kaiparowits, Panguitch, Kaibab, Uinkarets, Shivwits, St. George, Gunlock, Cedar, Beaver, Panaca, Paranigat, Moapa, Las Vegas, Chemehuevi. These are evidently small tribes, with territories averaging in area not far from 2000 square miles. However, Las Vegas with its historic Chemehuevi offshoot is disproportionately large: one-fourth of the total Southern Paiute area. Of the fifteen bands, eight agree with the Powell-Ingalls list (Hdbk. Am. Inds. 2:188), three are new, four cover the same area as twenty-three of Powell's (Kaibab two, Cedar three, Moapa seven, Las Vegas eleven), which accordingly represent subdivisions or mere localities. Drucker's 1935 element survey unites the Chemehuevi strongly with the Yuma in culture as against a Serrano-Cahuilla-Luisenlo-Dieguenlo unit farther west. 52 Kroeber: Cultural and Natural Areas of Native North Anmerica In short, about three marginal subareas are more or less authenticable for the Basin besides its main area: la. The cultural Basin area proper. lb. The Bannock and Shoshone of the Snake-Salmon drainage. lc. The non-Shoshonean tribes of the Klamath Lakes and Pit River. Id. The eastern border tribes recently influenced by those of the Plains, especially the Wind River Shoshone across the Rockies. All the tribes of the area are Shoshonean except those of subarea "e" and the Washo on the western border of "la." 2. CALIFORNIA Otis T. Mason made his California area include Oregon. Wissler makes it coterminous with California, except for excluding the southeastern corner of the state and including western Nevada. My classification gives southern Cali- fornia to the Southwest, the northwestern corner to the Northwest Coast, the northeastern, as just discussed, to the Great Basin, the eastern or trans-Sierra fringe also to the Basin. This leaves to the California area only the region which in earlier classifications, made with a local rather than continental view, I called Central California.' Essentially, this area consists of the Great (or Interior) Valley of California with the Coast Ranges and Sierra Nevada that flank it. Superficially it is a homogeneous unit ;1' but its plant cover is irregu- larly varied and difficult to classify. This is shown by the fact that no two of the vegetation maps agree closely, and that all of them recognize one or more vegetation types characteristic of the region and largely confined to it. Broadly, the region may be defined as a bunch-grass valley containing a core of marshland and surrounded by an inner belt of chaparral-covered hills and an outer one of pine forest. However, the pine encroaches on the chaparral in the north, vice versa in the south; and on the northern coast-range side, the redwood of northwestern forest type has intruded into the pine of western forest affiliations. Even the pine cover is somewhat specialized, being classed by Shantz and Zon as a separate local subtype of yellow pine-sugar pine asso- ciation (20c, map 4). So far as native habitat and utilization are concerned, all the plant-cover classifications are somewhat misleading. Californian subsistence was built up about the acorn; and the oak occurs more or less in all the vegetational areas. Even the densely shaded redwood belt includes the tanbark oak (Lithocarpus) in its typical association, and the acorns of this oak were most highly esteemed by the tribes that knew them. The Great and smaller valleys to which a grass- land cover is ascribed, contained, along the streams and in their moister por- tions, groves of the large valley oak, which yielded perhaps the heaviest of all the acorn crops. Other oaks pervade the chaparral and run up into the pine. In fact, what the map can only show as uniform chaparral is actually an intimate interdigitation of tracts of the smaller oaks and specific chaparral '8 BAE-B 78, fig. 73, 1925. ' Though the physiographers recognize three paralleling divisions, Sierra Nevada, Cali- fornia Trough, and California Coast Ranges (see Sec. XII and map 7), these are given a certain unity, in point of human utilization, by the central valley. The three divisions to- gether coincide rather closely with cultural California. 53 4University of California P1ublications in Am. Arch. and Ethn. (Arctostaphylos, Adenostoma, Ceanothus).' Shelford's lumping of every- thing below the higher-level pines into a single Broad-leafed Evergreen Semi- Desert of Winter Rains (map 3) is therefore not so crude a procedure as it may at first seem. It expresses at any rate the essential unity of the vegetation so far as native utilization is concerned. Only, it must be remembered that this winter rain semidesert includes southern California, which culturally has been reckoned with the Southwest. Southern California has already been de- scribed ethnographically as an area of characteristically Californian subsist- ence basis with a specific Southwestern culture content above the subsistence level. For the eastern side of the Great Valley and western gradual slope of the Sierra, C. Hart Merriam has shown' a neat correspondence to hold between his life zones and the ethnic groupings, which in turn correspond to minor cultural differences. This correspondence does not hold for the coast-range half of California nor for southern California. Here the life zones run over the map in endless irregularities with which the local ethnic and cultural cleavage lines mostly fail to agree. Historically the California culture area may be defined as a region lying between the Northwest and Southwest but not reached to any determining degree by influences from either. Influences from both can be traced: from the Northwest, chiefly along the Coast Ranges; from the Southwest, along the Sierra. The sitting cradle among the Pomo, the mourning-anniversary cere- mony and feather-stick offerings among the Maidu, serve as examples. Such imports, however, are few, relative to the totality of the culture. This culture, as set forth in the preceding section, evidently began as one similar to that of the adjoining Great Basin, and has never diverged very far from it. However, subsistence in California was so much easier that culture-surplus growths developed. These found a definite climax, though not a very high one, among the Pomo, Patwin, and Valley Maidu (Kuksu cult, Hesi ceremony, Pomo bas- ketry) about the center of the northern half of the area. The rest of the area is not classifiable according to broadly significant distributions, except into better-off valley and poorer hill tracts." "' Shantz and Zon, Atlas, p. 8; also fig. 5. "Science, 19:912-917, 1904. S S. Klimek, CED: I-The Structure of California Indian Culture, UC-PAAE 37:1-70, 1935, has approached the problem with a statistical analysis of the distribution of some four hundred traits. His map (p. 52) recognizes seven California provinces: Colorado River (including Chemehuevi); Southern California (including Chumash); San Joaquin (Yokuts and Mono); Central (Yuki to Miwok); Northwestern (my California-Northwest Transi- tion: Wailaki, Sinkyone, Wiyot, Chimariko, Shasta); Northwest Coast; Northeastern (Klamath-Modoe). The following are transitional: Wintun, Northwestern and Central; Achomawi-Atsugewi, Northwestern and Northeastern; Salinan, San Joaquin and South- Central; the mountain Maidu and Costano are sub-Central. From intereorrelation of elements, Klimek has also determined a dozen "culture strata," whose local strength he has mapped on his pp. 54-56. Five of these strata center in as many provinces, seven in the South-Central and Central provinces. Of the latter, four have their respective areas of characterization among the Pomo; the Patwin; the Miwok, Washo, and adjacent Shoshoneans; and the Central province generally; three in the South-Central, among the Chumash-Gabrielino, Cahuilla-Luisenlo-Diegue-o, and Serrano. Klimek's study opens up a new type of approach, but his determination of strata goes beyond what can be attempted in the present volume. 54 Kroeber: Cultural and Natural Areas of Native North America While the growth of the Californian climax culture may have been fur- thered by Northwestern and Southwestern influences, it is clear that these were not its primary determinants. The most specific manifestations of this climax are neither Northwestern nor Southwestern. A favorable ecological margin evidently brought about a cultural luxuriance, which, with but little material from the two greater centers available to work upon, because of re- moteness from both, fell back on native materials to elaborate. If it had been otherwise, Pomo basketry should show as a specialization of Northwest Cali- fornia basketry, the Kuksu society as a modification of the datura initiation of southern California; which it would be difficult to maintain reasonably. California, then, -differs from the other intermediate areas, especially the genetically related Great Basin, in that, owing probably to less stringent preoccupation with subsistence problems, it has throughout developed a some- what more richly characterized culture, and has even been able to mature a definite climax. It differs from the great expansive centers in that it never developed enough cultural energy to impart its products in any serious degree to other areas. As already mentioned, the groups from the Shasta to the Sinkyone, prob- ably including the western Wintu in Trinity drainage, are transitional be- tween California and the Lower Klamath subculture of the Northwest Coast. The classification of the area then is: 2a. (Main) California Area. 2b. California Climax, the lower Sacramento to the Russian River: Pomo, Patwin, Valley Maidu, and Nisenan. 2e. California-Northwest Transition: Shasta, probably Wintu west of the Coast Ranges in Trinity drainage, Chimariko, Athabasean tribes from Whilkut and Nongatl to Wailaki and Sinkyone. S. COLUMBIA-FRASER PLATEAU The two great drainages of the Columbia and Fraser rivers constitute the Pla- teau area of American ethnology, with which the Great Interior Basin has sometimes been included. As a matter of fact, not only is the Basin distinct vegetationally, ethnically, and culturally, with affiliations primarily toward the Southwest and California, but, as already shown, there is some warrant for classing the Snake portion of the Columbia drainage with it. This leaves the middle and upper Columbia, and the Fraser above its lowest region. These are the great salmon streams of the continent, south of Alaska; and they water the area in which the Northwest Coast culture is likely to have had some of its beginning and which at any rate still forms its hinterland. As expectable, influences from east of the Rockies have also penetrated this inter- mountain area; and as it failed to develop any great amount of culture of its own, it has long, and on the whole correctly, been regarded as a region marked by negative traits, by absences, except for its more immediate subsistence adaptations. Within the area, not only must the Fraser be distinguished from the Columbia, but also the latter must be separated into its middle and upper courses, making three provinces. 55 6University of California Publications in Am. Arch. and Ethn. 3a, the Middle Columbia area, is partly sagebrush-juniper and partly bunch-grass steppe, with pine forest on the higher levels. This is the recent Sahaptin area, with a few interior Salish tribes, such as the Wenatchi, Sin- kiuse, Spokane; and the Wailatpu.' The Sahaptin territory on the lower Snake and Salmon rivers is pine, interspersed with bunch-grass tracts. 3b, the Upper Columbia area, is wooded, forming, almost continuously, part of the western or mountain forest, though there is some grassland along the river valleys. This area holds the majority of the interior Salish tribes, from Methow and Okanagan to Flathead; besides them, only the Kootenay. 3c, the Fraser area, is the home of another group of interior Salish, the Lillooet, Thompson, and Shuswap, with the Athabascan Chilcotin, Nicola, and perhaps Carrier. With reference to plant cover, this area is variously classified. Malte (map 5) makes it a subprovince of the mountain (western) forest, with three grassland "dry belts" in the south. Shelford shows it mainly as steppe in his general map (map 3), but adds a special map of interior Brit- ish Columbia which gives the grassy areas in detail.' Harshberger (map 2) includes it in his Columbian division of the northwestern forest area; which evidently refers to species representation rather than to habitus or density of vegetation.= The common factors in these divergent classifications seem to be that the Fraser drainage is drier than the Upper Columbia, that its forest is sparser and more interrupted by stretches of steppe, and that its flora leans somewhat more toward that of the coast. This last factor is in line with its being a more specific cultural hinterland to the Northwest Coast than is either of the Columbia areas. The Fraser area has also been partly protected, culturally, against eastern influences by the Upper Columbia, whose forestation would filter out many specific Plains traits. It may therefore be reckoned as culturally nearest, of the three Plateau provinces, to the Northwest Coast. It was the Middle Colum- bia, with its prevalence of open country, that finally proved most receptive to Plains influences. Of the more special luxury manifestations of Plains cul- ture, like the coup system, the societies, the Sun dance, only fragments got over the Rockies; material adaptations like the tepee, the parfleche, and floral bead designs were largely accepted, and almost made the Middle Columbia culture over. The consequence was an unusually sharp cleavage at The Dalles, where alone Pacific Coast and Plains culture traits met in a conspicuous non- conformity. It must be remembered, however, that this is true of Plains horse Boas BAE-R 41, map, has the Salish before 1800 holding both banks of the Columbia as far down as the Chinook, and the Sahaptin, exclusive of the Nez Perce6, on the middle Snake, John Day, and Deschutes rivers. Two other tribal maps of the region have appeared since my continental tribal map was drawn: Spier's Tribal Distribution in Washington, Gen. Ser. in Anthr., no. 3, 1936; and V. F. Ray's Native Villages and Groupings of the Columbia Basin, Pae. Northwest Quart., 27, no. 2, 1936. Spier's map (p. 43) is for the early nineteenth century, Ray's (pp. 5 ff.) for about 1850. These maps, Boas's, and Mooney's all present discrepancies, due partly to shifts of groups. 2" J. Davidson in Shelford, flg. 8, p. 155; A. H. Hutchinson, p. 156. These grassy dry belts contain sagebrush and cactus. 25Harshberger, 599, recognizes a sage formation (Artemisia tridentata) in the middle valley of the Fraser as an extension of the Great Basin flora. 56 Kroeber: Cultural and Natural Areas of Native North America culture, probably not of the old culture of the Plains. In 1600 and 1700 the Middle Columbia was still a true transition area, an intermediate low-level zone. By 1800 the Plains influence had begun to come in; most of it probably fell within the nineteenth century; it continued operative in some degree after the beginning of Caucasian settlement; and at the base of the Cascades a little of it turned and flowed southward into a corner of the Basin area in north- eastern California, to the Klamath-Modoc and Achomawi. To what respective degree this late Plains influence reached the Sahaptin of the Middle Columbia through the Salishan tribes of the Pend d'Oreille branch of the Columbia, or through the Shoshonean Lemhi and Bannock of the Snake drainage, is not clear. It evidently did not come through the Great Basin Shoshoneans ac- tually in contact with Plains tribes, such as the Ute and Shoshone, else the effects would presumably have been passed on also to their westerly kinsmen the Western Shoshone and Northern and Southern Paiute, which was not what occurred. These remarks on recency do not mean that the Columbia and Snake did not serve at all as a channel of communications between the Pacific Coast and Atlantic drainage in prehistoric time. They must have done so. Only, the connections must have been far slighter before use of the horse; and the rela- tively poor subsistence conditions and consequent low level of culture along the Columbia and Snake would have strained out many of the more specialized traits, and most or all of the luxury developments, of both eastern and western culture. The ethnological position of the Carrier on the upper Fraser is not clear. Their communications with the coast seem to have been directly across the mountains with Tsimshian and Haisla, not through the Fraser Salish. Almost certainly, also, they maintained more intercourse with their Athabascan kins- men east of the mountains than did the southward-facing Shuswap. The Car- rier may therefore have to be reckoned as forming a separate subprovince, either of the Fraser area or, more likely, of the northwest Athabascan in- terior.' Farther north, inland from the Tlingit, live the Athabascan Tahltan and Taku-tine, on the Stikine and other Pacific rivers, the Taku-tine also partly in upper Yukon drainage. Both groups are part of the Nahane division, the rest of which holds Mackenzie drainage territory. The vegetation maps are not very definite or concordant for this poorly explored region. The Tahltan have been subjected to Tlingit influences. But on the whole it seems justifiable to include them with the other northwestern Athabascans. It is possible that with their interior neighbors, the Kaska, Etchao-tine, and Abbato-tine, they constitute a last, most northerly, intermountain culture group. Even so, how- ever, this would properly form a subprovince of the western or Athabascan division of the great interior Subarctic area.' I add in map 9 a reproduction of Livingston and Shreve's "generalized vege- tation map" of the United States, with a heavy line added to bring out more "See Eastern and Northern Areas, See. IX, 16e, p. 99 below. 27The same, 16d. 57 8University of California Publications in Am. Arch. and Ethn. 11 i VI,'' ?LW fl//Fl r;? w 00 N t % N, A 11 %"% I %? a 1. z I '. I.- ji I ?o 0 ro .4 5 P4I 440) +b 0e 4z; 'O o - oC o 0. CD M g .t,MU,. w 0) 4) cd Pko P4 (C) 58 Kroeber: Cultural and Natural Areas of Native North America 59 graphically a basic distinction which can sometimes be profitably made be- tween open and forested country in general, apart from the specific types of plant cover which constitute each. This map also correlates with that of Rus- sell's dry climates (no. 24). It shows as open country, either desert, steppe, grassland, or shrub, the great mass of the Southwest in the United States; most of the Intermediate-Intermountain territory, namely, all the Basin, most of California, and part of the Plateau; and the Plains and Prairie areas, which remain to be considered. With the exception of the tall-grass Prairies, these all evince ancient cultural interconnection: the Sonora-Yuma area with the Pueblo, this with the Basin, this again with California, Plateau, and true Plains. Only the Prairies lean culturally on the eastern forest into which they pass over a highly irregular boundary. 0University of California Publications in Amn. Arch. and Ethn. IX. CULTURE AREAS: EAST AND NORTH EASTERN AREAS TH REST of the continent north of Mexico, embracing nearly the whole of its Atlantic and Arctic drainages, constitutes a series of areas whose relations are different from those so far considered. The Eskimo, Northwest, and South- west cultures are highly defined, whereas those of California and the inter- mountain regions are low-level in characterization and transitional in content. East of the Rockies there is not a single native culture of as high a degree of characterization as occur west; nor, except in some regions near the minimum of subsistence potentiality, any as culturally uncharacterized as some of the western transitional cultures. In other words, the Atlantic side of North America is relatively uniform in its native culture. Its bent or direction is fundamentally similar everywhere. Once local subsistence adaptations and local culture imports are allowed for, there remains little in the way of local development; and, concomitantly, no great degree of difference in culture intensification. This lightness of cultural contour has its parallels in the environment. East of the Rockies there is not a single high or formidable mountain mass, not an elevated plateau. With all the range in latitude, summers are nearly every- where hot, winters either cold or at least punctuated by frosts and raw winds; seasonal variation in temperature is accentuated, precipitation fairly dis- tributed throughout the year. The plant cover is prevailingly forest, shading through parkland into open grass only toward the Rockies.' There is nothing like the wetness of the Northwest Coast, the deserts of the Southwest and Basin, or the winter rains of California; no extensive scrub nor shrub land. The vegetation areas are fewer, larger, more continuous, the differences be- tween many of them slight. As might be expected, segregation of the vast Eastern territory into cul- tural areas is difficult, and classification has varied. Mason recognizes six, Wissler only four areas. These have already been cited for their agreements, but their disagreements are equally significant. Mason's eastern areas are: Yukon-Mackenzie, defined as the transcontinen- tal coniferous belt, draining into arctic seas; St. Lawrence and Lakes, from Manitoba to northern New England; Atlantic Slope, Massachusetts to South Carolina; Gulf Coast, Georgia to Texas; Mississippi Valley; Plains. As against these, the Wissler eastern areas are Mackenzie, Eastern Woodland, Southeast, Plains. The difference is not only that Mason subdivides further. In fact, his Yukon- Mackenzie area sweeps across the continent to the Atlantic, taking in part of Wissler's Eastern Woodland. The rest of Wissler's Eastern (really North- 1 G. Friederici, Der Grad der Durchdringbarkeit Nordamerikas, etc., Petermanns Mit- teilungen, Ergainzungsheft 209, 216-229, 1930, argues that most of the eastern woodland, at least in the United States, was an open stand without underbrush, easily traversed even by vehieles, this condition being due to systematie firing by the Indians. He also discusses prairies, swamps, oak openings, groves, canebrakes, etc., features which may often have been of more importance for native occupants than the average composition of the prevail- ing timber cover. 60 Kroeber: Cultural and Natural Areas of Native North America eastern) Woodland, Mason assigns to his St. Lawrence-and-Lakes, Atlantic, and Mississippi Valley areas, which, however, also overlap into Wissler's Southeast and Plains areas. Wissler, after noting that the characterization of his Woodland culture is difficult, divides it into four types: a northern, non- agricultural and similar in material culture to the Mackenzie; an Iroquoian; a central Algonkin, west of the last named; and an eastern Algonkin, from Abnaki to Delaware. It will be seen that these four subareas do not correspond to the four parts of larger areas which Mason posits in place of their aggre- gate, the Eastern Woodland. The comparative diagram herewith, map 10, based on Wissler's schematized map < and Mason's text, illustrates / the degree of discrepancy. With this experience before us, it seems wisest to vary the \ ... 7 procedure of Wissler and that which has been followed here so far, namely, of blocking 3 - - out the grand areas and then 8 subdividing them; and in- stead, to begin with defining as small areas as justifiable. 0 I ; Of these, I recognize sixteen, plus some subdivisions: not all coordinate, almost surely, Map 10. The Ethnic Environment and Culture-area but yet difficult to subordi- Classifications of Mason and Wissler, superposed. nate to major d'iv.isions. k Mason: solid lines, roman numerals; Wissler, broken nate t maJordlnslos. In-lines, arabie numerals. deed, I confess myself unable to set up such a major framework satisfactorily for this large part of the continent. I have followed a quasi grouping into Eastern areas and Northern areas, corresponding more or less to those with and without agriculture or agricultural antecedents, respectively. But this is pretty summary. Within the Eastern group of areas, I accord preeminence to the Southeast; but this again does not take us very far. I do, at several points, discuss historic relations and cultural dependences. For the rest, I can only say that while my sixteen areas may seem seriated as if they possessed equal cultural weight and depth, they obviously are not equivalent. The culture of the eastern part of the con- tinent simply is harder to organize than the rest. 1. SOUTHEAST The Southeast is a long-recognized culture which unquestionably forms a valid unit, provided its area is not made to take in too much, but is limited to the Muskogian tribes and some of their immediate neighbors such as the Natchez and Tunica on one side and the Timucua and Yuchi on the other. This' is the area that must be accorded such cultural primacy as there was east of the 61 2University of California Publications in Am. Arch. and Ethn. Rocky Mountains. But it cannot be regarded as marked off by abrupt transi- tions of either cultural content or cultural saturation such as one encounters in passing out of the Northwest Coast or Pueblo areas. There is one thing, however, that corroborates the Southeast as culturally most advanced in the eastern half of the continent: it contained a distinguish- able climax or focus. This climax lay on the lower Mississippi, among the Natchez and their neighbors. What sets these tribes apart is,slight enough: their class system, with its emphasis on rank and sun symbolism. Their matri- lineate, litters, war captive sacrifice-torture, maize-harvest busk, ossuary and perpetual fire "temples," as well as everything that is known of their material culture, are found rather generally through the Southeast, and in part far beyond. It is the peculiar system of class exogamy by extremes, with death of the Stinkard on death of the Sun Spouse, and ranking of the children of Sun males in an intermediate class, that is distinctive. In fact, it is so decidedly unique that its authenticity might be doubted were it not for the corroboration of several reports. There is about this Natchez system something of the quality of a remnant: it is hard to conceive as a product growing out of the general Southeastern social structure. And it is clear that the French received the idea, in part from Natchez tradition itself, that the Natchez had dwindled from a previously more prosperous condition. But, whatever the origin, the system is peculiar and definite enough to fall into the category of a luxury product and therefore to be indicative of a climax condition, whether this was active or waning at the time of discovery. Captive torture on the frame is another trait that looks like a worn-down survival in the light of Mexican captive sacrifice, sometimes also performed on a frame, and with the Pawnee sacrifice of a girl on a frame occurring even farther north.2 This in spite of the fact that neither Natchez nor Muskogi seem to have been conscious that the torture was a sacrifice, and that torture ex- tended far beyond the Southeast. What may seem evidence of another climax, the successful Creek confed- eracy, must be interpreted as a formation which probably could not have arisen in native times. Not only were most of the Creeks provided with fire- arms, livestock, and fruit trees before white settlement reached them, but they had seen the Coast tribes, from South Carolina to Louisiana, one after another shattered or wasted under English, Spanish, or French contact. They had in fact received refugees from many of these tribes. Their confederacy was the quadruple product of these reenforeements, of an economic life full of Cau- casian absorptions, of pressure or consciousness of threat from the course of white settlement, and of a geographical situation that gave them more than a century of relative respite from fatal conffict with the invaders. Under purely native conditions, the Creek league would not have been so populous, cohesive, or permanent. To a less conspicuous degree, the same applies to the fortunes of the Cherokee, Choctaw, and Chickasaw; and it holds, perhaps even more notably, for the Iroquois in the north. The pattern of these confederacies was mainly or wholly native; their success and subsequent organization in 2Wissler and Spinden, Am. Mus. Jour., 16, 1916; Linton, AA 28:457-466, 1926. 62 Kroeber: Cultural and Natural Areas,of Native North America detail were the result of the coming of the whites and the misfortunes of other tribes. These other tribes probably had possessed a very similar social appa- ratus capable of development toward political integration, but were crushed before this development could grow beyond the native stage. In the light of all this, the "constitutional government" of the "five civilized tribes" was the last of three stages: first, unstable and loosely integrated leagues in native times; second, politically successful confederacies under white stimulus and pres- sure; and third, imitations of the American government after the loss of in- dependence. So far as this argument may be accepted, the inference follows that Creek culture was not quite so specialized as the Natchez at the time of discovery, and that therefore there need be little question that the focus within the area was situated on the lower Mississippi. The tribes includable in the Southeastern area are all the Muskogian peo- ples; the Yuchi; the Timucua, but none beyond them in Florida; the Siouan Ofo and Biloxi; the Tunican, Natchez, and Chitimachan peoples. The Atakapa seem to have been more or less transitional between this area and the South Texas area, in which they are here placed. The Quapaw-Arkansas may have belonged in the true Southeast, but have been tentatively reckoned as in the Red River area. The Timucua possibly were distinct enough to be considered as forming a subprovince. The Cherokee I exclude. The area, then, extends from the Savannah River to just across the Mississippi. Except for small areas of prairie and marsh grassland, the whole Southeast was forested. The prevailing cover was of the Southeastern Pines. There were also fairly large tracts of River-bottom, Cypress, or Swamp Forest; of Pied- mont Oak-Pine mixture or Transition Forest; and, especially on the north- west, of Trans-Alleghanian Oak-Chestnut Deciduous Forest. The Natchez and their neighbors lived in a habitat of River-bottom and Transition Forest, the Chickasaw largely in Deciduous; the Choctaw and Creek chiefly in the Pine, but also in the Piedmont Transition; the Timucua, Apalachi, and other coast tribes in Pine country studded with hardwood hammocks, traversed by a River-bottom stand along the streams, and fringed by shore marshes; the Chitimacha, and the supposedly Muskogian tribes downstream from New Or- leans, in a region of prevailing marsh grassland. These are the attributions in terms mainly of Shantz and Zon (map 4). The other ecological sources differ somewhat in detail, but give a similar picture. Shreve (map 5) and Harsh- berger (map 2) carry the Southeastern Pine Forest somewhat farther across the Mississippi, so as to include much of Caddo and Quapaw territory; which may be culturally significant. All in all, it is clear that the Southeastern cul- ture was not limited to one type of plant cover ;' but Pine Forest constitutes its largest block, and conversely most of the Southeastern Pine grew within the Southeastern culture area. Some centuries before the discovery, there flourished, most outstandingly in the Ohio Valley, but also in the region of the Great Lakes, the Mississippi Val- ley, and the Southeast, the culture or aggregation of cultures known by the 8 The pure pine stand is mainly attributable to local soil conditions. 63 4Univer8ity of California Publications in An. Arch. and Ethn. loose name of Mound Builder. This culture has similarities to that of the Southeast, and some sort of relationship is generally assumed. Whether in an earlier time the climax lay in the Ohio Valley region and the Southeast was dependent on this, becoming the climax only on the decay or retreat of the more northerly center; or whether the region of the lower Mississippi was already then the center, as its greater proximity to Mexico would make ex- pectable, and the Ohio Valley culture was a locally flourishing variant-this alternative cannot now be decided. After all, there has not yet been a serious attempt to integrate and interpret in broad terms the large mass of archaeo- logical material which for a century has accumulated east of the Mississippi. Since most of the foregoing was first written, Swanton has published a valu- able general paper, The Aboriginal Culture of the Southeast.' In this he enu- merates Southeastern culture elements as well as their distribution in the areal subdivisions, besides sketching the presumable development of the whole type of culture. Both his delimitation and his internal organization of the area differ from mine at a number of points; but, with all deference to his more thorough knowledge, I have decided to let my classification stand as written, for comparison. Swanton excludes from the Southeast the Calusa, Atakapa, Quapaw, and Shawnee, but, like Speck," extends the culture to the Potomac. His subdivisions are: 1, Algonkin tidewater of Virginia and North Carolina; 2, Eastern Siouan area, Piedmont and Coast; 3, Timucua; 4, Creek, with the Georgia coast, Yuchi, Cherokee, and Chickasaw as marginal; 5, Choctaw; 6, Natchez and allies; 7, Chitimacha; 8, Tunica; 9, Caddo. On these matters, the differences between Swanton and myself perhaps largely concern what might be called taxonomic order. Probably of greater historic import is his heavier weighting of inland as against coastal populations; and especially of the cul- ture of the Creek. By my standard, he is interpreting in the light of eighteenth- rather than of sixteenth-century conditions; but others must judge who is most nearly right.' Muskogian and Creek The problem of what constituted Creeks and what Muskogians remains rather obscure. Swanton's detailed researches have not yet made the fundamentals of the situation clear-in the main, it would seem, because the Creek confed- eracy was very different things at different times. Muskogian tribes that at one 4 BAE-R 42:673-726, 1928. 6 Cited below, under "Atlantie Coast Areas." 6 I believe I am not in fundamental conflict with Swanton in drawing the lines of my Southeast narrower than his. He defines his Southeast with reference to three or four other eastern culture areas; I with reference to fifteen. As between assigning the Caddo to the conventional Southeast or the conventional Plains, for instance, I would unqualifiedly follow him in the former course. Our differences appear to refer to frame rather than to specific relationships. Two important papers, illustrated by six maps, were read by Swanton at the December, 1932, Conference on Southern Prehistory at Birmingham, Alabama, under National Re- search Council auspices. I do not cite these, because the mimeographed report is marked "Not for publication"; but all anthropologists trust that both papers may soon be officially published in full or with extensions. In 1935 Swanton published Notes on the Cultural Province of the Southeast, AA 37: 373-385, in which he reviews various problems of prehistoric and historic culture and popu- lation in the area. 64 Kroeber: Cultural and Natural Areas of Native North America time were wholly independent of the Creek and perhaps hostile to them, later became reduced in numbers, moved, joined the confederacy, and gave up their proper dialect; just as did non-Muskogian tribes like Natchez, Yuchi, and Shawnee. At least three Muskogian dialect groups were represented in the Creek confederacy, besides fragments. These three are the Muskogi proper or Upper Creek; the Hitchiti-Apalachicola; and the Alabama-Koasati. Muskogi therefore denotes, in one sense, a relatively limited group which formed a fraction of the later Creeks; in another and later sense, the whole family of which the Creek along with the Choctaw, Chickasaw, and many others were members. Swanton's linguistic classification of the Muskogian family proper' (with- out Natchez, etc.) recognizes two grand divisions, a "Northern" or Muskogi, and a "Southern." The Northern or Muskogi division (A) has enumerated under it only a number of Creek "towns" like Kauita, Kusa, Eufaula, Tukabachi, Hohliwahali, nearly all of which are Upper Creek. The Southern division (B) has no fewer than nine subdivisions: 1, Choctaw-Chickasaw; 2, Alabama-Koasati; 3, Hitchiti; 4, Chatot; 5, Apalachi; 6, Osochi; 7, Guale- Yamasi; 8, Cusabo; 9, Tuskegee. The Hitchiti proper formed part of the Lower Creeks, as did the related Okmulgee, Oconee, etc.; but again, independent tribes like the Apalachicola are reckoned as part of the Hitchiti dialect group. The same may be said of the Alabama-Koasati group, some of which was, or became, Lower Creek, whereas at least some of it was originally non-Creek politically. The Choctaw-Chickasaw group was the largest of all, and included not only these two nations, which always remained independent, but also a series of tribes (Chakehiuma, etc.) on the Yazoo River; another (Houma to Acolapisa) on the lowest Mississippi and the Pearl River; and a third (Mobile, Pensacola) on the Alabama and western Florida coast. The Chatot, Apalachi, Osochi, Guale-Yamasi, Cusabo groups were smaller, and situated to the south and east of the later Creeks. I have plotted the approximate distribution of these dialect groups8 on map 11. It will be seen at once that the names "Northern" and "Southern" are wholly inappropriate for the two grand divisions of the Muskogian family, though they have some justification within the later Creek confederacy. The "Northern" or Muskogi proper division has "Southern" dialect groups on its east, south, west, and northwest. In fact, it is entirely surrounded by them except on the north, where its territory was bounded by that of the alien Chero- kee. It is also much the smaller group areally, occupying not more than a sixth of the total Muskogian territory. If, therefore, Swanton's classification of Muskogi proper as one of two coordinate main branches of the Muskogian stocks is linguistically sound, we are confronted by the very anomalous situa- tion that the most distinctive dialect group of the family lies almost sur- rounded by the others, and that the peripherally situated dialects are not the most aberrant. This raises a suspicion about the classification, namely, that it 7 BAE-B 73:11, 1922. 8 Except the Tuskegee, since I cannot gather from Swanton's account where he thinks their original habitat lay, 65 6University of California Publications in Am. Arch. and Ethn. may have been made primarily with the Muskogi proper (A) in mind, and that hence the nine B groups do not really agree among themselves so much intrinsi- cally as in all differing from the accidental starting point A. If Swanton's primary division into A and B is sustained by equalized comparison of the data, a reason ought to be sought for the distinctiveness of A, on account of the historical significance which the fact necessarily carries. In that event, two Map 11. Muskogian Dialect Groups; compiled from Swanton's data. A, Northern or Muskogi. B, Southern, comprising: 1, Choctaw-Chickasaw; 2, Alabama-Koasati; 3, Hit- chiti; 4, Chatot; 5, Apalachi; 6, Osoehi; 7, Guale-Yamasi; 8, Cusabo. possibilities suggest themselves, namely, that the Muskogi proper (A) have somewhat altered their speech through Cherokee contacts; or, more likely, that their habitat set them off somewhat from all their relatives. It is a hill country, where the Appalachian System breaks down, and was prevailingly covered with hardwood forest (oak-pine type) as against the Southeastern coniferous stand that dominated most of the remainder of Muskogian holdings. However this may have been, the Choctaw-Chickasaw-Houma-Pensacola group (Bi) is much the largest, covering about as much territory as all the rest together. The Hitchiti group (B3) was next largest, and, with the Ala- bama-Koasati (B2), it joined with the Muskogi proper (A) to form the greater part of the Creek nation in confederacy times. On the whole, A came to consti- tute the Upper and B2-B3 the Lower Creeks; but with certain notable excep- tions, such as Kauita and Kasihta, which spoke Muskogi proper, yet were the 66 Kroeber: Cultural and Natural Areas of Native North America leading war and peace towns of the Lower Creeks. The Muskogi proper, being best protected from English, Spanish, and French encroachment and demor- alization by their remote situation, probably came in time to occupy a prece- dence which they were far from enjoying in the wholly aboriginal period. The other divisions (B4 to B8) were small groups forming a southeastern fringe of the stock from the Gulf across the neck of Florida to the Atlantic. 2. SOUTH FLORIDA What is known ethnologically of the tribes of Florida south of Tampa Bay has been brought together by Swanton.9 It is evident that culture was of South- eastern type, but in a poorer phase: pottery seems to have been made, but agriculture is specifically stated not to have been practiced. The Atlantic Coast tribes in particular led a sort of beachcomber's life. Their gold may have been taken mainly from Spanish wrecks. The archaeological evidence at first seems conflicting, owing to the promi- nence of Cushing's famous but still only partly published findings at Key Marco. It is the preservation of wooden objects in muck that distinguishes this site, and Moore has shown that a deliberate attempt to find a second similar site would be nearly hopeless. Nor have other sites been discovered by accident, though a few wooden pieces from other spots in southern Florida have come to light and been described by Fewkes. These allow the ascription of a fairly developed carving art to the southern half of the peninsula at some time in its prehistory. The extensive explorations of Moore, however, confirm the ethno- logical data in showing that on the whole the ancient culture, like the historic one, was definitely meager south of Tampa Bay. Another fact which excava- tions seem to have established with fair conclusiveness, though more especially for Tampa Bay and the northern part of the peninsula, is a stratigraphic suc- cession from no pottery to plain pottery to ornamented pottery.10 All in all, Antillean influences are not so notable in southern Florida as might be expected from proximity. It seems that such Antillean features as occur in North America are characteristic of the Gulf Coast or Southeast as a whole rather than specific to a South Florida culture area.' This would sug- gest that connections were active chiefly at some time earlier than the discov- ery, and were followed perhaps by a period of dwindling of relations. In this connection it is no doubt significant that the climax of West Indian development lay in Puerto Rico and Haiti, and that those parts of the archi- pelago nearest to Florida showed a meager culture. This was especially true of western Cuba. The Bahamas also, with'their limited environment, can have possessed only part of the stock of Antillean culture. Wissler, Gower,U 9BAE-B 73, 1922, esp. 387-398. 10 Cushing, Proc. Am. Philos. Soc., 35:329-342, 1896; Fewkes, SI-MO 76, no. 13 (publ. 2787), 1924, 80, no. 9 (publ. 2960), 1928; 0. B. Moore, Jour. Acad. Nat. Sci. Phila., 11:352- 394, 1900, 11:421-497, 1901, 12:127-357, 1902, 12:364-394, 1903, 13:126-244, 299-325, 1905, 13:406-470, 1907, 16:515-577, 1918; J. Wyman, Mem. Peabody Acad. Sci., 1:1-94, 1875; S. T. Walker, SI-AR for 1879, 1881, 1883; N. C. Nelson, AMNE-AP 22:75-103, 1918; W. H. Holmes, BAE-R 20, 1903. W. H. Holmes, AA 7:71-79, 1894. 1American Indian, 257, 1922. 1" AAA-M 35, 1927. 67 University of California Publications in An. Arch. and Ethn. 1P d m O} 0 o o p4 0 * *? to.C P4 r o L m 0 e k mcvP re o o a~ +v *.4 Cd qGo d CN 0em -0O ? N 0R ?8 fra? ; z 0 zz m*0? fi Q,; > P4 68 G- 1-4 4B Kroeber: Cultural and Natural Areas of Native North America Lovene-the last on the basis of an intensive analysis-agree that West In- dian culture is fiundamentally South American. South Florida and west Cuba- Bahamas therefore were the poor peripheries of two areas whose centers in 1500 A.D. lay far apart-one on the lower Mississippi and the other on the South American mainland. This appears to be the reason why the tip of the peninsula, in spite of the fact that its climate and life were tropical, did not form an outright part of the Antillean culture area: mainland contiguity to the Southeast prevailed over environmental unison with the islands. At an earlier period, when cultural and ethnic relations were different, it may have formed part of the Antillean area.15 Southern Florida is a distinct natural area, though far from an ecological unit. The vegetation maps differ in terminology of characterization as well as in detail of area, but agree in marking off at least part of the southern end of the peninsula from the remainder of the southeastern United States. Harsh- berger (map 2) and Shelford (map 3) indicate the Antillean relations of the flora; Merriam puts the tip of the peninsula into the Tropical life zone. A small map of tree-species distribution, reproduced in map 12 from Livingston and Shreve, is an index of the particularity of the region. The outstanding climatic features are high temperature, due not only to latitude but also to warm ocean waters; and seasonal precipitation of savanna type--dry winter and wet summer. Land form, drainage, and soil cause the marked variations within this frame. The Everglades, for instance, alternate each year between being a lake and a prairie; surrounding them are swamp, scrub, tropical, man- grove, deciduous, and pine forest, and mixtures of these. Watson's classifica- tion of Florida plant covers,' though referring to the state as a whole, usefully supplements the somewhat schematic presentation of the maps, as table 3 shows. Map 14, below, also shows incisively the high specialization of South Florida in evergreen broad-leaved trees. In summary, it is clear that the southern end of the peninsula presents a distinctive environmental as well as cultural type. The ecology approaches the tropical, the culture is low-level. The environmentalist explanation would be that tropical environment retards or depresses culture through its physiologi- cal effect on the human organism. But this explanation leaves out cultural or historical factors, which are necessarily operative in all cultural phenomena, in order to build up a pseudo law by injecting remote, vague, and indirect physiological factors. A reasonably sufficient interpretation is given by the interaction of environment and history. The culture of South Florida, being mainly derived from that of the Southeast with its essentially temperate adap- tation, lost something and gained little by its transplantation to a different environment. That this environment was tropical is a mere incident: the Southeastern culture diminished equally in intensity northward in propor- tion as it extended into cooler temperate habitats. If the historic culture of 14Ueber die Wurzeln der Tainischen Kultur, pt. 1, Goteborg, 1924. As suggested by Fewkes, 1924, Conclusion. 1In Shelford, 427-440. Compare Harshberger, 227-232, 695-700. 69 0University of California Publications in Am. Arch. and Ethn. the Southeast had been primarily South American or Antillean in origin, tropical Florida would presumably have preserved it most fully and the rest of the Southeast have shown the impoverished form. This, then, illustrates how ecological considerations strengthen the historical conclusions toward which anthropologists have tended as a result of analysis and comparison of culture. TABLE 3 FLORIDA VEGETATiON TYPEs (After Watson) la. Grassy swamps, savannas, and marshes. Most of the southern third of the peninsula. Everglades. Cladium effusum saw grass. lb. Salt marsh. Spartina, and in extreme south Mariscus jamaicensis saw grass. ie. Mangrove swamp. 2a. Flatwoods. Open pine forest on level, poorly drained, acid soil, interspersed with bog vegetation. 2b. Cypress swamp. Depressions in flatwoods, stream and lake borders. Big Cypress swamp southwest of Lake Okeechobee. 3a. Scrub. On drier sands and dunes. Saw palmetto, evergreen oaks, Opuntia, Ilex; on dunes also cabbage palmetto, Agave, Yucca. 3b. Spruce pine, Pinus dlausa, on less dry sand, interspersed among 2 and 4. 4. High pine woods. Rolling, well-drained country. Open stand of long-leaved pine, inter- spersed with saw palmetto, scrub oak, lupin, chinquapin, short grass. 5. Hammocks. Hardwood forests, deciduous and evergreen. 5a. High-hammock climax. Evergreen magnolias, red bay, and holly dominant. 5b. High hammoeks. Deciduous trees preponderant. Most extensive toward northern parts of state, where it merges gradually into the eastern deciduous forest. Farther south, transitional between 4 and 5e. 5c. Low hammocks. On wet lands between 2b and 5e. Tupelo, ash, maple, hackberry, water oak and swamp oak, magnolia, cabbage palmetto. 5d. (=6). Tropical hammocks. Dense jungle, mostly evergreen, with lianas and epiphytes. Banyans, wild papaya, Swietenia, Fwius, Ocotea, hickory. 3. SOUTH TEXAS: NORTHWEST GULF COAST South Texas is an area which is little known. Every tribe in it has long been culturally extinct; some are absolutely so. Cabeza de Vaca found them poor and hungry; and so they seem to have remained. They were cannibalistic. They practiced no agriculture. They got bison too rarely to depend on them. They made little pottery. There was, no doubt, a subsistence differentiation between those immediately on the coast and those inland, but otherwise the culture seems to have varied little in fundamentals. The peoples involved were the Karankawa, the Tonkawa, and later in part the Athabascan-Apache- Lipan. The agricultural Atakapa'7 leaned toward the Southeast, but may be counted in cultural Texas. Part of the territory attributed to the Mescalero Apache on map 1 may once have belonged in. The Coahuiltec on both sides of the lower Rio Grande, and the so-called Tamaulipec to the south beyond, very likely were closely related in culture to the South Texas peoples. It is toward 1TBAE-B 43:35-36, 360-363, 1911. 70 Kroeber: Cultural and Natural Areas of Native North America the southern boundary of the Tamaulipec, a little north of the Panuco River, that agriculture and pyramidal structures appear and the South Mexican cul- ture may be said to begin. The South Texas area-or, better, the Northwest Gulf Coast area of which it is part-accordingly reaches from the edge of cultural Mexico almost to the border of the Southeastern climax. Expectably, this intervening area should manifest some traces of having been the medium through which the generally recognized connections between these two areas of higher-level culture passed. Instead, we have what Swanton aptly describes as a cultural sink.' Archaeo- logical exploration, which has never been systematically attempted there, may bring something to light; but nothing very notable is to be expected, else some indications should have appeared through cultivation and settlement before now. The problem is the more puzzling in that those Southeastern traits which seem most Mexican are generally not represented in the Southwest, and vice versa,19 so that a theory of circuitous diffusion around the South Texas area also seems contrary to the facts. If there were evidence of maritime movements along the shores of the Gulf or across it, one could more readily assume these as the mechanism of Mexican-Southeastern connections. Mexico City, Santa Fe, and Natchez form very nearly an equilateral triangle; and from Tampico to the mouth of the Mississippi is no farther, even by land, than from Mexico City to El Paso. The cultural backwardness of the Northwest Gulf area-or, at any rate, of its Texan portion-is also difficult to understand on environmental grounds. The rainfall ranges from 50 inches at the mouth of the Sabine to 20 at the mouth of the Pecos, the lower Rio Grande having about the mean. The precipi- tation-evaporation ratio ranges from semihumid to semiarid (map 13). Much of the area is agriculturally productive under Caucasian settlement. The plant cover is variously described, so as to suggest local peculiarities difficult to fit into broad schemes of classification. Shreve follows the Rio Grande down on the north side with a Texas Succulent Desert and Texas Semidesert (map 5), where Shantz and Zon assign Creosote Bush and Desert Mesquite Savanna, with areas of Desert and Tall Grass (map 4). Along the coast they are in not much better agreement. Harshberger (map 2) emphasizes a Mexican constituent in the flora as far as San Antonio and Matagorda Bay.' Map 14 shows microphyllous trees following the coast to the eastern edge of Texas and extending north into the Panhandle, and the lowest part of the Rio Grande Valley as the region of maximum accumulation of such tree species in the United States. Map 12 shows species of southeastern deciduous trees ex- tending, though in diminishing numbers, south to the Rio Grande and west to 1040. There does seem general agreement that from the Guadalupe or Nueces west and south the natural plant cover is xerophytic in spite of the consider- able precipitation-evidently on account of dry winters and rapid evaporation 18 ICA 20 (1922, Rio de Janeiro) :53-59, 1924. 19 Thus scaffold sacrifiee and ball courts, on the one hand; the metate, masonry, and masks, on the other. 20 Harshberger, work cited (see above, p. 14), pp. 659-660; also pp. 514, 528-531. 71 University of California Publications in Am. Arch. and Ethn. o I 0e B 0S ..0 0 E3-4 C>1 0) .~cq Cd o5 .,{ O o 0k)0 t9o PA C %-4 ce *- Cs *p m 0 LQ 4 Ca 0 _0d c3 5. r 0),, c. . -a5.4 I* 72 CD,, 0 2I CJ7 U o42s cra Kroeber: Cultural and Natural Areas of Native North America 73 o M a) 0 _2 ji S E~-4 0 j ! gX rd P4- Gs -ro roC 0 wq0 $E-,-4 C 0 S + Z p4*i 4D- a0 W- 0 4University of California Publications in Am. Arch. and Ethn. in summer; and that to the east of these streams savanna or scrub or oak woodland (cross timber) prevails over true forest. But there certainly is no vegetational unity underlying the cultural area. Perhaps it was not so much a culture unit as really a sump-a series of somewhat varying habitats none of which was favorable to the major subsistence patterns worked out in Mexico, the Southwest, and the Southeast. In contrast to Linton, who sees most specific Mexican traits that occur in the Southeast lacking in the Southwest, and vice versa,' Swanton is "inclined to regard most Mexican influences as having been introduced [into the South- east] via the Pueblos rather than by the more direct route [of- southern Texas]." Swanton's formulation of the limits of the "highest levels of the cul- ture of the Southeast" is also worth summarizing here with special reference to the suddenness of the transition toward the west :' The lower Mississippi Valley; "back from" the Gulf Coast eastward to the Atlantic, in- cluding northern Florida; formerly, most of the Ohio Valley; the Iroquoians forming a marginal territory. Along the Atlantic Coast the Southeastern culture shaded out much more rapidly. To the northwest, it extended "not much beyond the Mississippi"; to the west, "it ended rather abruptly with the Caddo tribes" of northwestern Louisiana and north- eastern Texas (the habitat of these tribes falling short of the Trinity River and not reach- ing the coast); on the Gulf, it "cannot be traced beyond Vermilion Bay, Louisiana." 4. RED RIVER AREA The Caddo group is usually considered transitional between the Southeast and the Plains. These people were subjected to strong Spanish and French influ- ences from the end of the seventeenth century, their tribal organizations have been partly dissolved and reconstituted, and their culture is much broken. They farmed, made pottery and wicker and twill basketry, lived in village settlements that were sometimes straggling or scattered, built large domed houses of thatch, erected mounds, kept perpetual fire burning in a temple or communal structure, acknowledged the authority of an intervillage or inter- tribal religious head, celebrated a first-maize and harvest festival, sometimes tortured or sacrificed captives on the frame. This culture obviously is basically Southeastern, with affiliations to the Natchez rather than the Muskogian tribes, but with some of the Natchez specializations lacking and with certain differentiations of its own, such as the predominant use of grass houses. Wiss- ler puts the modern Caddo, Kichai, Waco, Tawakoni into his Southeastern area, the Wichita into the Plains; the separation of the latter seems arbitrary, except perhaps for modern times. This was a deciduous forest area. By the Shantz-Zon classification (map 4), it lay prevailingly in the Oak-Pine Eastern Forest, partly also in Oak-Hick- ory, Southeastern Coniferous, and River-bottom Forest. In the nineteenth century some of the western tribes were in the prairie extending south through Dallas and Fort Worth and that about the Wichita Mountains. It is not wholly clear whether these were Caddoan habitats in native times, and, if so, whether 2'AA 28:464, 1926. '? ICA, 1924, as cited. This delimitation differs somewhat from the one in BAE-R 42: 673-726, 1928, which has been discussed above. 74 Kroeber: Cultural and Natural Areas of Native North America the tribes in them depended more on farming or on bison hunting. All the areas in question lie east of the hundredth meridian and are cultivable. The limits of this subculture are difficult to draw on the north and east. The Quapaw-Arkansas, the southernmost Siouan people west of the Mississippi, may have belonged either in this or in the Southeastern area: they used pali- sades, for instance.' The Osage are also difficult to place. Thanks to La Flesche, we know several of their rituals in detail; but these give relatively few indica- tions of the type of the culture as a whole. Osage organization into patrilinear clans and exogamic moieties is "Central" Siouan, but certain features, such as the relating of the moieties to peace and war, recall the Southeast. The situation of the central Osage settlements on the Osage River well inside of Missouri points to affiliation with the Siouan tribes rather than those to the south. The large extent of territory ascribed to them on the map, following Mooney, is probably misleading in this connection. The nucleus of Osage habitat was in woodland. Well to the northwest of the Osage, in the prairies of the middle Platte drainage, were the Caddoan Pawnee, who are particularly difficult to place. Usually reckoned loosely with the Plains tribes, they show a village organiza- tion, matrilineate, captive sacrifice, star symbolism, and similar traits which either relate directly to the Caddo-Natchez culture or at any rate set them off from both the Prairie and Plains tribes. There is some tendency to regard the Pawnee as the eastern tribe showing most relations to the Pueblos, and as possible intermediaries between the Southwest and the Plains and Southeast. But the general cast and emphasis of Pawnee culture are certainly very differ- ent from Pueblo, or even from that of the eastern Apache. In connection with this problem the question of the certain identification of the Pawnee and their territory in the period of discovery is important. Their nineteenth-century habitat centered in middle Nebraska, but their earlier territory has been placed in eastern Kansas. If the Caddo extended farther north into Oklahoma the two groups may still have been adjacent not many centuries ago. At any rate, there is no reason why contacts between them should not have remained open. If the authority for map 1-here, Mooney-is right, most of the inter- vening area was thinly occupied even at the beginning of the nineteenth cen- tury, constituting back country of the Osage and Kansa, the latter a small tribe. It seems altogether likely that Pawnee culture was basically a variant of Caddoan, but that on account of its more northwesterly situation, and per- haps relative freedom from exposure to Spanish and French contacts, it came more largely under Plains influences after the horse became common.' To summarize: provisionally the Quapaw may be reckoned as belonging n Or again, they may have belonged with the other "central" Siouans in the Prairie area. Little is available about them ethnologically, and my assignment of them is no more than a guess. " Much new light is shed by Strong's Introduction to Nebraska Archaeology, SI-MC 93, no. 10 (publ. 3303), 1935. He holds the Pawnee to have been long in Nebraska, and to have undergone a cultural floreseence in the prehistoric period from 1540 to 1682, decline setting in after the introduction of the horse. The Upper Republican archaeological culture is pre- sumably a prehistoric stage in Pawnee development, and is attenuated Southeastern. See especially Strong's pp. 9, 13, 15, 55, 245, 272, 273, 296. 75 6University of California Publications in Amn. Arch. and Ethn. with the Caddo (4a); the Osage, with the southern ("Central Siouan") Prairie tribes (6a) ; the Pawnee as forming a subtype (4b) of Caddo culture, with recent horse-bison culture overlay. 5. PLAINS AREAS The viewpoint from which the Plains are here treated has been previously out- lined in connection with a review of the cultural relations of the Southwest.' Essentially the view held is that the Plains culture has been one of the well- developed and characterized cultures of North America only since the taking over of the horse from Europeans, and that previously there was no important Plains culture, the chief phases in the area being marginal to richer cultures outside. In brief, the historic Plains culture was a late high-pressure center of culture in a region which previously had been rather conspicuously low-pres- sure. That there is nothing revolutionary in such a view is shown by the fact that as long ago as 1916 Sapir in a sentence analyzed the recent Plains culture into non-Plains origins.' The reason why he did not follow the matter farther is that his essay was concerned with method rather than fact. The Plains tribes, along with the Pueblos, Northwest Coast Indians, Cali- fornians, and Eskimo, are among the most intensively investigated in America. The reason has been the incentive to study extended by the saturation of their late culture, plus its preservation well into the nineteenth century. Even today it is possible to find informants who have experienced the old life and are able to give clear, vivid accounts of it. The returns being richer, more ethnological interest was directed to them. Specialization followed, and on that some inevi- table loss of perspective. This relatively rich culture, so much more satisfying to deal with than the remnants of that to the east or the meager ones of the Plateau and to the far north, began to be intimately dissected in some of its aspects-but mainly with reference to itself, not to its outward relations. Spier on the Sun dance,' Lowie on age societies,' Wissler on shamanistic and dancing societies,' analyzed historic developments within the culture as it was. How the culture as a whole came to be, was less and less asked. Wissler perhaps did most both to extend and to fix the concept of the Plains area, and to define its center.'8 He even went so far as to indicate that its culmination lay most probably among the Oglala Teton Dakota, with Arapaho, Cheyenne, and Crow participating next in order.' Another factor contributed to the essentially static conception. Wissler found that when the Plains tribes took up the horse they did not make their culture over." Travois transportation, the tepee, the bison hunt under control, UC-PAAE 23:375-398, 1928. "Time Perspective in Aboriginal American Culture, Canada Geol. Surv., Mem., 90 (Anthr. Ser., no. 13) :45, 1916. -7AMNH-AP 16:451-527, 1921. "Same, 11: 877-984, 1916. "Same, 11: 855-876, 1916. AA 16:447-505, 1914 (449-451 and map); The American Indian. "AA 16:473, 1914. AA 16:1-25, 1914. 76 Kroeber: Cultural and Natural Areas of Native North America had all been there before. The horse was simply put into the old patterns and made these more productive. It was easier for the tribes to do this than to evolve or adjust to a new set of patterns. As an analysis of cultural dynamics or social psychology, this was a valid demonstration. Too largely, however, it seems to have been tacitly interpreted also as a historical conclusion, that Plains culture after the horse went on much as before. Very little reflection shows that this could not have been so. Could any good-sized group have lived permanently off the bison on the open plains while they and their dogs were dragging their dwellings, furniture, provisions, and children? How large a tepee could have been continuously moved in this way, how much apparatus could it have contained, how close were its inmates huddled, how large the camp circle ? How often could several thousand people have congregated in one spot to hold a four or eight days' Sun dance? By the standard of the nineteenth century, the sixteenth-century Plains Indian would have been miserably poor and almost chronically hungry, if he had tried to follow the same life. Showy clothing, embroidered footgear, medicine-bundle purchases, elaborate rituals, gratuitous and time-cosuming warfare, all these he could have indulged in but little-not much more than the tribes of the intermoun- tain or southern Texas regions. In short, ethnologists have gradually become so interested in the specialized manifestations of Plains culture that they have forgotten that largely these are definite luxury developments possible only with the subsistence basis of life adjusted unusually favorably and dependably. That such an adjustment could have been made through the mechanism of dog traction by a migratory people dependent on a migratory animal for their food, is highly problem- atical. With the horse and all its culturally intensifying consequences taken away from the tribes of the western or true plains, such as the Blackfoot, Crow, Teton, and Arapaho, these have left but a meager stock of culture. The same subtraction from the agricultural Prairie tribes-Mandan, Santee, Pawnee, or Omaha-would leave them far more. In the sixteenth century, then, I be- lieve that culture within the so-called Plains area was richest and centered in the prairies, not the plains, and was not primarily but only incidentally based on bison subsistence. But the Prairie tribes show affiliations to both the South- east and the Northeast; and the Plains culture is thereby made doubly de- pendent. In the sixteenth century, instead of being a climax, it was not even subelimax: it was peripheral. If it seems unlikely that a ritual as elaborate as the Sun dance grew up in a few hundred years, the answer is twofold. First, many of its elements-tor- ture, painting, altar, bundle-occur in other associations and may be ancient, while the complex of elements that constitute the ritual is younger. Secondly, that ceremonial elaborations in this area can be highly unstable is evident from comparison of societies; for instance, the age-graded ones of the Arapaho and Atsina (Gros Ventre).' These are alike enough to make it certain that they represent, in the main, deviations from an original common system. The AMNH-AP 1:141-280, 1908; see esp. pp. 230, 260. 77 8University of California Publications in Am. Arch. and Ethn. two tribes are closely related in language, and the Arapaho regard the Gros Ventre as the northernmost of their five divisions. The two groups had separate ranges as early as 1750, but may well have been still a unit in 1600 or even 1650. With the ensuing geographic separation to help, the dialectic divergence between them could easily have been achieved by 1900, it would seem. The differences between the society systems of the two tribes comprise added or dropped societies, transfer of functions from one society to another, and trans- position of societies in the age order. What is an elderly, important group in one tribe, is a young group, near the beginning of the sequence, in the other. It is difficult even to imagine a mechanism by which a change like this could have taken place in a system after this had become based on the principle of seniority. It is much as if in some European countries Wednesday came after Thursday. Yet the change is there. If a now closed system could alter as this one has in two to three centuries, a new one could certainly crystallize as quickly, whether it be a society series or a Sun dance. What it is suggested happened is that not only ritual complexes, but indeed all sorts of cultural patterns, quickly blossomed out in the plains after the introduction of the horse had converted a strugglingly precarious or seasonal mode of subsistence into one normally assured, abundant, and productive of wealth and leisure. This development was strongest where the effect of the horse was greatest, in the true or western short-grass plains. Here, then, there rapidly grew up a new center-an active crater of culture, to use Wissler's figure. This in turn reacted on the agricultural tribes of the prairies, strongly influenced the nearer intermountain tribes as well as several at the edge of the northern forest, and about 1800 sent its influences down the Columbia to the Cascades. The new culture was not only active and intensive, it was still ex- panding when white settlement killed its roots. It is scarcely contendable that the western plains were wholly uninhabited before the horse was available. Agricultural groups from east and west prob- ably strayed in now and then and tried to farm. Small groups could make a living by combining bison and river-bottom hunting with berry and root gathering. But the population probably clung in the main to the foot of the Rockies, where wood, water, and shelter were more abundant, fauna and flora more variegated, a less specialized subsistence mechanism sufficient; and from there they made incursions into the plains to hunt their big game, much as the prairie and parkland and even forest tribes ranged in from the east in the historic period. Such habits would account for the dog travois and folding tent. They would give to the plains some human utilization and occupancy. They would not leave room for a specialized culture to center there. Wissler's views on the Plains have undergone decided changes. In 1907' he advocated substantially the position here maintained. He even spoke of the plains as uninhabited, and the moving out into them as due to the horse. In 1914' he held that the horse "is largely responsible for such modifications and realignments as give us the typical [western, Blackfoot to Comanche] Plains 34 ICA 15 (1906, Quebec) :39-52, 1907. 86 Same, 44, 45. 1 AA 16:1-25, 1914. 78 Kroeber: Cultural and Natural Areas of Native North America culture of the nineteenth century" ;' and that the "vigor and accentuated association of traits" of this culture could not have been achieved without the horse. On the other hand, "no important Plains traits except those directly associated with the horse [like saddles] seem to have come into existence" after its introduction; "all the essential elements of Plains culture would have gone on, if the horse had been denied them"; and "from a qualitative point of view the culture of the Plains would have been much the same without the horse."'While no "important traits, material or otherwise, were either dropped or added," yet "the relative intensities of many traits were changed, giving us a different cultural whole," and leaving to the horse its strongest claim "as an intensifier of original Plains traits."' Horse introduction is also held re- sponsible for "reversing cultural values," that is, causing old nomadic (Sho- shonean) cultures to "predominate" over the "previously dominant sedentary cultures of the Siouan and Caddoan tribes."' In short, a new culture grew up wholly out of old elements through the introduction of the horse. A later paper in 1914,1 and The American Indian in 1917 and 1922, go further in that they accept this new culture almost as if it were timeless. The purely horse-using tribes are described as forming the "center" of the area, and tribes like the Omaha and Pawnee as culturally less typical and dependent. This is of course a static interpretation of a historic moment. In short, Wissler's first approach was historical; his second, historical and analytic; his third, descriptively analytic. Returning to the primary consideration, we can summarize by saying that in the main, in the prehistoric period, the cultural emphasis of the conventional "Plains culture area" region lay on its borders; the plains themselves were a cultural margin. From this aspect, the so-called Plains area breaks up into several smaller areas. One is adjacent to the Southwest; another, to the intermountain regions farther north; on the east there can be recognized, besides the Caddoan or Red River area which is essentially Southeastern, a central Siouan, a north Siouan, a village, and a Canadian Prairie area. 5a. Southern Plains This is the area adjacent to the Southwest and more or less dependent on it. Its modern representatives are the Comanche, Kiowa, and Kiowa-Apache. The Lipan and possibly the Tonkawa may have belonged here rather than in the South Texas area, at some time in their career; so may part of what are now recognized as Apache, the Mescalero and Jicarilla. These Apache very likely represent rather well in some ways the status of the old Southern Plains cul- ture. The Spaniards called them, or related bands, Llaneros and Vaqueros. They were mountain tribes, marginally Southwestern, fronting on the plains and hunting bison. The Kiowa-Apache look like one of these eastern Apache bands, who, after they had the horse, committed themselves definitely to the bison and the plains, and on account of numerical weakness joined themselves to the Kiowa. The other eastern Apaches clung to their mountains, or were I" Same, 17. " Same, 16, 17. 89 Same, 18, 19. 40 Same, 25. 41AA 16:447-505, 1914. 79 0University of California Publications in Am. Arch. and Ethn. beaten back into them, continuing to use the plains as an auxiliary range. If we knew more about them and the Kiowa and Comanche, we should probably see many resemblances. Their style of bead embroidery is certainly similar; and, in its outlining quality, distinct from that of the more northerly Plains." Linguistic affiliations point the same way. Mooney accepts the Kiowa tradition of a northern origin; but the Kiowa language seems to be related to Tanoan.' Comanche is nothing but a Shoshone dialect. The tribes in the historic South-. ern Plains group thus appear to connect in origin with others in or beyond the Rockies. Even facially the Kiowa and Comanche resemble the Apache. This may or may not be due to common heredity. It certainly holds for the physiog- nomic expression, which argues a similar life. The Comanche seem not to have appeared in their historic habitat until about 1700. This lateness corresponds with the close similarity of their dialect to that of the Wind River Shoshone. These people, in turn, live in an area which belongs to the Rocky Mountains physiographically, with the Basin vegetationally: it is sagebrush, not grassland. Wind River culture must have been of pretty pure Basin type until the horse came in and they began to take on an overlay of Plains culture. It was about this time, apparently, that the Comanche moved south from them. The Comanche are much better known his- torically than ethnologically. A monographic study of them is perhaps the greatest desideratum, next to the publication of the full Murie Pawnee mate- rials, in the general Plains area." The ecological environment, especially of the Comanche, is not uniform. They extended from the true plains into desert savanna and scrub timber (maps 2-5), which again suggests a remnant of habits preceding their adop- tion of horse-bison culture. 5b. Northern Plains This is the area of the culture whose rapid and expansive development within the historic period has brought about the current concept of a large "Plains" culture area. Wissler considers that eleven tribes manifest the typical culture of the "Plains." Three of these are in the Southern area just discussed. The other eight are the Sarsi, Blackfoot (including Piegan and Blood), Atsina, Arapaho, Cheyenne, Crow, Teton Dakota, and Assiniboin. These in fact are the eight, or perhaps seven without the Assiniboin, which I would reckon as constituting the valid Northern Plains group. It was among them that the Sun dance apparently originated and certainly flourished most exuberantly. There is a good deal of evidence of flow into the area. The Sarsi are obvi- ously a northwest Athabascan tribe that left its kinsmen in the forest to attach "The difference will perhaps prove to be partly due to relative absenee in the south of antecedent porcupine-quill embroidery. " J. P. Harrington, AA 12:119-123, 1910. Mooney and Harrington may both be right, the tribe being southern in origin but having temporarily moved north and then south again, legend retaining only the last of the events. Mooney has them in contact with the Spanish frontier of New Mexico in the first half of the eighteenth century, in the Black Hills about 1775, on the North Platte in 1805. He puts a residence on the headwaters of the Missouri earlier than any of these habitats, but it may have fallen between the first and second. " Fortunately this is no longer true. G. Wagner has made such a study, and so has a Laboratory of Anthropology party under Linton. 80 Kroeber: Cultural and Natural Areas of Native North America themselves to the Blackfoot. The Crow are linguistically closest to the Hidatsa. They look, therefore, like an agricultural group that had early ventured to give up farming for the plains life-probably even some centuries before they got horses. The Assiniboin speak a Yanktonai (Prairie) Dakota dialect. The Teton Dakota, according to Mooney, did not begin seriously to push west of the Missouri until about 1750." For the Cheyenne there are traditions as well as records" of movement from the prairies into the plains.'7 This leaves only the Arapaho-Atsina and Blackfoot-Blood-Piegan without known indications of entry into the area. These groups are both Algonkin, but of speech highly diversified, as well from each other as from the great body of Algonkin; much more so than Cheyenne. Differentiation of such strength does not generally occur in languages that remain in geographical contiguity and intercommunication with the parent stock. It does often proceed with rapidity in languages that are subjected to contacts principally with alien idioms." " The lateness of this date may possibly be somewhat exaggerated, but the statement seems to be essentially true. Grinnell (passage cited in next footnote) would make the date even later-after 1800. " G. B. Grinnell, The Cheyenne Indians, 1:1-46, 2:382-384, 1923, has collected a mass of material. He holds, no doubt with reason, that the Cheyenne did not move as a unit, but by villages and bands, which successively caught up with or overtook one another; that some of them farmed until well after 1800; and that they met (reunited with) the Suhtai or Sutaio only after they had crossed the Missouri, in the Black Hills country. The farthest eastern point possibly attributable to the Cheyenne, but not authenticable, is Mankato, Minnesota. This is in timber, just east of the prairie. Yellow Medicine River (a tributary of the Minnesota) in southwestern Minnesota seems fairly authentic as a habitat, and already lies in prairie. Then follow the area west of Lake Traverse in South Dakota, the head of Maple Creek (western affluent of the James), and Sheyenne River (western tributary of the upper Red River). The last two are in North Dakota, and all three in prairie. All habitats from here on lie in short-grass plains. Next follow both banks of the Missouri, in the region of the mouths of the Cannonball, Grand, Owl, and Big Cheyenne rivers; thence up these rivers to and beyond the Black Hills, that is, the country to north and east of these mountains back to the Missouri. This was the main habitat in the period around 1800. Except for temporary movements of bands, there seems to have been no general drift to or south of the Platte until about 1826. Even this drift applied to only part of the tribe, since the division into Northern and Southern Cheyenne began as late as about 1830. Early enemies were the Assiniboin and Crow; friends, the Dakota, Mandan, and Arikara. In the Black HiUs region the Cheyenne were associated with the Arapaho, Kiowa, and "Comanche." So far Grinnell. The Black Hills evidently provided on a minor scale the same sort of ad- vantages of shelter, fuel, and small game as the foothills of the Rockies supplied to the early tribes of the western plains. The total Cheyenne migration was about four hundred miles, with a transient bend northwest at the beginning to include part of the Red River Valley but in the general direction of west; until the due south swing after the first quarter of the nineteenth century. Even in their earliest determinable habitat the Cheyenne were separated by Siouans (Assiniboin, Dakota, Iowa) from all Central and Eastern Algonkins (Cree, Ojibwa, Sauk, Kickapoo, Illinois). The upshot is: a prairie-farming people, separated and well differentiated from their ancient woodland kinsmen, yielding very hesitantly to the lure of the western bison after they had horses in the eighteenth century, and not wholly committing themselves to the "typical Plains" culture until well into the nineteenth. 67 Cheyenne speech is much closer to Central-Eastern Algonkin than is either Blackfoot or Arapaho. It is much more different, however, than it could have become during a separa- tion of only two or three centuries. The purely linguistic inference thus is that the Cheyenne, though recent in the plains, lived, before that, somewhat apart from the Central Algonkins of the woodland; therefore most likely in the prairies. This tallies with the historical in- ferences in the last preceding footnote. " This does not necessitate that form or even content is borrowed. It seems that the stimulus of alien contact is often sufficient to set up new processes, which go their own way. If taking over of vocabulary also occurs, it is evidently due to cultural rather than linguistic causes. The outright borrowing of grammar on any considerable scale is a putative phe- nomenon whose actuality remains to be proved. 81 University of California Publications in Ain. Arch. and Ethn. If the Arapaho and Blackfoot drifted to the base of the Rockies a fairly long time ago, we should have them fulffiling all the geographical and historical conditions which in theory would be needed to account for their set-off linguis- tic status. Moving them into their recent habitat since the introduction of the horse, or even a century or so before, would not allow time for the existing degree of diversity, according to all authentic precedent on the rate of altera- tion of speech. We may therefore regard these two groups of tribes as ancient occupants of the northern true plains, or rather of the foothills of the Rockies and the plains tributary thereto. The Blackfoot made much use of the moun- tains in the historic period; like the Mountain as distinct from the River division of the Crow. It cannot be asserted that the Blackfoot and Arapaho groups were the only ones formerly in the northern plains. They are the only ones who we can be reasonably sure were there. The Crow may have been with them. There may have been other tribes who have since disappeared or been expelled or absorbed. The Sutaio among the Cheyenne might possibly have been the remnant of such a group. Of the seven Teton subtribes, the Oglala seem to have been culturally the most vigorous in the nineteenth century. They were also the advanced outpost in the southwestward push away from the old Dakota prairie-and-forest habi- tat. This coincidence is evidently significant of the recent growth of cultural intensity in the plains proper. The Northern Plains subarea is one of short grass, with grama and buffalo grass characteristic.' It covers substantially all this short-grass territory ex- cept for parts within the Southern Plains and Village Prairie subarea. The stream bottoms contain cottonwood growth nearly but not quite to the Rockies. On the west, the short grass generally abuts on mountain pine. The one region in which the buffalo grass changes to sagebrush is in Wyo- ming. Here were the Wind River Shoshone. Their country is mostly open plains lying behind outlying broken ranges of the Rockies and draining through the Big Horn, Powder, and North Platte into the Missouri. But it is sagebrush- covered, like the habitat of all the Shoshoneans in the Basin.' This is an un- usually neat instance of ecological conformity. The Wind River Shoshone, in other words, belong to the Great Basin culture, with a recent veneer of North- ern Plains culture. Wissler virtually recognizes this-although he includes them and the Ute in a western border division of his Plains areae-when he mentions their basketry, mat houses as well as tepees, greater use of deer and small game and seeds than of bison, and half-hearted Sun dance. The natural or ecological boundary between Northern and Southern Plains may be conjectured to have lain nearly at Pike's Peak and just north of the "9 Shantz and Zon, 18. 50 The Northern Arapaho, in governmental times associated with the Shoshone on the Wind River Reservation, are known as "sagebrush people" among the former and present Arapaho divisions. 51 1922, p. 220. His map includes in this western "Plains" border, Gosiute, Bannock, Flat- head, with Nez Perc6 and Kootenay on both sides of the boundasry. His 1914 map leaves Kootenay, Flathead, and Nez Perc6 in the Plateau, but comprises the Bannock, Gosiute, and even Southern Paiute in the Plains. The Sarsi are mentioned in both publications as among the typical tribes constituting the culture center, but are not so indicated on the map. 82 Kroeber: Cultural and Natural Areas of Native North America Arkansas headwaters. From here south, the Shantz-Zon map shows a belt of pifnon-juniper woodland-a characteristic Basin-Southwest association-in- tervening between the grassland and the pine forests of the higher mountains. To the north, the forest meets the plains, except where the sagebrush extends out into the level land in Wyoming. If this ecological indication held for hu- man occupation, the southern limit of the Cheyenne and Arapaho should have lain a little farther north than is shown by the map, which is based on Mooney's reconstruction for 1832. If the upper Arkansas at an earlier time could be attributed to the Kiowa or some other Southern Plains tribe, the ecological- cultural fit would be exact. On the northeastern flank of the plains, Wissler recognizes the Plains Cree, Plains Ojibwa, and perhaps part of the Assiniboin as possessing many traits of the forest tribes.' A glance at the map shows the first two as mere border fragments of the great northern forest Cree and Ojibwa groups. Both are said to have pushed westward in the historic period, at the expense of Atha- bascans and Dakota. Their entry into horse culture was probably part of the same movement. The Cree and Ojibwa moved out into tall grass or prairie or poplar savanna, however, not into the true plains, and seem never to have lost contact with the woods and their kinsmen therein. The people whom they crowded were the Assiniboin. Even at that, half or more of the territory credited on the map as remaining to the Assiniboin was in the prairies. The Assiniboin, then, are a people only partly in the true plains in the recent period, and perhaps not at all in them formerly. This is confirmed by their close dialectic affiliation with the Yankton-Yanktonai Dakota, who are a prairie people. On the other hand, the Mandan, Hidatsa, and Arikara, the three "village tribes," farmed and built earth houses, but lived in the short-grass area. Their territories as shown on the map exaggerate the situation, since they are mostly hunting range. The settlements of the tribes lay on the Missouri, not far west of the prairie. Also, not far downstream, the prairie swings westward across the Missouri to take in most of the Niobrara. If the Mandan had come up the Missouri from a little farther than they have been traced,' or if they had come a short distance straight west, they would have come out of prairie. The Ari- kara, in the light of their close speech relationship to the Pawnee, may be assumed to have moved out of the prairie fairly recently. Here, then, we have something special: agricultural prairie tribes who entered the plains but re- tained their prairie culture. The cause is not clear, but it was evidently not the horse nor wholly the lure of the bison. It may have been hostile pressure from downstream or the east; or a mere experiment, before or after the horse. Certainly it was an only half-successful experiment once the neighboring tribes got their horse culture fully under way, if the rapid wasting away of the three village tribes after 1800 is an index. Also, the three village tribes did not need extensive farm land and planted in bottoms, so that it mattered little to them whether the rest of their range lay in short or tall grass. 1922, p. 222. 53 The mouth of the White River, in South Dakota. 83 4University of California Publications in Am. Arch. and Ethn. The true Plains areas, then, may be classified as follows: 5a. Southern Plains: Comanehe, Kiowa (including the Kiowa-Apache). 5b. Northern Plains: Arapaho, Cheyenne, Teton Dakota, Crow, Atsina, Blackfoot-Blood- Piegan, SarsL 6. PRAIRIE AREAS The prairie peoples are more difficult to classify than are those of the plains. Just as the prairie shades through river-bottom woodland eastward into park- land and deciduous forest, so with the culture. When the Northern Plains horse culture approached its nineteenth-century climax, reflexes from it penetrated the Prairie cultures, which were already crumbling under American pressure. General studies based on intensive ethnological field work deal almost wholly with Northern Plains tribes, virtually all of whom are monographed, whereas on the Prairie side there is practically but one-the Omaha. Roughly, the Prairie tribes correspond to the fourteen agricultural ones listed by Wissler as on the eastern "border" of the heart of the "Plains" area.' From these, however, the Wichita must be eliminated; the Osage, as already mentioned, are doubtful as a timber people with possible Southeastern lean- ings; Pawnee culture seems sufficiently distinctive to warrant its being set apart, as discussed under the Caddo. With the Osage counted in, this leaves twelve Prairie tribes or tribal associations, all of them Siouan with the ex- ception of the Arikara. These may be subdivided into three groups: one ("Center") consisting of the Santee and Yankton-Yanktonai Dakota; a second ("Village"), of the Mandan, Hidatsa, and Arikara; and a third ("Southern"), of the southern trans-Mississippi or so-called "Central" Siouan tribes.' Still farther north and northwest are the Assiniboin, Plains Ojibwa, and Plains Cree, already mentioned as not in the short-grass plains. The prairie here swings westward at the expense of the plains. Probably all Ojibwa and Cree were timber people in native times. The fur trade and firearms stimulated them to flow westward, the Cree penetrating far into Athabascan territory. Some got out into the plains with the horse and stayed there. These are our Plains Cree and Ojibwa. The Assiniiboin, too, seem to have flowed westward when they got horses.' It was evidently from them, and possibly from the Blackfoot, that the Cree and Ojibwa bands who had spilled into the open prairie got their tepees and other elements of "Plains" culture, while the more westerly of the Assiniboin in their turn were being affected by the active cul- ture developing on the northern true Plains. This northernmost prairie area is therefore in its cultural history directly marginal to the woodland, perhaps more dependent on it than are the prairie areas to the south. Moreover, the forest to which it clings is coniferous and unfavorable to maize; that with which the more southerly prairie areas were in relation is deciduous and gen- erally profitable under maize cultivation (map 27). To the east of the central and southern Prairie areas lay two others which u The American Indian, 1922, p. 220. 55 "Central" with reference to the stock as a whole, "Southern" with reference to the current concept of the Plains area. Im Boas, BAE-R 41, map, carries their territory before 1800 westward up the Baskatche- wan and Athabasca rivers to the Rockies. 84 Eroeber: Cultural and Natural Areas of Native North America were in close relation with them: Wisconsin and Ohio Valley. The former is the wild-rice district west of Lake Michigan. It happens that we possess good studies of three groups in this area, the Menomini, Winnebago, and Sauk and Fox. Their culture shows marked resemblances to the Prairie culture. The Ohio Valley area seems less similar. This is surprising, for several reasons. The Wisconsin area was wooded; the Illinois and northeast Indiana parts of the Ohio Valley area were prevailing prairie. Illinois lies between Wisconsin and the Southern Prairie area. Part of the Santee group of Dakota lived in the forested wild-rice area. It might therefore be expectable that the Central (Dakota) Prairie affiliated with Wisconsin, the Southern (Dhegiha, Chiwere) Prairie with Illinois; which seemingly is not what occurred. The legendary movements of the Dhegiha and Chiwere down and out of the valley of the Ohio would raise similar expectations. The factors concerned with these anom- alies will be touched upon again in connection with the Illinois-Ohio area. It is, however, clear that the prairie cultures three hundred years ago were connected more closely with the woodland ones to the east than with those of the plains on the west. Their bison hunting and tepees and travois were an- cillary. Many parts of the prairies contained a fair amount of woodland; some of the tribes reckoned as of the prairie group actually lived rather in the forest; and one of the woodland culture areas was part prairie. The tall-grass tracts, in short, were culturally associated with the woodland; no doubt be- cause the basis of both culture and subsistence had been worked out in pre- vailingly wooded territory, with agriculture. When bison exploitation through the horse developed a new primary subsistence type on the plains and caused a culture with new emphasis values to evolve there, the prairie tribes were af- fected because their habitat was sufficiently similar. Previously, the similarity in ecology had counted for less because the true plains were too extreme an environment for the thriving of cultures evolved in and primarily adapted to a generally wooded habitat and following farming. The situation in the Prairie area, then, is this: 6a. Southern Prairie or "Central Siouan" subarea: Kansa, Missouri, Oto, Omaha, Ponca, Iowa, perhaps Osage; Pawnee a separate unit with Caddo-Southeast relations. Deciduous park and bottom land; settlements and farms usually attached to this; houses earth covered; patrilineal, exogamie, totemic sibs and moieties, spatially grouped in theory; Sun dance mostly absent; well-defined tribes; notieeable resemblance to Wisconsin area culturally. 6b. Central Prairie subarea: Santee and Yankton-Yanktonai groups of Dakota. Aflia- tions of closely related ethnic groups, or tribes expanded into quasi confederacies still loosely cohering; social organization loose; resemblance to Wisconsin tribes in subsistence habitus rather than formal culture. 6e. Village Prairie subarea: Mandan, Hidatsa, Arikara. Compact village tribes, with earth lodges in palisaded enclosures, in the historic period in the plains rather than prairies, some of them matrilineal; agricultural; possessing age-graded societies; evidently an islet de- tached from its former habitat and cultural affiliations; of composite origin, Mandan and Hidatsa belonging to different Siouan divisions, and Arikara being Caddoan.57 6 The historic nucleus is undoubtedly Mandan-Hidatsa. Arikara speech is practically Pawnee. They must therefore be a recent Pawnee offshoot. Joining the Mandan and Hidatsa, they became somewhat assimilated to them, and probably even more associated in the minds of travelers and ethnologists than in fact. For instance, they have no age-graded societies. 85 8University of California Pusblications in Am. Arch. and Ethn. 6d. Northern or Canadzian Prairie subarea: part or most of the Assiniboin, and Ojibwa and Cree recruits-Algonkin timber people and a Dakota offshoot driven by Dakota hos- tility into Algonkin affiliations. Prairie adjacent to northern coniferous forest and poplar parkland, draining into the Arctic instead of the Mississippi; eastern relations closest with peoples who farmed little or not at all; late strong superficial influences from the Northern Plains. SUMMARY OF TRIBAL HISTORY IN THE PLAINS-PRAIRIES The outlines of tribal history in the plains and prairies, before the first Cau- casian influences made themselves felt, say about three to five centuries ago, may be tentatively reconstructed as follows. On the west, a series of tribes lived in the foothills and broken country in front of the Rockies, utilizing also the ranges behind and the plains before them, according to season, occupation, and need. Their primary cultural affilia- tions are likely to have been Intermountain. They consisted in the south largely of Athabascans. The Kiowa may have been among them, or northward. Still farther north, where the lower timber is pine instead of juniper or scrub, were Algonkins representing two drifts, both ancient, but the Arapaho-Atsina older and probably more southerly than the Blackfoot. The Sarsi may not yet have come out of the northern woods to join the Blackfoot. The Crow may already have left the Hidatsa to live at the foot of the western mountains; but this shift may not have taken place until somewhat later. In the sagebrush plains of Wyoming, behind the Laramies and Big Horns, and perhaps in the moun- tains to the north, were Shoshone. On the south, Caddoan groups extended up the Red and Canadian rivers far enough, probably, to abut, in the seasonally visited short-grass plains, on the Athabascans. South Texas groups like the Tonkawa were perhaps too pre- dominantly a woodland or scrub-timber people to participate with much im- portance in these contacts. Of the Caddoans, the Pawnee-Arikara branch had begun to drift northward, perhaps had already passed out of the woodland of Oklahoma-Arkansas-Missouri into the timber-streaked prairies of Nebraska, but maintained successfully the essentials of their rather complex culture. On the east there were mainly Siouan tribes. The Chiwere group-Iowa, Oto, Missouri-clung most rigorously to the woodland. The Dhegiha, if not already divided, split soon after, with the Quapaw and Omaha-Ponca as ex- tremes: the former hugging the forested Mississippi, facing southward, and reintegrating more closely with the Southeast-Lower Mississippi culture; the latter ascending the Missouri, trending westward into more open country, and beginning to diverge from their old woodland culture. The Mandan and Hidatsa were already in the open, perhaps less far north than later and still cultivating prairie rather than plains soil. Their specific tribal histories were diverse though roughly parallel and later joined and assimilated. The basis of their culture may have been southern-Pawnee-Caddo-in type, more than eastern-Central Siouan. They had perhaps been detached longest from the central body of the Siouan stock. North of the Chiwere were the Dakota: the Teton probably in timber-interspersed prairie, the other divisions mainly in the woods. The Assiniboin perhaps had not yet begun their quarrel with the 86 Kroeber: Cultural and Natural Areas of Native North America othler Dakota which ultimately led them into a separate history. Somewhere in the vicinity, more or less west of the Dakota and south of the Assiniboin, and presumably in prairie, are likely to have been the Cheyenne, already de- tached from the main Algonkin body in affiliations and probably in territory, and not yet in serious contact with Arapaho or Blackfoot across the other side of the plains. Cree and Ojibwa were still wholly woodland peoples. Some of these situations and conditions may of course have fallen earlier than others. It is impossible to assign any precise date for most of them. The intent is only to present the general pre-Caucasian picture. In the seventeenth century the horse began to come in; at first locally, and with little influence. By 1700 it had definitely affected some tribal cultures. By 1750 it had become in some measure universal,' and the historic plains- bison culture was getting into full swing. By 1800 it was flowing vigorously out of the plains and heavily overlaying both the Prairie and the Intermountain cultures, and even the margins of the Southwest. The peak may have been reached only as late as the early or middle nineteenth century. As soon as the horse made the plains desirable, a drift into them began from all sides. Contributing factors along the eastern front, at least locally, were the pressure of white encroachment, of tribes equipped with firearms, the west- ward shrinkage of the bison. Thus tribes that had previously met only at long range, perhaps not at all, were thrown into close and often intimate contact: the Teton and Cheyenne with the Arapaho and Blackfoot, for instance. The Arikara moved northwestward until they found a stay with the likewise seden- tary Mandan and Hidatsa. Roughly about these village tribes there revolved the greatest turmoil of new contacts, clashes, readaptations, and impartings. To these changes the villagers contributed, and they were not uninfluenced by them. As old settlers, they were not torn from their anchorage of maize fields, pottery, domed houses, palisades, matrilineate. But they became an increas- ingly smaller factor in the total situation as the new growths flourished around them. Farther south, the Pawnee, a larger unit, perhaps effected a better adap- tation, except for earlier demoralization by white contacts. Still farther south, the prairie narrows, and the culture of the woodland peoples had been too much undermined by French and Spanish contacts and conflicts for them to be able to shape anything notably novel. About 1700 a large part of the Sho- shone broke away from their Wyoming sagebrush, followed the front of the Rockies southward, and, as the Comanche, drove the eastern Apache back into the mountains or the Texas scrub, confirming them as marginal Southwestern- ers instead of the dominant southern Plainsmen which they might otherwise have become. In the far north, Cree and Ojibwa bands were evidently among the last tribes to try to enter upon a plains-prairie type of career. Of rituals, the Sun dance evidently represents a relatively recent develop- ment in the plains proper, which flowed eastward into the prairies with dimin- ished intensity, and crossed the Rockies late and to a still less degree. Whether the Sun dance is an agglomeration around an old Arapaho nucleus, or whether '" F. Haines, The Northward Spread of Horses among the Plains Indians, AA 40:429-437, 1938, gives the latest data, which roughly confirm my generalization. 87 88 University of Californmia Publicationw in Am. Arch. and Bthn. this people merely were the most active syncretists for a century or two, is harder to say. Age-graded societies appear to date back to the older stratum of culture among the village tribes and were taken into the historic Plains cul- ture by only a few groups that had long lived in or at the edge of the plains proper. The history of the ungraded society type of ritual organization is more obscure, but the region of development apparently was the southern prairies. The bison was exterminated by the Caucasian with Indian aid. Whether the Indian alone, but equipped with horses and guns, could have lived indefinitely off the animal, is an open question. It is entirely conceivable that even then he might have destroyed the species in a century or so. Once the balance turns against an animal, its decline, at first almost imperceptible, is known some- times to increase with almost incredible rapidity; especially has this been observed of game too large to seek hiding. Before the horse, difficulties of trans- port, water, and shelter in the plains allowed the Indian merely to nibble at the existence of the bison, so that the perpetuation of the species might have gone on indefinitely. It might easily have been different, however, with a very simi- lar species in a different habitat; say the foothills of the Rockies, which lacked, so far as purely native culture was concerned, the inhospitability of the open plains. A species adapted to such an environment might have met the fate of the historic buffalo of the plains almost as quickly in native times, once certain groups centered their subsistence on it. And such an event could as well have occurred a hundred as a thousand or ten thousand years before Columbus. That the Folsom bison belongs to an extinct species is, of itself, no reason for placing its human hunters into a past geological age. In its foothill range this animal might have been exterminated at a relatively late period by the very same populations whose descendants, with the help of horses, guns, and white men, terminated the plains bison. And with the animal gone, their culture would have had to end by altering or betaking itself elsewhere, thus perhaps appearing also to be more ancient than it really was. 7. WISCONSIN OR WILD-RICE AREA West of Lake Michigan in Wisconsin, and extending northward to Lake Supe- rior to include adjacent parts of Michigan and Minnesota, there lived at the beginning of the historic period an unusual number of tribes: the Siouan Winnebago and some of the Santee Dakota; and the Algonkin Menomini, Sauk, Fox, Kickapoo, Potawatomi and Mascouten,' and probably some of the Ojibwa. The Mascouten lost their identity, the Kickapoo and Potawatomi drifted or were driven out, the Sauk and Fox after a turbulent career moved into the central Siouan prairie; but the Menomini and Winnebago stayed and retained their numbers and old culture with unusual success, and the Ojibwa pressed increasingly into the northern part of the area. The general vegetation maps fail to show the cause of this concentration of population. They give the area as part deciduous, part coniferous forest, with " Some of these Algonkin tribes are said to have been originally between Lakes Michigan and Huron, but even if so, they were estab]ished on the Wisconsin side when the French reached them about the middle of the seventeenth century. Kroeber: Cultural and Natural Areas of Native North America patches of prairie. The coniferous forest is more of the pine type characteristic of the upper Great Lakes than of the spruce-fir association that predominates in the northern transcontinental belt (map 4). Much of the region evidently was covered with a mixture of pine and of the trans-Ohio and Mississippi type of oak association. Livingston and Shreve (map 5) designate most of it as evergreen-deciduous transition forest. As prairie was also present, this was a favorable enough native habitat; but not in any way extraordinarily so in its prevalent plant cover. It was not decisively superior, for instance, in general features to Michigan and Indiana, which were much more thinly populated. The cause of the population density, then, obviously, so far as it was en- vironmental, lay in something which the general vegetation classifications do not represent; and this was wild rice, Zizania, whose utilization Jenks has dis- cussed.' He estimates or quotes the Indian population of the wild-rice district, defined much as at the opening of this section, as 44,500M in 1764, that of Michi- gan, Illinois, Indiana, Ohio, and southern Wisconsin as 31,750. For 1778 the comparative figures are 32,000 and 14,150; for 1822,20,485 and 24,158. Zizania has a wide distribution, and its importance in the region in question must be due to cultural patterning as well as unusual abundance; but it clearly was a subsistence influence of the first order. Jenks believes that the supply becomes quickly exhausted, and that systematic use of the grain therefore could have begun only a short time before the first entry of the whites. However, with rice as a staple plus a fairly favorable mixed general plant cover, the area clearly has been utilized as a favorable Indian habitat since at least the sixteenth or seventeenth century. (It may have been so before. The prehistoric mound district of Wisconsin and the historic wild-rice district overlap, though they lie partly south and north of each othek. See map 15, p. 102 below.) The heart of the area was the Menomini-Winnebago-Sauk-Fox region bor- dering on central Lake Michigan. This is a district more favorable to agricul- ture, on account of a longer growing season for maize (map 27), than any to the west, and of course to the north. Physiographically (map 7), this same region around Green Bay and Lake Winnebago is reckoned as part of a rather uniform area extending through southern Michigan and Ontario to central New York, the "Eastern Lake section" of the Central Lowland. Immediately west lies the section called Wisconsin Driftless-and therefore relatively lake- less and riceless. The east Wisconsin hear-t thus added to the rice of other parts of its area a topography similar to that of favorable eastern regions, plus farming. possibilities superior to those of other districts in its latitude. The cultural affiliations of the area to the Central and Southern areas have been mentioned. 8. OHIO VALLEY This is the area of the drainage of the Ohio, plus Illinois and perhaps most of the southern peninsula of Michigan. In general, this stretch was as thinly populated at the opening of the historic period as the wild-rice district was densely settled. Parts of Ohio, Kentucky, and West Virginia are regarded as 00 BAE-R 19, pt. 2, 1900. 61 The Wild Rice area figures include some Dakota. 89 0University of California Publications in Am. Arch. and Ethn. having been uninhabited. The Illinois held Illinois; the Miami group, Indiana; the western Shawnee, parts of Tennessee and Kentucky. All three were Algon- kin. The first two suffered heavily early in the historic area. The western Shawnee moved northeastward across the Ohio. Delaware, eastern Shawnee, Huron, Kickapoo, Potawatomi drifted into the same general region of Ohio, Indiana, and Michigan in the eighteenth century. In other words, this pre- viously almost empty tract became a temporary refuge for tribes from all the surrounding regions (except due west) who were pressed by white or Indian enemies. That they made a stand here for nearly a century, and some of them held or increased their numbers, proves the habitat a potentially favorable one, and indicates that it was in a temporary depopulation when discovered. The legendary movements of the "Central" Siouans west across the Missis- sippi, and of the Delaware east across the Alleghanies, with the split of the Shawnee into two separated bodies, fall in with the concept of such a depopu- lation. So does the prehistoric Mound Builder culture, which definitely centers in Ohio drainage. In short, three stages are discernible in the history of this area: 1, relatively heavy numbers and an advanced culture of Southeastern affiliations, in Mound Builder time; 2, a scant population with an indecisive culture; 3, an inflow of tribes disturbed, directly or indirectly, by white con- tacts, and proceeding, temporarily, to evolve a partly new, assimilated, hybrid- Caucasian culture. This historic picture explains the chief causes of the apparently greater cul- tural resemblance of. the Southern Prairie to the Wild Rice than to the nearer Illinois-Ohio Valley area. The latter, at its discovery, was in a slump; later, it became a refuge of tribes from elsewhere. Both Prairie areas and the Wild Rice area remained relatively unaffected by these fluctuations and retained their common elements, at any rate until affected by the horse and firearms. Another factor probably is the sources of information. We have good mod- ern ethnological studies of the Omaha, Winnebago, Menomini; not one of any earlier or later Ohio Valley tribe. With comparable data, this area, especially in its western part, might seem less aloof. The Illinois would be particularly important to know something about in this connection, because their territory, and part of that of the Miami group, lay chiefly in prairie or parkland. Harshberger and Shelford designate most of Illinois as oak savanna, Livingston and Shreve as deciduous forest-grass- land transition, Shantz and Zon as prairie with broad oak tongues following the streams (maps 2-5). There may actually have existed the closer cultural relationship of Illinois with the Prairie areas which the ecological similarity would suggest. An exact scrutiny, from the modern comparative angle, of all available data on the Illinois might conceivably transpose them from the Ohio to the Prairie culture. It seems desirable, accordingly, for the early historic period, to divide this area into: a. Ohio Valley proper: Western thawnee, Miami, perhaps Potawatomi; later, other tribes. b. Illinois: the Illinois. 90 Kroeber: Cultural and Natural Areas of Native North America The prehistory of the Ohio Valley must have been one of the most interesting as well as important in North America. Unfortunately, most of the archaeo- logical work in this area has been done with rather little interest in broader culture-history problems. Consequently the rich data have been organized with reference to local interest, if at all, and when wider interpretations have been attempted they have been speculatively unsubstantial. As rich a culture as that of the Mound Builders must have embraced traceable variants of both district and period. The latter we cannot yet specify with certainty. Pre- sumably the basis of the culture type as a whole was related to that of the Southeast; but on this there grew fairly notable local superstructures, which temporarily equaled or surpassed the Southeastern development. When the population, ethnic organization, and luxury culture growths decayed in the Ohio Valley, the Southeast reemerged as dominant-perhaps was strength- end by the reflux. Some of the areas adjacent on other sides-Prairies, Wild Rice, Lower Great Lakes-also absorbed and retained some portions of Mound Builder culture, to their own enhancement. In the area itself, on the other hand, the destructive tendencies, once in the'ascendant, seem to have run their full course, until the heart of the old Mound Builder region was a low-pressure spot, culturally and populationally. The legendary southwestward movement of the Dhegiha-Chiwere Siouans may have been part of one of the last phases of this period of evacuation and decay. It is tempting to think of the Mandan, Hidatsa, and Winnebago as similar emigrants; but it would be speculative to follow this idea out until a clearer picture of Mound Builder culture is avail- able. At any rate, while Siouan tribes may have flowed out, by the time of dis- covery Algonkin ones had flowed in (or possibly remained), but in a thin layer, and, as an almost inevitable corollary, with a relatively uncharacterized, low- level culture. There is of course no implication in the foregoing of anything mysterious or abnormally advanced in Mound Builder culture. Its type and level, as already said, were in general those of the early historic Southeast. But the size of some of the earthworks, their configuration, the quantities of copper and pearls owned in certain localities, the quality of some of the decorative art, all argue that the culture, whatever its origin or level, at one time enjoyed a transient florescence of rather high degree. These matters will be reverted to in a subsequent section on Eastern archae- ology. 9. LOWER GREAT LAKES The Lower. Great Lakes area coincides with the main or northern Iroquoian block of our linguistic maps. It takes in all the tribes of this territory: Iroquois, Huron, Tionontati, Neutral, Erie, perhaps Conestoga-Susquehanna. Except the last, these are all in middle St. Lawrence drainage, whose watershed defines the area. The territory is that of the St. Lawrence River itself except at its mouth-from about Montreal up, in the period of settlement; Lakes Ontario, Erie, and St. Clair; and the southeastern shores of Huron. The area is a vegetational as well as physiographic unit: deciduous forest, 91 Univer8ity of California Publications in Am. Arch. and Ethn. in part with coniferous admixture, and shading in the north into prevailing evergreen. The several maps differ somewhat in their vegetation subclasses and in the allocation of these, but agree in regard to the general facts. It is noteworthy that the whole of what Malte calls the "Carolinean" province of Canada falls within this area. Even the most northerly tracts of the Lower Great Lakes area lie south of the great northern transcontinental coniferous belt. Their evergreens are hemlock and pine rather than the fir and spruce of the north. The area is the heart of Otis Mason's St. Lawrence-and-Lakes ethnic en- vironment, and one of the four subareas of Wissler's Eastern Woodland. Besides speech, culture is fairly differentiated. It is marked by emphasis on institutional rather than religious or technological developments: consistent matrilineate, strongly fulnctioning sibs, a tendency to co6rdinate and organize these as well as tribes into functioning quasi-political bodies. The Iroquois league was the most successful in historic times, perhaps largely owing to accidents of Caucasian relations. The purely native basis of this league was present in the other Iroquoian c6nfederacies, and lagged little if any behind the degree of development of the Southeastern confederacies. In material cul- ture there were Iroquoian specializations, none of a high order, in pottery, pipes, house types, and so forth; possibly a somewhat greater emphasis on farming than elsewhere in the same latitudes, on account of a somewhat longer and surer growing season (map 27). Resemblances between the Iroquoian and Wild Rice areas seem not to be specific so much as due to elements and trends common to the whole region east of the Mississippi. The position of the Conestoga is doubtful. Their habitat was in Middle Atlantic Coast drainage. They broke up so early, that their culture is only sketchily known. 10-12. ATLANTIC COAST AREAS As far north as the Muskogian tribes extended, a little beyond the Savannah River, the Atlantic coast can be assigned to the Florida and Southeast areas. Beyond, a new province is entered, as indicated not only by a change of prev- alent speech to Siouan, but also by the lower degree of coheMiveness and size of the ethnic units and consequently less successful resistance to Caucasian encroachment; although it must be admitted that the English attitude toward natives was also less tolerant than that of the French and Spaniards. Probably for the same reason, they were far worse ethnologists, with the result that, the native life having long since been crushed, we know comparatively little of the Atlantic Seaboard cultures. The whole region from South Carolina to the mouth of the St. Lawrence is fairly uniform as an environment except in temperature. The precipitation is much the same. There is neither high nor bold relief. The slope from the Ap- palachian ranges to the shore is of about the same width, and the length and size-of the parallel rivers therefore approximately equal. The coast, being low and tempered by the ocean, has generally a more southerly type of plant cover 92 Eroeber: Cultural and Natural Areas of Native North America than the piedmont, and this often differs in the same way from the Appa- lachian ranges. The vegetational belts thus stretch northeastward, and die away in a tapering strip as they meet the north-northeastward-trending coast. This is shown clearly on the Shantz-Zon map (no. 4), which carries finer dis- tinctions of the plant cover than the others. The southeastern pine extends along the coast as far as Cape Hatteras; the piedmont pine-and-oak forest, to New Jersey; the oak-chestnut hardwood forest of both sides of the Appalach- ians, to Rhode Island; the birch-beech-maple-hemlock association, to southern Maine; then comes the northern spruce-fir-although with deciduous admix- ture, since from the Canadian point of view Malte (map 5) reckons everything south of the Gulf of St. Lawrence as "hardwood" in contrast with the great Subarctic evergreen forest beyond. It will be seen that there is variation from prevailing coniferous to deciduous and back to prevailing coniferous forest, without any sharp breaks, and with probably a preponderance of deciduous character-though this deciduous character is not so marked as in the region betweem the Appalachians and the foot of the Rockies. What is constant is the forest cover. There is some marsh along the shores; but no natural true grass- land, even in patches of any considerable size. As might be expected, a setting as uniform as this produced no sharply differentiated cultures. The chief differences are in the intensity and success of maize culture, as this depends on length of frostless summer and conse- quently on latitude and nearness to the sea; the resultant density of popula- tion; and relative distance from more advanced cultural centers, especially the Southeast. It will be convenient to distinguish three cultural provinces. One extends north to the Potomac; another to New Hampshire or southern Maine; the third lies beyond. 10. North Atlantic Slope.-This is an Algonkin area, containing the Abnaki and Micmac, perhaps also the Pennacook, and about coterminous with Maine, Nova Scotia, and New Brunswick. The culture was simpler than in the next area, in dependence on its nonfarming subsistence basis. Maize was grown, but only to a subsidiary extent, being at the limits of its cultivability. 11. Middle Atlantic Slope.-The Middle Atlantic Slope tribes were also all Algonkin. They were the southern and central New England tribes from the Pennacook south; the Wappinger and Mahican; the Delaware; and perhaps the Conoy and Nanticoke. Of these, the Delaware evince some traditional, linguistic, and cultural indications of a western, trans-Appalachian origin. The Conestoga-Susquehanna may have belonged with this area or in the Iro- quoian Lower Great Lakes area. The inclusion of the Pennacook is doubtful. The Handbook of American Indians inclines to group them with the southern New England Indians. Michelson's Algonkin linguistic map puts them with the Abnaki. Their his- toric affiliations since warfare with the English settlers in the late seventeenth century have been with the Abnaki. These affiliations may disguise an earlier leaning toward the south. The Conoy and Nanticoke may belong with the next area. 93 JUniversity of California Publications in Ain. Arch. and Ethn. The culture of the Middle area was built around farming; but it was not one of intensive trends. If Mooney's computations are right, the population was rather denser than in the areas to the south and inland, and in the coastal stretch between New York and Boston it was heavier than anywhere east of the Rockies. 12. South Atlantic Slope.-This area includes the eastern Siouan tribes; a few Iroquoians, notably the Tuscarora and Nottoway; the little known North Carolina Algonkin; and the Algonkin Powhatan. Speck has set the latter off as constituting a distinct cultural subprovince.' The vegetation of their area is largely of piedmont type, although they lived in a tidewater district. The country of the Carolina Algonkin is one of swamp forest, marsh, estuaries, and wide, shallow sounds or bays. It is likely that they had modified the general culture of the region so as to make it accord with -their special subsistence requirements. The rest of the area, that of the Siouan and Iroquoian tribes, is probably divisible into a Lowland and a Piedmont subarea, fairly coincident with the southeastern pine and oak-pine ranges of Shantz and Zon. Too little is known of the culture to press the validity of these subareas, though they may be provisionally listed as follows: 12a. Piedmont 12c. Carolina Sound 12b. Lowland 12d. Virginia Tidewater On the whole, there is little to indicate strong specific influencing by the Southeast, although at the border culture probably shaded over continuously. Tribes as far south as the Catawba were in relations, though of hostility, with the Iroquois rather than with the Creeks. The Tutelo and Tuscarora sought refuge with the Iroquois. These facts indicate a northward outlook of the na- tive culture-a sense of community along the Atlantic slope rather than with the Southeastern area. So, too, there is little trace of Mound Builder resem- blances and influences; whereas as soon as Georgia is entered, these appear." Speck" classes the Powhatan culture definitely as Southeastern, and cites an impressive list of specific cultural resemblances. However, he analyzes the situation in terms of a contrast between a Muskogian-Siouan Southeast and an "older northern Algonkian" culture. He then has the Maryland-Virginia- North Carolina tidewater invaded by Algonkins from the north, who assimi- late the Southeastern culture and pass some of it on to their northerly kinsmen as far away as New England. This is a hypothesis involving a combination of ethnic and cultural considerations. One would expect Virginia culture to be more similar than Massachusetts or Maine culture to that of Georgia. But it seems an undesirable simplification of the situation to explain it wholly in terms of two original, contrasting cultures of Creek and Abnaki type. There seems no specific reason for believing that such a cultural discontinuity ex- isted more strongly in the prehistoric past than in early historic times. 2 AA 26:184-200, 1924. He includes the Conoy and Nanticoke of Maryland with the Pow- hatan culturally. I have hesitantly put them with the Delaware in the Middle Atlantic Slope. "8 The same. See also maps 15, 16, pp. 102 and 104 below. "The same. Swanton, as cited in the previous section on the Southeast, holds the same view. 94 Eroeber: Cultural and Natural Areae of Native North Americwa 13. APPALACHIAN SUMMIT The Cherokee are difficult to place. Their culture had hybridized through in- direct Caucasian absorptions before their territory was seriously penetrated. It seems to have been a rather anomalous culture. Specific Southeastern traits are not strikingly to the fore. The impression that the Cherokee are South- eastern appears to be partly due to the similarity of their and the Creek his- toric fortunes. Both groups prospered in comparative peace with the British until about Revolutionary times, fought the Americans stubbornly, and under- went analogous social and organizational transformations and removals to Indian territory. Nor do the Cherokee seem to show specially close relations with the Ohio Valley people nearest them, the Shawnee; with the Siouan tribes of the Atlantic slope; nor with their Iroquois kinsmen in the north.' Their situation evidently accounts for this aloofness. They occupied the southern and highest part of the Appalachian system, where this ends rather abruptly and falls into the piedmont and plain of the Gulf slope (map 17, p. 121). They are, with the possible exception of two or three obscure eastern Siouan tribe- lets, the only native people in the eastern United States that lived in a true mountain habitat. Their settlements, of course, were in the valleys among and about the mountains. But the way in which these settlements and the claimed territories clustered around the massif shows that this was the dominant ele- ment in their relation to the landscape. Among other eastern tribes, mountains were incidents, borders, hunting grounds, or waste areas in their territory; among the Cherokee, the mountains were the structural backbone of their habitat. The higher parts of their land have a vegetation cover characteristic of the latitude of central New York, with enclosed elevated islands of the type prevalent in Maine, according to the Shantz-Zon map (no. 4). It would be strange if the inhabitants of such a region resembled very closely those of the warm Gulf peneplain. While it is difficult to allot the Cherokee primarily to one or another of the three areas surrounding them-Gulf Slope, Atlantic Slope, or Ohio Valley- this very difficulty brings out a fact that is probably of historical significance: the importance of the Appalachian system as a secondary line of culture cleavage. NORTHERN AREAS The whole north of the continent except its shores and a belt of tundra is a great coniferous forest occupied by Algonkin and Athabascan peoples. These were perforce nonagricultural, the climate being subarctic and wholly un- adapted to maize. Subsistence was therefore by hunting and fishing. As early as the seventeenth century the fur trade began to bring a readaptation, which spread gradually westward. It became more and more profitable for bands to become dependent on trading posts. They gave furs and received traps, fire- arms, tools, trinkets, and provisions. Their meager specific culture was there- fore already affected when the first modern ethnological studies were made es Cf. Swanton BAE-R 42:712, 1928; though he classifies the Cherokee as culturally mar- ginal to the Creek. 95 6University of California Publications in Am. Arch. and Ethn. among them. On the other hand, the demand for furs encouraged them to maintain their hunting habitus. In the actual food consumption, flour and pork came to constitute a growing proportion; nevertheless, the long-run ef- fect of Caucasian contacts was to entrench these peoples more firmly in their occupation as hunters. The interior of Alaska was the last region to be reached by these influences: in some of its parts the miners' irruption at the close of the nineteenth century was the principal factor that determined the new order. But in the main the transmutation proceeded rather uniformly over the whole region. Underlying this recent uniformity was a considerable one of native culture, and, below that, of ecology. The northern forest is substantially one from Alaska to Newfoundland. Mason recognized the area as a unit definable in terms of this transcontinental coniferous belt. Wissler did the same when he set up the caribou food area; though he then proceeded to divide this between the Eskimo, Mackenzie (-Yukon), and Eastern Woodland culture areas. This scheme puts the Naskapi and Cree with the Iroquois and Winnebago, and Wissler has to set them off again in a northern subarea of the Eastern Wood- land admittedly very similar in material culture to the Mackenzie area. The awkwardness of this classification is obviated and the true relations are prob- ably best brought out if we follow Mason in basing culture on natural en- vironment and subsistence. A subdivision for convenience is provided by the line between Yukon and Mackenzie drainage and Hudson Bay and Atlantic drainage. This line ap- proximately coincides with the somewhat fluctuating Athabascan-Algonkin boundary. Another division is made by the Height of Land which separates the Hudson Bay from the Great Lakes drainage. 14. NORTHERN GREAT LAKES This is the area of the Ojibwa, Ottawa, and Algonkin proper as distinct from the Cree and Naskapi. It lies generally south of the Height of Land and drains into the Great Lakes and upper St. Lawrence. The Montagnais north of the lower St. Lawrence and Gulf should perhaps be included. This area knew some agriculture, though this was nowhere primary in the subsistence. It was also exposed to direct contacts with the agricultural areas on the south. These circumstances set it off from the more northerly Algonkin area. There is an ecological correspondence which is shown on some but not all of the maps. Shelford, for instance (map 3), includes the present area in his Northern Coniferous Forest. Harshberger (map 2), however, sets off a St. Lawrence-Great Lakes area which extends north to the Height of Land. Malte (map 5) distinguishes a (Canadian) Hardwood Forest province, extending between Lake of the Woods and Nova Scotia, from the Subarctic (Coniferous) province. The other Canadian source (map 4) recognizes first a Mixed Forest, and then an Eastern Coniferous Forest astride of the Height of Land, before the true, transcontinental Subarctic Forest is reached. Wissler's map' of cari- bou distribution points the same way: in the main, the present area is outside ea American Indian, p. 4, 1922; after Grant. 96 Kroeber: Cultural and Natural Areas of Native North America the range of the animal. According to Malte, most of Montagnais territory would fall into the northern vegetation; which agrees with the dialect-group distribution, though cutting across the drainage. There has been a southwestward drift in and near the area. The Iroquoian Huron abandoned the St. Lawrence between French discovery and settlement. Montagnais, Algonkin, and Abnaki flowed in. Ottawa territory now is west of the Ottawa River. The Potawatomi, traditionally of one origin with the Ottawa and Ojibwa, have moved about Lake Michigan in the historic period. The Ojibwa are always represented as having gained ground from the Dakota. If the several statements in the Handbook of American Indians may be ac- cepted literally, the prehistoric Ojibwa were wholly north of Lake Superior and Lake of the Woods, and their entry into the Wild Rice and Northeastern Prairie areas is recent. There seems also to have been a pushing of western Ojibwa northward into Cree territory rather late in the historic period, if the earlier references to the extent of Ojibwa territory can be taken at face value. J. M. Cooper67 gives the Algonkin groups between the St. ILawrence and Hudson Bay a distribution noticeably different from that of Michelson, Swanton, Skinner, the Handbook, and my map 1. He carries the Montagnais northwestward across the Height of Land to James Bay, so as to hold the whole of Rupert River and the lower parts of Nottoway, East- main, and Big rivers. They adjoin the Eskimo, and thus entirely cut off the Naskapi from the Cree. On the other hand, the T8te de Boule form a definite Cree island within Algonkin and Montagnais territory, more than two hundred miles east of any other Cree, and in St. Lawrence watershed. The Cree proper, Cooper has begin only at Moose River and stretch westward in a much narrower band than shown in map 1. For instance, on the Albany he puts them only below the Kenogami. Beyond longitude 900 or 920, their southern limit is not shown. The territory between their southern boundary and the Height of Land he as- signs to the Ojibwa, who extend eastward to the middle Nottoway River. The Abitibi he makes Ojibwa, not Cree. Cooper's line between Ojibwa and Cree coincides rather well with that in map 5 between the Eastern Coniferous and Subarctic forests. 15. EASTERN SUBARCTIC This includes the various Cree divisions, the Naskapi, the Beothuk of New- foundland, possibly the Montagnais. The Plains Cree represent a recent spill- over from the forest into parkland prairie. The boundary of Cree against Athabascan has been somewhat arbitrarily set between the Nelson and Church- ill rivers. This boundary the Cree have overflowed; and, wherever it originally lay, it has fluctuated in the historic period. Some of the Ojibwa have also worked northwestward. Skinner, for instance, puts the Northern Saulteau Ojibwa of today on the head of the Severn River.' It is of interest to compare Michelson's classification of the Algonkin lan- guagese with the cultural areas that have been reviewed. The inference is that whereas tribes occasionally moved into an entirely new habitat, dialect groups tended closely to conform to the cultural-ecological 67 Northern Algonkian Scrying and Scapulimancy, P. W. Schmidt Festschrift, 205-217, 1928; corroborated and extended by personal communication. " AMNH-AP 9:10, 1911. 69 BAE-R 28, 1912. The classification used is that given in the map, in which Swanton participated. The text classifies somewhat differently, with IV of the subjoined table split into a Central and an Eastern subtype. The Central subtype is made to consist of A1-2, 4-6, B, C, and D; the eastern, of A3. 97 University of California Publications in An. Arch. and Ethn. groupings. Subsistence being the same, habitats inclined to remain uniform; this made for close associations, which in turn held speech together. Speck has recently made avaluable addition to our knowledge of Montagnais and Naskapi band distribution and Labrador Eskimo territory, with maps for two to three hundred years ago and the last century.'" This study centers TABLE 4 ALGONKIN DLECT GROUPS AND CuTuRE AREAS Dialect groups I. Blackfoot (markedly distinct)................. II. Arapaho, Atsina (markedly distinct).......... III. Cheyenne, Sutaio (more similar to IV)......... IV. Eastern-Central Algonkin A. Cree type 1. Cree, Montagnais* . 2. Naskapi J . 3. Micmac, Abnaki, Pennacook............ 4. Menomini l 5. Sauk, Fox, Kickapoo |. 6. Shawnee................................ B. Ojibwa type Ojibwa, Ottawa, Algonkin l Potawatomi* J.............. Illinois, Miami............................ C. Massachuset type Southeast New England, Long Island....... D. (Delaware type), position uncertain Mahican, Wappinger, Pequot, Delaware..... V. Uncertain Nanticoke, * Conoy* } Powhatan, North Carolina Algonkin ....... Culture areas Northern Plains (long resident) Northern Plains (long resident) Northern Plains (newcomers) Eastern Subarctic North Atlantic Slope Wild Rice Ohio Valley {Northern Great Lakes Ohio Valley Ohio Valley Middle Atlantic Slope Middle Atlantic Slope {Middle (?) Atlantic Slope South Atlantic Slope I * Indicates that inclusion in the cultural area indicated by me is not certain, but they are included by Michel- son in the dialect groups shown. farther north and east than Cooper's account which has just been referred to, but on the whole agrees fairly well with it. Speck puts Montagnais and Naskapi into one group, as opposed to Cree, thus differing from Michelson's classifica- tion. The Eskimo have apparently receded, whereas Montagnais-Naskapi have advanced eastward and northward for several centuries.' 16. WESTERN SUBARCTIC This is the western half of the great northern coniferous forest. The limit to- ward the tundra is drawn somewhat variously; in many parts the forest be- comes low or sparse, and of course disappears in the higher mountains. In general, however, the tundra is assignable to the Eskimo, even where it extends 'I Montagnais-Naskapi Bands and Early Eskimo Distribution in the Labrador Peninsula, AA 33 :557-600, 1931. n Speck has gone farther in Inland Eskimo Bands of Labrador, in Anthr. Essays, UC, 313-330, 1936. Of particular interest is a list of traits shared by Montagnais-Naskapi and Eskimo. I I 98 Kroeber: Cultural and National Areas of Native North America well inland. The Athabascan tribes whose territories consist mainly or partly of tundra appear to be the Hare, Yellowknife, and Caribou-eater. These, or at least the first of them, seem to constitute a cultural subarea. There are areas of tundralike formation farther west, as between the Mackenzie and upper Yu- kon drainages, and again in Alaska; but these, being due to altitude, may be regarded as mountain hinterlands of tribal territories otherwise more or less forested. Toward the plateau and coast some border subareas have apparently to be set off. The Carrier in upper Fraser and the Babine in upper Skeena drainage have already been mentioned as of doubtful affiliation between the Fraser and the present region. The Tahltan and Taku-tine, back of the Tlingit, have been influenced by this people and appear in turn to have influenced especially the northern mainland part of the Tlingit. But they may tentatively be re- garded as constituting an Athabascan or Subarctic subarea rather than a northernmost Intermountain one. The Tahltan are in upper coast drainage- on the Stikine; the Taku-tine partly on upper Yukon waters. Both are shut off from the farther interior by the Rockies. They should therefore show some differentiation from the other Athabascans. But as the primary ecological boundary admittedly comes at the Coast Range, they will probably have to be reckoned ag in the main belonging culturally with the interior tribes. It may be added that most of the available plant-cover classifications (maps 2-5) agree roughly in assigning a Rocky Mountain type of vegetation to most of northern interior British Columbia. That is to say, the forest is Western Coniferous, not Northern.C' The ethnic habitats involved in this plant cover are Carrier, Babine, Tahltan, and Sekani, in part or whole. The tentative cultural classification is: 16a. Western Subaretic, main area. 16b. Interior Tundra (Hare, Yellowknife, Caribou-eater). 16c. Upper Fraser (Carrier, Babine). 16d. Northern Plateau Apex (Tahltan, Taku-tine). Addendum on Western Subarctic Osgood has recently given a classification of all northern Athabascansq' which is probably much better founded than my compilation as expressed in map 1. Besides the Sarsi, Nicola, Chilcotin, and Tsetsaut in the Plains, Intermoun- tain, and Northwest Coast areas, he recognizes twenty-one main tribes or na- tions in my Western Subarctic area, grouped into Arctic Drainage and Pacific Drainage major divisions on the basis of culture.7' The areas on his map often differ markedly from those of mine. New tribes appear, while some of those shown by me reduce to subtribes or bands. Though Osgood's essay is tentative, and will no doubt be modified in detail, it represents the first real attempt to organize ethnic knowledge on this vast area. 72 The Dominion map (4) divides the Tahltan and Taku-tine territory between North- western Coniferous and Subarctic forest, the line beginning at about 590 on the Alaska boundary and extending northwestward. 78 YU-PA no. 7, 1936. 7' Jenness, Nat. Mus. Can. Bull. no. 65, 1932, classifies culturally into a Mackenzie-Yukon and a Cordilleran area, with the Kutchin somewhat in doubt. 99 100 University of California Publications in Am. Arch. and Ethn. Arctic Drainage division Chipewyan. Includes my Caribou-eaters Yellowknife Dogrib: 4 groups Bear Lake: 5 groups on Bear Lake Hare, distinct from last, northwest of the lake, to west of the Mackenzie River Mountain, 3 groups, west of Bear Lake, both sides of the Mackenzie Slave, 4 groups, inel. Etehao-tine, on the Slave and Mackenzie rivers Kaska, a large area on the Liard, west of the last Bekani, upper Peace River, south of the last; 4 groups Beaver, lower Peace River, east of the last, south of the Slave (Sarsi, Athabasca River, south of the Beaver; in Plains culture) Pacific Drainage division Carrier, including Babine Tahltan, including Taku-tine. Stikine and upper Taku rivers Tutchone, a large area from 1400 to the continental watershed, and including most of the Taku-, Abbato-, and Etchao-tine territories of map 1, on the upper Yukon affluents Nabesna, on the upper Tanana Han, on the Yukon, 640-660 north latitude, comprise my Hun, but not Kutehin Kutchin, from 1300 to 1500, or from east of the lower Mackenzie to west of the middle Yukon. Elsewhere76 Osgood gives the true Kutchin tribes somewhat differently from Cad- zow, whom I followed in map 1: Nakotcho or Kwitcha, Tutlit, Takkuth, Vinta, Tranjik, Kutcha, Tennuth, Natsit. All other groups are denied as Kutchin, though they may have been so called. Tanana, on the lower Tanana and a stretch of the Yukon, southwest of Kutchin Koyukon, on the Koyukuk and lower Yukon. Include Yuna-khotana of map 1 Ingalik, Eskimo name, lowest Yukon and Kuskokwim: Kayu-khotana and Kalchana; also called Tena Tanaina, distinct from Tanana: the Cook Inlet Athabascans, my Khnaia-khotana Ahtena, Copper River (Tsetsaut, head of Portland Canal: Northwest Coast) (Chilcotin and Nicola, interior of southern British Columbia) An included tentative linguistic classification by Sapir puts eighteen of the languages into nine North Athabascan groups or divisions, seven being left unclassified for paucity of data: 1. Chipewyan, Yellowknife, Slave 2. Dogrib, Bear Lake, Hare 3. Kaska and Tahltan, on both sides of the continental watershed 4. Sekani, Beaver, Sarsi 5. Carrier and Chilcotin 6. Kutchin, the most divergent speech of all 7. Tanaina and Ingalik 8. Ahtena, perhaps distinet 9. Tsetsaut, probably most divergent after Kutchin Most of these divisions differ from one another as much as they differ from Navaho-Apache, it is stated. The New Mexico-Arizona Athabascans, and the Oregon-California ones, each constitute a single well-marked speech unit, to which each of the eight or nine or more northern ones is roughly equivalent in distinctiveness. 75 AA 36:168-179, 1934. Kroeber: Cultural and Natural Arecw of Native North America These facts about speech suggest strongly that the North Athabascans have occupied their territory long enough to diverge heavily from one another. The separateness of Tsetsaut is not surprising: they were a small group among aliens on actual salt water. The Kutchin, however, are surrounded by other Athabascans, except on the north, where they adjoin the Eskimo. Either con- tact with these latter set up disturbances leading to strong specialization, or the Kutchin must presumably once have lived in less contact with their fellow Athabascans or in greater exposure to some alien people. RTATIONS OF EASTERN AND NORTHERN AREAS It is an open question whether the Northern areas should be reckoned as part of the general Eastern tract or co6rdinate with it. They lie pretty solidly be- yond the practicable limits of maize agriculture. This environmental condi- tion has limited the population, stunted the culture, and kept it from making absorptions which otherwise would probably have taken place. It is in fact difficult to name traits specifically characteristic of the eastern areas proper which are also characteristically northern and limited to the two. Moreover, the door was ajar in the north to culture traits tending to seep in from sub- arctic Siberia: toboggan, snowshoe, birch-bark vessels, conical tent houses, cut and fitted clothing, scapulimancy (if not due to French Colonial import). These traits have generally worked across the continent throughout the sub- arctic or Hudsonian belt, but have not penetrated seriously the areas south of it, even where the environment permitted. On the other hand, the Northern areas do not show even a tendency toward a cultural center or culmination, and the transition between them and the Eastern areas is gradual, except for changes resulting from the impracti- cability of agriculture. Thus there is nothing against considering the Northern cultures as primarily a meager and undifferentiated form of the Eastern cul- tures which center in the Southeast. On the whole, this seems best to express the relation. EASTERN ARCHAEOLOGICAL AREAs On the side of pure archaeology there exist a number of distributional classi- fications which bear on the differentiation made in the foregoing pages between the Southeastern, Mississippi Valley, and Lower Great Lakes cultures, on the one hand, and those of the Atlantic slope, on the other. Thomas on mounds.-The first of these classifications is Cyrus Thomas' work on mounds of the eastern United States.76 In map 15 1 have tried to embody his principal regional findings. Wissler has previously condensed Thomas' main map of mound occurrence.' My reduction is somewhat less summary, in that it attempts to show with reasonable accuracy every area containing six or more mounds or mound groups separated from one another by not more than fifteen to twenty miles; more scattering occurrences are omitted.7' I have also added 76 BAE-R 12, 1894. 7 The Relation of Nature to Man, 12, fig. 5, 1926. " This map could have been added to, notably from the publications of Moore, but an exhaustive bringing of it up to date would be an exacting task, without, probably, mueh changing the general inferences to be derived from Thomas' work. 101 1University of California Publications in Am. Arch. and Ethn. the watersheds between Atlantic, Mississippi, and Great Lakes drainages. Further, my map embodies the "districts" discussed by Thomas in his text, namely: 1, Dakotan (or Northwest) with la (Wisconsin), subdistrict of Effigy Mounds; 2, Huron- Iroquois, from Lake Michigan to Quebee; 3, Illinois, including adjacent parts of Indiana, Iowa, Missouri; 4, Ohio, including eastern Indiana, northeastern Kentueky, southwestern West Virginia; 5, Appalachian, about coterminous with historic Cherokee territory; 5a, North Georgia, transitional between the last, the Gulf area, and the next; 6, Tennessee or Central district, including most of Kentucky; 7, Arkansas, down to the Red River, with 7a, Southeast Missouri, as a subdistriet; 8, Gulf, from the lower Mississippi east, with 8a, South Carolina, and 8b, Peninsular Florida, forming probable subdistricts. The delimitation of these districts is given somewhat unfortunately by Map 15. Mound Areas of the Eastern United States; simplified from Thomas. Groups of six or more mounds or mound clusters within not exceeding fifteen to twenty miles of each other shown in stipple; smaller groups and isolated mounds omitted. Mound areas and subareas: 1, Northwestern (Dakotan); la, Effigies or Wisconsin; 2, Huron-Iroquois; 3, Illinois; 4, Ohio; 5, (South) Appalachian; 5a, North Georgia, transitional; 6, Tennessee or Central; 7, Arkansas; 7a, Southeastern Missouri; 8, Gulf; 8a, South Carolina; 8b, Peninsular Florida. Division between Northern and Southern major "sections" shown by xxxx. Atlantic, Gulf, Mississippi, Great Lakes, and Hudson Bay Drainage watersheds shown by dotted line. 102 Kroeber: Cultural and Natural Areas of Native North America Thomas in terms mostly of states or counties, but in general is reasonably defi- nite. His classification is based primarily on the shape, structure, and function of the mounds themselves, but takes cognizance also of interments, pottery, etc. A basic classification by Thomas into a Northern and a Southern "section" cuts without explanation across some of the foregoing districts, southwestern Illi- nois and western Tennessee being thereby separated by him from the re- mainder of the Central district to go with the Arkansas and Gulf districts. This somewhat discordant major classification has been entered on map 15 by a line of crosses. The following conclusions result from Thomas' work: 1. The Appalachian watershed formed an important line of cultural cleavage. To the east, mounds were of shell or other refuse. A few spillings of mound groups eastward over the physiographic boundary fundamentally confirm the division, because in the main these exceptions lie close to the boundary. 2. The Great Lakes and Gulf drainage went with the Mississippi Valley. 3. The lower Great Lakes were set off from the Ohio Valley as a separate district or area. 4. West of southern Lake Michigan was an area of concentrated and specialized mound culture. This was continuous across Wiseonsin, without regard to the Lakes-Mississippi watershed, in contrast to the region east of Lake Michigan, where the watershed delimited cultural provinces. 5. The uppermost Mississippi mound culture extended in some degree to the Red River of the North and perhaps to the middle Missouri. 6. The western frontier of the intensive mound culture was approximately the edge of the forest, though in the north the mounds, and in the south the woodland, extended some- what farther west. The prairie areas of Illinois and Indiana (map 4) were comparatively moundless. 7. The heart of the mound area was the Ohio drainage, together with the immediate val- ley of the lower middle Mississippi. 8. The characteristic mound culture thinned out downstream, according to Thomas, com- ing to an end about Natchez. Lower Louisiana and coastal Texas are represented as outside the culture. This conclusion, however, can no longer be maintained." 9. The eastern Gulf states affiliated with the Mississippi-Ohio area. 10. This Gulf Drainage culture extended into the southerly part of the Atlantic slope, perhaps as far as the Great Pedee, though its most characteristic form ended at the Sa- vannah. 11. Peninsular Florida-the whole peninsula, not its southern half only-formed a dis- tinct subarea. 12. Another distinctive subarea was the South Appalachian district, the intermountain region of upper Tennessee River drainage. Except for the slump in the Ohio Valley from prehistoric to historic time, this archaeological classification agrees well with the ethnological one de- veloped in the present work, even to many details. Holmes on pottery.-Holmes's study of eastern pottery' is also so compre- hensive as to invite comparison. Again, I have taken his basic map, simplified it to dispense with the use of color, and added subareas from his text (map 16) . '9 As a result of recent exploration in Louisiana. In fact, Hopewell culture traits are now recognized in that state (F. M. Setzler, Jour. Wash. Acad. Sei., 23, no. 3, 1933, and USNM- R 82, 1933. See also J. A. Ford, Dep't of Conservation, Louisiana Geol. Survey, Anthr. Study no. 2, 1936, p. 219). Evidently, archaeological work on the lower Mississippi had not been prosecuted in Thomas' time. 80 BAE-R 20, 1903. 103 University of California Publications in Am. Arch. and Ethn. Holmes's areas, or "groups" as he calls them, represent the distribution of pot- tery types, and therefore, as is expectable, sometimes overlap. The subareas are regions or centers of characterization of special types, and are not sharply localized by Holmes. The relation of some of them to the primary groups is Map 16. Pottery Types of the Eastern United States; after Holmes. Major groups (from Holmes's map): 1, Middle Mississippi Valley; 2, South Appalachian; 3, Middle and North Atlantic Slope; 4, Iroquoian; 5, Northwestern (or Upper Mississippi Valley). Subgroups (from Holmes's text): la, East Arkansas-West Tennessee; lb, Southeast Missouri; lc, Cum- berland Valley; ld, Lower Mississippi Valley; le, (Southeast) Texas; If, Gulf Coast; Ig, Florida (Peninsula). Relation of ld, e, f to 1 not clearly defined. 3a, Pamlico-Albemarle; 3b, Potomac-Chesapeake; 3c-5c, Piedmont Virginia and Apalachee-Ohio; 3d, New Jersey; 3e, New England; 5a, Miami Valley; 5b LMS . ........ ............ ... a}......... ,: n t,,1~~~~~~~~~~~~~~~~~~~~~~~~A bY CULTURAL ARZA. ............ L .75- ............ ;: :: ::.:: . ...:::::. :. :: .....Rz2 .::. .:x .z::::: 2-50 1 7 DENSITY: : - ................. .. ... : E 12 S \ ~~~~~~~~~~~~~~~~~~~~~~~~~....... ;::..-...... ZX 50-75 I 0' - 75I. 0 775 Map 18. Native Population Delnsities by Cultural Areas. Compare maps 6 and 28. Kroeber: Cultural and Natural Areas of Native North Aimerica TABLE 7-(Continued) Culture | Tribes | Population Area in Density per Cutue Tie 100 kM.2 100 km.2 ARUTIC cOAST-(Continued) 2a Western Eskimo Mackenzie ........... . 2,800 800 3.50 Nuwuk, Kopak, Nunatak ....... 3,000* 2,036 1.47 Malemiut ........... . 1,600 555 2.88 Kinugumiut, Kaviagmiut ....... 2,800 338 8.28 St. Lawrence Island . ...... 600 50 12.00 Unaligmiut ........... . 1,600 210 7.61 Ikogmiut ...................... 400 278 1.43 Magemiut, Kaialigmiut......... 5,000 491 1.01 Nunivagmiut .......... . 1, 500 45 33.30 Kuskokwagmiut ............... 7,200 416 17.30 Togiagamiut, Chingik, Nushagak 1,300 665 1.95 Ogulmiut ...................... 3,700 511 7.24 2b Aleut Aleut ....................... 16, 000 247 64.70 2c Pacific Coast Eskimo Kaniagmiut .................... 8,800 287 30.60 Chugachigmiut ................. 1,700 262 6.48 Ugalakmiut .................... 800 40 20.00 NORTHWEST COAST la Northern Maritime Mainland Northern Tlingit ............... 2,500 250 10.00 lb Northern Maritime Archipelago Southern Tlingit ............... 7,500 742 10.10 Haida ....................... 9, 800 103 95.10 Tsimshian proper .............. 3,500 110 31.80 ic Northern Maritime River Niska, Gitskyan ............... 3,500 381 9.18 Haisla ....................... 1,2300 80 16.20 2a Central Maritime, Northern Heiltsuk ....................... 1, 400 80 17.50 Bella Coola .................... 1,400 150 9.33 Kwakiutl ...................... 4,500 211 21.30 2b Central Maritime, Southern Nutka ....................... 6,000 91 65.90 Makah, Quileute, Quinault ..... 4,000 62 64.50 3 Gulf of Georgia Comox, Pentlatch, Cowlitz, Lkungen, Seshelt, Squamish, Lower Fraser ................ 20,500 607 33.70 Nutsak, Lummi ............... 800 60 13.30 Klallam, Chimakum ........... 2,400 58 41.30 4 Puget Sound Skokomish, Nisqualli, Twana, Puyallup, Snoqualmi, Snoho- mish, Skagit ................. 6,000 357 16.80 * From here on Mooney gives only three Eskimo aggregates, of 8000, 17,000, and 15,000, for 1740; besides 16,000 Aleut. His total of 40,000 has been allotted according to his tribal figures for survivors in 1900. I I I I 135 1University of California Publicationm in An. Arch. and Ethn. TABLE 7-(Continued) Cuslture Tribes | Population | Area in Density per Cutrea Tries opuatin2 100 km.2 NORTHWEST CoAST-(Continued) 5 Lower Columbia Tlatskanai ..................... 1,600 27 59.20 Lower, Upper Chehalis, Owi- lapsh, Cowlitz ............... 1,200 182 6.59 Chinook ....................... 22,000 148 148.60 Tillamook ..................... 1,500 67 22.30 Yaquina, Alsea, Siuslaw ........ 6,000 83 72.20 6 Willamette Valley Kalapuya ...................... 3,000 334 8.98 7 Lower Klamath Southwestern Oregon Atha- bascans 1-8 .................. 8,800 184 47.80 Kus ....................... 2,000 20 100.00 Takelma ....................... 500 70 7.14 Tolowa (Cal. Ath. 1) ........... 1,000 21 47.60 Hupa, Chilula (Cal. Ath. 2).... 1,500 18 83.30 Yurok ....................... 2,500 19 131.00 Karok ....................... 1,500 32 46.80 Wiyot ....................... 1,000 13 76.90 SOUTHIWEST I. Pueblo Sphere Pueblo Hopi ....................... 2,800 70 40.00 Zufii ....................... 2,500 114 21.90 Keres ....................... 4, 000 120 33.30 Piro ....................... 9,000 85 105.80 Tano, Tewa, Tiwa, Pecos, Jemez 15,500 57 271.90 2a Inter-Pueblo Navaho ....................... 8,000 842 9.50 2b Circum-Pueblo Western, Eastern, Jicarilla Apache, incl. Mex ............ 6,500 5,588 1.16 II. Sonora-Gila-Yuma Sphere 3 Fuerte-Yaqui Lowland Yaqui, Mayo, and other Cahita (481t) 4 Sonora Opata . (847t) Pima in Meico J Papago, Mexico and U. S. 6,600 714 9.24 Gila Pima ..................... 4,000 150 26.60 5 Northern Sierra Madre Tarahumar .................... (715t) 6 Sonora Coast Seri, Guaymas, etc (306t) 7 Northwest Arizona Walapai, Havasupai ............ 1,000 261 3.83 Yavapai ....................... 600 405 1.48 t Areas in Mexioo. Mentioned here only to leave the list of Southwest areas complete. Not considered by Mooney. I I 136 Kroeber: Cultural and Natural Areas of Native North America TABLE 7-(Continued) Culture Tribes Population ~ Area in Density per_ Culture Tribes Population 100 km.2 100 km.' SoUTHWEsT-(Continued) 8 Lower Colorado River Mohave, Halchidhoma, Yuma, Halyikwamai, Kohuana, Coco- pa, incl. Mex ................. 11,000 361 30.40 Maricopa ....................... 2,000 55 36.30 9 Peninsular California E, W Dieguefio, Kamia, in U. S.. 3,000 166 18.10 Dieg., Kamia in Mex., Akwa'ala, Kiliwa, Cochimi, Waicura, Peric .(1,224t) 10 Southern California Desert, Mountain, Pass Cahuilla 2,500 63 39.60 Serrano 1-4 .................... 3,500 293 11.90 Luiseno, Juanenfo, Cupefto ...... 5,500 81 67.90 Gabrielino ..................... 5,000 77 64.90 Chumash ...................... 10,000 169 59.10 INTERMEDIATE AND INTERMOUNTAIN AREAS la Great Basin Ute, Gosiute ................... 4,500 2,917 1.54 Shoshone, W Shoshone, N Pai- ute, S Paiute ................. 7,500 3,062 2.45 Chemehuevi ................... 500 452 1.10 Panamint ...................... 500 236 2.11 Eastern Mono .................. 2,000 144 13.80 W asho ......................... 1,000 62 16.10 lb Snake-Salmon Drainage Bannock, N Paiute, Shoshone ... 3,000 2,886 1.04 lc Klamath Lakes-Pit River Klamath, Modoc ............... 1,200 249 4.81 Achomawi, Atsugewi ........... 3,000 171 17.50 Mountain Maidu ............... 1,000 81 12.30 Id Wind River Wind River Shoshone .......... 2,500 550 4.54 2a California Kato (= Athabascan 7) ......... 500 6 83.30 Yuki, Coast Yuki .............. 3,000 44 68.10 Wintu in Sacramento drainage.. 2,000 51 39.20 Wintun ....................... 2,500 74 33.70 Yana ........................ 1,500 48 31.30 Foothill Maidu (incl. Nisenan) . 4,000 138 28.90 Plains and Foothill Miwok (1-4) 9,000 190 47.30 Costano, Esselen ............... 7,500 163 46.00 Salinan ....................... 3,000 94 31.90 Valley Yokuts ................. 11,000 382 28.70 Foothill Yokuts ................ 7,000 65 107.60 Western Mono ................. 2,000 96 20.80 Tiubatulabal ................... 1,000 58 17.20 Kawaiisu ...................... 500 42 11.90 t Areas in Mexico. Mentioned here only to leave the list of Southwest areas complete. Not considered by Mooney. 137 13nierstity of California Publications in Am. Arch. and Ethn. TABLE 7-(Continued) Culture Tribes Population Area im De100tkm. areas 10k. Denit per2 INTERMEDIATE AND INTERMOUNTAIN AREAS-(Continued) 2b California Climax Pomo ....................... 8, 000 88 90.90 Coast, Lake Miwok (5-6),Wappo 3,000 47 63.80 Patwin ...................... 6,000 96 62.50 Valley Maidu (incl. Nisenan) .. 4,000 49 81.60 2c California-Northwest Transition Nongatl, Mattole, Lassik-Wai- laki, Sinkyone (Athab. 3-6) ... 4,000 71 56.40 Shasta 1-4, Chimariko ......... 3,000 88 34.10 Wintu in Trinity drainage . 1,600 51 29.40 3a Middle Columbia Klikitat,Yakima,Wanapum,Palus 11,200 390 28.70 Nez Perc6 ..................... 4,000 450 8.88 Tenino, Umatilla, Walla Walla.. 2,900 642 4.51 Wailatpu ...................... 500 93 5.37 Wenatchi, Sinkiuse, Peskwaus, Methow,Nespilim,Sanpoil,Col- ville, Spokane (part) ......... 3,500 313 11.20 3b Upper Columbia Wenatchi-Spokane group (part) 2,400 208 11.50 Kalispel, C.d'A., P.d'O., Flathead 2,800 1,861 1.50 Okanagan, Lake ............... 2,200 410 5.36 Kootenay ...................... 1,200 595 2.01 3c Fraser Chilcotin ...................... 2,500 197 12.60 Lillooet ...................... 4, 000 170 23.50 Thompson, Nicola ............. ,150 155 33.20 Shuswap ...................... 5,300 1,176 4.50 EAST AND NORTH I. East la Southeast Stono, Edisto, Cusabo, Yamasi, Guale ...................... 4,j400 113 38.90 Apalachi, Ap'ola, Chatot, Sa- wokli, Pawokti, Pensacola... 12,000 614 19.50 Mobile ...................... 2, 000 100 20.00 Creek ...................... 18,000 1,476 12.20 Yuchi ...................... 1, 500 130 11.50 Eastern Shawnee ............... 1,000 78 12.80 Chickasaw ..................... 8,000 866 9.23 Choctaw ....15,1.000........... 683 21.90 Tunica, Ofo .................... 2,000 206 9.70 Ibitupa, Chakchiuma, Taposa... 1,200 266 4.51 Biloxi, Pascagula .............. 1,000 88 11.30 Houma, Acolapisa, Washa, Cha- washa, Tangipahoa, Bayogula, Kinipisa, Okelusa ............ 5,400 314 17.10 Chitimacha .................... 3,000 94 31.90 l 138 Kroeber: Cultural and Natural Areas of Native North America TABLE 7-(Continued) Culture Tribes 1 Populatio 1 Area in Density per areas 0km2 10m2 EAST AND NoRTH-(Continued) lb Southeast Climax Natchez, Avoyel, Taensa ....... 5,300 277 19.10 le North Florida Timucua ....................... 8,000 678 11.70 2 South Florida Calusa ....................... 3,000 247 12.10 Ais, Jeaga, Guacara, Tekesta.... 1,000 295 3.38 3 (Northwest Gulf Coast) South Texas Atakapa ....................... 1,500 482 3.11 Karankawa .................... 2,800 282 9.92 Tonkawa ...................... 1,600 313 5.11 Lipan Apache .................. 500 980 0.51 4a Red River Caddo, Wichita, Kichai, Waco, Tawakoni .................... 13,400 2,577 5.19 Quapaw ....................... 2,500 680 3.67 4b Middle Platte Pawnee ....................... 10,000 1,306 7.66 5a Southern Plains Kiowa, Kiowa-Apache .......... 2,300 1,682 1.36 Comanche ..................... 7,000 1,400 5.00 5b Northern Plains Cheyenne, Arapaho ............ 6,500 2,111 3.07 Teton Dakota ................. 10,000 1,700 5.88 Crow ....................... 4,000 1,527 2.61 Assiniboin (part) ............... 2,000 343 5.83 Atsina ....................... 3,000 814 3.68 Blackfoot, Blood, Piegan ....... 15,000 3,464 4.33 Sarsi ....................... 700 937 .75 6a Southern Prairie Osage ....................... 6,200 2,260 2.74 Kansas ....................... 3,000 499 6.01 Oto ....................... 900 219 4.10 Missouri ....................... 1,000 552 1.81 Iowa ....................... 1,200 859 1.39 Omaha, Ponca ................. 3,600 300 12.00 6b Central Prairie Santee, Yankton, Yanktonai Dakota ..................... 15,000 2,996 5.01 6c Village Prairie Mandan, Hidatsa .............. 6,;100 225 27.10 Arikara ....................... 3,000 374 8.02 6d Northern (Canadian) Prairie Plains Cree .................... 3,000 1,567 1.91 Plains Ojibwa .................. 2,000 470 4.25 Assiniboin (part) ............... 8,000 1,371 5.84 I I I 139 1University of California Publications in Am. Arch. and Ethn. TABLE 7-(Continued) Culture Tribes Population Area in Density per areas 100 km.2 100 km.2 EAST AND NORTH-(Continued) 7 Wisconsin Winnebago ..................... 3,800 139 27.30 Kickapoo ...................... 2,000 155 12.90 Sauk and Fox .................. 6,500 312 20.80 Menomini ...................... 3,000 255 11.70 Ojibwa (part) .................. 3,000 600 5.00 8a Ohio Valley Miami ....................... 4,500 1,242 3.63 Shawnee (western) ............. 2,000 1,100 1.82 Potawatomi ................... 4,000 919 4.35 Uninhabited 1,381 8b Illinois Illinois ....................... 9,500 3,065 3.09 9 Lower Great Lakes Mohawk, Oneida, Onondaga, Cayuga, Seneca .............. 5,500 734 7.49 Conestoga ..................... 5,000 702 7.12 Erie ........................... 4,000 1,001 3.99 Neutral ....................... 10,000 592 16.80 Huron, Tionontati ............. 18,000 1,392 12.90 10 North Atlantic Slope Micmac ....................... 3,500 1,508 2.32 Abnaki ....................... 3,800 1,777 2.14 11 Middle Atlantic Slope Pennacook ..................... 2,000 267 7.49 Nipmuc ....................... 1,700 125 13.60 Massachuset ................... 13,600 129 105.40 Pequot ..2,200. t 2,200 2 29 75.80 Wappinger ..................... 5,600 192 29.10 Montauk ...................... 6,000 38 157.80 Mahican ....................... 3,000 271 11.10 Delaware ...................... 8,000 454 17.60 Nanticoke ..................... 2,000 122 16.30 Conoy ......................... 2,700 201 13.40 12a, b South Atlantic Slope, Piedmont and Lowland Monacan, Manahoac, Mohetan 2,700 311 8.68 Nottoway, Meherrin ........... 2,200 96 22.90 Coree ....................... 1,000 30 33.30 Tuscarora ..................... 5,000 95 52.60 Occaneechi, Woccon, Sara, Ca- tawba, Eno, Cape Fear, Pe- dee, Sewee, Santee, Congaree, Wateree, Tutelo, Saponi ...... 17,500 1,561 11.20 12c South Atlantic Slope, Carolina Sound Weapemeoc, Secotan, Pamlico.. 4,500 140 32.10 140 Kroeber: Cultural and Natural Areas of Native North America TA3LE 7-(Concluded) Culture Tribes Population Area in Density per area 100 km.2 100 km.2 EAST AND NoRTH-(Continued) 12d South Atlantic Slope, Virginia Tidewater I Powhatan ...................... 9,000 234 38.40 13 Appalachian Summit Cherokee ........ ....... 22,000 1,344 16.30 II. North 14 Northern Great Lakes Algonkin, Ottawa .............. 7,300 2,043 3.57 Ojibwa (other than in 6d and 7) 30,000 3,145 9.54 15 Eastern Subarctic Beothuk ....................... 500 1,242 0.40 Montagnais, Naskapi, TAte de Boule ....................... 5,500 12,550 0.44 Cree (except Plains Cree in 6d). . 17,000 11,885 1.43 16a Western Subarctic Chipewyan ................... 2,250 6,194 0.36 Beaver ....................... 1,250 524 2.38 Slave ....................... 1,250 892 1.40 Dogrib ..... ............. 1,250 1,418 0.88 Abbato-tine, Etchao-tine, Strongbow ................... 1,200 3,254 0.37 Sekani ....................... 3,200 3,218 0.99 Kaska ....................... 500 500 1.00 Kutchin tribes in Canada (4 plus 4 part) .................. 3,000 2,861 1.04 Kutchin tribes in Alaska (3 plus 4 part) .................. 1,600 2,464 0.65 4 Khotana tribes, Kalchana .... 4,500 4,750 0.94 Ahtena ....................... 500 621 0.81 16b Interior Tundra Hare ....................... 750 2,261 0.33 Yellowknife .................... 430 2,110 0.20 Chipewyan territory 750 Caribou-eater .................. 1,250 3,860 0.32 16c Upper Fraser Carrier, Babine ................ 8,500 1,125 7.56 16d Northern Plateau Apex Tahltan, Taku-tine ............ 2,500 2,142 1.16 MEXICO AND CENTRAL AMERICA Culture areas are given in table 11. Populations are not considered by Mooney, except Coahuiltec 15,000 (in U. S.?) POPULATION AND DENSITY BY AREAS I give in table 8 the population, size, and population density of each numbered cultural area, such as the Southeast, South Atlantic Slope, Prairies, Great Basin, California, with its lettered subareas merged in it. 141 University of California Publications in Am. Arch. and Ethn. TABLE 8 POPuLATION DENsrrIEs Or PRiNCIPAL AREAS OF CUTURE L Culture areas|Poulatin | Area in I Density per Culture areas Population i l0) kCM.2 [ 10 k3M.2 Arctic Coast 1 Eastern Eskimo ............ 30,900 15,057 2.05 2 Western Eskimo .58,800 7,231 8.13 Northwest Coast 1 Northern Maritime. 28,100 1,666 16.80 2 Central Maritime .17,300 594 29.10 3 Gulf of Georgia .23,700 725 32.60 4 Puget Sound .6,000 357 16.80 5 Lower Columbia .32,300 507 63.70 6 Willamette Valley .3,000 334 8.98 7 Lower Klamath .18,800 377 49.80 Intermediate and Intermountain 1 Great Basin .26,700 10,810 2.47 2 California . 84,000 1,941 43.30 3 Columbia-Fraser ............... 47,650 6,660 7.15 Southwest 1 Pueblo .33,800 446 75.70 2 Circum-Pueblo (Athab.). 14,500 6,430 2.26 4 Sonora (in U. S.) .10,600 864 12.20 7 Northwestern Arizona .1,600 666 2.40 8 Lower Colorado River .13,000 416 31.25 9 Peninsular Calif. (in U. S.) 3,000 166 18.10 10 Southern California .26,500 683 38.70 Eastern 1 Southeast ...................... 87,800 5,983 14.70 2 South Florida .4,000 542 7.38 3 South Texas .6,400 2,057 3.11 4 Red River (and Pawnee). 25,900 4,563 5.67 5 Plains .50,500 13,978 3.61 6 Prairies .53,000 11,692 4.53 7 Wisconsin .18,300 1,461 12.52 8 Ohio Valley .20,000 7,707 2.59 9 Southern Great Lakes .42,500 4,421 9.61 10 North Atlantic Slope .7,300 3,285 2.22 11 Middle Atlantic Slope .46,800 1,828 25.60 12 South Atlantic Slope .41,900 2,467 17.00 13 Appalachian Suimmit .22,000 1,344 16.30 Northern 14 Northern Great Lakes. 37,300 5,188 7.18 15 Eastern Subarctic (Algonkin) 23,000 25,677 1.11 16 Western Subarctic (Athab.) 33,930 38,944 0.87 I I I 142 Kroeber: Cultural and Natural Areas of Native North America Condensing still farther, into grand areas, we have the densities shown in table 9. The areas are arranged not geographically but in order of density. I have added in parentheses the three main subunits of the Intermediate- Intermountain area, because these are so diverse that the density of the whole area is only a statistical mean. For the same reason I have given the Eastern and Northern areas separately, though adding in parentheses their joint mean. TABLE 9 POPuLATION DENSITIES BY MAJOR AREAS Area in Density in Culture areas Population 100 km.2 100 km.2 (California ............................. 84,000 1,941 43.30) Northwest Coast ....................... 129,200 4,560 28.30 Southwest (in U. S.) ............ ........ 103,000 9,671 10.70 Intermediate-Intermountain .. .......... 158,350 19,411 8.10 (Columbia-Fraser ...................... 47,650 6,660 7.15) Eastern ................................ 426,400 61,328 6.95 Arctic Coast ........................... 89,700 22,288 4.02 (East and North ....................... 520,630 131,137 3.97) (Great Basin ........................... 26,700 10,810 2.47) Northern .............................. 94,230 69,809 1.35 Total, north of Mexico* ................. 1,000,880 187,067 5.35 Coahuiltec in the United States are omitted, Apache and Papago in Mexico included. The outstanding fact is the exceptional density on the Pacific coast-both Northwest and California. Next comes the Southwest, which extends to the Pacific coast. Even the Columbia-Fraser region, a Pacific Coast hinterland, more than holds its own against the fertile East. The Arctic coast, surprisingly enough, has a density more than half as great as that of the East, though this was mostly agricultural; and one approximately equal-on the face of the figures even slightly superior-to the agricultural Eastern and nonagricul- tural Northern areas combined. This means, of course, that the latter had much the lowest density of all. The figure for the continent, north of Mexico, falls somewhat below that for the agricultural East and somewhat above that for the Eskimo. COAST LAND AND FARm LAND Two generalizations are obvious: coastal residence did make for heavier popu- lation; agriculture did not by itself necessarily increase density. Before these propositions are analyzed more in detail with regard to their meaning, it seems worth while to express them in still more drastic figures. We can first set off the wholly nonagricultural Pacific coast; next, the essen- tially agricultural areas of the Southwest and East; and then treat the re- mainder of the continent north of Mexico as a unit. The Pacific coast may be conveniently taken as extending from the Malemiut Eskimo of Alaska to the Diegue-no and Kamia just short of the mouth of the 143 University of California Publications in Am. Arch. and Ethn. Colorado. The area is that of Pacific coast in the ordinary sense, not Pacific drainage. The entire Yukon, Fraser, and Columbia River areas are excluded, except for the Eskimo, Coast Salish, and Chinook at the mouths of these streams. California is included as defined as a native culture area, not as a modern political unit; so are the northwestern margins of the Southwest, namely, southern California and the United States fragment of the Peninsular California area. The agricultural region comprises the tribes in whose economy farming plays a significant rather than sporadic part. Excluded are the Walapai, Havasupai, Yavapai, Apache, Navaho,1' Ojibwa, Abnaki, and those of southern Texas and southern Florida. Included are Southwest areas 1, 8, and the Pima Alto and Papago of 4; and Eastern areas la-c, 4ab, 6 a-c, 7, 8ab, 9, 11, 12a-d, 13. This classification yields the results shown in table 10. TABLE 10 GRAND PoPuLATIoN DIVISIONS (Papago, Apache, River Yumans in Mexico included; Coahuiltec in United States excluded) Percentage Divisions Population Area 100 km.2 Density of total population Pacific coast, Bering Strait to the mouth of the Colo- rado ..................... 295,700 11,745 25.2 29.6 Essentially agricultural areas, East and Southwest. 404,600 39,884 10.1 40.4 Remainder, north of Mexico.. 300,580 135,438 2.2 30.0 1,000,880 187,067 5.1 100.0 In round numbers, the Pacific coast had three hundred thousand inhabitants out of a million north of Mexico, or 30 per cent of the population in 6 per cent of the area, with a density of twenty-five per hundred-square-kilometer unit; the farming regions, 40 per cent in 20 per cent of the territory, with a density of ten; the remainder, 30 per cent on nearly 75 per cent of the land, with a density barely exceeding two. If the tribal figures on which this summary is based seem loaded, it is only necessary to remind the critic that if Mooney's original Merriam figures for California had been used instead of the Kroeber ones, the share of the Pacific 14 The Navaho and part of the Apache should perhaps not have been excluded from the farming peoples. The former farmed not only sporadically but also for their main subsist- ence, according to what Gladys Reichard and W. W. Hill tell me. For the Western Apache we have G. Goodwin's data in AA 37:55-64, 1935. The inclusion of the Navaho and all the Apache among the farming tribes would, however, add only some 14,000 to the 404,000 population computed in the text. Also, their areal den- sity being low (2.26), their addition would somewhat decrease the density for farmers as a whole from the figure of 10.1, and further emphasize the heavier density (25.2) of the Pacific Coast nonfarmers. 144 Eroeber: Cultural and Natural Areas of Native North America coast would have been 40 instead of 30 per cent of the total, without increase of area. That among nonfarming natives a coast or coast-plain habitat was normally far more favorable than interior residence in conducing to an aggregation of population, is indicated not only by the much greater density in the Pacific areas, but also by two other facts: first, that the Arctic shore Eskimo are, by area, more numerous than their inland Athabascan and Algonkin neighbors; and second, certain density figures for adjacent Atlantic and Gulf tribal areas. Thus: Coast Interior Massachuset.. 105 Nipmu . . 14 Pequot ..76 Mahican ..11 Montauk .. 158 Iroquois. . 7 Powhatan .38 Monacan, etc. 9 Weapemeoc, etc ......... 32 Eastern Siouan .. 12 Coree .33 Stono, Cusabo, etc .39 Yuchi .12 Apalachicola, etc .20 Creek .12 Chitimacha .32 Natchez .19 Chickasaw. 9 Quapaw. 4 The only really low densities for coast tribes in this region are in southern Florida and southern Texas, which are nonagricultural districts. A sharp line of division between coast and interior cannot easily be drawn in this Eastern region, because tidewater in many places runs far inland and because tribal adhesions and territories are so often uncertain. I therefore do not venture on any statistical expression. But an inspection of the itemized tribal entries in table 7 will, I think, leave little doubt that on the whole the population density in the farming parts of the Atlantic and Gulf region was perhaps twice heavier on the coast, including habitats on tidewater or within a day's travel of salt water, than immediately inland thereof.' This means that for the continent as a whole, always unfortunately exclud- ing Mexico, coastal residence, inclusive of that on coastal plains or along the lowest courses of rivers, led to a populational density from five to ten times greater than in the interior as a whole, in nonagricultural regions; and prob- ably at least twice as great even in agricultural areas. This finding may be expectable; but that the nonfarming Pacific coast should overtop the farming areas with a two-and-a-half times greater density is certainly surprising, at least when modern agriculture is borne in mind. It means, obviously, that the relation to the land in terms of agricultural utiliza- tion by the United States Indian was fundamentally different from our own. He was not a farmer in our sense of the word. Not only did he derive perhaps 11 Swanton in AA 37:373-385, 1935, holds that in much of the Southeast agriculture was producing a drift of population from the coast to the interior even before white pressure began. 145 1University of California Publications in Am. Arch. and Ethn. half his subsistence through nonfarming; he utilized for his farming no more than a vanishingly small percentage of the land capable of being farmed. This is particularly true of the East; and the Southwest should be specifi- cally excepted in this connection. The agricultural total in table 10 breaks up thus: East, 347,200 souls, 3,799,762 square kilometers, 9.1 density per 100 square kilometers;16 Southwest, 57,400, 172,200, 33.3. Not only is the gross density nearly four times as great in the Southwest, but the larger part of the territory assigned on the map to the Southwestern agricultural tribes is desert or mountain and unfarmable, or actually unfarmed by ourselves. The native Southwesterners, so far as they farmed, therefore pushed the exploi- tation of the land to a much higher pitch than the Easterners. This fact implies a different history, and thus further justifies the current sharp segregation of the Southwestern and Eastern culture areas. These essentially different histories, in turn, reenforced by the nonagricultural geographic gap between the areas, indicate separate origins, or at any rate separate branchings from the same southern stem of maize culture. THE AGRICULTURAL EAST The basic situation in regard to native farming in the Eastern area may be made clearer by a comparison with our agriculture. The average yield of maize per acre today throughout the United States is between 25 and 30 bushels of 56 pounds of shelled corn. Maize notoriously increases its yield per acre but little under improved methods of farming. The improvements which we have made over Indian methods have been mainly in the direction of re- ducing production costs, especially in labor. The Indian therefore may be assumed to have derived nearly as many bushels from each acre of planting as we. He probably planted somewhat farther apart; but not unduly so, be- cause of the difficulty of clearing and cultivating unnecessary area with his tools. A yield of 15 to 20 bushels therefore seems a fair estimate. This is 840 to 1120 pounds, say 1000, or a little less than 3 pounds per day. This should more than sustain the average person in a community composed of men, women, and children. Beans and pumpkins would vary the diet as partial substitutes for maize without seriously affecting the acreage cultivated. The quantity of farm food consumed was probably less than here computed, because of the supplement of game, fish, mollusks, berries, wild seeds, and roots, which over much of the Eastern region is estimated to have contributed half the food supply.'7 However, let us keep to our figure of nearly 3 pounds of maize or equivalent in farm products per head. Since this involves only about one acre cultivated per person, and we reckon 347,200 population in the Eastern agri- cultural area, the total native plantations in this region aggregated in round numbers only a third of a million acres. Against this, we today plant a hundred 6 Spinden (cited below) computes, also from Mooney, 348,700 inhabitants in about 1,375,000 square miles, which comes to 3,561,000 square kilometers and a density of about 9.8. He appears to include the Navaho and southern Ojibwa as farmers. 17 The heavier population density in the Wisconsin wild-rice district as compared with ad- jacent areas suggests the influence, in a farming area, which even a single wild food plant might have if systematically gatherable. See the discussion of this area in Section IX. 146 Kroeber: Cultural and Natural Areas of Native North Aimerica1 million acres of maize alone in the United States-not all, but nearly all, within the native agricultural areas here called Eastern. We add another two hundred million acres in wheat, oats, cotton, and hay-many of these acres fairly suitable, though not profitable to us, for maize. True, part of our total lies outside the region of systematic Indian farming; but it is a minority part. It does not much matter whether our total is one or two or three hundred million acres and the Indian total one-third or two-thirds of a million: the conclusion remains that the eastern Indian cultivated less than 1 per cent of the area on which theoretically he could successfully have grown crops satis- factory to his needs and standards. My own opinion is that the figure was under rather than over one-half of 1 per cent. Here is another way of conceptualizing the situation. The Eastern agricul- tural density was 9.1 per 100 sq. km., a little under 9 souls-say 2 families- per township. We allot 144 quarter-sections to 144 families, or some 700 per- sons, in a township; and these earn through their crops not only their food but also their clothing, tools, vehicles, furniture, taxes, and luxuries. The ratio comes out about the same. It is clear that two things were fundamentally different in the Eastern Indian economics and ours: the land use, or relation to the land; and the place of agriculture in life. "Improvement" of land was confined to minute specks in the landscape. They were comparable in size to oases, although not in the least enforced by nature, being in fact simply selected by convenience or habit from among a hundred times as many sites almost equally well utilizable. In other words, there was a hundredfold surplus of potentially farmable land over farming population. (My colleague Sauer points out that this analysis omits one important fac- tor : because of his operating only with sticks and light hoes, the Indian avoided any but friable soil. I should have given this consideration more weight. For instance, breaking sod would have been very difficult with the native hand tools. Nevertheless there must have remained a great excess of land which under augmenting population pressure could have been farmed without draft animals or iron. No doubt this would have required some additional labor effort, but in some places less effort than plowing: in fire-cleared forest, for instance, where planting could be done between the unremoved stumps, as in the tropics. That in general the Eastern Indians did not have recourse to such devices suggests that their population remained so low from other causes that they could raise what they needed on the easiest and most fertile spots.) Second, while every native household in the area farmed, it becomes doubt- ful whether many of them did so from real necessity. If the Pacific coast from Bering Strait to the Imperial Valley desert could support 25 souls per areal unit without farming, it is not unreasonable to suppose that the uniformly fertile East could have supported 10 without farming. Agriculture, then, was not basic to life in the East; it was an auxiliary, in a sense a luxury. It made possible increased accumulation of food against the future, living in perma- nent sites and in larger groups, and therefore joint undertakings, whether of council, ritual, war, or building. It thus no doubt contributed somewhat toward 147 1University of California Publications in Amn. Arch. and Ethn. the enrichment of cultural life; but there is little to argue that the culture was leaning very fundamentally on agriculture. Does this mean that agriculture was a recent introduction in the East, not yet fully acculturated and its potentialities still mainly unconceived ? Theo- retically this might well be so; but it is not a necessary inference. As long as any other factors kept an originally light population light, the relation to the land, the part-only farm use of this, might go on indefinitely. The answer to the question of the age of Eastern agriculture should not be given deductively. The direct evidence to be considered is archaeological; the indirect, social factors bearing on population. As for archaeology, we are still handicapped by our disgraceful because probablyunnecessary inabilityto interpret Eastern prehistoric data in sequen- tial terms. Still, the gross fact remains that the Ohio and middle Mississippi valleys were found occupied, at the outset of the historic record, by an exceed- ingly thin and scattered population, but full of thousands of mounds and other structures which probably required a more concentrated population to erect. Allowing for all possible shifting about of this earlier farming population, and an abnormal readiness to leave one site as soon as its structures were com- pleted in order to begin over again elsewhere, a minimum of several centuries must nevertheless be allowed as the duration of the building; and to all major intents, this period was both past and forgotten when the first whites entered. Since the mound culture was agricultural, it is accordingly hard to see how fewer than 500 years, perhaps 1000 or more, could have elapsed between the introduction of maize and the coming of Caucasians into the East. If agricul- ture in itself tended automatically to produce a marked increase of population density, it was long enough in the land to have achieved this effect to a much greater degree than obtained at discovery. Rather, we see a positive thinning out of numbers, in at least part of the area. The indicated cause, then, is not mere shortness of duration of establishment of the agriculture, but "social" factors of some sort. Of social factors, the most direct may be considered to have been warlike habits. Reference is not to systematic, decisive war leading to occasional great destructions but also to conquest, settlement, and periods of consolidation and prosperity. Of all this the Eastern tribes knew nothing. They waged war not for any ulterior or permanent fruits, but for victory; and its conduct and shaping were motivated, when not by revenge, principally by individual de- sire for personal status within one's society. It was warfare that was insane, unending, continuously attritional, from our point of view; and yet it was so integrated into the whole fabric of Eastern culture, so dominantly emphasized within it, that escape from it was well-nigh impossible. Continuance in the system became self-preservatory. The group that tried to shift its values from war to peace was almost certainly doomed to early extinction. This warfare, with its attendant unsettlement, confusion, destruction, and famines, was probably the most potent reason why population remained low in the East. It kept agriculture in the role of a contributor to subsistence instead of the basis of subsistence. On the other hand, such farming as was practiced yielded 148 Kroeber: Cultural and Natural Areas of Native North America enough of added leisure, concentration, and stability to make pretty continu- ous warfare possible. A population of pure hunter-gatherers would probably, except on the immediate coast, have been too scattered in minute bands, too unsettled in a country of rather evenly distributed food possibilities, too occu- pied with mere subsistence, to have engaged in war very persistently. Just this seems to have happened among Montagnais, Cree, and Ojibwa, for instance, as compared with Muskogians, Iroquoians, and Siouans. The latter were caught in a vicious circle, which at the same time gave them a stable adjustment. Agri- culture made their wars possible; but their warfare kept the population down to a point where more agriculture was not needed. Behind all this must lie another, though negative, factor: the absence of all effective political organization, of the idea of the state. Effective, of course, means effctive from our point of view; it is not denied that the native organi- zation was effective so far as concerned its needs within the cultural system in which it found itself. Had controlling authority, in the form of a ruler, or of a cohesive, smoothly self-perpetuating group, ever developed in the East, war objectives other than revenge or personal status might also have developed: conquest, pacification, tribute, economic accumulations, further exploitation. From among many such beginings, no matter how humble in scope, there could sooner or later have emerged, through mutual eliminations, larger units, and from these, true states, stable, internally peaceful, capable of producing wealth, growing in population, and thereby increasingly productive and prof- itable. Just as something of this sort happened in China and Egypt, it hap- pened in Mexico and Peru; but it did not happen in any consequential degree in what is now the United States. The political systems of the Iroquois, Creek, Cherokee, Natchez either grew up mainly in historic times under Caucasian influence and pressure, or were, as appears possible, fragmentary remnants from the Mound Builder days of heavier population and quasi states. If there were such days, and it seems there were, it may well have been the introduction of agriculture that made their state system possible. But once the system crumbled, perhaps because of being a foreign import and not deeply enough rooted in the culture of the region, there would be a relapse to interminable, economically vain fighting, rendered, however, more persistent and wasteful than ever by the fact that agriculture gave an added margin allowing greater wastage. In the North, where farming could not be or was not introduced, the limitation of purely natural food sources was perhaps the main factor impos- ing an upper limit to the human population. In the East, where the combina- tion of agriculture and fertility made possible the comfortable subsistence by native techniques of a population many times greater, the causes must have been cultural; and of these the outstanding ones were the paired ones of high social premium on war for its own sake and the absence of value for political organization of more than a rudimentary kind. Incidentally, the cultural dependence of the Plains on the East, historically, is again indicated by the fact that the whole sociopolitical system and moti- vation of the Plains are, at large, a copy of those of the East. The acquisition of the horse gave the Plains tribes, while the buffalo lasted, a food margin and 149 0University of California Publications in Am. Arch. and Ethn. a leisure parallel to the agriculture of the East, and enabled them to duplicate the customs of the East with only minor modifications such as the replacement of torture by coup counting. We must, then, think of the East as agricultural indeed, but as inhabited by agricultural hunters, not by farmers, peasants, or peons. There were no economic classes, no peasantry to exploit nor rulers to profit from a peasantry. Every man, or his wife, grew food for his household. The population remaining stationary, excess planting was not practiced, nor would it have led to any- thing in the way of economic or social benefit nor of increase of numbers. Ninety-nine per cent or more of what might have been developed remained virgin, and was tolerated, or appreciated, as hunting ground, as waste inter- vening to the nearest enemy, or merely as something natural and inevitable. There was nothing to prevent a clan, town, or tribe from shifting its houses and fields to any one of dozens of near-by equally satisfactory sites in its ac- knowledged territory; or, if strong enough, to several hundred in land of its neighbors (subject to the qualification mentioned in the parenthesis on p. 147, above). There was as a rule nothing much gained or lost, other than for im- mediate considerations, by such shifts; and they were freely made-not per- haps mainly from sheer restlessness, but at least for trivial reasons. The consequence is the strange contrast of a relatively unstable, mobile agricultural population in the East and a rather highly sessile nonagricultural one on the Pacific coast. This point will be reverted to in a following section dealing with the relation between language groups and population. COMPARISON WITH MEXICO A comparison with Mexico seems worth while. There, conditions were differ- ent. It is known that population was denser, and that social classification and political organization were much more developed. However, there are only fragmentary general or gross estimates of the ancient Mexican population, and these vary.' We may therefore attempt to proceed by working backward from present conditions. The area of modern Mexico is roughly 750,000 square miles, or about 480,000,000 acres, of which a fourth, or 120,000,000, are con- sidered (or are nominally) cultivable, and 30,000,000 are actually cultivated, although for only about half of these 30,000,000 is a crop specified, so that the other half may be considered as in a condition of latent cultivation or given over to products like maguey or henequen. The largest area is in maize, 7.5 million acres in 1926. Next come beans with 2.2, wheat 1.2, cotton 0.6. The total is astonishingly small compared with the United States, whose maize acreage alone is more than three times as large as Mexico's total acreage in all crops. There is nothing to show that any considerable areas now unused were planted at the time of discovery. Rather have the hacienda system and modern engineering tended to add acreage. If we assume, as before, that an acre will support a person, the present total in maize and beans, if utilized to the limit, would have provided sustenance for 10,000,000 souls. The addition of other acreage now actively in crops would bring this up around 15,000,000, or the 18 They are considered below, in the subsection on Mexico. 150 Kroeber: Cultural and Natural Areas of Native North America present population. This is probably too high for the past. It would mean that the country at the time of discovery was settled up to the very limit of the population which it would support with the agricultural techniques at its command. Of this there is no indication. I would prefer to reduce the figure by three-fourths or more. Yet even this means that a fourth or a fifth of the most available farm lands, perhaps the majority of the best, was being worked. About the larger centers of population, as in the Valley of Mexico, there was probably little waste except of distinctly inferior tracts. The native historical records show that, in the Valley, farm land was at a premium, and either in the form of tribute in produce or by direct appropriation was a prize of con- quest. There existed here, then, a condition resembling that of modern civi- lized countries; and even in the less densely settled areas of central and southern Mexico, one approximating this. That the land was owned by towns or barrios or family aggregations instead of individually is socially and jurid- ically important, but does not affect the population and subsistence picture. Where the Eastern Indian farmed a fraction of 1 per cent of his available land, the Mexican farmed a considerable fraction of his total,' and in con- gested, politically dominant, and affluent areas, practically all of it. It was almost inevitable, therefore, that in Mexico there should be economic classes, political organization, large communal works, and war for profit. There were in Mexico the equivalents of peasantry and aristocracy. Without such classes, the population could hardly have accumulated as it did; and at the same time, its growth must have tended to make organization desirable if not necessary. However free in principle, the average Mexican of 1500 A.D. was no longer free as a Creek or Iroquois or Illinois was free. He could not farm if and where he pleased. He was bound by economic necessities of subsistence as well as by his state and rulers. The Spaniards perhaps found more peons in Mexico than they made. THE SOUTHWEST The Southwest was different from both Mexico and the East. It had maize as far back as Basket Maker times-less long than Mexico, no doubt, but longer than the East, where, though agriculture was evidently more than two or three centuries old, there is nothing to show that its importation goes back to the pre-Christian era. Population density in the Southwest also was inter- mediate, so far as genuinely agricultural peoples were concerned. The dis- tinctive feature of the Southwest is the presence in it, side by side, of two kinds of population: the fairly densely settled farmers, and the very thinly sown nonfarmers around and between them. How far back this condition goes historically it is difficult to say, because, as might be expected, the farmers have left abundant and striking archaeological remains, the gatherers few and scattered ones. The farming population of Pueblo type is- known to have been more widespread in Pueblo 2 time-say in the general period, 9Q0-1100 A.D. But there may have been nonfarmers near them, if not in immediate contact, 19 The reference is to the areas recognized as culturally Mexican in the present monograph, not to the modern Republic of Mexico, the northern half of which was much more thinly populated and in large part nonagricultural. 151 1Univer8ity of California Publication8 in Am. Arch. and Ethn. even then. We cannot say positively; but most of the Northwest Arizona area is devoid of Pueblo ruins or remains. The basis of this duality of the Pueblo-Southwestern economic system, whether it is relatively recent or ancient also, lies obviously in the nature of the land. The Southwest is an arid region, steppe and mountain or semidesert where not desert. Farming, with patience, can be made to yield a fairly reliable subsistence, but only in selected spots. The greater part of the surface of the Southwest was as useless to the Pueblos, for crops, as it is to us. They could and did farm many spots which we do not farm; but that was because they sought only their food, we a civilized living. Allowing, as before, an acre to a person, the 34,000 Pueblos whom Hodge and Mooney estimate for 1680 would have had under cultivation a total of only some 53 square miles-a township and a half. We may double the allowance of land per head to permit of wider spacing of planting or lower yield in the arid Southwest. We may enlarge the population somewhat to accord with the wider extent of the cul- ture in Pueblo periods 2 and 3.' Even this, however, brings the actually farmed land up to a total of only 100 or 200 square miles in 200,000 or 300,000. This is just about the ratio utilized in the East; but there most of the great unused remainder was farmable, whereas in the Southwest it was not. The Pueblo, then, resembled the Mexican in using for his crops, if not every inch of productive land, at any rate much of the best of it. This makes his subsistence appear more directly of Mexican origin, with but slight transmu- tations. Where he differed was in that so little of his land was cultivable, and that scattered. He could not become numerous. He therefore did not need states and rulers and a peasantry; the more so as the scattered distribution of his farmable land kept his communities small. But, once given a concentra- tion in towns, his agriculture became a necessity to him if he was not to starve. This in turn engendered an attitude, a lack of leisure and lack of sense of freedom and enterprise, which would keep him from plunging into chronic warfare as a social mechanism. His population was kept down not so much by being killed off or expelled and disrupted, as by clinging to a narrow shelf of subsistence mechanism without leeway or recourse. So far, discussion of the Southwest has been in terms of Pueblos and the nonagricultural tribes enclosing them. Populationally, this part of the South- west forms the smaller half of the Southwest within the United States: 48,300 souls out of 103,000. Pima-Papago, Lower Colorado Yumans, and Southern Californians alone, with 10,000, 13,000, and 26,500 souls, outnumber the com- bined Pueblo, Apache, and Navaho, even with the Pueblo counted at Mooney's high figure of nearly 34,000. Numerically, the preponderant half of the Ameri- can Southwest was the Gila-Yuma-California sphere, not the Pueblo one. In density the disproportion is even greater: nearly 20 for the former, against a little more than 7 for the Pueblo." It is true that the density of the pure Pueblo territory alone was the highest-around 75. But against this in the other half are figures like 31 for Lower Colorado and 39 for nonagricultural 20 However, Kidder cuts Mooney's 33,800 to 20,000 population. 2 The figures are: 54,700 in 279,500 square kilometers = 19.6; 48,300 in 687,600 = 7.2. 152 Kroeber: Cultural and Natural Areas of Native North America and semidesert Southern California. The Pueblo sphere density as a whole is brought down by the abnormally low density of the vast area occupied by Athabascans: 2.3. This expresses again the oasislike distribution of the im- portant population of the Pueblo sphere, and the contrast between town dwellers and mescal gatherers, which recalls so nicely in many ways the rela- tion of the town farmers and the herders in the Sahara, Arabia, and inner Asia. As against this, the Gila-Yuma-California sphere was much more evenly sown with population, irrespective of whether this was agricultural or not. In one sense, therefore, this area may be considered as having made a healthier adjustment with its arid environment than the Pueblo sphere. The archaeological evidence indicates that in the past, in Pueblo periods 1 and 2, say until about eight hundred years ago, the Pueblo proper popula- tion was much more widely and scatteringly distributed in numerous small settlements. In other words, its distribution then approximated that of the Gila-California area. This distribution began to be abandoned with the con- centration into larger towns in Pueblo period 3. This concentration may have been in part due to the pressure of preying Athabascans first intruding then. But whatever the causes-invasion, drought, inner cultural tendency, or a combination of these factors,-once the concentration had begun, it left ever larger areas open to the "nomads," that is, thinly sown mescal gatherers, and enabled them to establish themselves and their subsistence adaptation more firmly. The very flowering of Pueblo culture therefore tended to shrink its area, to embody it geographically in a culture of very much lower intensivity, and to put it on the defensive against this. Nothing like this occurred in the western Southwest, where farmers and nonfarmers remained in adjustment, and the whole of any given tract continued to be exploited more or less to the limit by whatever subsistence mechanism was most feasible, without notable "class" differentiation of its culture. The one exception was the Casa Grande type of concentration in the Gila Valley, when Puebloid polychrome pottery culture impinged on native red-on-buff; but this was evanescent, and, on its collapse, culture returned to its former adjustments. So far, we have been speaking, of necessity, in terms of the American South- west. If the Mexican part, for which we have no population data, could be included, the area and population of the Southwest would be increased by nearly half, and its Pueblo-type part would presumably shrink in numbers from a minor half to no more than a third; which goes to show again what dif- ferent historic concepts "Pueblo" and "Southwestern" are, and the need of their not being used interchangeably. CALIFORNI For California, with which my personal acquaintance is greatest, I add a special map (no. 19), which includes not only the California area proper, but also those parts of the Northwest, Basin, and Southwest areas which lie within the modern state. It will be seen that in general the population diminishes from coast to interior, rapidly dropping still further with the crossing of the interior ranges. The lower courses of large streams go with the coast. The one 153 1University of California Publications in Am. Arch. and Ethn. exception to the general trend is on the Sierra Nevada flank of the San Joaquin Valley. Here the aridity of the valley-Russell reckons it as desert climate," and it is waterless most of the year, except for the larger streams that transect it-made for a heavier population in the better-watered foothills belonging to the Yokuts and perhaps the Miwok. Further, the population in general, and with this one exception of the foot- s I ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I Map 19. Population Densities in Native California, by Gross Areas of Ethnic Groups. hill Yokuts, is conspicuously densest in the regions of cultural climax: the lower Klamath, the lower Sacramento and Russian, the Santa Barbara Chan- nel, the lower Colorado. In this connection these climaxes may of course not be compared among themselves, which would involve other considerations; but each does stand out as more heavily populated than the regions surrounding it. This fact would probably have been much more accentuated if the map could have been constructed in terms of actual habitation sites, for which the 2"Climates of California, UC-PG 2: 73-84, 1926. Compare map 24. 154 Kroeber: Cultural and Natural Areas of Native North America data are unfortunately still too partial, instead of gross tribal or ethnic terri- tories. Finally, the two heavier shadings on the map (45-70 and 70+) may be taken as indicating a density approximately equal to that of the Pueblo area proper (76). The fact that they cover so considerable a portion of the state shows once more that under native conditions even intensive agriculture did not necessarily lead to higher density than a favorable nonagricultural adap- tation. This map of California is almost the reverse of that in the Handbook of California Indians, page 887, giving the percentage of survival since Cauca- sian arrival. Where the Indian was numerous, the white man penetrated early and settled in numbers, so that in general it has been the densest native popu- lations which have suffered the heaviest decline. Apart from climax considerations, these two maps together probably pic- ture coast-inland population relations and maintenance much as they existed also on the Atlantic side of the continent, though probably to a somewhat less extreme degree there. Through their position, the Iroquois, Cherokee, and Creek, on this view, preserved or increased their population in the Colonial period much as the Achomawi and Mono maintained theirs relatively well in California. THE NORTHWEST COAST The figures for the areas within the Northwest coast also carry a story, though they must be used with a certain reserve because in some of the areas the land itself was so little or secondarily used that length of frontage on shore or river was evidently the decisive factor in regard to population.' Still, the areal densities mean something. They are: Areas Lower Columbia (Chinook, ete.) ................................ 64 Lower Klamath (Yurok, ete.) ................................... 50 Gulf of Georgia (Coast Salish) ......... ........................ 33 Central Maritime (Wakashan, etc.) ........ ...................... 29 Northern Maritime (Northern tribes) .............................17 Puget Sound (Coast Salish) ............ ........................ 17 Wilamette Valley (inland) ............ .......................... 9 Subareas Central Maritime, South (Nutka, Makah, Quinault) ...... ......... 65 Northern Maritime, Archipelago (Haida, Southern Tlingit, Tsimshian) ................................................ 22 Central Maritime, North (Kwakiutl, Heiltsuk, Bella Coola) ........ 17 Northern Maritime, River (Niska, Gitskyan, Haisla) .............. 10 Northern Maritime, Mainland (Northern Tlingit) ................ 10 The Willamette area is a wholly inland one. We do not know with certainty whether it should be reckoned as part of the Northwest coast or the Columbia- Fraser plateau. The Puget Sound area, although a salt-water one, also ex- tends its inlets far into the interior, and is quasi inland. Apart from these two minor areas, the other five range almost in geographical order, with density 23 Coast-line holdings and densities are considered in a separate subsection below. 155 University of California Publications in Aim. Arch. and Ethn. decreasing from south to north. The subareas within the two northern areas again show almost the same arrangement. Even if Mooney's computations for the Chinook and Gulf of Georgia Salish are taken as somewhat high, the gen- erally greater density of the south as against the north remains fundamentally unimpaired. On this point, too, shore-line density would not invert the situ- ation, the northerly areas having the more irregular, indented shore, the ratio of which to the already lighter population would go up faster even than their land areas. The difference seems to lie in this: The northern groups were essen- tially maritime, mostly lived fronting the beach, and made little use of the land which they owned. The southern groups lived on river and tributary as well as on the shore, perhaps more largely so, in fact,' and often made genuine use of their land holdings. Their habitat utilization and culture remained more generalized and simpler; those of the northern groups were more specialized and extreme. As in the Southwest, on comparison of Gila-California with Pueblo sphere, the more generalized method in the long run permitted of a heavier aggregate population. This set of facts also seems to reenforce the previously outlined interpreta- tion of Northwest Coast culture development. If the generalized southern areas represent, as seems reasonable, the survival of an earlier phase, it is the northern areas which have specialized away from this, and their type of culture must on the whole be the more recent. Whether this specialization was mainly the result of an internal development leading to a shift from river to inlet to ocean shore where the shore was most favorable, or was brought about by Eskimo or Asiatic or transoceanic contacts and influences, is another and difficult problem, but one which may prove soluble to investigators in a posi- tion to analyze intimately the entirety of Northwest Coast culture; though they also can hardly come to a final conclusion without taking into detailed consideration the geographic setting. For the present we can content ourselves with the findings that it is the southern half of this major area which is the more densely populated, more generalized in its subsistence adaptations, and more ancient in its type of culture; and that the full habitat and subsistence adjustments of the northern half, and the intensity of its development in art, ritual, and property distribution, are relatively recent. ESKIMO For the Eskimo areas, the range of land-area densities is: Aleut ................................................... 65.0 Pacific Coast (excluding Bering Sea) ........ .................. 19.0 Western (Bristol Bay to the Mackenzie River) ...... ............ 4.9 Central-Eastern ............................................. 2.3 Caribou ................................................... 0.4 Total Western ............................................... 8.1 Total Eastern .............................................. 2.1 XIn Handbook of California Indians, 117, I have computed a population per shore mile of salt water of 10 and 15 for Wiyot and Yurok, and of 20, 35, 25, and 30 per mile of navigable river for the same two groups and the Karok and Hupa; or a mean of 28 versus 12 in favor of river. All the groups are in northwestern California. 156 Kroeber: Cultural and Natural Area8 of Native North America Land areas mean particularly little in comparison with shore line to the Eskimo, whose life depends on water and ice far more than on what the land bears. Still, the figures probably give a crude approximation to shore-mile density, even if the Aleut population of 16,000 should prove too high. How far the higher latitude of the three low-density areas may be a factor must also be considered. Still, the figures on their face show this: Nearly a third of all the Eskimo lived on open Pacific Ocean frontage-27,300 Aleut, Kaniagmiut, Chugach- igmiut, and Ugalakmiut, out of 89,700. From the Malemiut south, that is, roughly, in Alaska from Bering Strait south, were almost 60 per cent of all members of the stock-53,000 out of 89,700. This is the region of masks and wooden houses and grave monuments and property-distribution festivals and war-fleet expeditions, traits which we are wont to regard as characteristic of the Northwest Coast culture. It is also the region where ice hunting of seals, the sledge and the snow house, and many other typical "Eskimo" traits, are lacking or nearly absent. In other words, "pure" or characteristic Eskimo culture obtains only among two-fifths of the members of the stock. Three-fifths live in a culture heavily charged with elements usually regarded as Northwest Coast or Asiatic and lacking much of the inventory of "typical" Eskimo life. It is obvious that our concept of what is Eskimo is due to a first approach from Greenland, and next Labrador, Baffinland, and the Central region. Had our knowledge begun in Alaska, where population centers and where the density is overwhelming, our most "typical" Eskimo would probably seem merely peripherally reduced and atypical. Just what this means for the origin and history of the culture it is hard to say. Most such evidence can be read two ways. The final word must be by specialists on the Eskimo. But the population distribution cannot be left out of account. And for a full understanding of this, reasonably reliable fig- ures of shore miles held by each Eskimo group are necessary. MEXICO AND CENTRAL AMERICA For Mexico and Central America there exists nothing like Mooney's complete group-by-group series of population figures. The contemporary and docu- mentary data seem never to have been gone over systematically, let alone assembled. The estimates which I give here therefore represent nothing more than my personal opinion as based on impressions, somewhat molded by com- parisons with the population size and density north of the Rio Grande. It is necessary first to set off the northwest Mexican districts which I reckon as of the Southwest. The rest, constituting my Mexican-Central American group of areas, I divide for present purposes into three parts: the region of higher culture, comprising the Mesa Central and adjacent parts of Mexico, together with Guatemala and Salvador; the region of lower culture to the southeast, about corresponding with Honduras and Nicaragua; and the mainly nonagricultural area of low culture to.the northeast of the Mesa Central. My areal measurements and population estimates run as shown in table 11. 157 5University of California Publications in An. Arch. and Ethn. TABLE 11 MExICO-CENTRAL AMERICA: AREAS, POPULATION, DENSITIES Areas T 100 km.2 100 km.2 Est. pop. Density SOUTHWEST AREAS Fuerte-Yaqui Lowland (Cd- hita) ..................... Sonora, total area.......... Northern Sierra Madre (Ta- rahumar) . Sonora Coast (Ser). Peninsular California. Total ..................... Less part of 4 counted with U. S. for pop..... 864 Less part of 9 counted with U. S. for pop..... 166 Net total in Mexico........ MEXICAN-CENTRAL AMERICAN AREAS NICARAGUA AND HONDURAS Atlantic Nicaragua- Honduras................ Pacific Nicaragua.......... Total ..................... REGION OF HIGHER CUILURE Salvador................... Upland Guatemala (High- land Maya).............. Yucatan Peninsula (Low- land Maya).............. Oaxaca-Tehuantepec ....... Guerrero................... Vera Cruz........... Southeastern Central Mesa. . Michoacdn (Tarasco)....... Jalisco Highland........... Jalisco Coast............... Northeastern Central Mesa (Otomf).................. South Sinaloa.............. Sierra del Nayarit.......... Total ..................... MINLY NONAGRICUITURAL REGION Central Sierra Madre....... North Mexican Interior Plateau.................. Tamaulipas (incl. Coahuil- tec in U. S.).............. Total ..................... Grand total............... 481 1,711 715 306 1,390 4,603 1,030 2,048 429 307 792 2,553 1,021 1,370 1,028 398 570 330 352 673 404 460 1,095 3,518 2,054 3,573 2,477 100,000 27.98 = 28 100,000 I 40.4 = 40 10,258 3,000,000 1 292.4 =300 6,667 22,975 100,000 3,300,000 15. = 15 3 4 5 6 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 I 158 Eroeber: Cultural and Natural Areas of Native North America The areal total of 2,297,500 square kilometers includes the part of Coahuiltec territory lying within the United States. This may be estimated at about two-fifths of the total of 149,900 km.2, or 60,000km.2, redueing the total to 2,237,500 km.2 On the other hand, Apache territory along the northern frontier of Mexico amounting to perhaps 100,000 km.2 has been measured in with the United States Southwest areas. On addition of this, we have about 2,337,500 km.2 as the area computed by planimeter measurement for Mexico, Guatemala, Salvador, Honduras, and Nicaragua combined, as against a trifle more than 918,000 M.2 or about 2,378,000 km.2 usually given for the five countries. We may now examine the estimates of population and density. I have allowed 100,000 souls for that part of the Southwest which lies in Mexico, but this should be regarded as a maximum. The area contains a trifle more than a third of a million square kilometers, as against nearly a full million in the American Southwest, of which the population, as based on Mooney, was 103,000. The density, then, would be nearly three times as great on the Mexican side, 28 as against 10.6. This seems liberal, both in view of the nature of the country and the tribes concerned. These are the Pima (other than Papago and American Pima), Opata, Tarahumar, Cahita, Seri, and the bands of Baja California. Wilicox" quotes my colleague Carl Sauer as believ- ing that Baja California contained one Indian to a square mile, and Sonora two. This would make the density of Mexican California greater than that which I have computed for American California. It would give Sonora alone 150,000 natives, or half as many again as the whole American Southwest in- cluding southern California. Of course, it may in the end be proved that Mooney and I have throughout cut figures much too low. However, our figures are itemized, and it seems sound to adhere to them as against general impres- sions or ratios based on densities of occasional spots. For the two-thirds of a million square kilometers of northeastern nonagri- cultural Mexico I have allowed 100,000 population, or 15 per 100-km.' unit. This is a higher density than in any of the grand areas north of Mexico, except the Northwest Coast; about the same as that of the agricultural southeastern United States; and five times as great as in nonfarming South Texas. This seems a very liberal estimate. The 3,000,000 native population which I allow to the region of higher cul- ture in Mexico and Central America may seem tame as against some of the figures currently mentioned; but it seems reasonable if our estimates to date have been approximately sound. The region, defined as Mexican areas 3-15, comprises a very little more than 1,000,000 km.2 out of a total in the continent of about 21,000,000, or not 5 per cent of the area; but 70 per cent of the popu- lation. Its average density of about 300 (292) is greater than that of the most populous and restricted tribal territory elsewhere in the continent. The den- sity is fifty-five times as great as the average north of the Rio Grande. It out- weighs the density in California seven times, on the Northwest Coast ten, in the American Southwest and agricultural East nearly thirty times. All these ratios are no proofs; but they do suggest that if our figures up to this point 15 Increase in the Population of the Earth and of the Continents since 1650, ch. i, pp. 3-82, in W. F. Willcox, ed., International Migrations, vol. 2: Interpretations (National Bureau of Economic Research, New York, 1931). 159 University of California Publications in Ain. Arch. and Ethn. have been tolerably reasonable, the allowance of 3,000,000 for cultural Mexico is also reasonable and perhaps liberal. The actual population in 1500 A.D. may have been more. But it may also have been less. Modern population figures need not shake confidence in this result. Accord- ing to Humboldt,' New Spain from latitudes 100 to 380 in 1793 contained 5,400,000 persons. The Mexican census of 1930 counted 16,404,000. To this the four northern Central American republics would add 5,500,000, bringing the total to 22,000,000. That is, the population has quadrupled in four to five generations, if the official estimates used by Humboldt were even approxi- mately correct. If it could be assumed that the increase could be continuously projected backward, we should be starting with little more than 300,000 souls in 1500 A.D.-and that in all New Spain, not merely cultural Mexico-Guate- mala. I do not wish to propose that this may have been so. But the illustration shows that we may not infer from present-day large populations to native large ones. And to assume that there was a large population, that this was reduced to a mere small fraction by the Conquest, and that then it built itself up again, is gratuitous. The Conquest no doubt did cause a shrinkage in num- bers; but in the well-settled regions this effect seems to have been transient, and probably began soon to be made good by an increase attendant on the new experience of internal peace under Spanish Colonial government. As a matter of fact, the population of New Spain seems to have fluctuated in rather un- accountable ways, apart from the effects of shock of Caucasian arrival. Will- cox' computes 5,115,000 in Mexico and Central America in 1650, but only 3,150,000 in 1750, and again, as does Humboldt, 5,400,000 in 1793. Why these fluctuations should occur in a long period of peaceful stagnation, and then be followed by a rapid and mounting rise in a period of progress but revolution and civil war, it is difficult to see. We evidently have not yet got solid ground under our feet in our knowledge of the historic population of Mexico. But for that very reason it is unsafe to do much reckoning for the prehistoric era by comparison with the present. One thing, however, is clear. If our 3,000,000 be accepted as anywhere near the truth, there has been a definite increase not only of total population, but also of Indian population in Mexico since aboriginal times. The 1900 census of Mexico rated 38 per cent of the population as Indian, 43 per cent as mestizo. For 1921 the respective figures are 29 and 60 per cent. Census classification must needs be on a social or linguistic rather than a biological basis; but it is generally admitted that this fact is more likely to result in an undercounting than an overcounting of the Indian element. If we allow the mestizos to be half-Indian in blood-a low estimate-and thus convert them statistically into half as many Indians, the result for both 1900 and 1921 is very elose to a 60-per cent Indian population in Mexico. In other words, the Indian blood is diluting by mixture, but not decreasing. Relative to purity, it is diminish- ing; relative to its place in the total population, it is holding its own. Now, 60 per cent of the 16,400,000 of 1930 is 9,840,000, or say about 10,000,000. Guate- ," Cited in Wilc0eox, 24. 27 Pp. 30, 38. 160 Kroeber: Cultural and Natural Areas of Native North America mala, with 60 per cent of "pure" Indians, and mestizos making up most the remainder of its total of 2,500,000, adds another 2,000,000. The total of 12,- 000,000" is four times our estimated 3,000,000 in 1500 A.D. Even if my estimate for native times should need doubling, a definite increase cannot be denied. This is the reverse of the universally admitted change in Anglo-Saxon America. The obvious cause of the difference, though there may be others also, is that in New Spain the settled Indian was fitted into the colonial and modern eco- nomic scheme,-in fact this was built upon him; whereas in Saxon America, broadly speaking, he did not fit into the economic plan and was thrust into negligible corners like more or less picturesque rubbish. One other consideration arises in connection with present population: how far the modern regional variations of density in Mexico correspond to ancient ones. For this reason I append map 20, which is a simplification of that by Cushing in the Geographical Review of 1921." It is at oAce evident that in general the district of present-day heavy density is the Mesa Central and regions to the southeast; in other words, ancient cultural Mexico, areas 5-15. Guatemala and Salvador, with nearly 30 souls per kilometer, belong to the same belt of heavy density. They correspond to areas 3 and 4, which, with 5-15, make up the region to which we have assigned 3,000,000 prehistoric population and a density approaching 300 per 100 km.' Against this, Hon- duras and Nicaragua, to which we have allotted the much lower density of 40, or 0.4 per km.', now are also much lower, with 6. It can therefore be inferred that, in general, prehistoric, historic, and modern populations in Mexico and Central America tend to be dense and sparse in the same areas. However, this fact does not preclude considerable shifts in relative ratios, as well as striking local ones, the former due perhaps to the introduction of domestic animals, the latter to growth of cities or intensive exploitation of mines. While from the basis of my assumptions for 1500 A.D. all parts of Mexico-Central America have increased in population, the ratio of increase has been heaviest precisely in those districts which once had the sparsest popu- lation. The Mesa Central has increased; the outlying areas have increased still more. A computation by districts instead of states would make the reckoning more accurate, but has been foregone because the comparands are after all only estimates and too sharp a definition of the increase ratios would consequently be misleading. It is clear that if my estimates are reasonably sound, mining and cattle in northwest Mexico, and mining and cattle plus local agriculture in northeast Mexico, have increased the population in these regions more than Occidental civilization in general has brought it up in the anciently semicivi- lized and intensively farmed parts of the country. Or, so far as this result is inherently expectable, the figures may be interpreted the other way round, namely, as indicating that my estimates for the ancient population of the three regions are reasonably sound in their respective relations to one another. The Cushing map (no. 20) may therefore be considered as giving an ap- Is This includes northern Mexico, but this would deduct barely 2,000,000. "The Distribution of Population in Mexico, 11:227-242. 161 1University of California Publications in Am. Arch. and Ethn. z 0 00!.