I: O-CA tkCORDS . s;s?; ,16 ABOEGNL POPULATION OF THE N JOAQUIN. VALLEY, CALIFORNIA BY S. F. COOK ( UNIVERSITY OF CALIFORNIA PRESS BERKELEY AND LOS ANGELES 1955 ~HE ABORIGINAL POPULATION pF THE SAN JOAQUIN VALLEY, ; ~CALIFORNIA BY S. F. COOK ANTHROPOLOGICAL RECORDS Vol. i6, No.2 I UNIVERSITY OF CALIFORNIA ANTHROPOLOGICAL RECORDS Editors (Berkeley): R. L. Olson, R. F. Heizer, T. D. McCown, J. H. Rowe Volume 16, No. 2, pp. 31-80 6 maps Submitted by editors October 8, 1954 Issued July 11, 1955 Price, 75 cents University of California Press Berkeley and Los Angeles California Cambridge University Press London, England Manufactured in the United States of America CONTENTS Page Introduction .............................................. .......... 31 The population of the San Joaquin Valley in approximately 1850 ................... 33 Contemporary estimates and counts for the entire region ................... 33 Analysis based upon restricted areas ...................................... 34 Stanislaus and Tuolumne rivers .................................... 34 Merced River, Mariposa Creek, and Chowchilla River ................ 35 The Cosumnes, Mokelumne, and Calaveras rivers .................... 36 The Fresno and the upper San Joaquin rivers ........................ 36 The Kings and Kaweah rivers ........... .......... ............. .......... 38 The Tulare Lake basin ............................................ 40 The Tule River, the Kern River, and the Buenavista Basin ............ 40 The aboriginal population ....................... . ........................ 42 The Tulare Lake basin ............................................ .... 42 The Kaweah River .... ............. .. ..45 The Merced River....... ....... ...................-.. .. 48 The Kings River . . . 49 The Upper San Joaquin, Fresno, and Chowchilla rivers and Mariposa Creek..50 The Southern San Joaquin Valley .... o o..... .. 54 The Northern San Joaquin Valley... . 56 The Miwok Foothill Area ...... .......... 68 Summary and conclusions ..... ....................70 Appendix .-0..0-000-0 ...... 00.............09-71 Bibliography ............. ... 72 MAPS 1. The San Joaquin Valley from the Cosumnes River to the Tehachapi .facing page 74 2. Habitat areas 1A-2: the southern Yokuts and peripheral tribes 75 3. Habitat areas 3A-4C: the basins of the Kaweah and Kings rivers 76 4. Habitat areas 5A-6B: the Yokuts, a part of the Mono., and the southern Miwok.7............. ........6.. o o ..........76 5. Habitat areas 7A-14: the northern Yokuts, central and northern Miwok . 77 6. The Lower San Joaquin River and Delta areas ....................78 [iii] THE ABORIGINAL POPULATION OF THE SAN JOAQUIN VALLEY, CALIFORNIA BY S. F. COOK INTRODUCTION cologically the great central valley of California a single unit. Nevertheless it is convenient for purposes of this paper to divide the entire area into | portions, north and south. The vast expanse from RBluff to the Tehachapi is too extensive to cover hographically in a single exposition. Moreover, the hern tribes, the Wintun and Maidu, are physio- Fphically clearly segregated from the southern by northern extension of San Francisco Bay and the of the rivers. Hence we shall consider here only ge peoples south of the Sacramento and American er watersheds. the area possesses definite natural limits but its ict boundaries must be to some extent arbitrary. On llnorth the line has already been indicated: the south of the upper Bay and the Sacramento River as far tream as a point five miles below the city of Sacra- ito and thence easterly along the El Dorado-Amador inty line into the high mountains. This follows Kroe- Is tribal boundary between the Maidu and the Sierra ok, On the west the line starts northeast of Mt. Ilo and follows the western edge of the San Joaquin ley to the Tehachapi Mountains. On the east we in- de the Sierra Nevada as far as was reached by per- aent habitation on the west slope. The southern ex- mity is represented by the crest of the Tehachapi. The region designated embraces the territory of the ha and Sierra Miwok, the Yokuts, the Western ho, the Tubatulabal, and the Kawaiisu. From the 2dpoint of habitat the area is diversified since it ex- s from the swampy valley floor through the oak mtry of the lower foothills into the transition life- v of the middle altitudes. Perhaps an ecological regation would be desirable. Such a procedure, how- would cut across tribal boundaries and make an urate evaluation of population difficult. On the ac- xpanying maps, areas are delineated, and numbered, marily for convenience of reference. At the same e they conform as closely as is feasible with the sral subdivisions of the territory marked out by ervalleys, lakes, plains, and mountains. It should stressed that they do not necessarily coincide pre- ely with the areas occupied by specific tribes or ups of tribes. The demography of the central valley is rendered more complex by the fact that the contact with the te race took place in a series of steps rather than a single overwhelming invasion. In central Mexico, to a somewhat lesser degree in northwestern Cali- ruia, aboriginal life continued relatively untouched 11 there occurred a rapid and catastrophic occupation the entire territory. As a result, the population was affected in a uniform manner throughout and a sufficiently clear line can be drawn between aboriginal and postcon- tact conditions. In the central valley the white influence was very gradual, beginning at or near the year 1770 with the entrance of the Spanish missionaries along the coast and the infiltration of a very few foreigners into the valley. The volume of invasion increased slowly over the next three decades, but the effect was intensi- fied by the escape of numerous mission neophytes into the valley. The years after 1800 saw repeated incursions by the coastal whites who overran the floor of the valley from the Sacramento River to Buena Vista Lake. Mean- while the foothill and mountain tribes were permitted to remain fairly intact. With discovery of gold, however, tnese groups iost trieir inmmuniuity aiici were rapidly de- stroyed. Therefore, even though we oversimplify, we may say that the aboriginal population persisted in the valley proper up to 1770, in the lower foothills up to roughly 1810, and in the higher foothills and more re- mote canyons of the Sierra Nevada up to 1850. Our sources of information cover only the period during which the demographic status of the natives was undergoing change. No written record exists that de- scribes conditions as they might have been found prior to 1770. The only possible substitute would be an exami- nation of the habitation sites left from prehistoric times, but archaeological research in the area has not yet pro- gressed to the point where an adequate quantitative esti- mate of population is available. There are three primary bodies of data to which we have access, all falling within the historical period between 1770 and 1860. The first of these derives from the serious effort on the part of the Americans, who between 1848 and 1852 were entering the region in large numbers, to determine the quantity of natives surviving in the central valley. This task was performed by such men as Sutter, Bidwell, and Savage, together with several Indian commissioners, and army officers sent out by the government. To their reports may be added the statements contained in the local county histories published in the era of 1880 to 1890, as well as in many pioneer reminiscences. A second major source of information consists of the ethnographic studies made within the past fifty years, among which should be mentioned the works of Kroeber, Merriam, Schenck, Gayton, and Gifford. These investi- gators depended principally upon informants who were elderly people in the decades from 1900 to 1940. Their memories, together with their recollection of what had been told them by their parents, carry back, on the average, to the period of the American invasion or just before it. Hence their knowledge of truly aboriginal pop- ulation would be valid for the hill tribes only; yet data [311 R, i, I., k I O' t 4 It, A? 9 ANTHROPOLOGICAL RECORDS derived from them for that region is probably more accurate than can be obtained from the general esti- mates made by contemporary white men. These two types of information, contemporary American accounts and modern ethnographic material, can thus be used to supplement and check each other for the era of 1850. For conditions in the valley before 1840 we have to depend almost exclusively upon the historical records left by the Spanish and Mexicans. These consist of a series of diaries, reports, and letters, by both laymen and ecclesiastics, together with baptism lists and censuses from the coastal missions. This array of documents is to be found in the manuscript collections of the Bancroft Library of the University of California at Berkeley. It will be clear from these considerations that the population of the San Joaquin Valley can be determine with some degree of accuracy at two stages in the history of the region. The later period is at the point of intense occupancy by the Americans, at or near tht year 1850, for here may be brought to a focus the dat from both contemporary counts and the research of modern ethnographers. The earlier is for the epoch just preceding the entrance of the Spanish into Califor or just before 1770. To assess the population at this period it is necessary to bring to bear information fro all sources, American and Spanish, and to utilize all indirect methods of computation which may be approp ate. As a matter of historical interest, as well as to provide a background for the estimate of aboriginal population, the state of the natives in the period of thi Gold Rush will be first examined. 32 THE POPULATION OF THE SAN JOAQUIN VALLEY IN APPROXIMATELY 1850 CONTEMPORARY ESTIMATES AND COUNTS FOR THE ENTIRE REGION General estimates for the population of the San Joa- nValley during the period 1848 to 1855 were made Vseveral individuals. James D. Savage, one of the Oliest settlers in the Fresno region, stated in 1851 Ft the population from the Tuolumne River to the km River was from 50, 000 to 55, 000. Elsewhere he hdlfied these figures considerably (Dixon, MS, 1875) I reported the total from the Cosumnes to the Kern 1 18,100, of which 14,000 were from south of the Stanis- iasRiver. James H. Carson, another pioneer, said in S52 that "the Indians of thp Tulare Valley number srly 6,000. About half this number inhabit the moun- The other portion inhabit the plains along he rivers and lakes. tIn 1852 the Indian commissioner, 0. M. Wozencraft, ktimated for the area lying between the Yuba and the ikelumne rivers a total of 40,000 inhabitants. He htes old residents as saying that four years previ- y(i.e., in 1848) the population for the same area Ibeen 80,000. At about the same time another agent, imJohnston (1853), estimated all the Sierra and leytribes as being 80,000 strong (including both Sac- ento and San Joaquin valleys). In general magnitude e figures correspond to those given by Sutter for [region bounded by the Yuba, the Stanislaus, the Sac- ento, the San Joaquin, and the line of the foothills: 873 (Sutter, 1850). Sutter's value definitely repre- ts conditions prior to 1847. Meanwhile H. W. Wessels orted in 1853 that from the Stanislaus south there re7,500 to 8,000 persons. In the same year G. W. rbour, another commissioner, referred to the reser- ion Indians as "seven or eight thousand hungry souls." 11856, agent T. J. Henly put the aggregate population jhe Fresno and Kings River reservations plus Tulare, riposa, Tuolumne, Calaveras, and San Joaquin coun- 3 as 5,150 (Henley, 1857). It is evident that the foregoing data represent two tinctly different types of estimate: broad generaliza- ibased largely upon subjective impression and apply- to the years preceding 1847, and more narrow semi- lmate derived during the years subsequent to 1849 some attempt to make an actual count. The figures ained from the first method are certainly too high, 'ticularly for the period centering around 1850. On other hand, it may be possible that the other method Ided figures which were too low. some check on the reliability of the estimates sup- d by the various commissioners and agents may be ained from two sources, neither of which constituted irect attempt to assess population. These comprise prts submitted concerning (1) vaccinations and (2) tribution of blankets. 'During the summer of 1851 Dr. W. M. Ryer was ployed to vaccinate those Indians in the San Joaquin lley who could be persuaded to undergo the operation. ch month Dr. Ryer submitted a voucher specifying the pber of Indians vaccinated during the preceding thirty r8 and also mentioning the tribes and areas covered. These vouchers are included with other documents in Senate Executive Document No. 61, 32nd Congress, first session, 1852 (pp. 20 to 23). Some question might be raised concerning the accuracy of the figures, but there is no indication in the correspondence of the period of irregularity or dishonesty. Dr. Ryer claimed that he had vaccinated, from the Stanislaus to the south shore of Lake Tulare, 6,154 persons. A somewhat smaller area was covered by four of the eighteen treaties concluded by commissioners McKee, Barbour, and Wozencraft 1 with the California tribes in 1851. These four treaties may be designated A, B, C, and N, following the order in which they are presented in the Senate Report. Under the agreements, one of the commodities which were to be furnished to the Indians by the government was blankets. The tribes included under treaties A, B, and C were to receive a total of 3,000. In treaty N (as also in several other treaties not concerned with this area) it was stated that the Indians were to receive one blanket apiece for every person over fifteen years of age, and presumably this ratio was employed universally in the issue of blankets. Under the conditions existing at that time it may safely be assumed that the persons over fifteen years of age constituted at least 80 per cent of the total population. Therefore the three treaties first mentioned (A, B, and C) must have covered 3,750 individuals. Regarding the group em- braced by treaty N it is explicitly stated that "they may number . . . some 2,000 to 3,000." If we take the mean, or 2,500, then the total for the area is 6,250. The area included under the four treaties extended actually only from the Chowchilla River to the south shore of Lake Tulare and the Kern River, whereas the territory covered by Ryer during his vaccination tour began with the Stanislaus. Within the treaty limits he vaccinated 4,449 persons. The discrepancy between his total and that of the treaties poses no difficulty since it is apparent that, as would be expected with any primi- tive group, fewer individuals consented to be vaccinated than made known their desire to receive gifts of blankets. Hence the figure derived from potential blanket distribu- tion is probably closer to the actuality than the vaccina- tion figure. If, accordingly, we correct Ryer's report of 1,705 persons vaccinated north of the Chowchilla River to conform to the ratio found south of that stream, we get 2,398. If we add this to 6,250 the total is 8,648 for the entire strip from the Stanislaus to the southern end of the San Joaquin Valley. In summarizing general estimates and counts we may discard the very high values submitted by Wozencraft, Johnston, and Sutter on the grounds that they were either mere guesses or applied to an earlier period than that I These treaties seem to have been concluded without proper authoriza- tion from the Federal government and were never ratified by the Senate. They were incorporated in Senate Confidential Documents, June, 1852, and remained unpublished for half a century. Finally they were ordered printed in 1905 as a Senate Reprint and are now available under the title of "18 Cali- fornia Treaties." [33] I I r f? II L ANTHROPOLOGICAL RECORDS which we are considering. There are left the follow- ing figures, which seem essentially valid. Ryer and the treaties (1851) ..... .......... 8,648 Wessels (1853) ............ 7,500- 8,000 Barbour (1853) ... ......... 7,500-8,000 Henley (1856) ............................ 5,150 Since the wastage of native population in the valle~ was exceedingly rapid during the decade of the 'fi these figures are remarkably consistent. As a pr liminary value, therefore, based upon the best g eral estimates, we may set the population in 1851 8,600. ANALYSIS BASED UPON RESTRICTED AREAS Further examination and correction are now in order. It will be noted that the estimates above do not include the area traversed by the Cosumnes, Mokelumne, and Calaveras rivers. Moreover, the federal agents con- fined their calculations to those natives who voluntarily or otherwise were incorporated in the local reservation system. That many Indians were overlooked, not only in the more remote foothills, but also in the valley it- self cannot be doubted. In order to assess the population in greater detail as well as to introduce new sources of information it will be advantageous to break up the entire region into smaller units and consider these units one by one. STANISLAUS AND TUOLUMNE RIVERS We may begin with the watersheds of the Stanislaus and Tuolumne rivers, since for this area reasonably complete information is available (see maps 1, 5, and 6, areas 7 and 9.) On May 31, 1851, the Daily Alta Cali- fornia reported the treaty made with tribes of this region and stated that they were 1,000 strong. This treaty (treaty E in the California Treaties) covered the courses of the two streams as far as their junction with the San Joaquin, on the one hand, and an indeterminate distance into the hills, on the other. Ryer vaccinated in the area during June of the same year and submitted a bill for 1,010 operations. He specifies 6 bands, rancherias, or tribes which were predominantly Siakumne and Taulamni, a fact which implies that he confined his attention prin- cipally to the inhabitants of the valley and the lower foothills. In the preceding discussion it was pointed out that Ryer's figures are probably too low and that a cor- rection should be introduced. If the same ratio is used as before, the value becomes 1,420. Adam Johnston, in a statement published in 1853 in- cludes a map (Johnston, 1853, p. 242). Along the rivers shown on this map he has placed figures for population. According to him there were 900 Indians on the Stanis- laus and 450 on the Tuolumne, or a total of 1,350. These are distinctly noted as reservation Indians and hence would not have included the entire population. Four years later, H. W. Wessels reported for the same area only 500-700 persons (Wessels, 1857). These were the Indians left on the reservations. At about the same period, James D. Savage gave as his opinion that there were 2,500 people on the Stanis- laus and 2,100 on the Tuolumne (Dixon, MS, 1875). In their report in 1853 Barbour, McKee, and Wozencraft refer to a statement by a chief named Kossus that under his jurisdiction were 4,000 persons and 30 rancherias from the Calaveras to the Stanislaus. Although these two estimates are widely at variance with those submitted by the officials, it must be remembered that both Savage and Chief Kossus may have been referring to a somewhat earlier date and that both included bands and settlements higher up the rivers than was actually reached by the commissioners. Hence, although the figure of over 4,000 is likely too high, 1,000 to 1,500 may have bee too low. With respect to the strictly lowland tribes there i but little doubt that by the year 1852 the northern Yol lying between Stockton and Modesto had practically d appeared. Thus the first state census, taken in 1852, showed only 275 Indians remaining on the lower Stan laus. George H. Tinkham states that in the same yet there were only 10 families (perhaps 50 persons) left from the tribe which formerly had inhabited the regi between the Calaveras and the Stanislaus and had ex- tended eastward along the latter stream as far as Ferry (Tinkham, 1923). The valley plains can conse ly account for no more than approximately 350 perso and it must be assumed that almost all the remaining natives were living along the border of the foothillsaX higher up in the mountains. One item of some significance is the discussion of Tuolumne River tribes by Adam Johnston, written in year 1860, definitely after the Gold Rush period. He there were six chiefs in command of six rancherias, names of which he gives. These rancherias "contain fifty to two hundred Indians, men, women and childr One of these bands, the Aplache, "resided further in mountains," from which one may infer that the other were also in the mountains. At an average of 125 per band, or rancheria, this means 900 people whose ex ence was known to Johnston as late as 1860. An equi ent number can be assumed for the Stanislaus, or 1, in all. The ethnographers have given us an imposing list villages for the area under consideration, derived en tirely from modern informants. There are three of l lists, those of Kroeber (1925), Merriam 2, and Giffo which merit careful scrutiny. Kroeber's (p. 445 of 4 Handbook) includes 49 names, which he says are of lages "that can be both named and approximately loc Merriam's "Mewuk List" has 28 hames of places loc on the Stanislaus and Tuolumne. Gifford shows 49 vi which he says are "permanent," in addition to perhap twice that number of "temporary" villages and camp Gifford's list is probably the most carefully compile the three. The geographical location is indicated by counties but since his field of observation embraces Calaveras and Tuolumne counties, it coincides territ ially quite exactly with the other two lists. 2 This village list and all others herein referred to under the named Merriam are part of the extensive file of personal manuscript material collected by the late C. Hart Merriam and deposited, through the kind. ness of his heirs, with the Department of Anthropology of the Universi8l of California, Berkeley. Merriam's village lists were very carefully a piled and for many regions of the state cannot be duplicated in any publ cations which have hitherto appeared. 31 am indebted to Professor Edward W. Gifford, of the Departmento Anthropology of the University of California, Berkeley, for the privileg of examining his list of Central Miwok villages, which was obtained so8 years ago through an informant and has remained unpublished. I I 34 COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY Certain villages are recorded by all three investiga- . others by two of them, and some by only one. trning the existence of the first two groups there b little, if any, doubt. Of those appearing on only list some question might be raised. On the other the care and conservatism exhibited by all three graphers makes it very difficult to doubt the essen- alidity of their data. The discrepancies are clearly to the differences between informants and the high ability that no single informant could recall all the bited places over so large an area. Lhave tabulated below the number of villages accord- river system and according to occurrence in the mentioned. Stanislaus Tuolumne roeber, Merriam, and Gifford .......... eber and Merriam eber and Gifford.. oeber only ........... s1fford only. Ierriam only .......... Total............... 8 2 6 6 5 1 28 13 3 5 8 12 1 42 hve therefore 70 reasonably well authenticated es in the hill area traversed by the two rivers. regard to the number of inhabitants, further data provided by Gifford. His informant gave for each ent place an estimate of the number of persons nt in the year 1840. Gifford secured his material proximately the year 1915 from a man very old at ime. If the informant was then seventy-five years e, he must have been born in 1840. Hence he could cely be expected to remember population figures a date much earlier than his childhood. The names ocation of the villages themselves were at least permanent and could have been derived from the mant's parents even if not from his own memory. e it is probable that the figure furnished to Gifford e nearly represents the number of inhabitants in than in 1840. The average value for all 49 villages 08 persons. Yet 7 villages are stated to have held rsons, 11 villages 10 persons, and 3 villages 5 or persons. Such a condition argues a rapidly declining tion, for no normal aboriginal settlement is likely ye contained less than 20 inhabitants. Gifford's age of 21 persons per village must, however, be pted as representing the closest we can get to the efor the period of 1850. This means a population 8 for the Stanislaus and 882 for the Tuolumne. The is 1,470 for the foothill region. Between 300 and may be added to account for scattered remnants g the lower courses of these rivers and on the San in itself, or 1,800 for the entire area under con- ration. To summarize, we have the following estimates for *Stanislaus-Tuolumne watershed at or about the year Savage (perhaps before 1851) ..... ........ 4,600 Chief Kossus ............................ 4,000 ;,Daily Alta California, 1851 ...... ......... 1,000 YTaccinations by Ryer ......... ............ 1,420 Adam Johnston's estimate, 1853 ........... 1,350 jAdam Johnston's estimate, 1860 ........... 1,800 BI. W. Wessels, 1853 ..................... 600 Village lists ............................. 1,800 The crude numerical average is about 2,070 but since the best of the above estimates, the village lists, shows no more than 1,800, it will be preferable to set 2,000 as a fair approximation. STANISLAUS-TUOLUMNE .... .......... 2,000 MERCED RIVER, MARIPOSA CREEK, AND CHOWCHILLA RIVER South of the Tuolumne are the Merced River, Mari- posa Creek, and the Chowchilla River, all within the territory of the southern Miwok (see maps 1 and 4, areas 5E, 5F, 6). The earliest of the midcentury counts per- taining to the region is probably that of Savage (Dixon, MS, 1875) who put 2,100 persons on the Merced but omitted reference to any other stream between the Tuol- umne and the upper San Joaquin. Ryer, in a bill submit- ted July 31, 1851, claimed to have vaccinated 695 per- sons along the Merced, principally on the lower course of that river. The value, corrected according to the system adopted previously, is 977. McKee, Barbour, and Wozen- craft in a report on May 15, 1851 (Wozencraft, 1851) described the proposed reservation No. 1 between the Tuolumne and the Merced and estimated the total num- ber of Indians on both rivers as 2,000 to 3,000, or let us say 1,250 on the Merced alone. The map of Adam Johnston, dated in early 1852, shows 500 persons on the Merced, but these were reservation Indians. The state census of 1852, as cited by the Sacramento Union for November 17, 1852, gave 4,533 persons for Mari- posa County, a figure which no doubt included all the natives from the Tuolumne to the Fresno River. H. W. Wessels on August 21, 1853, wrote that there were 500 to 700 Indians on the Stanislaus and Tuolumne, 500 to 600 on the upper San Joaquin and that the entire area contained 2,500 to 3,000 (Wessels, 1857). The Merced- Fresno region therefore accounted for somewhere be- tween 1,000 and 1,700. A rough average for all these rather haphazard estimates would be 1,000 natives on the Merced watershed and another 1,000 on the Mari- posa and the Chowchilla, or 2,000 in all. We may now turn to the village lists. Unfortunately, Gifford did not work south of the Tuolumne but we have the list given by Kroeber in the Handbook (1925) for the southern Miwok and two manuscript lists of Merriam (entitled "Mewuk Village List" and "Indian Village and Camp Sites in Yosemite Valley and Merced Canyon"). For the middle Merced Valley, from a point some ten miles below El Portal to the base of the foothills, Kroe- ber and Merriam both list 14 villages, to which Merriam alone adds another 10. From El Portal to a point six or seven miles downstream Merriam has found no less than 15 villages. In Yosemite Valley itself he has located 33 villages, of which 12 are qualified as either camps or summer villages, leaving 20 which he presumes are permanent. On the upper Merced, above Yosemite, and the headwaters of the Chowchilla, Kroeber has found the name of one village and Merriam one. Clearly this area has never been investigated exhaustively. For the well- known portion of the river, therefore, there are 59 lo- cated villages. Of the 35 village sites in Yosemite and below El Por- tal, Merriam says 10 were large and 6 small. The rest are not qualified but were presumably medium to small. Gifford's average for the central Miwok of 21 persons per village in 1850-1852 may be applied directly, giving a population for the Merced Valley in the hills of 1,239. I IPA 35 ANTHROPOLOGICAL RECORDS To this may be added, according to Ryer and to Johns- ton, 50 to 600 for the lower river, making a total of 1,800. Mariposa Creek and the Chowchilla River have never been as thoroughly investigated as the Merced. Mer- riam's "Mewuk List" mentions 13 sites on each of the two streams, including the 6 given by Kroeber in the Handbook. At 21 persons per village this would mean a population of 273 for each or 546 for both, a value which appears rather low. Another approach to the problem is by way of terri- torial comparisons. There are under consideration, including those previously discussed, five small river systems, those of the Stanislaus, Tuolumne, Merced, Mariposa, and Chowchilla. Physiographically and eco- logically they are very similar since the rivers all descend the foothills of the Sierra Nevada and traverse the plain to the San Joaquin through the same life zones and at nearly the same latitude. There are, to be sure, some local differences between them with respect to how much of their course is favorable for village sites, but in the aggregate the similarities outweigh the differ- ences. It is of interest, therefore, to estimate the vil- lage density along each watercourse. This value can be computed with a fair degree of accuracy by measuring on a large-scale map the length of each river and its principal affluents from the edge of the plain to the upper limit of known permanent habitation. The village numbers can be derived from the lists of Kroeber, Gif- ford, and Merriam. Estimated River length (mi.) Villages Stanislaus Tuolumne Merced. Mariposa . Chowchilla . 85 105 125 40 65 28 42 59 13 13 Villages per river mi. 0.33 0.40 0.47 0.32 0.20 tion must have been very small in the period of the ea 1850's owing to extreme attrition suffered from the Si ish and particularly from the gold miners. Kroeber only 20 villages on all three streams, most of them a the Mokelumne. Merriam adds another 3, making 23 all. At Gifford's population value this means 480 pen The official sources are of little help since none of tb agents or commissioners reported specifically on the area. Evidently there were too few survivors among natives to warrant the trouble of placing them undert reservation system. Savage assessed the population on the Cosumnes, Mokelumne, and Calaveras at 1,000 each (Dixon, MS, 1875) but it is likely that he was thinking in terms of I days before the Gold Rush. F. T. Gilbert (1879, p. 1 says that the Mokelkos, by which he means all the Ind between the Mokelumne and the Cosumnes in the hills as far as Stockton on the plain, had 12 rancherias of to 300 each and numbered about 3,000 in all. He, how ever, was referring specifically to the period "before the advent of Sutter." Likewise J. D. Mason (1881, p. 256) ascribed to the same tribe "nearly a score of toi with a total of 3,000 to 4,000." In amplification Gilbert says that in 1850 rancherias lined both banks of the Mokelumne from Ahearn's (near Lodi) to Campo Seco (near the present Pardee Reservoir), and that they no bered then about 2,000. In 1852, however, there were only 4 rancherias left, with 390 inhabitants. Gilbert was referring explicitly to the lower cours of the rivers, whereas the villages cited by Kroeber, definitely above this region in the foothills. We maya cept Gilbert's figure of 390 on the lower Mokelumne, which may be added 110 for the lower Cosumnes and Calaveras and 480 for the upper villages, making a t of 980 or, let us say, 1,000. COSUMNES-MOKELUMNE-CALAVERAS ..... Mean .. 0.34 The figures, considering physiographic differences and varying coverage by ethnographers, are quite con- sistent. Only that for the Chowchilla appears unduly low and this in turn may be referable to an incomplete count by Merriam. It is reasonable to concede this possibility and assume an actual count of 0.30 village for each mile of this stream. On 65 miles of river front there would thus have been 19.5 villages. This consequently means, using Gifford's population average of 21 per village, 273 inhabitants on the Mariposa and 410 on the Chow- chilla. These may be added to the 1,800 calculated for the Merced, making a total of 2,483. The very approximate value derived from general estimates was 2,000 persons. The village data are prob- ably more accurate and may be rounded off to an even 2,500. MERCED-MARIPOSA-CHOWCHILLA ... 2,500 THE COSUMNES, MOKELUMNE, AND CALAVERAS RIVERS The northern Miwok held the upper reaches of the Mokelumne plus most of the Cosumnes and Calaveras (see maps 1, 5, and 6, areas 10, 11, 12). The popula- THE FRESNO AND THE UPPER SAN JOAQUIN RIVERS We next turn south and consider the valleys of the Fresno and upper San Joaquin rivers (see maps 1 and areas SB, 5C, 5D.) There are three counts or estim pertaining to this area specifically. The first is that Savage, who does not mention the Fresno but puts 2, persons on the upper San Joaquin. The second source the May 29, 1851, issue of the Daily Alta California, which carried a letter written by an unidentified offic who was with the Indian commissioners and in fact m have been G. W. Barbour. This officer refers to the treaty made with the natives between the Chowchillad the Kings rivers and says that "the total is probably' 3,000 Indians.' The third is Adam Johnston, who on map ascribed 1,200 people to the Fresno and 1,000to San Joaquin (Johnston, 1853). The average of the thr~ estimates is 2,633. W. M. Ryer submitted three reports for the terril below the Merced and north of the Tehachapi Mountai4 In each he mentions the tribes vaccinated (Ryer, 185~ There are 45 in all, but 8 tribal or rancheria names indeterminate and there are many duplicate names at the rest. Putting all three lists together we can get 2 recognizable tribal names, of which one is southern wok, four are Mono, and the others Yokuts. The tot vaccinations performed numbered 4,451, or, correct to conform to the figures based on blanket distributio I - -W Ii I I I I I I I 0 1 I 0 3 3 6 COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY 1,255, an average of 232 per tribe. To allow for the ntribal and unrecognizable names on Ryer's lists his value may be arbitrarily reduced to 200. Ryer etions in the Fresno-San Joaquin area the following: owchilla, Chukchansi, Heuchi, Pitkachi, Goshowu, ma, Dalinchi, Pohinichi (Miwok), and Posgisa no). The Pohinichi should be excluded since they e already been considered in connection with the thern Miwok. The other nine, reckoned at 200 per- 3 per tribe, would represent an aggregate of 1,800. wever, Kroeber (1925, p. 481, and map, p. 526) w four other Yokuts subdivisions within the same ritory: Hoyima, Wakichi, Kechayi, and Tolichi. tough Ryer may have included these under other 1 names they perhaps ought to be included here, making the total 2,600. For villages there are two sets of sources. The at pertains primarily to the Yokuts, covers a terri- r substantially coterminous with that seen by the temporary observers mentioned above, and is found he work of Kroeber (1925), Gayton (1948), and Latta 99). The second set of villages is confined to the no and is derived from Gifford (1932) and Merriam. The first group of authors list villages for the 13 s mentioned in the preceding discussion, 49 in all an average of 3.77 per tribe. With respect to size re is reason to believe that the settlements in this a, even in the early 1850's, were considerably larger those described by Gifford for the central Miwok. eestimate of Adam Johnston of an average of 125 per cheria on the lower Tuolumne has already been men- ed. H. W. Wessels in 1853 wrote that the Pitkachi the Noo-to-ah, a Mono group, had 500 to 600 souls asels, 1857). Half of these, or 300, may have been achi, a tribe for which Kroeber lists 3 villages. would have meant 100 per village. Merriam credits age with the statement that in 1851 the Kechayi had 0 people. Kroeber, Gayton, and Latta list 6 differ- villages for this tribe or, according to Savage's es, 167 persons per village. Ryer's total of 2,600 rated among 49 villages, would yield 53 persons Although it is probable that the values computed the statements of Johnston, Wessels, and Savage too high, that derived from Ryer may be somewhat low. An intermediate figure of 70 inhabitants per e for the valley and lower foothills would perhaps e as close as we can get to the truth. This, with 49 sge, gives 3,430, somewhat more than the 2,633 as the average of the general estimates. Ihabiting the higher foothills and extending to the r limit of habitation from the San Joaquin to the ah rivers were the Western Mono. This tribe just above the Yokuts and at points was in very e association with them. As a whole the Western constitute a racial and ecological unit and as such probably preferable to consider them as a single ation entity than to segregate them by rivers, as been done for the Miwok and the Yokuts. It will be asary, therefore, to digress for this purpose and equently return to the discussion. e classic ethnographic work on the tribe, and the work which contains any numerical data, is that of rd (1932) on the North Fork division of the Mono. is supplemented by Merriam's manuscript entitled nache Tribes, Bands, and Villages." Gifford gives names (text and map) of 67 North Fork villages, or, b prefers to call them, hamlets. These were quite e either those of the Miwok or of the valley Yokuts, very much smaller and subject to an extraordinary turnover in inhabitants. Gifford makes it very clear that each family was accustomed to move every few years from one settlement to another and that sites were being continually occupied and deserted. The 67 names are therefore no criterion for population. For the time of the American occupation Gifford estimates the number of persons in the group or subtribe as approximately 300, which, divided directly by 67, would give the absurd average of 4 persons per hamlet. However, a more de- tailed analysis is possible. Of Gifford's 67 names, 2 may be deducted as being only camps, leaving 65 which at some period were per- manently occupied. In his Appendix A (pp. 57-61) he lists the sites, together with the number of houses in each and the number of males and females inhabiting them. From these data may be computed the total num- ber of families and the mean number of persons per family. There were 227 families in all. However, 36 of these are listed two or more times by virtue of moves made from one hamlet to another, which were remem- bered by Gifford's informants. This would leave 191 families for the subtribe, provided Gifford recorded all the moves. But Gifford clearly implies that he did not, since his informants could not remember them all. Hence the number of families must be further corrected. In Appendix A, 15 out of a total of 65 hamlets were con- cerned in the moves recorded. These 15 hamlets were inhabited at different times by 61 families but many of these, owing to frequent change of residence, are repeti- tions. Actually there was a total of 24 different families rotating among the 15 villages. Now if in the other 50 hamlets the same process was going on, although Gifford was not able to record the moves, it is legitimate to apply the same ratio as is in fact found for the 15 ham- lets. The crude total of 227 families must therefore be reduced to 89. From Gifford's complete list it can be determined that there were on the average 4.93 persons per family. This gives a population of 439 for the period remembered by the informants. On general grounds it is to be expected that the con- ditions reported by Gifford's informants were not entirely aboriginal. This is also indicated by the value of 4.93 persons per family, which is somewhat too low for a stable prehistoric population. Moreover, Gifford him- self states that there were formerly 44 more houses than there were in the time referred to by the informants (figures given individually for the hamlets in App. A). About 1850 there were 227 houses, and if 44 are added, the aboriginal number would have been 271. Each house may be assumed to have held one family but the houses were probably occupied in rotation. The crude estimate of 271 houses or families, each containing (according to aboriginal standards) a possible 6 persons, would mean a total of 1,626 for the subtribe. If, however, we apply the correction factor for family moves we must reduce this estimate to 640, a far more reasonable figure. For the North Fork Mono, therefore, we may accept as the best estimate obtainable a population of 440 for the period near 1850 and of 640 for precontact time. The other subtribes of the Mono provide no data com- parable with those available for the North Fork group. Some method of extrapolation is thus called for. The village method is very unsatisfactory. Kroeber says substantially nothing on this score and Merriam, although he lists 19 villages for the North Fork Mono, gives no more than one or two or, at the most, half-a- dozen names for each of the other groups. Tribal dis- tinctions are also very confusing. Kroeber in the Hand- book mentions 6 Mono subtribes: North Fork group, 37 ANTHROPOLOGICAL RECORDS Posgisa, Holkoma, Wobonuch, Waksachi, and Balwisha. Merriam subdivides to a much greater extent. His grouping may be expressed essentially as follows: 1. Pogesas ..........equivalent to Kroeber's Posgisa 2. Nim ...... . synonymous with the North Fork subtribe 3. Kwetah .... . included in Kroeber's Holkoma 4. Kokoheba. . . . included in Kroeber's Holkoma 5. Holkoma ... . included in Kroeber's Holkoma 6. Towincheba . . included in Kroeber's Holkoma 7. Toinetche. - . . included in Kroeber's Holkoma 8. Tsooeawatah . included in Kroeber's Holkoma 9. Emtimbitch * * classed by Kroeber as a Yokuts tribe 10. Woponuch. . . . equivalent to Kroeber's Wobonuch 11. Wuksatche . .. equivalent to Kroeber's Waksachi 12. Padoosha. . . . equivalent to Kroeber's Balwisha Nos. 5 to 8 inclusive are consolidated by Merriam as smaller groups within a main group or subtribe called the Toohookmutch. Concerning these Merriam says: "Large tribe on King's River. On both sides but largest area on north side. Contains many ran- cheria bands.' Using Merriam's nomenclature, the Nim are general- ly conceded to have been the largest single subtribe. For this we may take as a working base line the previous estimate of 440 persons and Merriam's list of 19 villages. Elsewhere Mlerriam mentions the names of the following: Toinetche 3 villages, Holkoma 4, Woponuch 9, Emtim- bitch 2, Waksache 1, Kokoheba 1, and Toohookmutch 10. The total is 30. By direct proportion the inhabitants should have numbered 695 but this would leave five of Merriam's groups with no population at all. If we con- sider that the Toohookmutch complex plus the Kokoheba and Kwetah are the equivalent of Kroeber's Holkoma we find 18 villages, which implies 416 people. Merriam cites 9 villages or, at the same ratio, 208 persons for the Wobonuch. The total for these three of Kroeber's subtribes would then be 1,064. If we guess that the re- maining groups contained 500 persons, the figure for the Mono in 1850 would reach the vicinity of 1,600. In view of the paucity of the village data for all sub- tribes except the North Fork group it is proper to fall back on area-density comparisons. The territory actu- ally inhabited by the Mono is vague, particularly on the eastern border approaching the high mountains. Never- theless Merriam's villages furnish a fair guide in out- line, since his findings, while very incomplete, can be regarded as a reasonably well distributed sample. More- over, his descriptions of tribal boundaries and village locations appear to be very accurate. When we plot the latter on a large scale map, therefore, the outlines of the Western Mono area become sufficiently distinct. There are two possible variants of the method, one by computing stream distances and the other by measur- ing areas. Both must of course rest for their basis on the data for the North Fork subtribe. This in turn may entail some error, since the North Fork group may have been not only the most populous but also the densest. For the North Fork territory the distribution shown by Gifford on his map (1932, p. 18) is used plus the area of Bass Lake, since Merriam has found that there were once villages there. The southern and eastern boundary is taken as the San Joaquin River, because the North Fork Mono apparently did not cross to the left bank of the river. Several miles on Little Fine Gold Creek must also be included, according to Gifford's map. In this region there were approximately 60 miles of streams, including the San Joaquin River itself. With population of 440 this means 7.33 persons per stream mile. The stream mileage for the San Joaquin system as a whole within the Mono boundaries amounted to 1 miles. Hence the population in the same ratio wouldb 733. The analogous values for the Kings River syste are 150 miles and 1,100 persons and for the Kaweah drainage 75 miles and 550 persons. The total populati4 would then be 2,383. If areas are calculated from the township lines on map, that covered by the North Fork Mono is appronx mately 150 square miles and that of the Mono collect is 1,090 square miles. Equating the North Fork popul tion to the entire area gives for the Mono as a whole 3,195. We may now return to the consideration of the Fre San Joaquin region. For the lower courses of these ril mainly in Yokuts territory, three values were derive 2,633 from general estimates, 2,600 from Ryer's vac tions, and 3,430 from village lists. We may accept th average, 2,890. For the Mono of the upper San Joaq the best estimate, as given above, is 733. The total io 3,623 or, rounded off to the nearest hundred, 3,600. F RE SNO - SAN JOAQU IN . ..........3, i THE KINGS AND THE KAWEAH RIVERS The Kings and Kaweah watersheds may be conside at this point in their entirety (see maps 1 and 3, are 3 and 4). If we deduct 730 persons for the San Joaquin basin, the estimates for the Mono on the two former streams was estimated by the village method as 870, the stream mileage method as 1,653, and by the area method as 2,465. If one regards some of these figures as too high, he should bear in mind that the natives o the Kings and Kaweah rivers were exposed to more i tense contact with the white race for a longer period before 1850 than those on the relatively sheltered No Fork, and that their extermination proceeded with tre mendous velocity after that date. This fact may well account for the inability of either Kroeber or Merria to find more than a few villages on the Kings and Kaw as compared with the success of Gifford on the North Fork. The more exposed villages may simply have di peared before the era reached by the memory of mod informants. If this is so, the stream mileage and are comparisons may be more accurate than otherwise be supposed. Considerable evidence for a rather high population this region at the midpoint of the nineteenth centuryis to be derived from contemporary accounts and from statements obtained by Merriam. Among the papers in his collection is a clipping from the Stockton Record February 21, 1925, containing an article by WalterF of the United States Park Service. Included is an acc of early days on the Kaweah by Hale D. Thorpe, obt by Mr. Fry in 1910. Mr. Thorpe says: When I first came to the Three Rivers country 1856, there were over 2,000 Indians living alongth Kaweah River above Lemon Cove. Their headqua camp was at Hospital Rock . . . There were over Indians then living at the camp. I 38 COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY g The Indians were mostly Mono, of the Patwisha tribe. Merriam evidently consulted Mr. George W. Stew- concerning this matter, since the file also contains etter from Mr. Stewart written to Dr. Merriam on crh 29, 1926, stating that this camp was occupied l during the summer and that there were several ermanent rancherias along the stream. Mr. Thorpe's ge of 2,000 probably refers to Indians of all tribes, Wce by 1856 all the natives from the delta region had n driven up the river. The 600 at or near Hospital k were undoubtedly Mono. Inhis manuscript entitled "Ho-lo-ko-ma, Cole Spring, e Ridge," Merriam has the following to say: Ben Hancock, who has lived in this country about 40 years [in 1903] tells me that when he came here there were about 500 Indians (Ko-ko-he-ba) living in Burr Valley, a few on Sycamore Creek, 600 or 700 at Cole Spring (Hol-ko-mahs) and about the same i number (also Hol-ko-mahs) in Fandango Ground and in Haslet Basin.. .. He says a very large village .was stretched along the south side of King's River two or four miles below the mouth of Mill Creek and for half a mile the dome grass-covered houses nearly touched. There were also large villages on Dry Creek and one above the forks of King's River .some miles above Dry Creek. The tribe at the forks is now extinct.' (There is only one survivor of the Burr Valley kibe) Although the numbers may be somewhat exaggerated, kere is no reason why the essential correctness of this :count should be questioned. This is particularly true -view of the circumstantial detail with which it is re- Drded. The Kokoheba must be regarded as having a Fpulation of at least 500 and the Holkoma of 1,200, aking 1,700 for the Kings River Mono. If there were 3Oon the upper San Joaquin and 600 on the upper hweah and if 500 are added for the Emtimbitch- fobonuch group, the total is 3,530, not much more m was calculated by means of area comparisons. For the Kings River as a whole the estimates of 50 to 1853 indicate a substantial Indian population. vage (Dixon, MS, 1875) sets the number as 2,000, a pmarkably low figure for him. G. W. Barbour and lam Johnston (Sen. Ex. Doc. 4, 1853, pp. 253-256) th state that for the purpose of consummating treaties 0 Indians came to Camp Belt on the Kings River in 51. Lt. George H. Derby in his careful account of the uthern part of the central valley in 1851 says that hre were 17 rancherias on Kings River, "numbering -all about three thousand including those situated long the hills in the vicinity" (Derby, 1852). Many of se were Choinimni, but at least half must have been bono. If we accept Derby's count of 17 villages for 3,000 rsons, the average number of inhabitants per ranch- iawould be 177. For the area farther north the equi- 1ent number was taken as 70. There is reason to 'eve that for the basins of the Kings and Kaweah rby's figure of 177 is a closer approximation. Ben neock's description of the village on the Kings below llCreek is very graphic and explicit (see citation ve.) If the "dome-grass covered houses nearly ehed" and stretched along the river in only a single , and if each occupied 50 linear feet, then there st have been 52 houses in half a mile. Allowing 5 *rons per house, in accordance with Gifford's data for the North Fork Mono, the inhabitants must have numbered 260. One of the rancherias seen by Derby was Cho-e-mime which had 70 "warriors.' Reckoning the "warriors" as half the males the population would have been 280. Derby says the village of Notonto (of the tribe Nutunutu on the south bank of the lower Kings) had 300 inhabitants. These places were of course relatively large and important and do not represent the general average. However, the village of Notonto must have reached fully 150 persons. Apart from the Mono, the tribes located on the Kings River were all Yokuts, as follows: Aiticha, Apiachi, Wimilchi, Nutunutu, Wechihit, Toihichi, Chukomina, and Choinimni. For these the modern ethnographers Kroeber, Gayton, Latta, and Stewart have been able to locate and identify 25 villages inhabited during the youth of informants. Since this covers a somewhat larger terri- tory than was seen by Derby, the correspondence in num- ber of rancherias is reasonably close. At 150 persons per village the population would be 3,750. If we add 1,700 for the Kings River Mono, the total is 5,450. However, there may have been some overlap, so this figure may be reduced to 5,000. It should be noted that the area em- braced within this estimate includes the Kings River basin as a whole, together with that of all its affluents. The Kaweah River from Lemon Cove to the town of Tulare diverges to form a delta, which originally con- tained a very large native population. At the time of the American occupation there had occurred a material re- duction, which was accelerated by the fact that the region provided excellent farming land for the entering Ameri- cans. Hence the value for the population in 1850-1853 must be relatively low in comparison with preceding decades. In May, 1851, according to G. W. Barbour (1853, pp. 253-255) there were 7 tribes on the Kaweah, and 1,200 people came to treat with the commissioners. These tribes included the following: Chunut, Choinok, Wolasi, Telamni, Gawia, Yokod, and Wukchamni. Of these, the first, the Chunut, inhabited the shore of Lake Tulare and should not be included as a Kaweah River tribe. The estimated population of the remainder would, therefore, be approximately 1,000, if the figure of the commissioners is to be taken without qualification. With respect to the individual tribes there are a few scattered bits of information. Derby (1852) mentions three rancherias or bands in the area: Cowees (Gawia) with 200 people, Thulime (Telamni) with 65 men, or roughly 200 people, and Heame-a-tahs (Telamni) with 200 people. Merriam in his "Yokuts List" cites an in- formant who said that the Wukchumne "used to number" 5,000 and occupied the valley now called Lemon Cove and up and down the Kaweah River. Clearly this is an extreme overestimate, unless the informant was re- ferring to the period prior to 1800. Finally Merriam cites a letter by Lt. N. H. McLean, which states that the "Four Creeks Country" included the "Cahwiahs, Okuls, Choinux, Wicktrumnees, Talumnies" and in 1853 had not over 1,200 souls.4 It thus appears quite evident that the six Yokuts tribes, except perhaps the Wukchumni, had no more than 200 persons apiece during the era under consideration. From modern informants Kroeber, Gayton, and Latta have ob- tained for the Choinok, Gawia, Telamni, Yokod, and Wolasi collectively the names of only 8 villages. Assum- ing the Kings River value of 150 persons per village, 4 Merriam's manuscript entitled "Yokuts List" mentions a report from Lt. N. H. McLean, dated July 12, 1853, to H. J. Wessels, on file in "Old Files Division," Adjutant General's Office, Washington, no. H369. As far as I am aware, this letter has never been quoted elsewhere. I 39 ANTHROPOLOGICAL RECORDS which seems to be confirmed by Derby for the Kaweah River also, this means 1,200 persons for the five tribes. Gayton and Latta, however, find 15 names for the Wukchumni, which would indicate a population of 2,250. Such a figure is highly unlikely. It is probable that earlier times are referred to by the informants or that there is confusion among tribal affinities. Alterna- tively, the Wukchumni villages may have followed the style of the hill-dwelling Mono and have been very much smaller than has been indicated by Derby for the valley- inhabiting Yokuts. Since we cannot resolve the difficulty with the data at hand, it is better to accept the practi- cally unanimous opinion of contemporary white observ- ers that the population below Lemon Cove did not exceed 1,200 in 1851. To these must be added the 600 Mono previously discussed, making a total for the Kaweah River as a whole of 1,800 persons. If the two river basins are considered jointly, the method of area comparisons as applied to the Mono, estimates by government officials, accounts by early pioneers, and the village lists secured from modern informants all apparently agree that the population of the region reached several thousand as late as 1850 and 1851. We may therefore accept the total of 6,800, or 5,000 on the Kings and 1,800 on the Kaweah. KINGS-KAWEAH ....... 6,800 THE TULARE LAKE BASIN The shores of Tulare Lake (see maps 1 and 2, area 2) were aboriginally inhabited by three tribes, the Tachi, Wowol, and Chunut. In close proximity on the northeast were the Nutunutu, but since the latter have been in- cluded with the Kaweah River tribal group, they must be omitted from consideration here. Savage allocated 1,000 Indians to Tulare Lake (Dixon, MS, 1875). Mc- Lean said there were 1,000 Indians "on the lakes" in 1853, 500 of which were "Notontos," leaving 500 fbr the "Taches" and "Tontaches" (Merriam collection). The most reliable account is that of Derby (1852). However, Derby in his terminology confused the Tachi with the Chunut, in which mistake he has been followed by Mer- riam (under title "Indians of the Tache Lake Region in 1850," MS). Derby makes it clear in his account that he found the village of Sintache (population 100) at the northern side of the then nearly dry Lake Tontache, that is to say on the southern shore of the big Lake Tache (Tulare). These were probably Chunut. There was also a small rancheria which he called Tinte-Tache at the south side of the same lake, i.e., Tontache (popu- lation 50). These are likely to have been Wowol. The tribe known to ethnographers as the Tachi were north of the big lake (i.e., Lake Tache or Tulare). Their chief told Derby that they had 800 people and that their principal rancheria was northwest of the lake (popula- tion 300). Since Derby also applies the name of Tinte- Tache to the northwest village, it is clear that there were two rancherias of this name included in his ac- count. Kroeber and Gayton mention a total of 8 villages for the Tachi. If one of these had 300 people, as Derby states, then the average population of the other seven was approximately 70. This agrees with Derby's two southern rancherias of 50 and 100 persons respectively. For the Chunut Kroeber, Gayton, and Latta all mention the village of Chuntau. Kroeber mentions one other, Miketsiu. This would indicate a population of nearly For the Wowol the ethnographers give three villagei an implied population of, say, 220. The total for th4 lakes would then reach 1,170, or very close to the eral contemporary estimate of 1,000. The figure 1, may be accepted as a compromise. TULARE LAKE BASIN ....... TULE RIVER, KERN RIVER, AND THE BUENAVISTA BASIN The remaining Yokuts territory is large in area relatively small in population. It includes the water sheds of the Tule and Kern rivers together with tho of the small creeks between (Deer, White, and Posc creeks) and Buenavista Basin south of Bakersfield maps 1 and 2, areas 1F and 1G). The tribes placed Kroeber in the region are the Koyeti, Yaudanchi, Bc ninuwad, Kumachisi, Bankalachi (Shoshonean), Pali yami, Yauelmani, Hometwoli, Tuhohi, and Tulamni G. W. Barbour (1852), in a letter dated July 28, said that the area bounded by Buenavista Lake, Tuli River, and Paint Creek contained a population of ab 2,000. Savage (Dixon, MS, 1875) said there were 1,' on the Kern River and Barbour (1853) stated that, fc treaty-making purposes in 1851, 1,700 congregated Paint Creek below Tule River. The villages listed b Kroeber, Gayton, and Latta for the various tribes a as follows: Bokninuwad 2, Hometwoli 3, Koyeti 8, Kumachisi 6, Paleuyami 7, Tuhohi 1, Tulamni 3, Y danchi 8, and Yauelmani 7. The total is 45. The vi size indicated by Derby for the Tulare Lake Basin adjacent valley territory is 60 or 70; that for the }l regions is undoubtedly smaller. If we take 40 persa as the average village population, the aggregate for region would be 1,800 and if we take 50 persons, iti 2,250. We cannot be far in error in setting the popu tion at Barbour's value, 2,000. TULE-KERN-BUENAVISTA 2 On the basis of gross estimates and semicompro sive counts for the entire region the population for San Joaquin Valley and neighboring foothills in 1851. tentatively set at 8,600 (p. 34). The detailed consid4 tion of the seven subdivisions of the entire region, above, leads to an estimate of 19,000, as set forth the following recapitulation. Stanislaus-Tuolumne ..... .............. 2,000 Merced-Mariposa-Chowchilla ... ........ 2,500 Cosumnes-Mokelumne-Calaveras ....... 1, 000 Fresno-San Joaquin .................... 3,600 Kings-Kaweah ......................... 6, 80 Tulare Lake Basin.......... a ....... S.. 1,100 Tule-Kern-Buenavista ..... ........... 2,000 Total .......................... 19,000 It is believed that this total is more reliable tha previously given for several reasons. In the first it is derived from a careful consideration of all av able sources in detail. In the second place, the prel inary estimate was weighted heavily by the reportsi government officials, who saw principally those Int with whom they were able to make treaties or who U 40 COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY ey were able to collect on reservations. That this ems to represent less than one-half the natives in territory is not surprising. In the third place, ent investigations by ethnographers have brought to light many local groups which were overlooked by contemporary observers, official and civilian alike. We may therefore accept the figure 19,000 as the popu- lation of the San Joaquin Valley surviving in 1852. 41 THE ABORIGINAL POPULATION In order to estimate the aboriginal population of the San Joaquin Valley it is necessary to rely very heavily on the accounts furnished by the colonial Spanish and Mexicans. These were primarily ecclesiastics and military men who entered the territory for purposes of exploration, to seek new converts to the missions, or to chastise stock raiders. The more responsible of these left circumstantial and, as a rule, fairly accur- ate narratives and diaries. Unless there is in a partic- ular case some reason for doubt, their statements may be accorded considerable confidence. At the same time two circumstances often render the interpretation of the data derived from these docu- ments difficult. The first is the lack of consistent designations for places. During the process of opening up the area it was inevitable that rivers and villages should be assigned different names by one explorer after another and that the same name should be applied to more than one locality. The second is that during the early phases of exploration some localities were visited repeatedly, whereas others were overlooked perhaps entirely. Hence the information available to us is very uneven; it permits us to achieve a reason- ably clear idea of the population of one region but leaves another almost completely blank. As a result extrapolation by area is almost unavoidable. It must also be constantly borne in mind that the Spanish records themselves do not give us an absolutely undistorted picture of aboriginal conditions. It is very evident from the reports of the earliest official pio- neers, like Garces in 1776 and Martin in 1804, that from 1770 onward and perhaps even before white men had straggled into the valley and had consorted with the natives. There is reason to believe that these un- known interlopers may have introduced diseases which adversely affected the population and may have initiated a process of general social disruption. The best we can do is get as close to the prehistoric condition as the records allow. Two other demographic consequences arise from this very early white contact. In the first place, the documentary record, if we ignore Garces for the moment, runs nearly continuously from 1804 to approx- imately 1840. During this long period an uninterrupted change was going on among the native population: the population was continually decreasing. Hence later re- ports tend to deviate from earlier ones, and indeed may show an entirely new state of affairs arising within a very few years. In the second place, the deteriora- tion in certain areas took place so rapidly in the first part of the nineteenth century that any information secured from informants alive since 1900 is com- pletely useless. Unless very good documentary evi- dence is available for such areas, there is no recourse but to fall back on the method of extrapolation and area comparisons. The principal Spanish accounts upon which we must rely include a few which have been published. Most of them, however, are to be found in manuscript form in the Bancroft Library, University of California, Berkeley. Some of them were translated for an unpublished manu- script by the Late Professor Herbert I. Priestley and several were translated for Dr. C. H. Merriam. M riam's translations are on file in his manuscript co tion. The citations to these accounts, published and published, are given in the manuscript section of the Bibliography. In this text they are referred to, witho further citation, by the author's name and date. THE TULARE LAKE BASIN We may commence detailed consideration of the i original population with the Tulare Lake Basin, whid was inhabited in 1800 by three Yokuts tribes, the Wo Tachi, and Chunut (see maps 1 and 2, area 2). The official visitor to the area was Father Juan Martin w entered the valley in 1804 in search of new mission sites. He found the principal village of the Wowol, w he called Bubal. This rancheria, he said, contained less than 200 children. It was visited again in 1806 b Moraga, who found 400 inhabitants. Eight years late Father Cabot passed the site and found 700 people. S sequently, it was visited by Ortega in 1815 and Estud in 1819 but these writers gave no population figures., Since no other village was ever recorded by name in' territory of the tribe, it is safe to assume that there was no other, at least of permanence and reasonably! large size. Gifford and Schenck (1926), in their discussion of history of the southern valley, conclude that because village was reported as having 400 persons in 1806 a 700 in 1814 there was a real increase in population db ing the intervening eight years. This they ascribe to fugitives from the coastal missions who entered the valley as refugees. The opinion expressed by these authors may serve as the starting point for discussio of certain general problems which are encountered i attempting to estimate the aboriginal population of thi valley. In 1804 Martin saw 200 children. If we knew the of children to adults, we could easily compute the tol number of inhabitants. The age of "children" was va ously estimated in colonial New Spain, indeed all the way from seven to fifteen years. The early Californi, missionaries used approximately fourteen years for males and twelve for females. In 1793, however, th system was standardized for doctrinal purposes. In ans, both gentile and converted, were designated as children if they were under ten years, i.e., in the a bracket from 0 to 9 inclusive. Hence all the clergy formed to the method in so far as they were able an unless they specified otherwise. There are certain data available which permit us estimate rather closely what proportion of the popula tion in California should be regarded as falling withi the category of children. Within the missions the anr censuses enable us to compute with accuracy that th individuals under the age of ten years, between the 1782 and 1832 averaged 21.4 per cent of the total po tion (Cook, 1940). This value is relatively high and not conform to gentile, or aboriginal, conditions. Wi regard to these we have information from archaeolo sources. In the Museum of Anthropology at Berkeley [42] i I II I I'Ii I I II I I r COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY there are several hundred skeletons excavated from -habitation sites in central and northern California, the wes of which have been determined and which consti- btte a fair cross section of the native population during bJhe centuries immediately preceding invasion by the ,,white man. Of these skeletons 22.6 per cent represent !"persons dying under the age of twenty years, and per- haps 10 or 15 per cent persons dying under the age of iNten, Further light is shed by the baptism records of the iWssions San Jose and Santa Clara (these are discussed in greater detail in a later paragraph) which list gentile baptisms according to village and distinguish between men, women, and children. In the two missions, from pproximately 1805 to 1833 there were baptized a total o5,217 persons from villages in the valley region. Of ;these 930, or 17.8 per cent were children and 1,939, or 37.1 per cent were listed as men. The sex ratio is 0.826. Evidently the natives captured and brought to e missions do not give us a completely true picture fthe composition of the aboriginal population, despite e large sample at our disposal. It is highly probable ht (1) the natural sex ratio was nearly unity and (2) nany of the men were killed in warfare or escaped the 4lutches of the convert hunters. Therefore we are justi- ledin setting the number of men equal to that of the women. If we do this, the population represented by the 1,217 conversions was actually 5,626, of which men dwomen each constituted 41.8 per cent and children 6.4 per cent. Finally, we have figures from Zalvidea (MS, 1806) "with respect to villages at the extreme southern end of theSan Joaquin Valley. (These are discussed subse- quently in connection with the population of that area.) ttwo of these, after adjusting for disturbed sex ratio, iefound respectively 13.5 and 9.6 per cent children. |owever, Zalvidea's account states specifically that in sese villages he carries the age of childhood only rough the seventh year. If he had counted as children sih0e under ten years of age, the percentages would iaturally have been higher. The data just set forth render it abundantly clear Ibatthe children constituted between 10 and 20 per cent the aboriginal population. Since the exact value can ler be ascertained, it is wholly reasonable to estab- hthe arbitrary figure of 15 per cent. If we apply this ctor to Bubal the result is not less than an aggregate 1,333 persons, much greater than the value set by oraga in 1806. With respect to the sugge-stion of Gifford and Schenck the number of inhabitants of Bubal had been aug- nted between 1806 and 1814 by refugees from the ssions the following points may be noted. In the first ce, it has been possible to show (Cook, 1940) by ans of the mission censuses that in 1815 the cumula- ye total of fugitives reported by all the missions in the lony amounted to 1,927 persons. Of these a great many o ran away in the earlier years were deceased. Many er went to the valley at all and the remainder were tributed from Sacramento to Bakersfield. It is highly ely that as many as 300 would be concentrated at village such as Bubal. In the second place, the jority of the fugitives who did reach the village or its cinity were former inhabitants of the locality who were erely returning to their old homes rather than coastal Mans, who would have constituted real refugees. On whole, therefore, and this conclusion applies through- the valley, true increase of population by immigra- n of foreign fugitives was negligible. A further problem of importance illustrated by our data for Bubal is the extent to which population estimates for villages were affected by local fugitivism or tempor- ary scattering of the natives at the advent of the Spani- ards. Very frequently the explorers left notations that the inhabitants of a certain rancheria had fled, or that many were absent. It seems clear that even by the year 1800 the natives were all too well aware of the purpose of the missionaries and soldiers and took measures to defeat that purpose. For this reason, remarkable as it may appear, the largest estimates are likely to have been the most accurate. Returning now to the population of Bubal we find Martin counting "no less" than 200 children in 1804, indicating a total number somewhere in the vicinity of 1,300, although most of the adults apparently had ab- sconded. In 1806 the same situation arose and Moraga found only 400 left in the village. In 1814 Cabot esti- mated that the village contained 700 people, despite the fact that some may have been missing. The apparent increase in 1814 can be very simply explained by the assumption that fewer natives had fled the village than had done so when Moraga arrived. Cabot's figure may be quite near the truth for the year 1814 since we must concede a drastic overall reduction of population in the area between 1804 and 1814. Certainly the population can never have been less than 700. The weight of the evidence at hand thus indicates that the estimate based upon Martin's account, i.e., 1,300 persons, is essen- tially sound. Further evidence of collateral importance is derived from consideration of the location of the village of Bubal. Gifford and Schenck (1926, p. 27) place Bubal on Atwell's Island, near Alpaugh, in T23S, R23E, that is, on the east side of Lake Tulare. Neither Martin (in 1804) nor Moraga (Mufloz diary of 1806) locates the rancheria with any precision but Cabot (1815) left San Miguel on Octo- ber 2, 1814, and on October 3 traveled over an immense plain, arriving late in the day at Bubal, on the shore of a big lake. This can have been only Lake Tulare and the west shore thereof. The next year Ortega (1815), approach- ing from the north or northwest, passed through Sum- tache (i.e., Chunut) and went on to Bubal, where he arrived late at night, not having been able to find the village ". . . por haverse mudado de su sitio propio. Estudillo was the next visitor who has left us a detailed account of this area. On October 22, 1819, he went from near Cholam to a place called Los Alisos near the edge of the foothills of the coast range. On October 23 he went across the plain and on October 24 arrived at Bubal, ob- viously from the west, and found it deserted, adding the comment that the village ". . . manifesto aver ya dias q. se fueron a otra parte." The following day he pushed five leagues south through tule swamp and found the settlement on the bank of the lake although his soldiers had to wade waist deep for two leagues farther in order to catch most of the inhabitants. Apropos of this incident he says regarding Bubal: "Esta es la rancheria de gen- tiles mas immediata a las misiones, y la q. con mayor frecuencia se hacen cristianos en la de San Miguel.' From these accounts it is very clear that the original site of Bubal was on the west, not the east, shore of the lake and that because of the depredations of the Spaniards the inhabitants fled into the lake itself, where they made at least temporary settlements. That these became their permanent home is attested by the fact that no later than 1826 Pico stated that Bubal was situated on an island in the lake. Subsequently contemporary writers as well as the modern ethnographers agree that the principal village a 43 ANTHROPOLOGICAL RECORDS of the Wowol was on Atwell's Island. From the demographic point of view the chief justi- fication for tracing the migration of Bubal in the first two decades of the nineteenth century is to indicate how the constant pressure of the Spaniards, through incessant military expeditions, could affect the popu- lation. Through a series of years, their native village site having become untenable, the people of Bubal were forced to seek precarious and inadequate shelter where- ever they might find it in the depths of the tule swamps until ultimately they could establish themselves in a new home, an island fortress where they might remain relatively undisturbed. Starvation, casual massacre, and disease coupled with exposure must have strongly reduced the total number. Hence a 50 per cent decrease in ten or fifteen years-from Martin to Cabot and Estu- dillo-is not at all surprising. The Chunut were first visited by Martin in 1804, who designated their principal rancheria Chuntache but gave no population figures. Two years later, in 1806, it was seen by Moraga, who called it Tunctache and said it had 250 people. Cabot in 1814 said there were 700 persons and Ortega in 1815 found 20 males. Estudillo in 1819 found 103 young braves ("indios gallardos mozos") and 200 women, old men, and children. How- ever, he also states that the captain and "la mayor parte de la gente" were away on a visit toward Lake Buenavista. The estimates of Cabot and Estudillo appear to be quite reliable. Cabot describes Bubal and then passes on to Suntache. The latter place he says had a popu- lation "about the same as the preceding," or 700 persons. Since Estudillo took the pains to count the young men precisely, his remaining estimate must be fairly cor- rect. The total thus is 303 persons present plus more than the same number of absentees, or approximately 700. Since the location and history of Tuntache was very similar to that of Bubal and since in the period 1815- 1819 the population was nearly the same, it is very probable that there was a reduction in population at the former village analogous to that seen at the latter. Al- though we have no concrete data, such as Martin's report for Bubal in 1804, which may be applied to Tuntache, it may be assumed with safety that the ab- original inhabitants of this rancheria numbered at least 1,200. The third lake tribe was the Tachi. This tribe, or its principal village, was first recorded by Martin in 1804. He gives no direct figures but implies that there were 4,000 inhabitants, although he may have been referring to the entire lake area. The next visitor of consequence was Cabot in 1814 who stated that Tache " segun presenta y por la caseria que la compone had 1,000 souls. At a distance of two leagues he found another rancheria, Guchame, which may have belonged to the same tribe, which ". . . segun presenta y tnformes tomados no pasara de 200 almas . . . " The next year Ortega attacked the rancheria but the people had been warned and had all fled when he entered. They had not returned, moreover, in 1819, when they were seen by Estudillo. They must have been in bad straits, because Estudillo found them living deep in the swamp, in a "gran Bolson de Tule, sin poder tener lumbre.' Estudillo gives no figures but he makes the interesting comment that the Tachi had four chiefs and that the rancheria (or tribe) had several "parts,'" each at some distance from the others. This raises the question whether Cabot saw the only rancheria of the tribe or one of a number. The village he saw he examined su ciently carefully to enable him to count the houses. f an arrangement is incompatible with rancherias "ea at some distance from the others.' Furthermore fA chiefs would imply four more or less equal subdivis or four rancherias and possibly 4,000 inhabitants. A first sight this appears preposterous. However, the following facts should be noted. 1. The area held by the tribe extended across th4 north and west shores of Lake Tulare from the pres town of Lemoore to Coalinga close to the western fd hills. This comprises a greater area than the Wow and Chunut together. 2. Modern informants have been able to-give the ethnographers Kroeber, Gayton, and Latta the narn of 3 villages for the Wowol, 2 for the Chunut, and 8, the Tachi. Although the number of villages has no at quantitative significance, it does indicate the great e size of the Tachi. 3. As mentioned previously, Derby in 1850 foun Tachi tribe to contain about 8000 individuals, of w 300 lived in the principal rancheria. In view of the great attrition to which all the open valley tribes h been subjected between Estudillo's visit in 1819 and of Derby in 1850 it is almost incredible that the Ta should have diminished only from 1,000 to 800 dur that period. It is much more reasonable that the p pal village should have declined from 1,000 to 300 would be indicated by the figures of Cabot and Der so, then the tribe as a whole must have once cont much more than 1,000 people. 4. Father Martin in the description of his trip plies that there were 4,000 people living in the vici of Tache. It has generally been assumed, and is so stated by Gifford and Schenck (1926, p. 22), that was referring not only to the borders of Lake Tula also to the lower reaches of the Kaweah and Kings This is simply an assumption and rests upon no un vocal evidence. 5. Cabot's quite careful estimate for the princi rancheria shows that it was larger than Bubal or T tache in 1814. Martin's data for Bubal showed that town must have contained fully 1,330 persons in 18 If we disregard any shrinkage prior to that year, contemporary population of Tache would have reac at least 1,600 if Cabot's estimates for the two vill in 1814 are to be credited. On the basis of all these facts the author believ that the Tachi aboriginally possessed one village least 1,600 inhabitants and that Cabot's figure for village was reasonably accurate. In addition, the ments of Estudillo in 1819 and Derby in 1850-and of these observers were trustworthy persons-poil nitely to the existence of at least three other villa These were undoubtedly smaller than the principal cheria. In default of any concrete data each may estimated as half the size of Tache, or 800 person apiece. The total for the tribe would then be 4,000 nearly twice as much as for the Wowol and Chunut bined. An aggregate of 6,500 natives for precontact ti seems to be indicated in the Tulare Lake basin. figure 1,100 was obtained for the period of approx 1850-1852. The reduction would then have been to value of 16.9 per cent of the aboriginal level. If seems excessive, it should be borne in mind that area was subjected to the ravages of disease, both demic and venereal, from 1770 forward, as is at or implied by both Garces in 1776 and Martin in 1 I I I I 44 COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY AIt was overrun by clerical and military expeditions in S 1804, 1812, 1814, 1815, and 1819, not to mention an ndefinite number of private raiding parties which have pJeftno trace in the documents. From 1820 to 1850 it -was entered repeatedly by ranchers from the coast, ,iAmerican trappers of the Jedediah Smith variety from e southwest or north, and by New Mexican bandits. A these took a toll in the form of mission converts, oatt1e casualties, burnt food stores, and disrupted dillage life. Finally, it should be remembered that the Wdry and arid plains of modern Kings, Tulare, and Kern counties bear no resemblance to the former region of h'ivers, sloughs, swamps, and lakes which once sup- ported uncounted millions of game birds and animals, together with a luxurious vegetation capable of sup- porting a very dense human population. TULARE LAKE BASIN 6, 500 THE KAWEAH RIVER Together with the Tulare Lake Basin the lower aweah River and its delta from Lemon Cove to below htown of Tulare was probably one of the most dense- populated spots in California, or possibly even north the Valley of Mexico (see maps 1 and 3, area 3). The peated comment of the missionaries with respect to "infinidad de gentiles" to be found there creates a bjective impression which is borne out by the numeri- data we possess. There seem to have been two rather indistinctly sepa- ted divisions of the region. One, centering around salia and occupying the delta and sloughs, contained ee tribes, the Telamni, Wolasi, and Choinok, of ch the Telamni were the most important and num- Us. The other, centering around Lemon Cove and bably extending some distance into the lower foot- s, included the Wukchamni, Gawia, and Yokod, the gest group being the Wukchamni. Martin entered the delta in 1804 and called the ple Telame. Moraga in 1806 explored it more thor- ly, According to the Munioz diary (Oct. 19-20), the y noted Telame with 600 souls, together with a g rancheria" one league east and the rancheria hochs two and one-half leagues east. In addition there Ire "otras varias rancherias" in the vicinity. The ge list appended to the diary gives Telami I ("ten- segun corto computo 600 almas"), Telame II with souls, Uholasi with 100, Eaguea with 300, and hochs with 100. Uholasi is no doubt Wolasi, and ea and Cohochs are probably respectively Gawia Yokod. If the last two are omitted, it is evident Moraga saw or knew about four rancherias, Telame ndII, Uholasi, and the unnamed big rancheria. To se must be added the "otras varias rancherias," ch may have amounted to another four, or eight in A population of 2,000 to 4,000 is certainly indicated. Cabot in 1814 was the next visitor who left a record. referred to the "Roblar de Telame Rio," which in- ded Telame, the largest rancheria in the Tulares. t's Telame may well have included both the villages which this name was ascribed by Morgan. If so, on raga's figures it must have contained a minimum of persons. A higher number is more probable, how- r, in view of the fact that it was the largest in the a.o In 1816 Father Luis Antonio Martinez passed through the region and left a circumstantial account of his visit. Starting from Bubal, he approached the Telame area, reaching first the village of Gelecto, where ". . . encon- traron no mas el cementerio: se habia destruido por las guerras. . . ." These wars apparently were raids and skirmishes in which refugees from the missions and other Indian villages participated. From Gelecto the party went to Telamni ". . . al llegar alli los divisaron de Lihuauhilame el grande . . . done al dia anterior habian tenido una gran refriega cuyo resultado fue dar muerte a uanos 8 hombres . . . ." The captain of the latter rancheria sent a messenger to Martinez with the report the place contained "como de 300 casados.' Gelecto was one league from Lihuauhilame and since the latter village could be seen from Telame the distance between the two could not have been more than a league. Martinez then went six leagues south to Quihuama, be- fore proceeding westward on the way home. Lihuauhilame contained 300 married men, or heads of families. The aboriginal social family consisted of at least five persons, and even after the disruption suf- fered from 1804 to 1816 must have amounted to four. The total population, according to this assumption, must have reached fully 1,200, with a probable pre-invasion value of at least 1,500. Martinez therefore gives us four sizable places: Gelecto (depopulated), Telame (minimum 800 according to Moraga and Cabot), Lihuauhilame (1,200), and Quihuama. Subsequent visitors (e.g., Estudillo, 1819, and Rodri- guez, 1828) mention Telame but give no data with respect to size nor do ttey specify any other rancherias in the immediate vicinity. For basic population data we are consequently forced to depend upon Cabot, Moraga, and Martinez. In the discussion of Bubal mention was made of the attrition of population due to war and disease during the period following the first entry of the Spaniards in or about the year 1800. That these factors were very seri- ous becomes even more evident from the accounts of the Telame region. Martinez describes the total obliteration of Gelecto, which he ascribes to the "wars.' Elsewhere in his report he refers to much internecine fighting among villages and between natives and fugitives from the mis- sions. Moreover, the Spanish accounts repeat ad nauseam the statement that this or that village was attacked or destroyed in the course of various expeditions, or that village after village was deserted by its inhabitants be- cause of fear of the soldiers. It is highly probable that there is a great deal of lost history pertaining to the central valley during this period and that tremendous destruction was inflicted upon the native villages which was never recorded in the official documents. Hunger and disease were likewise rampant. Clear indication of this condition is contained in the sentence of Ortega, in 1815, with respect to Telame: " . . . en- contrando esta grande rancheria toda desparramada por la mucha mortandad que havian tenido, y la much hambre que padecian . . . " With regard to the cause of the "mortality" it is clear that a part was due to the killing by the Spaniards and other Indians during the "wars," a part was due to famine, and very likely the remainder was due to disease. Although this factor is not specifi- cally mentioned, the word "mortandad" was widely em- ployed by the Spaniards and Mexicans to connote the effects of an epidemic. Furthermore, the absence of disease would be more difficult to explain than its pres- ence in view of the wide intercourse between the peoples of the southern valley and those of the coast at a time when the Indians of the missions were dying by thousands 6.500 45 46 ANTHROPOLOGICAL RECORDS from measles, dysentery, and other contagious mala- dies introduced by the whites. The whole picture is one of ruinous devastation in the Kaweah delta just prior to 1816, with accompanying disorganization of the local economy and reduction of population. The effect of war, disease, and starvation cannot be emphasized too strongly, nor can mention be made of them too often. On account of their debilitating influence the populations seen in the Kaweah delta and reported in the documents cannot possibly be overestimates of the aboriginal number. On the contrary, they undoubt- edly represent too low, rather than too high, a figure. Reverting now to the villages reported, Moraga mentions eight places, four of them by name or other specific reference. Martinez mentions four, all by name. Cabot refers to Telame as the largest village in the Tulares. Elsewhere (MS, 1818) he states that be- fore reaching Telame there are five rancherias, includ- ing Quiuamine and Yulumne. Quiuamine is no doubt the Quihuama of Martinez. Telame was one village, according to all observers except Moraga (actually Munioz, who wrote the diary). Moraga ascribes 600 people to the first Telame and 200 to the second. The first estimate, be it noted, was "segun corto computo," or according to a short count. The estimate must therefore on Moraga's own admis- sion be increased, certainly to 1,000 and perhaps more. In view of the size of the well known rancheria Bubal, fully 1,300, Telame must have contained 1,200 persons. In addition to the two Telames Moraga mentions a "big rancheria" one league to the east. Hence there were three villages which comprised what may be termed the Telame complex. No figures were given by Moraga for the unnamed rancheria, since it was entirely deserted. However, since it was regarded as "big," there must have been several hundred inhabitants, say 500. The total for the triad then would have reached nearly 2,000. The Martinez description is apparently somewhat at variance with that of his predecessor. Martinez saw, cites distances for, and mentions by name three ran- cherias: Telame, Lihuauhilame, and Gelecto. They were located within a radius of one league of each other and must correspond to the three seen by Moraga. Gelecto was in ruins, with only the cemetery still in evidence. Hence Gelecto may very well have been the big, deserted rancheria of Moraga. Martinez gives no population data for Telame but says there were 300 heads of families in Lihuauhilame, which was, therefore, without much doubt the largest of the three. According to Moraga's figures, Telame I was the largest. Hence the concordance seems to be that Telame, Lihuauhilame, and Gelecto of Martinez correspond respectively to Telame II, Telame I and the "big" rancheria of Moraga. As pointed out previously, the total inhabitants to be deduced from 300 heads of families, under the condi- tions existing in 1816 was 1,200. This is twice the esti- mate of Moraga. An important point arises here with respect to Moraga's estimates. At Bubal, it will be remembered, Martin found evidence of 1,300 people in 1804 whereas Moraga reported only 400 in 1806. At Lihuauhilame Martinez found according to the statement of the village chief 1,200, although Moraga had reported ten years previously only 600. Furthermore Cabot, at Bubal eight years after Moraga, found 700 persons. For these two important villages therefore Moraga differs flatly with three other competent authorities by a factor of two or three. Similar instances may be found elsewhere in I which Moraga's population figures are far too low, seems difficult to escape the conclusion, consequen that Moraga (or Mufioz) consistently underestimated native population. The reason is not immediately ap ent, although several possible suggestions may be 0 fered. Moraga personally had little interest in such matters. Although he himself did not write the acco of the expedition to the Tulares in 1806, he did writ that of his expedition to the Sacramento Valley in 18 The latter diary shows very clearly, through the ex- treme paucity of its population data, that Moraga ei made no direct counts or estimates, or considered too unimportant to mention in his manuscript. For 1806 trip the estimates were all supplied, obviously, Mufioz. There is no reason to impugn either the jud or veracity of this missionary. However, if one ex ines his account, it becomes evident that Mufioz bas his figures either (1) on statements of gentiles or (2 the number of natives seen by him. The former sou might or might not be accurate. The latter was alm certain to yield too low values because the Moraga e pedition was notoriously hostile to the natives and at nearly every village approached the inhabitants fled they could possibly do so. Mufioz therefore consiste saw only the residue, a fraction of the actual numbe For the above reasons the writer believes that a rection factor should be applied to the Moraga-Mufo data, and unless there is specific reason to believe otherwise, the figures should be regarded as indica only about 50 per cent of the true population. Such a correction should not be applied to the figures of oth explorers, like Cabot or Estudillo, who were far mo careful in their methods of estimate. If, now, we apply a correction factor of 2, Moral estimate for Telame I becomes 1,200, or the same that found by Martinez for the same village (Lihuau} lame). On the same basis Telame II (Telame of Mal ez) would have had 400 persons. Gelecto (unnamed l Moraga) was "big" but probably not as big as Telam Hence we may assume an intermediate value, say 80 The total for the Telame complex, or the triad of vi would have been 2,400. In addition to the triad we have Uholasi and the "o4 varias rancherias" of Moraga. Since Moraga gives 1 for Uholasi we may increase that number to 200. Al the other rancherias we have Quihuame (or Quiuami and Yulumne, which were noted by later visitors. Ml aga, however, in saying "otras varias" clearly mea more than two, probably at least four. It is pertinen note in this connection that some of these may have appeared during the turmoil of 1806 to 1816 and that surviving inhabitants may have been absorbed by oth larger villages. Such an explanation would account i the failure of Cabot and Martinez to refer to them. assume four villages at the time of Moraga's expedi (and of course the aboriginal number would have bee less), it is safe to consider them as having been rel ly small. According to the size scale of the Kaweah lages as a whole 200 inhabitants could reasonably be cribed to each of them, or 800 for the group. The aboriginal population of the Telamni and the Wolasi may therefore be set as closely as we can g 3,200. The Choinok appear to have had only one ran At least there is one and one only which recurs repe edly in the Spanish documents. This is Choynoque (] aga, 1806), Choynoct (Ortega, 1816), Choinoc (Cabot 1818) or Choijnocko (Estudillo, 1819). Moraga gave as the population, as did also Estudillo. The two val are comparable, if we remember the attrition occurr I COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY tween the years 1806 and 1816. We may then apply correction factor of 2 and get 600 as the most Qbable number in 1806. Such a value is also consis- Et with the status of the Choinok as an independent bal entity of the Kaweah basin, although it does not ce into account any reduction in population prior to expedition of Moraga. There was doubtless such a duction, but since we have no direct evidence bear- upon the matter it will be better to let the figure stand. The total for the Kaweah delta group (Telamni, 1asi, Choinok) is 3,800. This is indeed surprising tthe figure perhaps is corroborated by the state- nt of the Franciscan President for the California u8ions, Father Payeras-made in support of the blishment of new missions in the valley-that the e district alone contained 4,000 unconverted |ten. 'The middle Kaweah above Visalia was inhabited by -Gawia, Yokod, and Wukchamni. The Gawia are esented in Moraga's account by Eaguea (300 inhabi- ht) and the Yokod by Cohochs (100 inhabitants). The chamni were by far the most numerous and for an ellent account of them we are indebted to Estudillo. officer, in addition to being a competent field nder, appears to have been a scholar and a W eman. His report on the Wukchamni village of {icha is unquestionably the most complete and ae- Fwate left us by any of the Spanish explorers and as oh is worth discussing in detail. Istudillo was the first white man to see Chischa. On ;s point he is very explicit: . . su capitan joasps, ni su gente jamas havian visto tropa, siendo esta la primera vez q. havilan Ilegado alli, pues hace mucho tiempo paso por abajo (este fue D. Gabriel Moraga en el reconocimiento q. Lhizo en 1806) y solo noticia tubo por sus amigos de Telame. a8equently, allowing for possible communicable lease, Chischa was in its aboriginal state when htdillo saw it. VChischa was 5 leagues east of Telame and 3 leagues m Choinocko. This places the village, according to .maps of Kroeber and of Gayton, at or just above mon Cove in the territory ascribed by these ethnog- phers to the Wukchamni. Estudillo measured off the Densions of the village by pacing. The shape was cilunar, crescentic or approximately that of the ctor of a circle. The short side ("por su frente") was varas long and the long side ("por la espalda") was varas. A figure plotted on co5rdinate paper to scale sthat the area was 80,000 square varas. On the Wumption that the Spanish vara equaled a yard, and an average city block measures 300 feet on a side, village of Chiseha would have covered eight city Estudillo caused the Indians living in the village to a line before the town, with the men in a single and the women and children massed in front of them. counted the men and found that there were exactly warriors ("jovenes de arma") and "como 600 mugeres Xs4' According to the translation made for Merriam rin his collection): Then I went opposite where the invited guests were lodged, and as they all, men and women and boys and girls were presented to me in a confused mass, I could not count them as I did those of Chischa, but there were perhaps 600 men.' He specifies the 600 men as 'jovenes" and adds that there were 200 "mugeres jovenes." He then describes going behind the village to the arroyo, where he saw more than 100 "mugeres de mayor edad," washing seeds for atoles for the celebrants of the fiesta, and an even greater number of "jovenes moliendo en piedras dhas semillas.' The extraordinary care with which Estudillo con- ducted his investigation can leave little doubt of the accuracy of his figures. He saw 437 "jovenes de arma" in front of the village together with 600 women and children, plus 100 "mugeres de mayor edad" and more than 100 "jovenes" behind the village preparing the meal. Even allowing for some duplication of individuals the population must have reached at least 1,250. The solidity of this evidence for Chischa renders even more probable comparable figures for Bubal and the other large villages of the general area. Estudillo saw 600 young men and 200 young women who were visitors. If we use the same ratio of young men ("jovenes de arma") to total population for these groups as for Chischa, then the 600 young men repre- sented a total of 1,700 persons. These were all, says Estudillo, from the "roblar,"' or the Kaweah basin. When he arrived at the village, he was met by seven chiefs (who were already on the scene), two from Telame, one from Choynoco, and four from other rancherias of the "roolar' near tne sierra. We may assume that the seven visiting chiefs were accompanied by approximately equal retinues, or 114 persons each. If two of the chiefs and 228 persons came from the Telame district and one chief with 114 persons from Choynoco (i.e., Choinok), then the remainder, 458, were from other tribes. By the same proportionality factor these represented a total of 980, or let us say 1,000, Indians. The Wuk- chamni and their satellites must therefore have num- bered 2,250 individuals in the year 1819. Estudillo him- self says that the population of Chischa and its neighbors was 2,400, but he may have included some Telamni among this number. On the other hand, the visitors to Chischa on the occasion of the fiesta could scarcely have included all the inhabitants of the villages whence they came. Some, for one reason or another, must have remained at home. Hence the estimate of 1,000 is prob- ably under the true value. Now it is important that Estudillo was in the "roblar" in 1819. In view of the severe disorganization, "mor- tality," and "famine" of 1814 to 1816, the population of the Wukchamni must have undergone a serious decline before Estudillo saw the tribe. Despite the absence of any specific figures the documents give the impression that the reduction of population around Tulare Lake was almost complete by 1819 and that the valley tribes along the margin of the foothills had lost fully half their num- ber. It will be proper therefore to ascribe a one-quarter reduction to the Wuchamni, Gawia, and Yokod. If we accept Estudillo's estimate of 2,400 for the year 1819, the aboriginal population for these groups would have been 3,200. In the meantime the Mono of the upper river had scarcely been touched, save possibly by epidemics of which we have no record. It is significant that at the great gathering at Chischa there appeared, near the middle of the day, a chief with 69 men and 42 women from a rancheria called Apalame in the interior of the Sierra Nevada. These natives, probably Balwisha or I 47 ANTHROPOLOGICAL RECORDS Waksache, had never seen troops. To arrive at the population of the entire Kaweah basin in aboriginal or proto-aboriginal times these tribes must be included. Their strength, as previously estimated, was of the order of 600 persons. Computing now the total for the Kaweah river and delta as first described by white men, we find an aggre- gate of 7,600 inhabitants. As set forth previously, the survivors in 1850 numbered about 1,800 or 23.7 per cent of the aboriginal (or early historical) value. Ex- cluding the relatively undisturbed Mono the comparable value for the lower river and delta is 17.2 per cent. These percentages are in close agreement with those found for the ecologically similar area bordering Lake Tulare. KAWEAH RIVER 7,600 THE MERCED RIVER It will be convenient at the present juncture to con- sider the watershed of the Merced River, although this area lies at a considerable distance from that just ex- amined (see maps 1 and 4, area 6). In the preceding section it was concluded that only 500 to 600 natives still remained in 1850 on the lower portion of the river below the foothills, whereas the population of the southern Miwok in the foothills and higher ranges amounted to approximately 1,250. The latter figure was based principally on Merriam's village lists and the population counts obtained from informants by Gifford for the Miwok farther north. The question must now be propounded whether these data, which appear to be fairly accurate for the year 1850 or even 1840, can be taken as showing the population under substantially aboriginal conditions, let us say those obtaining prior to the intense Spanish invasion of the valley in the decade 1800 to 1810. 1. As a matter of generalization it can be stated that the environment as remembered by the oldest informant or even his parents can scarcely reach into pre-Spanish times. Hence the village populations and distributions as reported in good faith to Gifford or Merriam must have been subjected in some measure to the disruptive effect of the white man. The great disturbance in the valley itself, which was manifested by the entire ex- tinction of whole Yokuts and Plains Miwok tribes, must have had repercussions in the near-by hills through disease, kidnaping, and minor dislocation of food supply, even though the actual territory of the natives was not physically invaded by the newcomers. Hence, a priori, one might anticipate that the populations as derived from ethnographic sources would be somewhat less than truly aboriginal. 2. In the discussion of Gifford's data on the North Fork Mono it was shown, on the basis of persons per family and houses per village, that the population in the memory time of the informants was about 440 where- as the precontact value must have been nearer 640. The population residue in 1840-1850 would then have been 68.8 per cent of the aboriginal level. 3. For the upper Tuolumne and Stanislaus Gifford's population figures were based upon the values given by his informants for 49 villages. The average was 20.8 persons per village, a number which was ac- cepted as valid for the period of 1850. The distri- bution of population for the villages is as follows: Inhabitants per Village 60 55 35 30 25 20 15 10 5 0 Total Nunmber of Villages 1 1 3 6 8 9 6 12 1 2 49 Number Perso 60 55 105 180 200 180 90 120 5 0 995 Now it may be assumed that under normal condit few if any villages would contain less than 20 personi and that those listed by Gifford with 15 or less were victims of a general decline in numbers. Hence to tl latter may be ascribed a minimum of 20 persons. AO the same time the other villages must have suffered some reduction. Although there is no positive eviden bearing on the matter, it would not be excessive to a five persons to each of the others. Making these cor tions the total becomes 1,340 instead of 995. The re' due in 1850 would then be 74.2 per cent of the abori level. Incidentally, the inhabitants per village would, then be only 27.35, a value by no means excessive f prehistoric times. Some confirmation for these assumptions can be tained by further consideration of Gifford's study of North Fork Mono. As previously mentioned, GiffSord shows the number of houses and hence the number of families living in the hamlets of this tribe. For mar hamlets two or more sets of houses are given, imp4 ing consecutive, not simultaneous, occupancy. The average number of houses per hamlet occupied at on time is 2.7. However, informants were able to recol an additional 44 houses, which had been formerly usi Including these, the average number per occupied ha let is 3.21. Gifford's family number is 4.89, a value which may be increased to 6.0 to cover aboriginal co ditions. Thus the mean size of an active prehistoric Mono hamlet may be taken as 19.25, or let us say 20 persons. Since the Mono villages were intermittentll inhabited whereas those of the Miwok were permanet and probably somewhat larger, the average value of 27.35 for the latter seems in no way excessive. From the above considerations the conclusion is warranted that for the northern Mono and the Miwoki population as derived from good modern ethnographi data is about 70 per cent of the precontact value. Tb estimate for the upper Merced, derived from Merr village lists was 1,239. If the factor of 70 per cent i applied, the aboriginal population becomes 1,770. For the lower Merced Valley we are dependent es tirely upon the account of Moraga's visit in 1806. Co ing from the west, he crossed the San Joaquin River September 27 and moved three leagues north to cams or near Bear Creek in T8S, R1OE. The following day, September 28, Moraga divided his expedition and se4 one group north and another northeast to explore. B groups found a great river, with many natives, all ~ whom fled on seeing the white men. At least one rat eria was found, because Moraga "adquirio la noticia otras 5 rancherias sitas en el rio fuera de aquella et que se hallaba del parte de 250 almas, segun el infoi de los gentiles:' On the 29th the camp was moved th I 48 COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY agues ENE (more probably NNE) to the river, the erced. There were two rancherias on the river bank, e people of which had fled through fear of the white den. On the 30th a party went up the Merced and found m natives "sin duda de sus 5 rancherias.' Moraga then went north and returned to the Merced mhOctober 7. The Spaniards saw many natives and were visited by 79 warriors from the rancheria "del Itro lado del rio," i.e., on the south bank. The 8th of Detober the expedition visited the rancheria just etioned; to judge by the number of men (the women ing fled) the rancheria had 200 souls. This place pis called Latelate, and there was another village $ear by, called Lachuo, with the same number of in- mbitants. The next day the expedition moved on south- Moraga evidently saw two villages and heard of about lOe others. The two which he saw, Latelate and Lachuo, re said, on the basis of the warriors seen, to have tained 200 persons each. Since warriors of one vil- ige, Latelate, numbered 79, the estimate of 200 total habitants, or a ratio of 2.5 to 1, is entirely reason- 6le. If the other five villages had the same number, Feaggregate for the river would have been 1,400. bwever, some of the others may have been larger. fthe list of rancherias appended by Mufioz, the ap- roximate sequence of the journey is followed. Five ncherias can be ascribed logically to the Merced: bineguis, Yunate, Chamuasi, Latelate, and Lachuo. hineguis follows Nupchenche in the list, Nupchenche sung 250 souls and Chineguis the same population. 4kewise, Yunate and Chamuasi have the same "segun ~mpute regular." Latelate and Lachuo are given 200 ich, thus corresponding to the text of the diary. The ther two villages are not mentioned by name in the bAt but it may be presumed that they were of approxi- mately the same size, let us say one of 250 souls and be other of 200. Thus the Mufioz-Moraga count gives J 1,600 inhabitants. It will be remembered that the figures cited by lDraga for the population of villages in the Kaweah- Iiare region were uniformly at variance with those lother observers and were always too low. Hence a piestion may be raised with respect to his data for the lerced valley. The villages in this area, by all sub- wquent accounts, were smaller than in the heavily ppulated territory farther south. Furthermore, Mor- p's was the first expedition of which we have record idch explored the Merced Basin. These facts would pad to indicate that Moraga's figures may be reason- bly accurate. On the other hand, the repeated state- ments that the Indians fled on the approach of the white ien and the fact that estimates had to be made from ienumber of warriors seen leave the possibility open bat there actually were more people than Moraga thought. kence it will be reasonable to ascribe an aboriginal spulation of 250 to each of the seven rancherias, giving a total 1,750 for the lower Merced River. The population of the entire valley then would have Een 3,520, or, rounding off to the nearest hundred, ~500. The survivors along the lower river amounted bapproximately 550 in the year 1852. If the population Moraga's time was 1,750, then the reduction from 106 to 1852 was to 31.4 per cent of the original level. Lview of the somewhat more remote position of the kerced, this figure checks quite well with the values ,Wnd on the Kaweah River and Lake Tulare. THE KINGS RIVER The next region to be considered is the basin of the Kings River. Like the Kaweah, this stream may be divided into three sectors. The first comprises the delta and slough area southwest of Kingsburg and was the home of the Yokuts tribes, Apiachi, Wimilchi, and Nutunutu (area 4A). The second includes the valley margin and foothills, with the tribes Wechihit, Aiticha, Choinimni, Chukamina, Michahai, and Emtimbich (area 4B). The third is in the higher foothills and embraces the territory of the Mono groups, Wobunuch and Holkoma (area 4C). The Kings River sloughs were first described in 1804 by Martin, who mentions the tribe, or rancheria, of Notonto (Nutunutu) but gives no population data. The next visitor was Moraga in 1806. In the diary of the expedition, written by Father Mufloz, no mention is made of Notonto but in the appended "List of rancherias visited in this trip and the one in April" are included Notonto I with 300 persons and Notonto II with 100. Estudillo saw the region in 1819 and said that Notonto (only one village of this name is mentioned) had 303 men "todos gente robusta y de armas.' He also saw a few old women and children. Since the men are of the same type ("robust warriors") and were carefully counted in the same way as at Chischa, the same ratio of warriors to total inhabitants may be used. A popula- tion of 866 is thus indicated or, in round numbers, ubO. Estudillo also says there were four chiefs, one each of the "Notontos," Gumilche, Guchetema, and Tateguy. The Nutunutu are thus clearly segregated from the Wimilchi (Gumilche). The other two names cannot be traced and may indeed have been those of individuals. The "guimilchis," in the meantime, had been seen in 1815 by Pico, who says that they had at least two rancherias. From the ethnographers we get indication of six villages: of the Apiachi, the village of Wohui (Kroeber, Gayton, Latta); of the Nutunutu, the villages of Chiau (Kroeber, Gayton, Latta), Hibekia (Kroeber), Honotau (Gayton), and Kadestiu (Latta); of the Wimilchi, the village of Ugona (Kroeber, Gayton, Latta). If these villages actually existed in the early years of the nine- teenth century, they can scarcely have held less than 250 persons apiece and the population would have been in the vicinity of 1,500. From the Spanish accounts we find evidence of at least four villages: originally two (perhaps later one) of the Nutunutu and two of the Wimilchi. One of the latter may have been in fact the principal village of the Apiachi. The Nutunutu, whether as a single village or as a tribe, seem to have amounted to fully 850 persons at the time of Estudillo. Since these groups had been exposed to expeditions beginning in 1804, it is very probable that they had undergone considerable attrition before they were observed by Estudillo. This point of view is supported by Estudillo's remark that he re- quested the warriors of Notonto to meet him without their weapons because this rancheria "es la mas veli- cosa y terrible de los Tulares.' Hence it is quite prob- able that the aboriginal population reached 1,200. A value of 500 may be assigned arbitrarily to the other villages or tribes, for Estudillo mentions three chiefs apart from the Notontos and Pico says that the Wimilchi had at least two rancherias. The probable aboriginal population for the entire area is therefore 1,700. By the year 1850 the tribes of the Kings River delta were represented, according to the account of G. H. I MERCED RIVER ......... 3,500 49 ANTHROPOLOGICAL RECORDS Derby, only by the rancheria of Notonto which then had 300 inhabitants. The population had thus fallen to 17.6 per cent of its former value. A footnote to the decline of the native inhabitants in this region is the fact that within a year or two after Derby's visit the village of Notonto was attacked by American cattlemen and farm- ers. The rancheria was devastated and 200 of the 300 people present were massacred in cold blood. For the second sector of the Kings River we are dependent primarily upon the record of the Moraga expedition. Moraga and Munioz evidently covered the river from the vicinity of Reedley to, or nearly to, the junction of the main stream and Mill Creek. The villages mentioned by them belonged principally to the Aiticha and the Choinimni. The Wechihit and the Toihicha may have been included but the Chukamina, Michahai, and Emtimbich seem to have been overlooked. Hence the figures given by Moraga are undoubtedly incomplete. On October 16, 1806, having arrived from the San Joaquin River two days previously, Moraga sent out two scouting parties. One went upstream and found a rancheria of "como de 60 almas," called Ayquiche (or Aycayche). They were no doubt among the Aiticha, above Sanger. Here they heard about, but did not see, six other rancherias "sitas a la orillas del rio por la parte de la sierra." The other party went downstream and found three villages close together on a spacious plain along the banks of the river. They had a total of 400 inhabitants, but most of the people had fled. The "List of rancherias visited in this trip and the one in April" gives the names of these villages: Aycayche, which "according to the Indians" had 200 people, Ecsaa with 100, Chiaja with 100, and Xayuase with 100. In addition there was Capitau, which was very small and a 'sugeto" of Xayuase. It had about 10 people. Appar- ently in October Mufioz and Moraga found only 60 Indians left in Aycayche, whereas in April they learned that it really contained 200. The difference must be ascribed to fugitivism. The three downstream villages are credited by the "List" with 100 inhabitants apiece, but the diary states that there was a total of 400. The latter figure is more likely to be correct. Thus, with Aycayche, Moraga saw in this sector four villages and 600 persons. The other group of villages, six in number, was farther toward the mountains and no particular information concerning them is given in the diary. The "List," however, is more explicit. Under Aycayche it is stated: Aqui hay otras 6 rancherias que no se pudieron re- conocer y son todos, segun la noticia de los indios de esta rancheria como del porte de almas de Pizcache. Pizcache is said to contain 200 souls. An aggregate of 1,200 persons is therefore indicated or, for the entire region seen by Moraga, 1,800. The middle course of the Kings River has been dis- cussed in the preceding section and it has been pointed out that in the middle of the nineteenth century this region was relatively heavily populated. The accounts of several contemporary observers indicate that in 1850 or thereabouts somewhere between 2,000 and 4,000 natives were still to be found between the remnants of the Nutunutu on the west and the foothills Mono on the east. The ethnographic data supplied by Kroeber, Gay- ton, Latta, and Stewart show approximately 25 villages remembered by informants. If we use the fairly con- servative average of 150 persons per village, the total is 3,7 50. To assume 3,500 is merely to stay within the bounds of the existing evidence. If we accept tentatively 3,500 as the number of India on the middle Kings River in midcentury, then we are confronted with the problem of backward extrapolation. For the Tulare-Kaweah region the probable decline fro 1800 to 1850 was probably to the level of approximatel 20 per cent of the original value. Direct application of, this factor to the Kings River gives a value for 1800 of 17,500. This is manifestly far too high. For the Mono and the Miwok in the upper foothills many facts point t a population decline to approximately 70 per cent of the prehistoric value. Application of this factor gives 5,00 for the Kings River, a high but not impossible figure. Other considerations are worth mention at this poin In his diary of 1826 Jos6 Dolores Pico describes his adventures on the Kings River in January of that year. He was chasing stock thieves and trying to recover stolen animals. From January 10 to January 14 he bea back and forth along the Kings River, from the sloughs to the foothills, attacking every Indian in sight. The results were discouraging. He captured no animals, killed not over a score of natives, and was completely. outmanoeuvered by the combined forces of the Wimilch the Notontos, and Chukamina. The entire tenor of the document suggests an active, competent, and quite powerful local confederacy of tribes. This diary of Pic describes the only expedition to the Kings River of whi we have documentary knowledge between 1306 and the coming of the Americans. These facts suggest, first, that there was a sizable population which managed to maintain itself reasonably well for several decades along the Kings River. Secon they suggest that there may perhaps have been a slow migration of the more exposed valley people, like the Nutunutu, higher up the river. Both these factors woul tend to keep the population decline to a minimum. In view of the confusion surrounding the evidence in this area and in view of the apparent inadequacy of the Moraga figures the aboriginal population of the middle Kings River may be set at 5,000, with the full realiza- tion that this value represents the best guess under the circumstances. The upper river was inhabited by the Mono groups, Holkoma and Wobonuch, for which an 1850 population 1,700 was computed. The decline to 70 per cent mayb accepted here without serious reservation; hence the original number would have been 2,340. Adding the values for the three sectors of the river we get 9,130 or, estimating to the nearest hundred, 9,100. KINGS RIVER ......... 9,100 UPPER SAN JOAQUIN, FRESNO, AND CHOWCHILLA RIVERS AND MARIPOSA CREEK The area between the Merced and the Kings rivers (see maps 1 and 4, area 5), which includes the course of the upper San Joaquin, the Fresno, and the Chowchil rivers, together with Mariposa Creek, is very poorly represented in the early documentary sources. The central valley itself, as far as the foothills, was ap- parently traversed by numerous expeditions and raids, and the population was largely missionized, killed, or dispersed. The written record is, however, quite in- adequate. It is therefore not feasible to consider each of these river systems separately, as was done in the I - l I I 5 0 COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY discussion of the population about 1850. It is preferable to discuss the entire region as a unit and, when neces- sary, pass to indirect methods of estimate. The Pitkachi on the San Joaquin are mentioned in 1806 by Moraga, who allows 200 persons to their ranch- Deria. The tribe appears again in the baptism record of aoledad Mission (MS in the Bancroft Library, Berkeley) aeeording to which 205 Indians from "Picatche" were baptized from 1821 to 1824 and another 18 in 1831. An additional 23 came from rancherias in the vicinity, a total of 246. Another rancheria, Capicha, is referred Uby Pico in 1815, who said it was uninhabited at that Ume, the inhabitants having fled to the mountains. As late as 1853 Wessels said that the Pitcache, together withthe Noo-to-ah, a Mono group, numbered 500 to 600 ouls. Kroeber mentions three villages remembered by modern informants. If 246 Indians were baptized in one mission, the tribe a whole must have numbered at least four times as many, or 1,000. If two fair-sized rancherias are men- toned by the Spanish observers, the entire tribe may 1ell have possessed four or five, which again implies apopulation of 1,000. If there were approximately 300 srvivors in 1853, by comparison with other open alley areas the original population must have been ullythree or four times as great, or perhaps 1,200. three rancherias were known to modern informants, they must formerly have been important places with anywhere from 200 to G00 people, again indicating a total of 1,000 for the tribe. Concerning the Hoyima there are two references, one by Pico in 1826 and one by Rodriguez in 1828. Pico tes merely that he attacked the rancheria and cap- tred 40 gentiles and 1 Christian, a fact which in itself uld not furnish a very significant clue to population. e also noted "mucha guesamenta y cueros casi frescos de caballada que habian matado.' The account by Rodriguez is more circumstantial. this soldier went along the San Joaquin River in late pril of 1828. On the 24th he sent a group of men to cout the "rancheria de los Joyimas, que es adonde an 'omido la caballada." At dawn the next day they attacked evillage, "que estaba en medio de los dos brazos del rio" (the San Joaquin west or northwest of Fresno). He captured 26 Indians and 27 animals (horses). Another 10or 80 horses escaped "en el monte." At about this e a gentile captain came from a rancheria designated uche or Getche, depending upon how one deciphers the dwriting of the manuscript. He "vino a los Joyimas comer caballo.' The rancheria named here is prob- ly that of the Heuchi on the Fresno River. This gentile dthere was another rancheria "mas arriba" at which ere were horses. Rodriguez sent Simeon Castro to estigate. He found no one at the rancheria mentioned twent on 2 leagues to another rancheria, likewise serted but containing the carcasses of 100 dead horses, ich had been slaughtered and were about to be eaten. was noted by Rodriguez that: "Estas 3 rancherias son misma que es la de los Jaimes." It was also re- rked that the rancheria was divided when the horses rived in order to eat with less fear of detection. From s account it is clear that the Joyimas had at least ee villages. Allowing somewhat over 300 persons ch, the population of the group would reach 1,000. The slaughtered horses open up an interesting field speculation. It is clear that by 1828 large segments of eaboriginal population had entirely given up the seden- ancestral mode of life in favor of an existence based n stock raiding. To do this it was necessary to recast village life completely-as is suggested by the fact that the rancheria was "divided" when the horses arrived. In order to catch the horses for food other horses were essential for rapid transportation to and from the coastal settlements. New arts and skills had to be learned, and new categories of labor had to be evolved. Rodriguez found among the Hoyima as a whole 87 to 107 live horses (27 captured, 60-80 in the wilderness), which were presumably about to be killed and eaten, together with 100 animals already slaughtered. The total thus reached approximately 200. The question now is pertinent: how much food can be obtained from 200 horses? If we assume that each of these relatively light range animals weighed 800 pounds, we may deduct about 25 per cent to account for bones, hide, certain of the viscera, and other inedible parts, leaving 600 pounds which the Indians could and did consume. The aggregate is 120,000 pounds of meat. If this meat was dried and preserved, according to general practice, it was suffi- cient to supply 329 persons the equivalent of one pound of fresh meat per day for one calendar year. If it had to be consumed immediately or within a few days, and if every man, woman, and child ate 20 pounds apiece, it was adequate for 6,000 people. If the entire tribe, not merely one rancheria, divided the meat into equal shares, and if the tribe numbered 1,000 persons, then the share of each individual amounted to 120 pounds. Whether these figures are strictly accurate is irrelevant. They merely emphasize that a quite sizable group must have been concerned. We may therefore regard the Hoyima as being as large a tribe as the Pitcache, and estimate that the population was at least 1,000. The remaining two tribes in the valley proper, as listed by Kroeber and others, were the Heuchi and the Chauchila. They occupied the north bank of the Fresno River and the distributaries of the Chowchilla River. The ethnographic data include no more than one or two villages for each tribe. The Heuchi are referred to by Rodriguez, who says that the rancheria of the "Jeuche" was completely deserted. However, since it was the principal tribal village, it must have contained at least 200 persons. The Chauchila were also noticed by Rodri- guez, who says that at "Chausila" he "captured" 142 people and "killed many.' If we concede that as many escaped as were captured or killed, there must have been fully 400 in all. The Nupchenches, although they are merely mentioned as a possible tribe by Kroeber (Handbook, p. 485) and are doubtfully recorded by Schenck (1926), occupied an important position in the early nineteenth century. In- deed, the failure of Kroeber and Schenck to consider them seriously makes it necessary to set forth in some detail the information about them contained in the Span- ish reports. These natives were distributed along the San Joaquin River from its big bend near Mendota to approximately the mouth of the Merced (see map 4, area 5A). The first mention of them is by Moraga in the diary of 1806. He found two rancherias, Nupchenche with 250 people and Cutucho with 400 souls which was "junto a la primera llamada Nupchenche." This means that Cutucho was close to but at that time not necessarily part of Nupchenche. From the description in the diary Nupchenche was situ- ated at or near the mouth of Santa Rita Slough in T9S, R12E, and this is almost exactly where Schenck places it on his map (Schenck, 1926, p. 133). The next visitor who left a record was Jos6 Dolores Pico in 1815. On November 7 he left San Luis Gonzaga in western Merced County (in approximately T1OS, R8E) and went east to the 5 1 ANTHROPOLOGICAL RECORDS Tulares at "Arroyo nombrado San Jose," which was close to the rancheria of the Cheneches. At dawn of the 8th he attacked the village and captured 66 persons, but ". . . la mayor parte de esta gente se fue pr estar dha rancheria en mal parage." The gentiles said that 4 leagues up the San Joaquin River was Nupchenche, thus placing Cheneches on the river in the southern part of T8 S, RllE. This location checks well with the statement made elsewhere in the diary by Pico that Cheneches was near the junction of the San Joaquin and "Las Mariposas," or Mariposa Creek. If Pico captured 66 persons but "the majority" escaped, the total num- ber must have reached from 200 to 400, if not more. Pico then scouted Nupchenche and learned that all the inhabitants had fled. He therefore by-passed the village and went 8 leagues southeast up the San Joaquin to the rancheria Copicha. This rancheria, which by the way must not be confused with the Cutucho of Moraga, was thus located on the river several miles north of Firebaugh, probably near or in Tl1S, R13E. As a check on this location is Pico's further statement that Copicha was in the valley of the San Joaquin "junto del Tecolote," or the Chowchilla. On November 10 he moved 8 leagues southeast from Copicha and saw haorses from the ranch- eria Tape, which, from the distances, was near Men- dota. This view is supported by Estudillo, who saw the region in 1819 and says that the spot ". . . donde Tape tenia su rancheria" was 24 leagues south of Cheneches and 25 leagues north of Notonto. Actually, Mendota appears to be approximately halfway between these two points. Pico mentions one other village, Malim, which he places near Cheneches. Confirmation is found in a letter of Fr. Marcelino Marquinez (MS) on May 25, 1816, stating that the Cheneches recently have killed two Christians from Malim. The latter rancheria there- upon allied itself with Notoalh and Luchamme. No other trace of the two last-named villages is found. Other writers who mention the Nupchenches group include Fr. AntonioJaime, who mentions Cutuchu (MS, 1816) as a rancheria from which Soto brought back gentiles, and Ortega, who, in his 1815 diary, mentions Cupicha as having been attacked by Pico. Finally Inocente Garcia in his manuscript of 1878 records an expedition against the Nupuchineches under Ignacio Vallejo. The rancheria, even in the 1830's was "muy Populosa." The expedition captured 100 warriors and 300 of all ages and sexes, arguing a population of over the 300 claimed as captives. From these accounts emerge six rancherias, each of which is mentioned independently by at least two writers. From north to south they were: Cheneches and Malim, Nupchenches and Cutucho, Copicha, Tape. Moraga says Nupchenches had 250 people and Cutucho had 400. From Pico's statement concerning captives we may ascribe a minimum of 300 to Cheneches, and Copicha, Malim, and Tape can scarcely have been much smaller. Hence the entire group can have numbered no less than 1,800 in 1816. At Tape on November 23 Pico found 16 live horses and mules recently killed together with "mucha carne enterciada." If we neglect the meat, 254 whole animals, dead or alive, were actually counted. From November 25 to 28 the party traveled steadily from Tape to Cheneches. From Tape to Cheneches inclusive they saw 500 dead horses. It is not clear whether the 238 animals seen at Tape were included in this figure. If, however, assuming that they were, we use the same ratio of dead horses to inhabitants as was discussed with respect to the Hoyima, these villages should have contained 2,500 persons. This figure is quite reasona} if we grant that the horses were to be consumed by th4 entire group of villages, rather than only one or twoa them, and may be provisionally accepted. On the basis of the records presented, a probable population value for the valley floor between the Merci and the Kings rivers in the decade 1810-1820 was 5,11 But this may well be an underestimate and be represe tative of the aboriginal population. Evidence pointingi this direction is the almost complete obliteration of these tribes before 1850. That very serious attrition was going on among these exposed people is evident ft the records of all the explorers. The massacre andk naping described by Pico is itself significant. In addil we have the discussion by Estudillo in 1819, who foun( almost the entire surviving population of Tape sick at dying. He also points out that at the moment there wer no less than four expeditions, including his own, rang up and down the open valley, bent upon destruction. T4 explore the problem further indirect methods must be employed. We may therefore turn to estimates based upon stream distances. If minor local variation is disregarded, the habitat provided by the Merced and the Kings rivers from the lower foothills out to the center of the valley is in no essential respect different from that characterizingth Mariposa, the Chowchilla, the Fresno, and the San Joaquin throughout its length below the foothills. The native villages were spaced more or less uniformly along the larger rivers. Hence an approximate propor tionality should have existed between riverbank distan and the number of inhabitants. No high degree of pre- cision can be expected from calculations based upon these premises but the method yielded rational resulti for the period centering around 1850 and from it the correct order of magnitude should be obtainable. Airline distances are used for the rivers. The gen course of all the streams is substantially straight and numerous small meanders are uniform in size and oc rence throughout the area. Three river sectors are u as a basis: the lower Merced River, the middle Kingi River from and including Mill Creek to Kingsburg pl1 the principal tributaries, and the lower Kings from Kingsburg to Lemoore. The data are compiled briefly as follows in tabular form. River Sector Miles Persons i in Length Population River Mih Lower Merced... 32 Middle Kings.... 75 Lower Kings .... 20 1,750 5,000 1,500 55 67 75 Despite the uneven nature of the basic information these figures show considerable internal consistency, The mileage of the San Joaquin, Fresno, Chowchilla, and Mariposa amounts collectively to approximately 11 miles (the four streams west of Kroeber's line of the valley Yokuts and down the San Joaquin as far as the mouth of Bear Creek). At 65 persons per mile (the ap proximate mean of the three values cited above) the population would be 12,350, or, let us say an even 12,000. This is more than double the number indicated directly by the Spanish accounts. It has been pointed however, that these accounts are incomplete with res to the villages seen and recorded. Furthermore the records demonstrate a condition of severe disorganizi tion on the part of the native society. Hence the indire I e I I I I I 52 COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY computed figure may reflect more closely the aboriginal population level. The population in 1850 for the part of the Yokuts territory here being discussed was considered in a pre- viOus section. The best estimates were found to be -1,000 for the Mariposa and Chowchilla and 2,900 for the Fresno and San Joaquin. The total, 3,900 is 32.5 per cent of the estimated aboriginal population and rep- ,resents, therefore, a reduction of the same general extent as was demonstrated for the Kaweah-Tulare Lake region. The foothill region drained by the four rivers being (discussed includes the extreme northern Yokuts tribes, ~the North Fork Mono, and some of the southern Miwok. the consideration of the 1852 population it was not advantageous to segregate river sectors as has been done for the earlier data. This is because, with certain exceptions, the data pertaining to the later period cover a8 a rule the entire stretch of each river, rather than te central valley plain as distinct from the foothills. Nevertheless it is possible to arrive at the result de- sired indirectly. For the Yokuts on the middle Fresno River it was ,concluded that the average number of inhabitants per 'village was 60. This value was based on village numbers and general estimates for the period of 1850 and in- cluded also the assumption that the villages had been much reduced in size by that year. For precontact rtimes it is quite justifiable to maintain that the average 8ize was of the order of that demonstrated for the Kings and the Merced, or let us say 150. The tribes on the Fresno and San Joaquin not seen or at least not reported y the Spanish writers are the Gashowu, Wakichi, Ke- chayi, Dumna, Toltichi, Dalinchi, and Chukchansi. The otal number of villages recognized for these seven ibes by Kroeber, Gayton, and Latta is 36. This total of course rests on the memory of informants and per- ains to conditions in the period 1840 to 1850 or perhaps 1860. There is no proof whatever that the village num- r in 1800 was the same, yet the whole history of *an-white contact in the valley region leads one to lieve that it can hardly have been smaller. Since there no evidence to the contrary and since the hypothesis inherently reasonable, we may concede 36 villages of t50 persons each or 5,400 in all. For the southern Miwok on the upper Mariposa and Chowchilla, calculated by means of village counts and ifford's average of 21 Indians per village, the values 273 and 410 respectively were obtained. The factor a reduction to 70 per cent of the aboriginal popula- tion may be here applied, yielding a total of 975 for the o streams. The figure for the North Fork Mono in rehistoric times has already been placed at 640. If we now add 12,000 for the valley and marginal okuts, 5,400 for the foothill Yokuts between the Miwok rder and the Kings River, 975 for the southern Miwok uthe Mariposa and Chowchilla and 640 for the North Fork Mono the total becomes 19,015. The validity of this figure can be subjected to a check rough comparison by area. This method cannot be ex- cted to show up minor or secondary errors but it will ring to light any fundamental or serious discrepancies. e may block out four major regions: the Kaweah-Tu- e Lake, the Kings River, the Merced River, and the gment between the Merced and the Kings. Each of see represents fundamentally the same type of envir- ent, i.e., a rough strip extending southwest to north- at, beginning with the lakes and sloughs of the central alley axis, passing across the valley floor to the foot- hills, and reaching ultimately the middle altitudes of the Sierra Nevada. Four cross sections are thus obtained, differing in width but fairly uniformly including the habi- tats represented. It should be noted that the water sur- face of Lake Tulare as it existed in 1860 has been de- ducted from the area of the Kaweah-Tulare region; also that the two northern regions include a relatively greater expanse of uninhabitable mountain territory than do the two southern regions. The western boundary has been drawn along a line approximately five miles west of the San Joaquin River and the prolongation of its axis toward the lake. The westward extension of the Tachi toward Coalinga had to be neglected since there are no clear tribal boundaries in this area. The number of square miles was computed by township lines and the error of estimate must be considered at least plus or minus 20 per cent. The results follow: Population Region Kaweah- Tulare ..... 1,880 Kings ............. 1,530 Merced ............ 1,400 Mariposa- San Joaquin ..... 3,760 Area density (sq. mi.) Population per sq. mi. 14,100 9,100 3,500 19,000 7.12 5.85 2.50 5.05 The density of the Mariposa-San Joaquin area is quite close to that of the Kings River Basin. The Kaweah- Tulare territory has a somewhat higher density, but this finding is compatible with the known enormous concentra- tion of population around Tulare Lake and in the Kaweah delta. The value for the Merced strip is unduly low. The discrepancy can be accounted for on two grounds. The first, already mentioned, is that this river, throughout its length, passes through a greater area of uninhabitable mountains than do many of the other streams. The second is that our estimates for the lower Merced are insuffi- cient. They rest in essence on the single report by Moraga, who, as has been shown, tended to underestimate and who did not see, or at least did not report upon, the entire course of the lower river. Moreover there is no report at all from Spanish sources with respect to the San Joa- quin between the mouth of the Chowchilla (Nupchenche group) and the mouth of the Tuolumne. That villages did exist throughout this region is attested by the illuminating account of J. J. Warner, who was a member of Ewing Young's expedition to the great valley in 1832 and 1833. (I use the text as quoted in Warner, 1890.)He says (p.28): In the fall of 1832 there were a number of Indian villages on King's River, between its mouth and the mountains: also on the San Joaquin River from the base of the mountains down to and some distance below the great slough. On the Merced River from the mountains to its junction with the San Joaquin there were no Indian villages; but from about this point on the San Joaquin, as well as on all of its principal tributaries, the Indian villages were num- erous; and many of these villages contained from fifty to 100 dwellings. It is noteworthy that Warner saw no villages on the lower Merced, precisely at the spot where Moraga in 1806 had recorded no less than seven. All of these must have been obliterated during the intervening twenty-six years, striking testimony to the devastation being wrought among the open valley peoples. But from the junction of the Merced and the San Joaquin rivers, along the main 5 3 ANTHROPOLOGICAL RECORDS axis of the valley the villages were numerous, some of them containing 50 to 100 houses or at least 250 to 500 people. What happened to these villages is graphically told in Warner's own words. On our return, late in the summer of 1833, we found the valleys depopulated. From the head of the Sacramento to the great bend and slough of the San Joaquin, we did not see more than six or eight Indians; while large numbers of their skulls and dead bodies were to be seen under almost every shade-tree near water, where the uninhabited and deserted villages had been converted into grave- yards; and on the San Joaquin River, in the immedi- ate neighborhood of the larger class of villages, which, in the preceding year, were the abodes of a large number of those Indians, we found not only graves, but the vestiges of a funeral pyre. At the mouth of King's river we encountered the first and only village of the stricken race that we had seen after entering the great valley. This was the pandemic of 1833, concerning which, in comparison with some accounts, Warner's description is a model of conservatism. It is evident that a combination of circumstances prevents us from making an adequate assessment of the aboriginal population of the lower Merced River and adjacent segments of the San Joaquin. Our density figure is about half the expected value. If we had the full facts, we could perhaps double the estimated popu- lation. Under existing conditions we can feel reason- ably sure of the value given for the area between the Mariposa and the San Joaquin rivers. MARIPOSA-SAN JOAQUIN ......... 19,000 THE SOUTHERN SAN JOAQUIN VALLEY The southern end of the valley, beyond Tulare Lake and the Kaweah River, can best be considered in three parts. The first is the foothill strip from the Kaweah to the Tejon Pass, which was inhabited by the Yokuts tribes Koyeti, Yaudanchi, Bokninuwad, Kumachisi, Paleuyami, and Yauelmani (maps 1 and 2, area IG). The second comprises the lower Kern River together with the former Buenavista Lake basin. This area was held by the Yokuts tribes Hometowoli, Tuhohi, and Tulamni. The third in- cludes the peripheral fringe of relatively high foothill and mountain country of the southern Sierra Nevada and Tehachapi and was inhabited by non-Yokuts people: Tubatulabal, Kawaiisu, Kitanemuk, and the Tokya branch of the Chumash (maps 1 and 2, areas 1A to 1E). Only the Koyeti are described by the Spanish authori- ties hitherto consulted. Moraga mentions the rancheria Coyahete with a population of 400 in 1806. Estudillo in 1819 found a rancheria, which he called Arroyo de Copai- pich, with 200 and one called Canyon Agspa with 400 people. The latter may perhaps be Moraga's Coyahete. If so, the tribe had a population of at least 600 in 1819, but it must have suffered some decline prior to that year. Latta's informants were able to remember 8 vil- lages. Moreover, the tribe was oriented ecologically toward the Kaweah delta and oak forest, although it was actually situated on the lower Tule River. Thus an esti- mate of 800 persons would not be too much for the pre- contact period. The Yaudanchi on the upper Tule River also, according to Kroeber and to Latta, had 8 villages and covered considerably more territory than the Koyel Hence the same population may be ascribed to them. T7 Bokninuwad were evidently a smaller group, since Kr4 ber reports for them only two villages and Latta none, It would not be safe to allow them more than 200 persos If we do so, then the tentative estimate for the three tribes must be put at a total of 1,800. For the remainder of the territory held by the Yokut there are only two documentary references, the diaries of Garc6s in 1776 and Zalvidea in 1806. Both these writers give population data which have been subject to considerable controversy. For the Buenavista region the four pertinent village are mentioned by Zalvidea and are as follows: Village and Tribe Houses Men Women Children Tot Malapoa (Tulamni) Buenavista (Tulamni) ... Sisipistu (Hometwoli) 28 5( Yaguelame (Yauelmani) ... 29 22 8 59 36 144 38 218 )-60 ... ... ... 92 ... ... 300 From even casual inspection it is apparent that Zal. videa did not see the complete population of any one of these villages and that many of the inhabitants had been removed by previous expeditions or were in hiding. Th village of Malapoa is small but presents no serious dem graphic discrepancies. The number of children was lol but as has been pointed out in a previous discussion Z videa was counting as men or women everyone over the age of seven years. The children, calculated according his method, amounted to 13.5 per cent of the total. At Buenavista he found only 36 men to 144 women, a incredible situation unless most of the men had fled or had been killed. Under normal conditions the numbero men should at least approximately equal that of the women. Therefore in order to reconstruct the probabli population we are forced to assume the presence of at least 144 men. This gives a total of 326 persons of whi 8.6 per cent would have been children. For the other villages only the number of men is given, no doubt the men actually seen. Indeed at Yaguelmane Zalvidea "counted" the 92 men he specifies. Significantly, how- ever, he counted men "from 7 to 40 years" and infers that the village had a population of 300. If for Yaguelmn we allow 10 per cent of children seven years old or younger the adults would number 270. If the sex ratio were near unity, then, with 92 men 40 years or younge there must have been 47 men over that age and 135 women of all ages. If the same ratios are applied to Sisipistu with 55 men from 7 to 40 years of age, the population would be 180. This figure is quite consisten with the number of houses, 28, for the number of per- sons per house would then be 6.43. The four villages (Malapoa, Buenavista, Yaguelame, and Sisipistu) con- sequently must have had populations of 59, 326, 300, and 180 respectively. The average of the four is 191 persons. Since there are no other historical data pertaining the lake region, it is necessary to utilize the village lists of Kroeber (1925) and Latta (1949). These invest gators, through their informants, have located 3 villag I I I I 54 COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY for the Hometowoli, 1 for the Tuhohi, 3 for the Tulamni, and 2 for the Yauelmani of the lower Kern River, making 9 in all. As suggested with respect to other areas the (umber of villages was undoubtedly as great in 1806 as In 1840 or 1850. Hence we can be assured of at least 9 in 1806. For size it is proper to use Zalvidea's average of 191 inhabitants, thus giving as the population of the Buenavista basin 1,720. For the southern foothills we must rely upon the Siary of Garces. Gifford and Schenck (1926) discuss '*iW document at length, concluding (p. 21) that the pop- 41tion actually seen by Garces north of the slopes of ~the Tehachapi was 750 and that the total population r'South of the Tule River" was 1,000 to 1,500. Since the esent writer must differ from these authors, it is rth while to review once more the evidence furnished Y the Garc6s account. In so doing the exact route of the bzplorer must be made plain. * On May 1, 1776, having previously descended the swthern mountains to the valley floor, Garc6s broke 4p: i Having gone one league northwest I came upon a large river which made much noise, at the outlet S (al salir) of the Sierra de San Marcos and whose waters . . flowed on a course from the east through a straitened channel. ues, ed., 1900, pp. 280-281). The river of course s the Kern and the spot was without question the point which the river suddenly breaks out onto the plain m its canyon. The water was here swift ("made much 3e"). It literally "sallied forth" from the mountains, its course from the east was through a narrow el. This place is about 14 miles east-northeast of ersfield on California State Highway 178. Garces then went downstream "a little way" and d a rancheria (no. 1) on the right bank. After going ttle way farther he saw a rancheria (no. 2) on the bank and another (no. 3) "to the west." He went stream no more than 2 or 3 miles, otherwise, as his invariable custom, he would have specified his ces in leagues. Three rancherias can therefore ocated on the Kern between the last abrupt slope of eastward hills and just below the mouth of Cotton- Creek. These correspond on the map to Kroeber's ges Altau and Shoko of the Paleuyami and Konoilkin he Yauelmani, although the actual identity is by no s assured. After crossing the river with difficulty Garc6s struck west "and a little north" for 3 leagues. This brought to a stream where there was a rancheria (no. 4). m a point 3 or 4 miles below the entrance of the Kern r canyon a line running northwest by north extends nally about 7 miles across T28S, R29E to reach Creek near the northern boundary of the township. Mter passing the night at the rancheria mentioned 4), Garces went straight north for 4-1/2 leagues. the way he went by some deserted rancherias. These es were not temporarily deserted, with the inhabi- in hiding. They were "rancherias despobladas," si, permanently depopulated or abandoned. It is sreting to speculate on the cause of this phenomenon, the depopulation can have been due only to intertribal are or disease. We know nothing of any native wars ficient magnitude to have destroyed several whole ges. On the other hand, as Garc,6s himself later ts out, Spaniards had already penetrated the region. Fages was in the southern valley in 1772 on his way to the Colorado and Garces found at least one desert- ing soldier living with the Indians. It is quite possible that decline of population had already begun as early as 1776. After traveling 4-1/2 leagues Garces found another rancheria (no. 5), at which he spent the night of May 2-3. This must have been somewhere near the hamlet of Woody at the southern boundary of T25S, R29E. On May 3 he moved another 2-1/2 leagues, still north, to reach the White River near or slightly to the west of the village of White River in T24S, R29E. Here he camped at a rancheria (no. 6) of 150 souls. On May 4, having reached his farthest point north, he visited another rancheria (no. 7) half a league east. At rancheria no. 6 he found an Indian who was a fugitive from the coast and also heard that two Spanish soldiers had been killed for molesting Indian women. The contact with the whites was therefore clearly established. Stephen Powers (1877), who was in the San Joaquin Valley in the decade of 1850 says that "on White River there are no Indians, neither have there been any for many years." Here again is an indication of depopulation at a very early date. On May 5 Garces started to retrace his steps south- ward, reaching at 2-1/2 leagues the previous rancheria (i.e., no. 5). From here he must have diverged some- what eastward of his northbound trail for at 2 leagues he saw another rancheria (no. 8) "to the east" which he had not seen on the way up. This probably was toward the eastern side of T26S, R20E. Tnen, he says, he went southeast 3 leagues to Poso Creek. This would put him on Poso Creek near the center of township T27S, R30E, a point about 9 miles airline above his place of crossing on May 2. Here he found a rancheria (no. 9), the chief of which told him about another rancheria (no. 10) to the east where a Spanish renegade lived with an Indian wife. The following day, May 6, he started out again south or southwest and got lost in the hills of upper Poso Creek. In these hills between Poso Creek and the Kern River he found another rancheria (no. 11) of "more than 100 souls." This was probably in the northern part of township T28S, R30E. Finally on May 7 he reached the Kern 1 league above his first crossing. His first crossing had been accomplished 2 or 3 miles below the mouth of the canyon hence he must have come out very close to the mouth. He then went downstream to the rancheria where he had crossed (no. 1) but he did not stop here. He continued down the river for 2 leagues to a rancheria he had not seen before (no. 12) and which had "some 150 souls.' Two leagues downstream from rancheria no. 1, or about 3 leagues below the mouth of the canyon would have put him at a point roughly 5 to 6 miles east-north- east of Bakersfield, not at the site of the city, as is supposed by Coues (1900, p. 299). On May 8 he went 3 leagues south-southwest, then turned and traveled 6 leagues southeast and east to the Tehachapi. These dis- tances and directions plotted on the map place him just at the mouth of Tejon Creek. To summarize the rancherias mentioned: Garc6s saw four villages on the Kern in territory of the Paleu- yami or Yauelmani (nos. 1, 2, 3, 12), six on Poso Creek or minor watercourses to the north thereof (nos. 4, 5;8, 9, 10, 11), all Paleuyami, and two on White River (nos. 6, 7) in the territory of the Kumachisi. The size of these villages has been subject to some debate. Garc6s cites two with 150 persons and one with 100, but Gifford and Schenck think that he specifies popu- lation only for the largest places. The other nine would therefore be smaller. These authors, however, put the average village size at about 60 (750 people in 12 villages). 56 ANTHROPOLOGICAL RECORDS Deducting 400 for the three rancherias specified, the average of the other nine would be 39 which seems much too low. If Zalvidea's figures are any criterion, the villages on the Kern should have averaged at least 100 inhabitants, and it must be noted that Garc6s found two rancherias in the hills with 150 and 100 persons respectively. Thus it seems reasonable to allow an average of 100 rather than 60. If so, the population seen by Garces was in the vicinity of 1,200. Now it is evident that Garces did not see all the villages in the region. He covered about 10 or 12 miles of the Kern below the canyon, a good deal of upper Poso Creek, and perhaps 5 miles of White River. He never reached the lower stretches of the rivers at all. It is fair to assume that there were as many rancherias which he did not see as there were seen by him. If so the estimate of the population should be doubled, making 2,400. One secondary piece of evidence is at hand. Garc6s saw 8 villages of the Paleuyami (6 in the hills, perhaps 2 on the Kern). Now Zalvidea in 1806 says that the Pelones (Paleuyami) had at that time 13 rancherias. Allowing for shrinkage in the intervening thirty years, this is twice the number seen by Garc6s. We may at this juncture have recourse to river mile- age estimates. It was found previously (p. 36) that for the Stanislaus, Tuolumne, Merced, Mariposa, and Chowchilla there was in 1850 0.34 village per mile of stream, with the Chowchilla having the lowest value, 0.20 village. For the Merced and the Kings rivers below the foothills in the first years of the nineteenth century it was calculated that there were on the average 65 per- sons per river mile. Assuming that the average village size was 150 inhabitants, there would have been 0.44 village per river mile. The southern streams were probably more sparsely inhabited than those just men- tioned. Hence it is reasonable to apply the factor found for the Chowehilla, 0.20 village per mile, to the White River, Poso Creek, and the Kern River. There are about 150 miles of stream in these systems east of a line running from Porterville to Bakersfield, a line which Kroeber takes as the approximate westward limit of the foothill tribes. This means a probable 30 villages. If the average of 100 persons per village is used, as suggested above, this means a population of 3,000. The direct documentary approach thus gives 2,400 and the indirect method 3,000. A fair figure would be the mean of the two, or 2,700. The peripheral hills on the southeast and south were held by several tribes. The entire upper Kern River, above the present village of Bodfish, belonged to the Shoshonean group, the Tubatulabal (area 1E). Kroeber thinks they may have reached a population of 1,000, which seems a reasonable figure. From the Kern and Walker's Pass south to Sycamore Creek (area 1D) were the Kawaiisu, a tribe, according to Kroeber, of 500 persons. In the southeastern corner from Sycamore Creek to Poso Creek were a few Yauelmani and the Kitanemuk. Pastoria Creek and Alisos Creek were occupied by a northward extension of the Alliklik, and from Alisos Creek westward to Bitter Water Creek were found the Tokya group of the Chumash. For the groups beyond the Kawaiisu there are no population data of any kind. Even Kroeber fails to make an estimate. If we say 1,000 for them all in aboriginal times it will be a pure guess, but one which may be somewhere near the truth in view of the extent and character of the terrain involved. The total for the peri- pheral region would then be approximately 2,500 and that for the southern end of the valley as a whole 6,920, ori round numbers 6,900. SOUTHERN SAN JOAQUIN VALLEY ....... 6,900 THE NORTHERN SAN JOAQUIN VALLEY The remaining portion of the Yokuts-Miwok territor lay in the valley and foothills north of the Merced Rivet This area (see maps 1 and 5, areas 8-13 inclusive), pa ticularly the delta of the San Joaquin and Sacramento rivers, was entered relatively early by the Spaniards and by the year 1820 had been almost completely swept of its native population. The names of many whole trii have been lost and the exact locations of many othersa, now almost impossible to ascertain. Of village names only thos'e few are known to us which were preserved, often by chance, in the mission records and accounts ol expeditions. Several attempts have been made to recon struct the aboriginal human geography but none has bee entirely successful. Kroeber's account, which accom- panies his discussion of the Plains Miwok and northern Yokuts in the Handbook of California Indians, is mani- festly incomplete. Merriam's paper on the Mewan Stocd of California (1907) is helpful, but probably the best work of the modern investigators is that of Schenck (191 The early nineteenth-century accounts for this regiona also less satisfactory than for the central and southern: parts of the San Joaquin Valley. Moraga's record is us ful only for the Tuolumne River, and the delta is cover only by Abella and Duran. It is true that both Sutter anO Gatten give figures for villages south of Sacramento bu their information pertains only to the badly depleted natives of the 'forties. Hence their censuses are of litt value for assessing the aboriginal condition. One source not available for other areas is the mis8 records. The converts from the delta and lower San Joaquin Valley were brought almost exclusively into th4 San Francisco, San Jose, and Santa Clara missions. Tl baptism books of these missions have been preserved, and two copies have been made. The first, of the San Francisco Mission, was made by A. Pinart in 1878 and is at present in the Bancroft Library in Berkeley. The other records, copied by S. R. Clemence in 1919, in- clude the records of all three missions and are now to be found, in typed form, among the manuscripts in the file of C. H. Merriam. The baptism books set forth thl name and village of origin of every native in the missia as well as the date of baptism. Newly converted gentile are readily distinguished from infants born in the miss itself, since the origin of the latter is ascribed to the mission and not to a village. In addition to the names o villages, not all of which can be located with certainty, the dates of baptism constitute almost conclusive evi- dence. If the baptisms from San Francisco and Santa Clara are tabulated by village and date, it is very clea that the villages of local tribes were cleaned out before the year 1805. At this point an entirely new set of naa appears, most of which are undoubtedly in the T'ulares, Hence, if the name of a village does not correspond to any now known to ethnographers and no baptisms are ri ported from it prior to 1805, the conclusion is warrant that the village was actually situated in the central vall The same assumption may be made with somewhat less certainty concerning the San Jose records. This missic was founded in 1797 and its earliest converts were drai from the Costanoan tribes on the east shore of San Fra I - - - i t I I j- I I COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY cisco Bay. The reduction of this region may not have been complete by 1805 and Tulare Indians were coming In by that year. Hence there is a chance of overlap. This source of error, however, may be excluded for all practical purposes if no doubtful village which con- tinued to furnish converts after 1810 is included in the 'list, for the reduction of the Costanoans was certainly 'complete by that time. Concerning village size various items of information are available. In the diary of Ramon Abella in 1811 he mentions that the Cholbones had three rancherias with a population of 900, or 300 per rancheria. That of the oyboses had 180 and that of the Tauquimenes 200 men md 60 houses. The population of the latter tribe, if we pply the ratio found by Zalvidea at the southern end of evalley, should be 650. This ratio, it will be remem- red, is based on Zalvidea's statement that he counted men all males between the ages seven and forty. If, the other hand, we assume that Abella referred to males except small children and further that the sex atio was unity, the adults would have numbered 400 d, if 15 per cent of the village were children, the tal would be approximately 470. However, in the rthern end of the valley we have much more solid ta with which to work than at the extreme south. The baptism records of the missions of San Jose Santa Clara to which reference is made above in- de for each gentile village a breakdown of men, men, and children. These data have been already scussed in connection with the rancherias on Lake lare and it has been shown that, if proper correction made for the sex ratio, men and women each con- buted 41.8 per cent of the population and children .4 per cent. It is clear that in the north the Francis- employed their standard system of calling children persons under the age of ten years (not seven years) including as males all men above the same age. videa's system was used only by himself. Conse- ntly, a village with 200 men would have contained 3 persons in all. For the village of the Tauquimenes with 60 houses average would have been 9.38 persons per house. at this number is not excessive is demonstrated by account of the village of Chuppumne contained also Duran's diary. This rancheria had 35 houses, some which were 40 to 50 paces in circumference. Since pace is roughly a yard the diameter of such a house uld be 43 feet, amply sufficient to accommodate 9 per- Chuppumne would thus have had a population of 5, Duran also mentions a rancheria of the Ochejamnes *ch had 40 houses, or 360 inhabitants. LuisArgulello (MS, 1813) describes an expedition er the command of one Soto, whose party was at- ed by Indians in the marshes of the delta. Schenck 26, p. 129) locates the scene as in T5N, R4E, near ut Grove and designates the tribe as the Unsumnes Cosumnes. Now ArgSello states that the expedition pt up on the Indians overnight and attacked at dawn. ey were surprised to find that their coming had never- ess been detected and that the Indians had sent away women and children. The Spaniards were met by a ce of warriors, which Soto placed as his best esti- te at 1,000 persons. These were drawn from four erias in the vicinity. One may always exercise pticism with reference to these estimates of enemy ces, particularly in this instance, since the Spaniards e roughly handled and suffered several casualties in tion to being forced to withdraw. On the other hand, invaders consisted of 13 well armed Spaniards and 100 Indian auxiliaries. Nothing like an equal number of natives could have withstood them. Soto's estimate may be cut in half but at least 500 warriors must be allowed, or 125 for each of the four rancherias. Now the fighting population, even in a great emergency, does not coincide with the total male population. If there were 500 warriors, there must have been fully 300 young boys, invalids, and old men who were not present. Hence we must concede a male population of no less than 800 for the four villages. If the percentage values established previously are used, the mean village size was approximately 475. To the villages just described may be added the one seen by Moraga on the Stanislaus River in 1806, which had 200 inhabitants. These twelve villages thus yield an average of 362 inhabitants each. Although throughout the territory many rancherias were doubtless small, it is equally probable that some were very large, approaching the magnitude of Chischa and Bubal in the south. Hence, unless in some particular instance there is clear reason to believe otherwise, 300 cannot be regarded as an excessive esti- mate for the average village of the delta. In considering in detail the population of the delta (see map 6, area 13), it is convenient to segregate groups according to tribal distinctions rather than strictly ac- cording to geographical points. The reason lies primarily in the fact that the early writers and the mission records were relatively explicit with respect to names of villages and groups but were badly confused with respect to locali- ties. In the densely populated but physiographically homo- geneous delta region, with its scores of small streams, sloughs, and islands, explorers found it very difficult to establish clear landmarks by which the inhabitants might be oriented. A state of confusion has arisen of a kind to generate many controversies among ethnographers, con- troversies which are not pertinent in the present connec- tion and which it is desirable to avoid as far as possible. In order to adopt a more or less uniform system with respect to tribal nomenclature and arrangement it is proposed to follow here the practice of Schenck (1926), who has made an exhaustive study of the area. Bolbones (syn. Cholbones, Chilamne, Chulame).- This large group occupied the sloughs of the lower San Joaquin west of Stockton. Schenck, on his map (1926, p. 133) shows their territory as being bounded by the main stream of the San Joaquin River on the east and by the channel now known as the "Old River" on the west. This delineation of their habitat is supported by the diaries of Abella and Viader. Schenck classifies the subtribes or divisions of the main group as follows: Cholbones ..................... a group Pescadero ................... a village Jusmites or Cosmistas ........ a village plus Fugites or Tugites ............ a village plus Tomchom, under Fugites ...... a village Nototemnes ................... a village Although these natives are mentioned frequently in the correspondence of the period, the first recorded exploration of their area was that of Fr. Jose Viader in 1810. This missionary left Mission San Jose on August 15 and went by way of Pittsburg and Antioch to the mouth of the San Joaquin, whence he traveled southeast to Pescadero, ". . . la rancheria de los Cholvones." Leav- ing the rancheria he went on up the river. Viader's second expedition was carried out during the month of October of the same year. This time he went directly fromSanJose to Pescadero, which he sayswas 15 leagues I 9 11 e v 57 ANTHROPOLOGICAL RECORDS northeast to east-northeast of Mission San Jose. The account at this point is not particularly lucid. The entry for October 20 states that at Pescadero the gen- tiles were having a dance (bayle). That for the follow- ing day begins with the statement that at dawn Viader's party attacked ". . . asaltamos una rancheria de este lado del rio y solo escapo un Christiano . . . ." Then they attacked another rancheria on the other side of the river and captured 15 Christians and 69 gentiles. From the context it may be inferred that the first rancheria attacked was the one at which the dance was being cele- brated on the evening of the 20th, that is to say, Pes- cadero. If it was, then there was another, quite sizable, village just across the river. If the first village was not Pescadero, then there were two other villages in close proximity to it. The next visitor was Fr. Ramon Abella, who left San Francisco by boat on October 15, 1811. Passing Sherman Island on the 18th and wandering erratically through the swamps he reached the "tierra de los cholbones" on the next day. On October 20 he reached the village of Pescadero but made no comment on it in his diary. After examining the territory of the Cosmis- tas and Boyboses 5 to 15 miles to the east, the party turned about 8 to 9 miles (3 leagues) northwest, follow- ing the general trend of the river downstream. At this point they found a rancheria of 900 persons "divididas en tres rancherias, alguna distancia una de otras. No vimos que la una: Se presentan como 150 personas y nos enseniaron al desembarcadero y las mismas casas que havia duplicado gente . . . ." Abella's dis- tances are extremely inaccurate but it is apparent that the three villages mentioned were north or northwest of Pescadero. The key village in this complex is Pescadero, a rancheria to which repeated reference is made in the documents of the period and whose identity neither Viader nor Abella could have mistaken. That it be- longed to the Bolbones is attested by Viader's expres- sion ". . . la rancheria de los Cholvones." Viader saw at least one and perhaps two other villages near by be- longing to the same tribe. Abella clearly states that he saw three rancherias in addition to Pescadero. One of these may have been the one attacked by Viader, and if so, the entire group included a minimum of four villages. Otherwise, there were at least five. Abella's count of 900 persons for the three villages appears accuirate and reasonable. On the other hand, Pescadero was evidently regarded as the most important rancheria of the area and probably was more populous than any other. Hence it must have contained no less than 400 persons. The sum of the four villages would then be 1,300. Between 1806 and 1811 the mission records show a total of 200 baptisms ascribed to the Cholbones, most of them at San Jose. In addition, there were 81 baptisms from 1821 to 1828 designated Chilamne. At the time of Abella's visit, therefore, the area had been subject to repeated raids for the purpose of securing converts and must have undergone serious social and economic dis- turbance of the type noted throughout the entire San Joaquin Valley. Merely adding the 200 missionized natives would bring the population estimate for the Bolbones up to 1,500, and the aboriginal value was probably even higher. The Jusmites, or Cosmistas, are credited by Schenck with "a village plus," meaning certainly one and probably two or more. Viader, on his second expedition, found 'los indios Jusmites" about 2-1/2 leagues southeast of and up the river from the second village, which he at- tacked on October 21. This places them in the locality shown by Schenck on his map (1926, p. 133), i.e., in northwestern TIS, R6E. No further information is give4 by Viader. The next year Abella found "la rancheria di los Cosmistas" in approximately the same region, but gave no data regarding size. Neither author implies in any way that there was more than one village. At San Jose 86 converts were baptized from "Jossmit," a numr ber which suggests a village of fully 300 inhabitants. Viader on his first expedition, on August 20, went south-southeast from Pescadero for 3 leagues and reached a village "cuyo capitan se llama Tomchom,' He then went 2-1/2 leagues southeast from the JusmiteI and reached "los indios Tugites." Both Tomchom and Tugites therefore appear to have been in the same gen area. For this reason Schenck has placed the Tugites, as a tribe, directly south of the Jusmites and has call* Tomchom a village of the tribe. It is perhaps more lil that there were two villages involved (rather than a tr and an included village), designated respectively Tom-] chom and Tugites. This view is substantiated by the baptism data. Of the entire group 268 were baptized, rather equally distributed between San Jose and Santa Clara. Over half the conversions occurred in the year 1811. The San Jose book lists 126 from "Tamcan" and 7 from "Tuguits." The Santa Clara book has 125 from "Los Tugites" and none under any other designation. It may therefore be concluded that two villages, or sub- tribes, were involved, one of which was taken to San Jose and the other to Santa Clara. A total of 268 con- verts would imply a population of at least 500 persons at the time of conversion and probably more aboriginz The village of Nototemnes is mentioned only by Du in his diary of 1817. In the night of May 22-23 he pass "la rancheria de los Nototemnes," but did not actually the village or count its inhabitants. However, the ran eria furnished 97 converts to Mission San Jose. It mu therefore have contained at least 200 people. Schenck shows the Nototemnes as covering nearly two townshipi in the northern delta region and calls them "a village plus." He cites, however, no authority for this view other than Duran, and Duran, as mentioned above, refe only to the rancheria of the Nototemnes. There is no reason, consequently, for assuming more than one viT for the tribe or group. In summary, the Bolbones tribal complex consisted fully eight medium to large villages. Those belonging the Bolbones proper, four in number, were estimated contain 1,500 persons. The Jusmites were allowed 300 persons, the Tugites 500, and the Nototemnes 200. Th' total is 2,500, and the average village size slightly ove 300 persons. (Bolbones ....... 2,50 Leuchas.-Schenck shows this tribe as living east of the San Joaquin River 10 to 15 miles south of Stockton. He implies that the tribe contained two villages, Coy- boses and Pitemis (Aupimis), in addition perhaps to other settlements. The mission books mention all three names and show baptisms (figures in parentheses), wm. may be tabulated as follows. Baptisms, San Jose Baptisms, Santa Cla Leuchas. .. "Leucha" (26), 1805- 1812 (88 per cent in 1805- 1806 "Los Leuchas" (81) 1805-1809 (85 per cent in 1805) I I i 58 COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY Baptisms, Santa Clara (60), 1814-1831 (98 per cent in 1814-1816) None To judge by the three separate periods in which the Osjority of the baptisms occurred there were three ,oups of people: the Leuchas, who were brought into gefold primarily during 1805 and 1806, the Coybos, Fincipally in 181 1-1812, and the Pitemis, converted Fo or three years later. The Leuchas were taken to i missions, but the Coybos were brought only to San e and the Pitemis only to Santa Clara. Abella said tin 1811 the village of Coybos had 180 inhabitants, a Ire which has been used in computing the average age size. But the aboriginal population was probably ater, This view is substantiated by the events which ceded Abella's visit. In 1805 Father Cuevas of San e Mission went on an unauthorized expedition to the uchas-the best account is that by Jose Arguiello (MS, 5)-in search of converts. He was badly treated and U of his men were wounded by the natives. This and 'punitive expeditions which immediately followed no btaccounted for the wave of conversions in 1805 and B, But at the same time the entire aboriginal group estionably suffered heavily from battle casualties economic disturbance so that the population five years r must have been seriously reduced. It is thus justifi- to assume that originally there were three villages that each was of average size. The population may efore be set at fully 900 persons. Some further information is derived from the recollec- of Jose Maria Amador (MS, 1877). This pioneer, received his facts second-hand from his father, men- (pp. 13-15) the campaign of 1805 against the "Loe- ' who, he says lived 4 to 5 leagues from Livermore. would put them west of the San Joaquin River, south e Bolbones, in T1S, R5E, not on the east bank as by Schenck. Amador then goes on to say that after Cuevas affair the Leuchas ". . . se habian ya cambiado o de San Joaquin a una rancheria que se llamaba de Pitemis.' They were all captured and taken to San Jose. thus reasonably clear that the Leuchas originally did west of the river, and crossed over to the east side result of the punitive expeditions of the Spaniards. ermore, the village of the Pitemis was already in tence at this time, probably at or near the spot shown henck. Coybos undoubtedly was another village within same area. This region, therefore, at the time of a's visit in 1807 contained the established villages of nis and Coybos plus a residue of unconverted, fugitive chas who had taken refuge in them. adors assertion that the Leuchas were all captured taken to San Jose is not borne out by the baptism fig- which show only 23 Leuchas enrolled at Mission San in 1805 to 1806. Many more, actually 73, were bap- at Santa Clara in 1805. The total is 96, and scarcely seents the entire personnel of the group. Nevertheless, add the casualties of battle, disease, and exposure to baptized in the missions, and allow for the dispersion e remainder, the sum will amount to no less than the assumed above for the Leuchas. There are numerous other letters pertaining to this matter in the same of the Provincial State Papers. As for the Pitemis, Viader, on his first expedition, left Pescadero on August 20, 1810, and traveled south- southeast at some distance from the river. Within 3 leagues he passed ". . . en frente de una rancheria Aupemis." Schenck says (p. 141): "Pitemis is a village of the Leuchas and it seems that Aupimis is to be identi- fied with it." This cannot be true because Viader is high- ly explicit to the effect that he was west of the river and Amador is equally emphatic in stating that Pitemis was across the San Joaquin from Leuchas, i.e., to the east of it. Since Viader's visit was in 1810, after the Cuevas affair, there must have been three rancherias of the Leuchas and their allies: Aupimis, Pitemis, and Coybos. Parenthetically, and for the record, the present wri- ter would like to offer the comment that certain modern writers tend to assert the identity of Spanish or Indian names without adequate evidence. Schenck's opinion that Aupimis and Pitemis were the same place could have been based upon no more than a fancied resemblance in the names. Also, on page 141 of his paper he says: "The Leuchas might possibly be identified with Kroeber's Lakisamni (Yokuts) on the Stanislaus river." A brief examination of the mission records, apart from any other evidence, shows conclusively that two separate and distinct tribes were recognized by the contemporary missionaries. (Leuchas et al ......... 900) Ochejamnes.-This tribe is placed by Schenck on the east bank of the Sacramento River near the mouth of the Cosumnes. Kroeber refers to the village of Ochehak and considers it a "political community.' He shows it on his map (1925, p. 446) as lying on the Mokelumne, due north of Stockton. Duran, in his diary, May 21 (MS, 1817), describes how he followed the main stream of the Sacra- mento, i.e., the left branch, on his way back from his stopping point above Courtland. He reached the rancheria "llamada de Oche jamnes," which, although it contained 40 houses, was deserted. Quite soon thereafter ("a poco rato") he reached "la punta de la isla llamada de los Quenemsias,' which has been identified definitely as Grand Island. Clearly, therefore, in 1817 the Oche- jamnes had a village on the Sacramento higher up the river than is shown by Schenck. According to Duran the village had 40 houses, which would mean 360 persons without reckoning possible sub- sidiary rancherias. The name is mentioned for only one mission, San Jose, at which 428 Ochejamne, or Ooche- ganes, were baptized between 1829 and 1836. This is prima facie evidence that Duran, who saw them in 1817, was referring, as he implies, only to one rancheria and that the tribe was actually larger. This idea is supported by the account of Jos6 Berreyesa in 1830 of severe Indian fighting in the delta (Berryesa, MS, 1830). The Oche- jamnes and the Yunisumnes with certain American trap- pers were arrayed against the Californians, who had gathered together 450 auxiliary fighters from the Cosum- nes and other tribes. No value is placed upon the number of Ochejamnes but it must have been considerable. It was probably as a result of this campaign that 428 mem- bers of the tribe were baptized at Mission San Jose. Even with a relatively complete conquest many of the natives must have escaped; hence in 1830 their total number must have reached 500. But this was in 1830, after a generation of expeditions and petty warfare. The aboriginal number must have been considerably greater, let us say 750. (Ochejamnes ........ 750) Baptisms, San Jose ~Pitemis... None Coybos.... (94), 1808-1826 (71 per cent in 1811-1812 w -: . F, 59 ANTHROPOLOGICAL RECORDS Guaypem.-This group is thought by Schenck to have been simply a village but Merriam (1907, p. 350) re- gards them as a tribe called the Wipa, located on Sher- man Island near the Sacramento River estuary. Duran in his diary says that Guaypens is 6 leagues south and southeast of the fork of the river below Courtland. Allowing for his usual exaggeration of distances, this puts the rancheria near the mouth of the Mokelumne, in the vicinity of Walnut Grove. He speaks of the ranch- eria "de los Guaypens" and saw only a few people. Thus neither size nor locality supports the contention that Guaypem was synonymous with Wipa. The tribe was not converted until relatively late, 41 converts being taken to San Jose between 1821 and 1824. By that time the tribe had been subject to severe attrition. Thus the evidence points to an aboriginal group consisting of one village of average size, or close to 300 inhabitants. (Guaypem ....... 300) Quenemsias.-These people, who lived near the two preceding tribes, are designated a "group" by Schenck (p. 136). They covered, according to him, "the south- ern part, or perhaps all, of Grand Island." The ecclesi- astical diarists make no mention of them save the refer- ence by Duran to the "isla llamada de los Quenemisias.' One other citation is worth mentioning, however. In the Bancroft Transcripts is a document dated January 31, 1796, entitled "Informe en el cual el teniente Hermdo Sal manifesta lo que ha adquirido de varios sugetos para comunicarlo al Gobernador de la Provincia," which gives a description of the lower reaches of the San Joa- quin and Sacramento rivers and the delta and mentions the natives (Sal, MS, 1796). In detail, the account is extremely inaccurate. However, one of the Indians ". . . dio noticia de las naciones Tulpunes, Quinensiat, Tau- nantoc, y Quisitoc: los primeros son de la orilla del estero; los 20S estan del otro lado de los rios ...." Although no numerical data are given, the mention of the Quenemsias (Quinensiat) as a "nacion" in the delta region establishes them as a group of more than aver- age importance. The mission books show 185 Quenem- sias baptized at Mission San Jose. Roughly double the number of baptisms may be taken as the aboriginal popu- lation, i.e., 400. (Quenemsias ........ 400) Chuppumne, Chucumes.-Schenck places these two settlements, which he calls villages, on the Sacramento River near the mouth of the Cosumnes. Most of our documentary information concerning them is derived from the accounts of Duran and of Luis Arguiello. Luis Antonio Arguello accompanied Duran on his ex- pedition and wrote a report to the governor in the form of a letter, dated May 26, 1817, the original of which is preserved in the Bancroft Library (library no. fm- F864A64; also typed copy). The existence of this letter evidently was not known to either Kroeber or Schenck. It is less complete and less detailed than the diary of Duran but it is of value in checking the statements made by the latter. On May 16 the party reached the foot of Grand Island and on May 17 proceeded up the left-hand (i.e., western) watercourse. The village of Chucumes was found 8 leagues (leguas) upstream, according to Duran, 13 miles (millas) according to Argilello. The latter estimate is probably closer, since Duran is notoriously inaccurate (usually on the side of overestimate) in his computation of distances. Here Duran counted 35 houses whereao Arguello says 36, a sufficiently close correspondenc As indicated previously, a population of 315 persons probable. Continuing their journey, they went on for miles (Arguiello); Duran says approximately 3 league There they stopped at a rancheria, "arruinada" acco ing to Argiuello, although Duran makes no mention of this. On May, 18 the party went on upstream, making d the day 4 leagues (Duran) or 16 miles (Arguiello). states that after going 1 league they got back into the main stream of the Sacramento. This was clearly at head of Grand Island, close to Courtland. At 1 leagu beyond this point, on May 19, they found the rancher Chuppumne, which was deserted. The location there fore was very close to that shown by Schenck on his (p. 133) and, if we can put any credence in the Duran Arguiello account, a good many miles north of Chucu Near Chuppumne Duran saw three other rancherias i the distance (inland ?) but could not get at them. On May 20 the expedition pressed on upstream for 5 mi (Argiuello) or 4 leagues (Duran), at which point they-' turned around and began the return trip. On May 21 Argulello says that they passed "algunas rancherias," all deserted, which may well have been those mentio by Duran on May 19. On the river frontage covered from May 17 to Ma the expedition saw a minimum of 6 villages, 2 of whi are mentioned by name (Chucumes and Chuppumne) for 1 of which the houses were counted. If all these 5 lages were of comparable size-as they may have be aboriginally-then the total population represented wd have been 1,800. This estimate would of course not i clude other villages which the expedition did not see. The mission records show for San Jose a total of persons baptized from Chucumne and Chuppumne, of- whom 322 were converted during 1823 and 1824. We predicate, therefore, a residual population of 700 to just prior to those years. That the area had suffered severely before that is attested by the deserted and "ruined" rancherias seen by Duran in 1817. It is quiti probable that the aboriginal population reached a val of 1,500. (Chucumes, Chuppumne . ....... 1 Chupunes (Chupcanes), Tarquines (Tarquimenes, Tauquines), Julpunes (Tulpunes) and Ompines.-Thisi constellation of tribes is best considered collectively first, because there are no direct estimates of their population, and second, because they occupied a rela tively unified area. Schenck places them along the south shore of Sui8 Bay from the east entrance of Carquinez Strait and through the slough region between the Sacramento an San Joaquin rivers as far upstream as Isleton on the Sacramento. However, he points out that there is gre uncertainty with respect to their exact location, an u certainty which is emphasized by the wide divergence between his views and those of Merriam. Even the So ish accounts present numerous discrepancies. In vied this state of our knowledge Schenck makes the very r sonable suggestion that the lower delta tribes may ha been so greatly disturbed and shifted around during t period from 1775 to 1810 that the aboriginal location were forgotten. It is worth while to examine in some detail some of the evidence on this problem. We may begin with examination of the area at and just east of' Carquinez Strait on the south shore of Suisun Bay. Th I I ---- ---- - --- -L I - -- - -- - -1 II 4, 4 I I 7 60 COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY consideration entails a preliminary discussion of two small groups, the Aguastos and the Huchium (syn. llabastos, Quivastos, Juchium, Huchimes, Tuchimes, etc.). This tribe or group of tribes, which must have been of some importance, is not mentioned by name by Kroe- ber or Schenck, but there is a brief set of typed notes in rthe Merriam collection in which the location is discussed (US entitled: "On the East Side San Francisco Peninsu- h"). The multiplicity of synonyms, however, as well as large number of neophytes involved, indicates that tese tribes were very familiar to the missionaries. The Merriam notes (pp. 5 and 6) point out the follow- considerations. 1. "Abella's diary (1811) speaks of present Point San blo as the Point of the Huchunes and says their terri- r extended on the mainland from this point to Pt. San dres (Pt. Pinole)." 2. Several rancherias belonging to this tribe are men- ned as being on the east side of the bay. 3. "The mission books locate the Habasto tribe 'on other side of the Bay from the Mission of San Fran- co toward the estero which goes to the rivers (Suisun y).' Abella's diary calls Point San Pedro the Point of e Abastos.' Merriam therefore was strongly of the opinion that se tribes inhabited the south shore of San Pablo Bay d did not extend farther than Carquinez Strait. On the other hand, the item in the mission books ted by Merriam (par. 3, above) indicates Suisun Bay her than San Pablo Bay. Moreover, there is another tement in the baptism books alongside the designation guastos o Huchum" to the effect that this tribe was 16 18 leagues by water from San Francisco. This distance uld place them close to the site of the modern town of ttsburg, that is, on the southern shore of Suisun Bay. this area is assigned by Schenck to the Tarquines perhaps the Julpunes, tribes which are also clearly toned by name in the mission records. If the Aguastos extended from Richmond to Crockett thereabouts, they were Costanoan and strictly bay ple; hence not pertinent to this study. If they lived agSuisun Bay, regardless of their ethnic affiliation may be included for demographic purposes among delta tribes. Some further light can be thrown upon problem by an analysis of the dates shown for bap- ms in the San Francisco Mission records. If the baptisms of gentiles are tabulated according to ge and year, it is seen immediately that the con- sions in the first year, 1777, were all from local cherias. This group was extended during the follow- decade until the San Francisco peninsula had been mpletely covered. However, after the year 1792 all tion of the peninsula abruptly and entirely ceases. early as 1778 on the other hand baptisms are listed m a village (Halchis) specified as being in the "sierra ente de la otra banda." In the succeeding years villages ribed to the "otra banda" become more frequent and h a peak between 1790 and 1795. Subsequent to 1800 conversions from these places diminish rapidly and ppear. Now we know by following the documentary unts of expeditions that during the decade 1790 to Othe great effort of the San Francisco Mission was nded in securing neophytes from the east shore of Francisco Bay as far north as the Carquinez Strait. re are no baptisms of gentiles whatever listed in the Francisco books for the years 1797, 1798, and 1799. refore it is reasonable to assume that the supply of anoans from the east bay had been exhausted. Fur- thermore, village names qualified by the term "otra banda" and appearing in the baptism record for the first time prior to 1800 must certainly refer to villages in this region. Among these are rancherias stated as be- longing to the "nacion Juchium" together with the sep- arate designation "Tuchimes." Thus it is clear that the Huchium lived, as Merriam believed, on the east shore of the bay. After the inactive period at the end of the century a flood of neophytes began to pour into,the mission to- gether with a completely new set of names. One of the first of these is Habastos, a rancheria which contributed 137 converts in 1800 and 1801 and which is now stated, for the first time in the mission book, to lie "acia el estero de los rios." Later, the variants Quivastos and Aguastos are used. Conversions from this tribe con- tinued untll 1810, after which the name disappears from the lists. The sharp segregation of dates of conversion are clear evidence that, whatever the racial or linguistic affiliation, there were two groups of Indians, one con- verted before 1801 and living along the shore of the bay generally south and west of the Carquinez Strait, the other converted between 1801 and 1810 and living at the east end of the strait and along Suisun Bay. There prob- ably was no clear separation of the two in the minds of the Spaniards; hence the confusion of names. We are concerned here with the second group, the one uniformly designated Aguastos, which inhabited the approaches to the delta. With respect to the aboriginal population of this group we have no direct evidence whatever. On the other hand the record of the San Francisco Mission shows 396 bap- tisms. This immediately sets a lower limit to the num- ber of Aguastos for there certainly can have been no fewer members of the tribe than were baptized. Regard- ing the upper limit it can be pointed out only that the group was completely obliterated at the time of conver- sion and its name never appears again in either contem- porary or modern records. Hence it is safe to assume that substantially all the Aguastos were taken to San Francisco and that the baptisms include the entire tribe. We may thus ascribe to them a population of approxi- mately 400 persons. We now encounter the Chupunes (or Chupcanes), con- cerning whom Schenck (1926, p. 143) has this to say: The Chupunes (Chupcanes), apparently a group, were located along the southern shore near the east end of Carquinez strait. West of the strait, also on the southern shore-in the Pinole region of San Pablo bay-were the Huchones. The earliest documentary reference is to the diary of Abella, in 1811. On October 16 he went through Car- quinez Strait by boat. Then he says that the strait ". . . remata en la tierra de los Chupunes, porque hay ya en- s'ancha. . . ." The "ensancha" or widening begins at Port Costa and continues to Martinez. This, then, is the boundary of the Chupunes. On October 28, discussing the Suisunes on the north side of the bay, he says that "La rancheria citada de los Suisunes cahe al nordeste de los Chupanes, tierra adentro del Cerro de los Kar- quines . . . .' The Cerro de los Karquines is, of course, Mt. Diablo. In his account of the expedition of 1817 Duran tells how he arrived at noon of May 14, by boat from San Francisco, at the "remate" of the "estrecho de los Chu- canes' at a point 14 leagues northeast of San Francisco 6 1 ANTHROPOLOGICAL RECORDS and 17 leagues north-northeast of San Jose. The ranch- eria of this name, he states, is now Christian, at San Francisco and San Jose. The mission books show a total of 105 baptisms at the two establishments. It is reasonably plain that the Aguastos and the Chu- punes occupied more or less the same territory-along the south shore of the eastern end of Carquinez Strait and the western end of Suisun Bay. The diaries and the baptism records both indicate that the original inhabi- tants were the Aguastos, who were missionized and removed. Their place seems to have been taken by another group of natives known as the Chupunes, who also were gathered into the fold at some period between the visits of Abella and Duran. Subsequent to the 1817 diary of Duran there is no further mention of this tribe. With respect to population we have only the record show- ing 105 baptisms. Since the conversion seems to have been quite complete, we may set the aboriginal value at no more than 150. Let us now consider the Ompines. This group is placed by Schenck on the north bank of the Sacramento River at and above the junction of the river and Suisun Bay. Schenck also (p. 137) discusses the possibility that the Ompines and Julpunes composed a single group. In spite of an assumed similarity in names the Spanish accounts are unequivocally explicit to the effect that there were two groups, not one, hence Schenck's hypo- thesis may be disregarded. With respect to location the later Spanish accounts bear out Schenck's contention that the tribe was situated north of the river. In his entry for May 14, 1817, Duran says that his expedition stopped at the mouth of the San Joaquin River, whereas another boat (that of Argiello) stopped opposite "en tierra de Ompines." The next day they all went up the Sacramento River to the "remate de las lomas de los Ompines.' Meanwhile Arguiello, in his entry for May 15, says that they went along the north shore and stopped "donde termina la tierra de los Ompines." This puts the eastern edge of the Ompines at the east side of the Monte- zuma Hills in T3N, R2E, approximately as shown by Schenck. Altimira describes an unauthorized raid by Fr. Duran on the tribes north of Suisun Bay, among them f. . . otra rancheria aislada llamada los Ompines" (Altimira, MS, 1823). A few of the earlier documents, on the other hand, contain statements which raise the possibility that the Ompines were not always confined exclusively to the north shore. In his diary of 1811 Abella describes how, on October 17, his party entered a big bay (Suisun Bay) and, after 5 leagues, following along the south shore, began to find estuaries and numerous islands covered with tules. They continued into the west channel of the San Joaquin and stopped at an island on which large trees were growing. At this point, somewhere near Antioch, there was a "pescadero" of the Ompines. It is evident, therefore, that in 1811 the Ompines had at least temporary fishing spots on the south side of the estuary, in an area usually ascribed to the Julpunes or Tarquines . The San Jose baptism book shows the conversion of 108 Ompines. Those from San Rafael and Solano do not mention the tribe. The fact that a tribe situated north of Suisun Bay does not appear in the records of either of these missions is noteworthy, since during the 1820's and 1830's the north-bay groups were brought to them in large numbers, and since we know from Altimira's comment on Duran's raid that the Ompines were still in existence in 1823. Furthermore, the Ompines must have constituted more than a single small village, for Arguiello and Duran both refer to the "tierra" of the Ompines. The hypothesis is possible, although admitte ly there is no real proof, that the Ompines may have originally occupied the slo-ughs and islands at and abov Antioch, that they may have been pushed north at an early date by Spanish intrusion from the south and wes and that they may have been further dispersed, or exte minated without extensive conversion, prior to 1830. I such a theory in any way represents the course of their decline and disappearance, then it also follows that the aboriginal population was considerably greater than the baptism number would lead one to suppose. To turn now to the Julpunes, there seems to be littl difference of opinion regarding their original location. This was as Schenck pictures it: the south shore of the San Joaquin estuary from Antioch to the line between R3E and R4E. The "Informe" of Hermengildo Sal, writ ten in 1796 and previously referred to, specifies the "Tulpunes" as a "nacion" living on the "orilla del este Fourteen years later in 1810 Viader went 7 leagues fro Pittsburg to the "old river" west of Stockton. He was: . . . esta tierra es de los Tulpunes." Duran, May 24, 1817, on his return journey downstream reached the region of the Julpunes at 8:00 A.M. and joined the othe boat at 6:00 P.M. of the same day at Carquinez Strait. Schenck (1926, p. 137) points out that Kotzebue, who was in the area in 1823, implies that the Julpunes were living on the north bank. Merriam (1907, p. 348), says that the Hulpoomne "occupied the east bank of the Sacr mento River from a few miles south of the mouth of American river southward . . . ." Schenck's explanatia of the discrepancy appears to the present writer entire sound: the Julpunes retired across the estuary to the north bank and then upstream nearly to Sacramento. In so doing they may very well have carried the surviving Ompines with them. The San Jose record lists 148 bap- tisms of Julpunes but the name is absent from the reco of San Francisco, Santa Clara, San Rafael, and Solano missions. Along with the Ompines the Julpunes must h' escaped the active proselyting effort of San Rafael, and particularly Solano, between 1824 and 1834, by a rapid retirement so far up the river as to elude the parties sent out from the missions. The converts at San Jose must have been captured by the Viader, Duran, Arg5eU and similar expeditions before the migration upstream, The Tarquines are claimed by Schenck to have been a single group. It seems to have stretched from east to west entirely across the marsh area between th main channels of the Sacramento and San Joaquin river and then to have extended along the southern shore of Suisun Bay" (pp. 134-136). Schenck's belief in this re- markable distribution is based upon three documentary- references (at least he cites no more than these threefi his tabulation on p. 135). The first of the three documents, chronologically, is the first expedition of Viader, in 1810. In his entry for August 17 Viader says that, having spent the preceding night near the present location of Pittsburg, he recon- noitred these lands which " . . . son de los Tarquines, que lo mas, 6 casi todos son Cristianos de San Fran- cisco." After noting the mouths of the two rivers, he goes on to mention a spot on the estuary ". . . en donde dicen estaba la rancheria de los Tarquines" (emphasis mine). Let it be emphasized that in 1810 the Tarquines are almost all Christians in San Francisco, and Viader saw there the rancheria which was, or had been, thatof the Tarquines. The San Francisco baptism book shows 18 "Talquines" converted in 1801 and 63 more in 1802, making a total of 83. This number could well be the I I I 62 COOK: ABORIGINAL POPULATION OF SAN JOAQUrN VALLEY majority, or almost all, the inhabitants of a moderate- sized rancheria. Schenck is therefore technically cor- rect in placing the tribe on the south shore of the east- mrn end of Suisun Bay. The second document is the diary of Abella in 1811. OnOctober 25, in the course of the return trip down- stream, some distance below the junction of the chan- ` els of the San Joaquin, he found a rancheria of the tauquimenes, one part of each side of the river, which #as 30 to 40 varas wide. This point was apparently at near the head of Sherman Island. The rancheria had houses. He saw 200 warriors. He then crossed rough the sloughs to the Sacramento River and on or pposite Sherman Island saw one rancheria of 14 houses ,ad several of 2 to 3 houses. He says that all they passed this day was . . . parte de una isla" (i.e., Mherman Island). Furthermore en todo este dia andubimos como unas 12 leguas [overestimate] y podra haver gente, como 200 almas, todavia puede que haiga mas, porque en la primera [rancheria] habraumas 1,000, segun lo grande que por aqui son las casas, tienen un circuito de 28 o 30 varas, con su orcon en medio This account deserves comment on several grounds: ith relation to Viader's visit of the previous year and iebaptisms at San Francisco it is evident that whereas [e southern extension of the Tarquines' habitat, what- er its size, had been swept clear prior to 1810, never- eless the tribe persisted on the estuarine islands in uly large numbers. Moreover, since there is evidence no more than one rancheria on the south shore, it pears that the territory in that region allotted by :henck to the tribe is too large and should be restricted la small area of the southeastern corner of Suisun Bay. With respect to population, Abella's figures are quite edible. It has been suggested that one of the huge uses found in this region could accommodate 9 persons Wlout difficulty. Then the large village should have had L0 inhabitants. Allowing 24 houses for the other villages been, 216 persons should be added, making a total of 756, figure not far from Abella's guess of 1,000. The final reference to the tribe occurs in the diary of ran. During the night of May 22-23, 1817, he went up main channel of the San Joaquin, in T3N, R4E, and sed the Tauguimenes on the left, that is to say, on e east bank. Schenck thinks that the group covered the dtre strip from Pittsburg to the east bank of the main ver contemporaneously. Now it has been pointed out probable that the southwestern outliers were mission- ed, or pushed back into the swamps, as early as 1801. is equally possible that the island communities de- ribed by Abella in 1811 were pushed, in the next five six years, off the islands altogether and clear back tward to the far bank of the main river. Of consider- ic significance is the fact that whereas both Viader and ella mention the Tarquines as being in the estuary gion, Duran, who covered this area thoroughly, is mpletely silent with regard to their presence. It is ly unlikely that, had there been any of the tribe left their former habitat, he would have failed to note em. The details are very obscure but the main outlines of ents in the first three decades of the nineteenth cen- can be perceived. Aboriginally and perhaps till rly 1800, there was a dense population of natives nding from Port Costa along the southern shore of sun Bay and up the rivers for fifteen miles beyond Antioch. Among them were included tribal groups, or rancherias, called Aguastos, Chupunes, Ompiries, Jul- punes, and Tarquines, belonging very likely to different ethnic and linguistic stocks. Under the pressure of the Spanish military power, which was the real force behind missionization, portions of these groups were extermin- ated, other segments gave ground and shifted habitat, and occasional remnants persisted in the old localities. Thus each visitor in turn found a different geographical organization, until the entire native society was oblit- erated. An accurate assessment of aboriginal population in this area is impossible. The best we can do is try to make an intelligent guess. Several methods are avail- able for this purpose-group comparisons, mission fig- ures, area comparisons. Throughout the plains of the lower San Joaquin and Sacramento valleys the native social units appear to have resembled rather uniformly the political organiza- tion of the Yokuts in the central and southern San Joa- quin Valley. There were aggregates, or communities, consisting of perhaps one, but usually more than one, village, and occupying a more or less clearly defined territory. These groups, as they may be called, can be identified by the plural names which are ordinarily attached to them-the Bolbones, the Leuchas, and so forth. Naturally these groups varied considerably in size, and concerning no single one of them can we be absolutely sure of the number of their people. Never- theless, if we had data concerning enough of them, tne variations due both to inherent difference and to inac- curate estimate would tend to cancel out and an approxi- mate average could be secured. No pretence can be made that we have enough estimates to establish a mean which would be statistically satisfactory. Nevertheless, as so frequ ently happens when we are dealing with data of this character, we have to employ the information available to us or forsake the problem entirely. We have hitherto considered a number of the local groups mentioned above and have estimated their popu- lation as follows: Bolbones (restricted group, see p. 58), 1,500; Jusmites, 300; Tugites, 500; Nototemnes, 200; Leuchas, 900; Ochejamnes, 750; Guaypem, 300; Quen- emsias, 400; Chucumes and Chuppumne, 1,500. The average for the nine groups is 705 or, in round numbers, 700. If we consider that the Aguastos, Chupunes, Om- pines, Julpunes, and Tarquines were groups of the same character as the foregoing, then their total population may be taken as 3,500. The total baptisms shown in the mission books of the five northern missions (in fact, only San Francisco and San Jose) for these groups is 911. In previous instances we have estimated the aboriginal population by doubling the baptism number. This procedure is admittedly purely arbitrary and based upon the general consideration that, except for small local populations relatively close to the mission, it was impossible for the missionaries and soldiers to prevent the escape of a sizable fraction of the people. Of the five groups here discussed, the Aguastos, it is evident, were completely missionized or at least obliterated. A much greater proportion of the other tribes survived, as is attested by their prob- able migrations up the rivers. Hence for the entire population it is doubtful if even one-half received bap- tism. Using the value of one half, the aboriginal number would have been approximately 2,000. Linear distances along streams are useful as a basis for comparison in country where the rivers are similar ecologically but are clearly separated spatially and where I p 63 ANTHROPOLOGICAL RECORDS the human population is concentrated along the stream banks to the exclusion of the interfluvial hinterland. Where a territory is marked by a network of creeks and sloughs, and the intermediate land is marsh, the linear comparisons become impossible. Areas must be substituted. In relation to the present problem three such areas may be delineated. The first comprises the territory of the Bolbones (including all the subordinate villages) and the Leuchas. Following Schenck's map, it em- braces all the land between the channels of the San Joa- quin plus a strip approximately two miles wide east of the main river in Tl and 2S, R6E which accounts for the Leuchas. The area, as projected from a large-scale map onto co5rdinate paper, is 775 square miles, the population 3,400, and the density 4.39 persons per square mile. The second comprises the home of the Ochejamnes, Guaypem, Quenemsias, and Chucumnes- Chuppumne. For the habitat of these groups we have followed Schenck as far as possible. Our line runs actually from the junction of the east and west channels of the Sacramento at the foot of Grand Island southeast to the main channel of the San Joaquin, thence northeast and north to just east of Walnut Grove and then, at a distance of about 2 miles east of the eastern channel of the Sacramento, to a point 4 miles north of Courtland. Here the line crosses the river and continues downstream, 2 miles west of the river, to the starting point. This strip of the western bank of the western branch of the Sac- ramento is included in order to take in the Chucumes, who may have lived on the west side of the river. The area of this territory is 330 square miles, the popula- tion 2,950, and the density 8.94 persons per square mile. The third area is the one shown by Schenck as belong- ing to the Chupunes, Tarquines, Julpunes, and Ompines, with the exception of the region east of the San Joaquin attributed to the Tarquines. For reasons stated previously the author does not believe that the Tarquines occupied this spot aboriginally. A strip 2 miles wide is included on the north shore, however, between Rio Vista and Collinsville, in the probable land of the Ompines. The eastern boundary is formed by the borders of areas one and two. In area three there are 600 square miles. The mean of the densities of the other two areas is 6.67 per- sons per square mile. Hence the population would have been 4,002 persons. No significance should be attributed to the third and probably also the second digit in these numbers. They are used only for purposes of estimate. The three methods employed have yielded respec- tively 3,000, 2,000, and 4,000 as the most likely popu- lation of the five groups here being discussed. In default of any other evidence we may take the average 3,000. (Chupunes, Tarquines, Ompines, Julpunes ... 3,000) Adding the totals for the tribes known to inhabit the delta region of the great rivers and the southern shore of Suisun Bay, we arrive at a total population of 9,350. Delta area 9,350 It is now preferable to depart from a strictly tribal sequence and revert once more to a classification based upon river basins. Three areas of this type are suffi- ciently clearly marked out; those corresponding to (1) the Cosumnes River, (2) the Mokelumne River, and (3) the lower San Joaquin River from just below the Merced to the head of tide water near Manteca. The inhabitants may be designated village or tribal groups in accord- ance with the river system where they were located. The Cosumnes group.-On the river of this name the large and important aggregate of peoples known larly as the Cosumnes, which included a restrictedt let or subgroup also called Cosumnes. Ethnically apI tion of the Plains Miwok, they extended from Sloughl close to the foothills, along the lower course of the , sumnes River to its confluence with the Mokelumne Thornton, and from that point northwestward to the 4 ramento. The tribe as a whole was divided into eithe villages or tribelets, the names of many of which ha come down to us from the Spanish records or have be ascertained by informants from ethnographers. As I be expected, there is considerable confusion among different sets of names. The mission documents are replete with village a tribal names but the number of baptisms was not as large as might be anticipated from what must have be a very populous aggregate of natives. The reason pr ably lies in the fact that missionizing expeditions to Cosumnes were preceded by exploratory and punitive expeditions which, to be sure, brought home a few c verts but which were chiefly preoccupied with milita objectives. The Cosumnes, together with the Mokelu and other peoples of the lower San Joaquin Valley, h the time and the opportunity to develop great facility the raiding and stealing of livestock and consequently for many years were in a state of uninterrupted war the coastal settlers. The bitter hostility thus genera together with the aggressive psychology which accom- panied successful physical opposition to the Spaniards made extensive conversion to Christianity very diffic As a result the relative proportion of the natives baF was unquestionably much lower than among the bay a delta tribes previously considered. The baptisms whi appear in the mission records follow. Tribe or Group Date of Conversion Cosumnes (Tribelet) ... ....... 1826- 1836 Juni sumne (Anizumne, Unsumne) ........... 1813-1834 Lelamne (Llamne) ... ........ 1813- 1836 Gualacomne ... ........ 1825- 1836 Amuchamne (Mackemne) ........ 1834- 1835 Sololumne .... ......... 1828- 1834 Locolumne .... ........ 1826- 1834 Baptism 84 363 128 158 13 6 52 Total .804 If we apply the general principle used with the de]tA groups and double the baptism number, the population; becomes 1,608, a figure which is much too low. The Lelamne, with 128 baptisms, comprises the group attacked by Soto in 1813, at which time we have esti- mated that there were four villages of 475 persons e involved in the battle. This calculation implies a tota of 1,900 for the Lelamne alone. On the other hand, the account is not entirely clear as to whether or not there were members of the Cosumnes tribelet concerned. If so, we may be dealing with both the Lelamne and adja- cent neighbors who were designated locally Cosumnes. If we include the baptisms of all those under both name we have 212. Furthermore, the Junisumne (or Unsurn or Anizumne) were often confused with the Cosumnes. If the 363 baptisms listed under the Junisumne are a I i i i t I I 64 COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY get 575 and, multiplying by 2, the population of the e divisions collectively would have been 1,150. This mate also appears too small and leads to the con- sion suggested above on historical grounds that a sm factor valid for the delta would not be appli- e to the Cosumnes group as a whole. Another documentary source is of interest in this -ection. This is the account by Jos6 Berreyesa in 30 (MS) of an affray along the lower Sacramento er in which Americans participated under Ewing g. Christian fugitives from the missions had been tected by the Yunisumenes (Junisumne), who had ed with the Ochejamnes. They were opposed by the icans and their allies, the Sigousamenes (Siakumne), Cosomes, and the Ilamenes. These last tribes had ered an army of 450 "Gentiles auciliares." The isumenes, Cosomes, and Ilamenes are, of course, isely the three subtribes discussed in the preced- paragraph. Now if the Sigousamenes, Cosomes, Ilamenes contributed 450 men collectively, they h may be considered to have furnished 150 men. ce the opponents were fairly well matched, it is ely that the Yunisumenes supplied a similar number. can assume that for routine fighting of this sort, icularly where two of the tribelets were ranged h the Mexicans instead of against them, the armies luded no more than the strictly military population, not in excess of half the males over the age of ten xs. Hence, if the sex ratio was unity and the young Idren constituted approximately 15 per cent of the ulation, the aggregate number of the three subtribes uld have amounted to 1, 920, or almost the same as s estimated from the Soto report in 1813 for the amne (Ilamenes) above, or perhaps the Lelamne mented by some of the Cosumnes tribelets or sub- se. The Berreyesa episode occurred in 1830, after these groups had suffered twenty years of attrition ng to perpetual minor warfare, disease, and starva- o Hence the population of the three tribelets jointly, isumne, Cosumnes, and Lelamne, must have reached kUy 3,000 in 1813. The baptism factor, consequently, 1ould not have been 50 per cent, but 575 divided by 3,000, or 19.2 per cent. Three other villages or tribelets which can be identi- Ped in the mission records as being closely associated vith the Cosumnes are the Amuchamne, Sololumne, and ocolumne. The first two probably correspond to Mer- am's Oo-moo-chah and So-lo-lo, which in later times least were rancherias. Assuming all three to have en villages, we may consider that each contained an werage number of 300 inhabitants. The respective bap- 8rmnumbers were 13.6, and 52. In relative terms the ptisms amounted to 4.3, 2.0, and 17.3 per cent. The last division listed above is the Gualacomne, wnonymous with Merriam's Wah-lah-kum-ne. Merriam llewko List, MS) places them between the lower Stanis- lus and the Tuolumne rivers, but quotes Hale, who saw hem in the 1840's, as saying that they lived on the lower *st side of the Sacramento River. Hale' s statement is trongly supported by the fact that they appear in J. A. aatten's census of 1846 (MS, 1872). Gatten ennumerated Oly the tribes along the lower Sacramento. Whether beGualacomne can be affiliated with the Cosumnes ithnically is doubtful but it is reasonable to include them pith this group demographically. Of the Gualacomne 158 were baptized in the missions. Piat the group was fairly large is attested by the fact batGatten reported, under the name Yalesumne, that 85 were alive in 1846. Since no open valley group could possibly have retained more than one-third of its former members in 1846, it does not seem excessive to ascribe 1,455 persons to the tribelet. The baptism factor is 10.8 per cent, and the average of the five values secured with the Cosumnes group is 10.7, or, let us say 10.0 per cent. The total population on the lower Cosumnes and adjacent Sacramento rivers, according to the discussion above would be 5,355 souls. We may approach the problem from a different direc- tion if we start with the villages compiled by Merriam (1907, p. 349). He mentions sixteen villages on the Cosumnes River system from Sloughhouse nearly but not quite to the Sacramento. It is extremely probable that there were other villages on the Sacramento River itself. Nevertheless, let us take Merriam's list as it stands. The upper seven villages lie between Sloughhouse and the junction of the Cosumnes River with Deer Creek, the remainder below that point. Of the lower nine we may consider that four correspond to those seen by Soto, which were quite large. It was estimated that they contained 475 persons apiece. The other five lower villages, although perhaps not so populous, must have held fully 300 inhabi- tants each. The upper seven were no doubt smaller but still should have reached the values given by Moraga for similar stretches of the Tuolumne and Merced, i.e., approximately 250 persons. The total would then come to 5,150, very close to the previous estimate. It will be both adequate and conservative to establish the popula- tion at 5,200. Cosumnes group ........ 5,200 The MoquelLumne group.-Here are included the Indians living on the lower course of the Mokelumne River, the Calaveras River, and the plain between the two. Five tribes mentioned by the Spanish writers fall within this category: the Moquelumnes, the Siakumne, the Passa- simas, the Yatchikumne and the Seguamne. The exact territorial status of these tribes has been a subject of considerable disagreement among ethnographers. The original Moquelumnes of the Spaniards were un- doubtedly located on the Mokelumne River itself from Campo Seco nearly to the junction with the Cosumnes at which point they adjoined the Cosumnes tribe. Accord- ing to George H. Tinkham, in his History of San Joaquin County (1923), they extended in a north-south direction all the way from Dry Creek to the Calaveras River, but by the middle of the nineteenth century they may have spread out from their original habitat. The Yatchikumne are shown by Schenck as filling the space between the lower Mokelumne and the lower Calaveras and extending westward to the San Joaquin River. Merriam (Mewko List, MS) quotes F. T. Gilbert to the effect that they occupied the Mokelumne River basin, but if they did so, it was because of the displacements during the mining era. The Passasimas are placed by Schenck on the left bank of the Calaveras River at, and for several miles upstream from, its junction with the San Joaquin River. The Siakumne and the Seguamne are subject to some confusion. This difficulty arises partially from the similarity in name. The Siakumne are called Si- a-kum-ne by Merriam and Sakayakumne by Kroeber. In Gatten's census of 1846 they appear as Sagayakumne. In the San Jose baptism book we find Ssicomne, Zi- comne, Siusumne, and Sigisumne. The Seguamne, on the other hand are designated Seguamnes and Saywa- mines by Merriam and Sywameney or Seywameney by Sutter in his New Helvetia Diary (1939). Gatten calls them Sywamney. They appear in the San Jose I F t 65 ANTHROPOLOGICAL RECORDS record as Secuamne, Seguamne, Seyuame, and other variants. The Siakumne lived somewhere between the Cala- veras and Stanislaus rivers according to Merriam, who places one of their villages at Knights Ferry on the Stanislaus. Schenck doubts Merriam's location and Kroeber puts the rancheria Sakayakumne as far north as the Mokelumne. Sutter (1939, p. 88) says that some of these people came to work for him, an unlikely event if they had been living as far away as the Stanislaus. It is probable that the lower Calaveras River is as close as we can place them. The Seguamne are not mentioned at all by Schenck. Merriam (Mewko List, MS) says they were a "tribe or subtribe on E. side lower Sacramento River" and may have been a subtribe of the Bolbones. Sutter and Gatten both refer to the tribe, and the sphere of activity of these men did not extend much below the Sacramento River itself. Hence, although there are grounds for including the Seguamne with the Bolbones or the Cosumnes, no serious error will be committed by placing them in the Mokelumne group. The Moquelumnes were unquestionably quite numer- ous. In Spanish and Mexican times they were the most aggressive and belligerent of all the valley tribes and gave the coastal settlers a very rough struggle. Never- theless, in spite of their detestation of the missionaries they furnished 143 converts between 1817 and 1835. At a ratio of 10 per cent this would mean a population, prior to the mission period, of about 1,400 souls. J. M. Amador (MS, 1877, p. 43) says that once, during the later colonial period, they furnished 200 auxiliaries, a fact which would argue fully 1,000 people at the time. Gatten in his census of 1R46 gives them a total of 81 persons but G. H. Tinkham says that in 1850 or there- abouts they possessed four sizable villages with four chieftains. This may have meant between 200 and 400 persons, a really considerable number of survivors for a tribe which had suffered so extensively in the pre- ceding three decades. These indications, and it must be admitted that they are only indications, would lead one to infer that the aboriginal population reached at least 1,500. Precisely because the Moquelumnes were so brutally handled in the colonial era the modern ethnographic accounts of villages are very incomplete. Neither Mer- riam nor Schenck gives us any list. Kroeber puts three on his map (1925, opp. p. 446): Mokel (-umni), Lel- amni, and Sakayak-umni. I think we are now in a posi- tion to state that these names represent former tribes and if they were applied to villages by informants, it is because the component units had shrunk to very small size. Stream density comparisons are of value for the Mokelumne group. On the Cosumnes River, from Slough- house to Thornton, Merriam shows thirteen rancherias (omitting those close to the Sacramento River). As was proposed above we may ascribe from 200 to 400 inhabi- tants to each of these, say on the average 300. Now there is no reason to suppose that the Mokelumne River from the San Joaquin-Calaveras county line to just west of Lodi was less heavily populated than the Cosumnes. If so, the number of villages per linear river mile must have been very nearly the same. For the stretches under consideration there were 24 miles on the Cosumnes and 22 on the Mokelumne. Thus we would get 12 villages and 3,600 persons living on the Moquelumne River. The Yatchikumne and, if we are to credit Schenck, the Passasimas occupied a position on the Calaveras River comparable to that occupied by the Moquelumnes on the Mokelumne. Schenck regards the Yatchikumne a tribe equal in importance to the Moquelumnes, and county historians speak of them as a large group. Th river frontage is equivalent to that of the Moquelumn For these reasons we would be justified in ascribing the Yatchikumne and Passasimas the same population as the Moquelumnes, i.e., 3,600. The evaluation of other two groups from the geographical standpoint is difficult, owing to the uncertainty of their location. Siakumne may be regarded as living somewhere on th lower Calaveras and, if so, must be included with th Yatchikumne and Passasimas in the estimate for the Calaveras. The Seguamne may or may not have inha the banks of the Mokelumne and Calaveras rivers. In view of our ignorance on this point it may be well to omit them from consideration in this connection and leave the estimate with the existing total of 7,200. We may attempt some direct tribal comparisons. considering the northern San Joaquin Valley and delta tribes and tribelets have been examined, namely: Ag stos, Bolbones (4 tribes), Leuchas, Ochejamnes, Gu pen, Quenemsias, Chuppumne, Chupunes, Tarquines, Julpunes, Ompines, and the Cosumnes group (7 tribe For all these the average population calculated has b very close to 700. If this figure is applied directly to the Moquelumne group, its population becomes 3,500. However, some adjustment is necessary. The Moque lumnes by all accounts, Spanish and American, were an unusually large tribe, probably reaching at least 1,500. The Yatchikumne may not have been as numero but were apparently above the average size, let us sa4 1,200. The Passasimas, despite the fact that Schenck thinks they were a "group plus" may be regarded as smaller, perhaps no more than average. For the Sia. kumne and Seguamne we must also assume the avera~ figure, 700. With these adjustments the total reaches 4,800. The baptism books give us a record of the followin conversions. Tribe San Jose Santa Clara Moquelumnes ........ 143 Yatchikumnes ........ 118 Passasimas ......... 145 Siakumne ............ 22 Seguamne ........... 47 116 The Passasimas, Siakumne, and Seguamne were ated in the vicinity of the San Joaquin River and hence were more exposed to the Spanish expeditions than thel tribes along the lateral streams. Hence the proportio of those taken for conversion may have been higher t the 10 per cent of the aboriginal population found for Cosumnes, although it would not have attained the va of 50 per cent characteristic of the more westerly del tribes. We may take an intermediate figure, 20 per e This would give the Passasimas a population of 725, Siakumne 110, and the Seguamne 815. The great disp between the figures for the last two tribes may wellb due to confusion of names in the mission records. Th total for the three is 1,650. For the Yatchikumne on Calaveras River no more than 10 per cent baptisms c be assumed, yielding a population figure of 1,180. If only geographical location were considered, the same factor could be used for the Moquelumnes but this tril resisted missionization with extraordinary tenacity. Hence we are not justified in using a factor of more Us 7 per cent, from which we may infer that the populatia I .I I i II -1 i j I -1 I i I 6 6 COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY Us 2,040. The baptism data would then give us a total the group of 4,870. According to the estimates furnished by pioneers 4 government officials for the period just preceding FGo1d Rush the population ran into the thousands. he census by Savage (Dixon, MS, 1875) puts 4,000 on FCosumnes, Mokelumne, and Calaveras and 2,500 on Stanislaus, F. T. Gilbert (1879, p. 13) says that "before advent of Sutter" there were 2,000 on the Mokelumne as far as I can ascertain, he implies that on the umnes and Mokelumne together there were fully tO0 These figures were undoubtedly greatly exagger- I but nevertheless indicate a very large population in area just before the discovery of gold and subsequent tedestructive epidemics of 1833-1835. Even if we these estimates in half, there would remain in mid- :ury approximately 2,000 persons in the basins of OMoquelumne, Calaveras, and adjacent San Joaquin ers. A residue of 2,000 in 1850 means certainly an Anal population of three times as much, i.e., 6,000. To recapitulate the estimates for the Moquelumne Iup, we find: k By stream densities ........................ 7,200 :VBy adjusted tribal averages . ................ 4,800 By baptism data ............................ 4,870 tBy extrapolation from American estimates .... 6,000 Mean ............ ................... 5,720 The mean, 5,720, appears entirely reasonable for the ab- nal population of such a vigorous and important group. Moquelumne group ......... 5,720 The lower San Joaquin River group. -Here are included f convenience the tribes and fragments of tribes inhabit- the banks of the San Joaquin River from the habitat of | Leuchas, in the vicinity of Manteca, to just below the guth of the Merced, together with those living along the Wer courses of the Stanislaus and Tuolumne rivers (see sps 1, 5, and 6, area 8). The San Joaquin villages or bes appear to have been Cuyens, Mayemes, Tationes, iApaglamnes. The first two are regarded by Schenck villages only and the latter two as "villages plus." The ~iy Spaniard who described the area was Viader, in the ecounts of his two expeditions of 1810. On his first expedition, having left the village of Tom- 6m, he went south-southeast up the river for 2-1/2 leagues -another village ". . . cuya capitan se llama Cuyens." 118 was very close to section 1 0, in T3S, R6E. After a mrney of another 2-1/2 to 3 leagues he found another vil- te, whose captain was Maijem (sec. 8, in T4S, R7E). tens after 2 leagues, still another village, whose captain s Bozenats (in sec. 34, in T4S, R7E), was seen. Three 4gues farther in the same direction brought him to the cheria ". . . cuyo appelido es Tationes' In the mean- he had seen 30 gentiles from the Apaglamnes. The iones were located close to section 27, in T5S, R8E. During his second expedition, on October 22, Viader nt from Pescadero southeast up the river for 5 leagues 'los indios Tugites." Three leagues farther on he was tby Indians from Cuyens, who went with him to the ancheria de Mayem," another 4-1/2 leagues farther on. en, having forded the river to the east shore, they Wut still another 2 leagues to a rancheria "que se llama Taualames." The Rio Dolores (Tuolumne) was sup- ised to be 2 to 3 leagues north. However, Viader went stream on the east bank 6 leagues to the Rio Merced, ving in the meantime passed "en frente de . . . los 8s Apelamenes y Tatives." The distances on both trips are very consistent and the village locations check closely with those shown on Schenck's map, except that only the Taualames should be placed on the east bank of the river. Viader is very explicit in saying that all the others were on the west bank. Cuyens, Mayem, and Bozenats are beyond doubt villages, since each was named after its chief, or cap- tain. The Tationes and Apaglamnes are given in the plural: "los indios Apelamenes y Tatives." They may well have possessed more than one rancheria each, as is supposed by Schenck. Schenck thinks that Cuyens and Mayem were transient parties from Kroeber's Miwok villages, Chuyumkatat and Mayemam, which were on the Cosumnes. Aside from the possible similarity in names there is not the slighest evidence in Viader's diaries to support such a theory. Viader definitely speci- fies rancherias, and the missionaries of that period were able to distinguish rancherias from fishing parties. From the record we have in this area five villages certain and at least one other probable. For six villages of average size (there is no indication that they were smaller) the population would be assumed as 300 persons each, or 1,800 in all. The mission records show for baptisms: Tribe or Village Dates of Conversion Cuyens ............... 1811-1813 Mayemes . 1813- 1823 Apaglamnes .......... 1818- 1824 Tationes ............. 1805-1811 Number of Baptisms 88 91 48 243 The total is 470. These were San Joaquin River natives, not from the delta and marsh region. On the other hand they were less remote from Spanish influence and attack than the tribes which extended up the lateral streams. Hence the proportion of baptisms was probably intermedi- ate between the value of 50 per cent assumed for the very exposed bay and delta people and that of 10 per cent ascribed to the Cosumnes. An estimate of 25 per cent would be reasonable, yielding a population value of 1,800. The two methods of calculation coincide, and the result, 1,800 inhabitants, may be allowed for the area. For the lower Stanislaus and Tuolumne rivers the only tribes mentioned in the Spanish documents are the Tauhalames (or Taulamnes) on the Tuolumne and the Lakisamne (or Lakisumne or Laquisemne) on the Stanis- laus. Kroeber (1925, p. 485) writes: "the Tawalimni, presumably on Tuolumne River . . . the Lakisamni on the Stanislaus. . . ." Schenck says (p. 141): The villages of Taulamne and Taualames are both definitely placed, the former on an inaccessible rock on the Stanislaus river in the foothills, the latter at the ford of the San Joaquin just below the mouth of the Tuolumne river . . . . This seems to establish the region between the lower Tuolumne and Stanislaus rivers as Taulamne territory. Merriam agrees in assigning the same region to the Tuolumne. Schenck's only reference to the Lakisamne is on the same page: "The Leuchas might possibly be identified with Kroeber's Lakisamni (Yokuts) on the Stanislaus river." But the mission records and all other documents clearly distinguish between the two groups, rendering Schenck' s hypothesis entirely untenable. Some of the confusion may derive from the account of I ';. 6 7 ANTHROPOLOGICAL RECORDS Munioz. In his diary of the Moraga expedition he tells how, on October 1, 1806, the party left the Merced River and proceded northwest for 7 to 8 leagues, reach- ing finally a river which they called the Dolores (i.e., the Tuolumne, probably near Modesto). There were no Indians, but signs of "varias rancherias," the inhabi- tants having all absconded. On October 2 they went northwest again and at 4 leagues, in the middle of a very large oak park, they came upon another river, which they called the Guadelupe. This could only have been the Stanislaus, probably somewhere east of Ripon. On the next day, October 3, they went up this river, and at the end of 6 leagues reached a rancheria called Taulamne. It was situated in "unos empinados voladeros e inacesibles por unas encrespadas rocas." They could not get at the Indians but estimated the population as 200, on the basis of the people they could discern. This village, be it noted, was situated among "steep cliffs, inaccessible because of certain rough rocks"-not on an inaccessible rock in the river. This spot, judging by both the distances and the description, was along the limestone bluffs which steeply border the south bank of the Stanislaus for several miles opposite Knights Ferry. The Indians said that there were six other rancherias upstream. From this point the expedition moved the next day again northwest toward the Calaveras River. We gather little concerning tribal names from Moraga's account but we learn that there was a considerable popu- lation along the Stanislaus which demonstrated sharp defiance to the Spanish invaders. In the later documents there is little if any reference to the Taulamnes but much discussion of the Lakisamni. There are repeated allusions to this group as being very hostile, bad raiders, and the object of several military campaigns, particularly those against the great Indian rebel chief, Estanislao. The fighting was undoubtedly on the Stanislaus River and the Indian protagonists were frequently allied with the Cosumnes and Mokelumnes. From the context of the documents they would seem to have been as numerous, or at least as bellicose, as either of these two tribes. Jos6 Sanchez in 1826 refers to his bitter battle with Estanislao, which took place on the "rio de los Laquisi- mes" (MS, 1826). Joaquin Pifia describes a military expedition under Guadelupe Vallejo in 1829 (MS, 1829). The objective was two "rancherias," one of the Laquisi- mes and the other of the Tagualames, on the "Rio de los Laquisimes," or the "Rio Pescadero." The campaign was inconclusive since nearly all of the Indians escaped. From the citations above it appears probable that the Taulamnes and the Lakisamne were two distinct tribal groups and that their home was on both the Tuolumne and Stanislaus rivers. It is also likely that in the tur- moil and confusion of the period between 1800 and 1830 the original spacing and distribution of the tribes be- came irreparably lost and that the surviving fragments of both amalgamated and reconstituted themselves with reference to their Spanish enemies rather than with refeience to their aboriginal social organization. Hence they may have come to be concentrated more on the Stanislaus than on the Tuolumne. The only direct population estimate we have for them is that of MuRoz, who claimed 200 persons for the village of Taulamne, among the cliffs. Assuming that 50 per- sons were not seen, the village would have had 250 in- habitants, which is more or less standard for the gen- eral area, according to Moraga's account. If the other six villages had an equal population, the total would have been 1,500. But this estimate does not include the portion of the Stanislaus below Taulamne which was covered by Moraga in his march of 6 leagues upstre No villages are mentioned in connection with this ma but they could scarcely have failed to exist. Hence w4 may add another 500 without much fear of exaggerat making a total of 2,000 for the course of the river fr the San Joaquin to several miles above Knights Ferry On the Tuolumne "varias rancherias" were seen, all deserted by their occupants. However, Moraga also marked that the lower Tuolumne resembled the lowe Merced. On the latter were 8 rancherias, hence the may have been an equal number on the Tuolumne. At conservative 225 persons in each, the aggregate wou have been 1,800. The sum for the two rivers wouldb 3,800. The baptism lists show 151 conversions for the La samne and 263 for the Taulamnes, or 414 in all. In' of the notorious hostility and the successful resistano these groups opposed to the white men, evident eveni Moraga's day, we are justified in setting the baptism factor as low as for the Mokelumnes, or 7 per cent. This gives a potential aboriginal population of 5,920. The midcentury American estimates would indical more than this number. H. W. Wessells (1859) claim 500 to 700 on the Stanislaus and Tuolumne in 1853. Adam Johnston (1853) put 1,350 on his map of the sar area in 1852. W. M. Ryer vaccinated 1,010 on the twu rivers in 1851. The Daily Alta California for May 31, 1851, said that the Indians were 1,000 strong between the Stanislaus and the Tuolumne, and Savage, for an earlier period, put them at 4,600 (Dixon, MS, 1875). On the other hand, it must be remembered that as a result of Spanish and Mexican, not to mention Ameri aggression most of the strictly San Joaquin River peo had long since retreated up the lateral streams. Hen the natives seen by the commissioners between 1850i 1853 included the residues of all the river tribes frog Manteca to Merced. For the southern part of the San' Joaquin Valley it was determined, in a previous disc4 sion, that the population remainder in 1850 represent approximately one-third of the aboriginal population. the estimates just cited the most reliable is that of H Following the suggestions presented in the consider of his activities, we must make a correction to accos for persons who missed vaccination. Such a correctic would bring the number to 1,420. Then application of the factor one-third gives an aboriginal value of 4,73C The three modes of estimate yield respectively a population of 3,800, 5,920, and 4,730, with an averag of 4,817. We may use a slightly greater value and ca the population 5,000. To this must be added the 1,800 persons estimated to have lived along the San Joaquin River itself. The lower San Joaquin River group as a whole, therefore, may be assigned a population of 6,0 Lower San Joaquin River Group NORTHERN SAN JOAQUIN VALLEY ....... . .... 6,80 27,071 THE MIWOK FOOTHILL AREA Above the central valley itself and occupying the foothills from the Cosumnes to the Tuolumne lived th4 northern and central Miwok. This region was not reached by the Spanish expeditions nor were many, if any, of the inhabitants incorporated in the missions It is therefore necessary to rely exclusively upon the I - I I I 6 8 COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY reports of the ethnographers. In a preceding discus- don of the central Miwok, who lived on the upper mtanislaus and Tuolumne, there were cited the data iecured by Gifford, Kroeber, and Merriam for 70 vil- ages. This area in 1850 was estimated to contain a iopulation of 1,470. There are no data comparable to ford's for the rivers farther north, largely because natives on the upper Cosumnes, Mokelumne, and lav eras were thoroughly dispersed during the Gold zh and village names and locations have become lost the memory of Indian and white man alike. It is pos- le, however, to get a reasonable estimate of the ulation indirectly. T The territory of the northern Miwok, from the eco- cical standpoint resembles closely that of the central Miwok. Hence stream mileage and area comparisons are justified. If we use the boundaries of the two groups substantially as given by Kroeber in the Handbook (map, opp. p. 446) and plot rivers and areas on a large-scale map, the equivalent aboriginal population for the north- ern Miwok by stream mileage and area is 2,480 and 1,535, respectively. The discrepancy in the two esti- mates is due to the greater frequency of streams and creeks in the northern area. The average of the popu- lation calculated by the two methods is 2,008, very close to that found for the central Miwok. The total for the foothill strip is then 4,138 or in round numbers 4,150. MIWOK FOOTHILL AREA..... 4,150 69 --- ------------I SUMMARY AND CONCLUSIONS From the data presented in detail in the last section we may now derive the aboriginal population of the San Joaquin Valley as a whole. Region Population Tulare Lake Basin . .............. 6,500 Kaweah River . . 7,600 Merced River . .3,500 Kings River . .9,100 Mariposa, Fresno, Chowchilla, upper San Joaquin . . 19,000 Southern San Joaquin Valley ..............-.6,900 Northern San Joaquin Valley Delta area ...... ....... 9,350 Lower Cosumnes ....... 5,200 Lower Mokelumne . 5,720 Lower San Joaquin, Calaveras, Tuolumne, and Stanislaus ....... 6,800 .. 27,070 Foothill strip (central and northern Miwok . ............. 4,150 Total .83,820 The total, 83,820, is more than four times as large as the population estimated to be surviving in 1850 (19,000) and much exceeds any previous estimate ad- vanced by modern students of the California Indians. Dr. C. Hart Merriam in 1905 computed the popula- tion of the entire state of California as 260,000, of whom perhaps one-fifth may have occupied the San Joaquin Valley, although Merriam does not attempt to assess the population of this area as such. Kroeber discusses the matterat length in theHandbook (pp. 488- 491, 880-891) and concludes that the population of the whole state was 133,000. Of these the Yokuts had I the Miwok (Plains and Sierra) 9,000, the Wester about 1,000, and the peripheral tribes in the sou haps 2,000, a total of 30,000. Schenck is more since for the delta region he allows for a spread tween 3,000 and 15,000 persons. The present es for the same area, as closely as it can be deter is in the vicinity of 13,000, or within Schenck's although toward his upper extreme. Since the data and reasoning upon which the p figure of 83,820 is based are set forth in detail preceding pages there is little value in repeating nor will anything be gained by attempting a reb the arguments presented by Kroeber. At the same the author may be permitted to recapitulate three wherein he thinks many modern scholars have bee led. 1. All available information from the Spanish Mexican sources must be consulted. To confine an ment or an estimate to a single account, such as Moraga, may lead to a false impression. Kroeber4 to have been thus deceived in his discussion of the lation of the Yokuts. w 2. It must be remembered that in the centralv as contrasted perhaps with an area like the Kla River, no informants speaking since 1900, and pa larly since 1920, can possibly have furnished a picture of conditions prior to the Spanish invasion decade following 1800. 3. The depletion of population in the San Joa Valley between 1800 and 1850 was far greater tha been appreciated, although the basic facts have a been recognized. Warfare, massacre, forced co sion, starvation, and exposure all took a treme toll of life but the sweeping epidemics of the 1 were even more devastating. Together these fq destroyed in the aggregate fully 75 per cent of t original population. [70] I I APPENDIX After this manuscript was completed, the writer had a opportunity to examine those documentary files of the ffice of Indian Affairs and of the War Department which we at present in the National Archives at Washington. bveral letters in the files containing information on the stive population of the San Joaquin Valley have never, b far as could be determined, been published. Since the ita thus procured are fragmentary and since they do not parently invalidate the conclusions set forth in previous ges, they have not been incorporated in the body of this ser. These items, however, have some intrinsic inter- and therefore merit specific mention. They are brief- abstracted as follows. War Department orordrGroup 98. 10th Military Dept. Letters received Calif., Document no. K 21. E. D. Keyes, Camp lMagruder, June 17, 1851. a 8 tribes on the Kaweah, with whom a treaty was concluded on May 30 contain 1,240 individuals. e 4 tribes on Paint Creek with whom a treaty was concluded on June 3 contain 1,660 persons. ord Group 98. Letters received Calif., 1854. Enclosure to document no. W 2. John Nugent, Camp Wessells, Dec. 31, 1853. e Four Creeks region (Kaweah) from the Sierra Nevada to Tulare Lake will not contain more than 1,000, all told. rd Group 98. Letters received Calif., 1854. Enclosure to document no. W 12. H. W. Wessells, ort Miller, March 7, 1854. Indians under control of Fort Miller include those an the Fresno, San Joaquin, Kings, and Kaweah vers. They are much reduced in numbers, owing tothe recent sickness. sno River: 400 persons, including 100 able men. Joaquin River: 350, including 80-90 able men. 8 River: 1,100, including 250 able men. eah River: 800, including 200 able men. Office of Indian Affairs rdGroup 75. Letters received Calif., 1854. Enclosure to document no. H 758. D. A. Enyart, Fresno Reservation, Nov. 3, 1854. Indians on the Fresno Farm include: 30 Chowchilla, 220 Choot-chances, 90 Pohonicha, and 100 Potohanchi. Indians in Mariposa, Stanislaus, and Tuolumne counties do not exceed a total of 2,000. river system he breaks them down thus: 300 on the Iterced, 350 on the Tuolumne, 250 at Plant's Ferry nthe Stanislaus, 100 elsewhere on the Stanislaus, sad 100 scattering through the country. rdGroup 75. Letters received Calif., 1855. closure to document no. H 1050. Report of D. A. yart, Fresno Reservation, Aug. 22, 1855. dthat there are at least about 1,000 to 1,500 Indians the River (i.e., San Joaquin) . . . . This does not [7 include the 'Mono' tribe which is the most numerous of any tribe . ..." Record Group 75. Letters received Calif., 1859. Enclosure to document no. M 66. M. B. Lewis, Fresno Agency, Aug. 30, 1859. A report on the 22 tribes which recognize the Fresno Agency as their headquarters. Abstracted as follows: Wel- leelch-um-nies: the most northerly tribe; is "temporarily" on the Tuolumne River because of displace- ment by the whites. ..... ................. 85 Poto- en- cies: have abandoned their native land, the Merced Valley and are now on the Chowchilla. ....... ...................... 110 Noot- choos: "a union of the remnant of other tribes," including some Yosemites. Now on the north fork of the Chowchilla . .............. 85 Po-ho-nee-chees: on the headwaters of the Fresno. .... ...... 105 Chow- chillas: have moved from the Chowchilla to the Fresno River . ........................... 85 Cooc- chances: the largest "unbroken" tribe in the agency, originally on Coarse Gold Creek; some still there, some at agency. .... .......... 240 How-ches: once large; always have been on the Fresno. ......... ........................ 18 Pit-cat-ches and Tal-linches: (two distinct tribes); native habitat was the San Joaquin River; still near Fort Miller. ..... 150 Coss-waz: "to some extent identified with the Pit- cat-ches"; native land is Deer Creek ..... 88 Monos: on Fine Gold Creek and the upper San Joaquin River ..... 535 War-to-kes, Itee-ches, and Cho-pes: all on Kings River; "constitute one nation" but have separate heads (on Wartoke Creek) ..... 290 Wat- ches: since 1854 have been on Kings River Farm. ..... 75 No-to-no-tos and We-melches .. ... 190 Tat-ches and Wo-wells: these four tribes are native to the lower Kings River and Tulare Lake. They were recently driven to their homes on the Fresno Farm ..... 165 Cow- willas: their home is the mouth of the Kaweah at the foothills ..... 110 Tel- em- nie s: on the Kaweah, near Visalia ..... 105 Total ................................ 2,436 71] BIBLIOGRAPHY PUBLISHED WORKS Barbour, G. W. 1852. 32nd Cong., 1st sess., Sen. Ex. Doc. 1, pt. III. 1853. Report to the Indian Commissioner. 33rd Cong., spec. sess., Sen. Ex. Doc. 4, pp. 249-264 [Ser. no. 688]. Barbour, G. W., R. McKee, and 0. M. Wozencraft 1853. Report to the Indian Commissioner. 33rd Cong., spec. sess., Sen. Ex. Doc. 4, pp. 56-59. Carson, James H. 1852. In San Joaquin Republican (Stockton, Feb., 1852), as quoted by S. P. Elias, Stories of Stanislaus (Modesto, 1924), p. 196. Chapman, Charles E. 1911. Expedition on the Sacramento and San Joa- quin Rivers in 1817, Diary of Fray Narciso Duran. Publ. Acad. Pacific Coast Hist., Vol. 2, No. 5. Cook, S. F. 1940. Population Trends among the California Mission Indians. Univ. Calif. Ibero- Americana 17. Berkeley. Coues, Elliott, ed. 1900. On the Trail of a Spanish Pioneer. (The diary of Francisco Garces.) Trans. and ed. by El- liott Coues. New York. The parts pertaining to the San Joaquin Valley are in 1:281-300. Derby, Lt. George H. 1852. A Report of the Tulare Valley. 32nd. Cong., 1st sess., Sen. Ex. Doc. 110, pp. 4-16. Farquhar, Francis P. 1932. The Topographical Reports of George H. Derby, California Hist. Soc. Quarterly, 11:99, 247, 365. Gayton, A. H. 1948. Yokuts and Western Mono Ethnography. Univ. Calif. Publ. Anthro. Rec., Vol. 10. Berkeley. Gifford, E. W. 1932. The Northfork Mono. Univ. Calif. Publ. Am. Arch. and Ethn., 31:15-65. Berkeley. Gifford, E. W., and W. Egbert Schenck 1926. Archaeology of the Southern San Joaquin Valley, California. Univ. Calif. Publ. Am. Arch. and Ethn., 23:1-122. Berkeley. Gilbert, F. T. 1879. History of San Joaquin County, California. Oakland, Calif. Henley, T. J. 1857. Report to the Commissioner of IndianAff accompanying Ann. Rept. Sec. of the Inte for 1856. No. 100, pp. 236-246. Johnston, Adam 1853. Report to the Indian Commissioner. 33rd Cong., spec. sess., Sen. Ex. Doc. 4, pp 241-247. 1860. In H. R. Schoolcraft, Archives of Aborig Knowledge, 4:406 ff. Kroeber, A. L. 1925. Handbook of the Indians of California. Bug Amer. Ethn. Bull. 78. Washington. D.C. Latta, F. F. 1949. Handbook of Yokuts Indians. Bakersfield, Calif. Mason, J. D. 1881. History of Amador County, California. O0 land, Calif. Merriam, C. Hart 1905. The Indian Population of California, Amer Anthropologist, n.s., 7:594-606. 1907. Distribution and Classification of the Me Stock of California, American Anthropola n.s., 9:338-357. Powers, Stephen 1877. Tribes of California, Contributions to Nor American Ethnology. Washington, D.C., Ryer, W. M. 1852. Vouchers for vaccination. 32nd Cong., 2M sess., Sen. Ex. Doc. 61, pp. 20-23 [Ser. no. 620]. Savage, James D. 1851. Letter in the True Standard, reprintedin Sacramento Union, Apr. 10, 1851. Schenck, W. Egbert 1926. Historic Aboriginal Groups of the Califor Delta Region. Univ. Calif. Publ. Am. Ar and Ethn., 23:123-146. Berkeley. Sutter, John A. 1850. Letter to H. W. Halleck, Dec. 20, 1847. Cong., 1st sess., H. R. Ex. Doc. 17. 1939. New Helvetia Diary; a Record of Events K by John A. Sutter and His Clerks at Newi vetia, California, from September 9, 1841 to May 25, 1848. San Francisco, Calif. [72] I COOK: ABORIGINAL POPULATION OF SAN JOAQUIN VALLEY Tinkham, George H. 1923. History of San Joaquin County, California. Los Angeles, Calif. United States Treaties 1905. Message from the President . . . communi- cating Eighteen Treaties made with Indians in California . . . . [1851-1852, by G. W. Barbour, 0. M. Wozencraft, and Redick McKee.] 32nd Cong., 1st sess., Sen. Con. Doc. Reprint of 1905. Washington, D. C. arner, J. J. Description of 1832 Epidemic among the Indians of the San Joaquin Valley. In An Illustrated History of San Joaquin County, California . . . , pp. 28-29. The Lewis Publishing Co. Chicago. Wessels, H. W. 1857. Report on the Tribes of the San Joaquin Valley, 34th Cong., 3rd sess., H. R. Ex. Doc. 76, pp. 31-32. Wozencraft, 0. M. 1851. Letter dated July 12, 1851. 32nd Cong., 1st sess., Sen. Ex. Doc. 1, pt. III, pp. 488-490 [Ser. no. 906]. MANUSCRIPTS All manuscripts are in the Bancroft Library, Univer- sity of California, Berkeley, unless otherwise stated. bella, Ramon Diario de un registro de los Rios Grandes, Oct. 31, 1811, San Francisco. Santa Barbara Archive, IV:101-134. Also original manuscipt. ftimira, Jose Letter to Prefect Jos6 Senan, July 10, 1823, San Francisco. Archbishop's Archive, IV (2): X21-26. ador, Jose Maria Memorias sobre la Historia de California, 1877. Original manuscript C-D 28. giello, Jose Letter to Governor Arrillaga, May 30, 1805, San Francisco. Provincial State Papers, XIX: 42 ff. i-, 11 a guello, k I . L I Luis Antonio Letter to Governor Arrillaga, Oct. 31, 1813, San Francisco. Provincial State Papers, XIX: 345- 349. Carta al Gobernador Don Pablo Vicente de Sola . . . May 26, 1817, San Francisco. Original manuscript (no. fm F864A64); also typed copy. rryesa, Jose Dated July 15, 1830, San Jose. Departmental State Papers, II:135-137. bb,Juan Expedicion al valle de los Tulares, Letter to the Padre Presidente, Apr. 7, 1815. Santa Barbara Archive, VI:67-72. Letter to De La Guerra, May 23, 1818, De La Guerra Documents, VII:88. Dn, H. California Indians. 1875. Duran, Narciso Diario de la expedicion de reconocimiento hecha en el mes de Mayo de 1817.... Original manuscript. (See also Charles E. Chapman, 1911.) Estudillo, Jos6 Maria Diario que formo yo el tente dn Jose Maria Estudillo en la campafia . . . emprendo pa el reconocimiento y visita de las rancherias situadas en los tulares . . . . Nov. 10, 1819, Monterey. Original manuscript; also typed copy. Garcia, Inocente Hechos Historicos de California, 1878. Ori- ginal manuscript CC-D 84. Jaime, Antonio Letter to Governor Sola, March 30, 1816, Soledad. Archbishop's Archive, III(1):23-24. Marquinez, Marcelino Letter to Governor Sola, May 26, 1816. Arch- bishop's Archive, III(l):41-42. Martin, Juan Visita a los Gentiles Tularefios, Apr. 26, 1815, San Miguel. Santa Barbara Archive, VI:85-89. Martinez, Luis Antonio Entrada en las Rancherias del Tular, May 29, 1816, San Luis Obispo. Archbishop's Archive, III(1 ):42-45. McKinstry, George Documents for the History of California, 1846-9. Presented by Dr. George McKinstry of San Diego, 1872. Merriam, C. Hart Manuscript collection in Department of Anthro- pology, University of California, Berkeley. [73] : L? t 7 3 A., _ ANTHROPOLOGICAL RECORDS Moraga, Gabriel Diario de la tercera expedicion echa por el Alferez Don Gabriel Moraga . . . a los rios del norte; verificada en el mes de septiembre de el afio de 1808. Original manuscript; also two typed copies. Mufioz, Pedro Diario de la Expn echa por D. Gabriel Moraga de la Compania de San Francisco a los nuevos descubrimientos del tular . . . Nov. 2, 1806, San Francisco. Santa Barbara Archive, IV: 1-47. Ortega, Juan de Diario que forma el SargtO DistdO Dn Juan de Ortega segun los sitios qe por orn. del Sr. Govor de su mando registrar . . . Dec. 2, 1815, San Juan Bautista. Original manuscript; also typed copy. Pico, Jos6 Dolores Diario formado pr el SargtO Jos6 Dolores Pico de la expedicion que a echo pr dispocion del ciudadano . . . Jos6 Estudillo, Jan. 31, 1826. Original manuscript. Pinia, Joaquin Quaderno de las Novedades Hoccuridas diaria- mente en la expedicion que marcha a las or- denes del . . . Guadelupe Vallejo, June 13, 1829, Monterey. Original manuscript; also a copy in the California Manuscript series, no. E-88. Rodriguez, Sebastian Diario que forma yo el SargtO Sebastian Rodriguez de la Campana nombrada el dia 17 de Abril de 1828 [dated May 8, 1828]. Ori manuscript. Diario formado pr el Sargento Sebastian guez desde el dia 26 de Mayo . . . una cion al Tular por el rumbo de S. Miguel, 22, 1828, Monterey. Manuscript. Sal, Hermenegildo . . .Informe en el cual el teniente Hermi Sal manifesta lo que ha adquirido de vari sugetos para comunicarlo al Governador frrovincia, Jan. 31, 1796. Provincial Sta Papers, XIV:14-16. Sanchez, Jos6 Letter to Ignacio Martinez, May 10, 1826 State Papers, Missions and Colonization, II:15-20. Savage, James In H. Dixon, California Indians. MS 1875. Viader, Jos6 Diario, o noticia del viaje que acabo deh . . . desde el 15 hasta el 28 de Agosto d. Aug. 28, 1910, San Juan Bautista. Santa bara Archive, IV:73-84. Diario del P. Jose desde 19 hasta 27 de de 1810. Letter to the Padre Presidente, 19, 1810, San Jose. Santa Barbara Archii IV:85- 94. Zalvidea, Jos6 Maria Diario de una expedicion tierra adentro, Santa Barbara Archive, IV:49- 68. 74 -12 -swil, tk 17A s)-l ILLE K-, I YUBA Cl 'elk LA KE G. I 1?,o Ki 7-AHOE k.; -Of. < ....... Mc.K... C',Pp.,G. XE J J " - - , I !? - T R- C', N_ ?_k 7 50 Y. -7- LA E. OODLI" 0 L MAIIIIIJI K ?Q F--N" Y A F, To T I.- k w 7,1_ RO?? 7/' %: oN S. SO P-1 E C'. V.-L k iw. G,- 0 t Wd sh.1 c;i h -y S W.,d NO- Lod Y alleio Do. SAN P. 0 V# S4 BAY -7, S. R d Q p", J, AOit_ A Y A sop 11_1_?'_ Y I L.h.p., yt- J. "'? - 0 0 I i A N A, F i Q -u Ll- . ; - 6 - 1 5 17if 20 2 1 .-d Pt B. C?I.b S 1, k Golde Pt L-.b- SAN FRANCISC ed [StO C. S.. cleefi r."y DY' 11=14 L PIS.llP.d,,_,- Ott... at. -- -L 4 e Mc Clure 4 E I G,'A,,. PAI.,pt _4? wowme woo - ll,, CIT 117 C I Ao(?lced 0 r j Al- t., /r C-: vt JOSE S -t f dAl cairse u el th 006 C S-4 Colo. -0,0 jA r -Pwt" hi. 4 S-nV6 N DGE?ORT- N 0 LAK 28 0 3 24 29 I s co c 3 ..Moll am* Labsk 4 PIA 1) i6lAN I RE ERVATIO 00, OF w mk J H""t I A- t?h 10 C. s s _Jy j- '?,G R AA Pt pi.. B I T A SOY I 6 XACLU D-1 15 17 7`2 N )E 18 19 m 2 0 7-T-- KI I N G M.-Y 2 23 24 T, SW. So. L P%d- SI.- POW 7 p u dt. lb 30 R-d. Pt IF 31 Pt a's 2D 21 8- ----- 77-- H 1. This map covers the entire area under discussion, ng from the Cosumnes R'ver to the Tehachapi. The Pt S.1 er divisions, denoted by numbers and letters, represent 0,- r bitatareas considered in detail in the text. The succeed- C aps, drawn to larger scale, show these same areas with ibal divisions entered as far as possible. A A B R kA s 114P 0 5 10 20 30 /F- S-WT MILES Pt A,V-U. 1406 accompany Cook, "Aboriginal Population of San Joaquin A." Univ. of Calif. Publ. , Anthro. Ree. Vol. 16, No. 2. W d C., Ri.- 4 wh't-y HITM Sp,, "g r 44 u Y? A R K 38 24 26 2 30 35 L--? Eld.,W-d OWENS Ri- LAKE N C. 61 ....... uAoe, H., Li.d- 0 Tipl- '21 7 INDIAK 2 -C, RESERVATI T 14 23 uto. Ith 26 00C.M., Ric lk ELD t9 --r-710 my Sktd1IiI, cr 7f-'. 0.k. L Om 2 26 7 28 29 31 3? 38 1 12 z 3 10 L.b- S..d . ..... -t L.kio H ?c U-*, 15 ltl-k L 4 PI, -TS. t~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Win "eb C 'el-tt :g I 4A~~~~~~~~~~ IT N~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 0 5 10 20 30 v_ . . ~~~~~MILES - Map 2. Habitat areas 1A-2: the southern Yokuts and peripheral tribes. S ~~~~~~~~~~~~~~[75] I 4 ? I Map 3. Habitat areas 3A-4C: the basins of the Kaweah and Kings rivers, including the Yokuts and part of the Mono. Map 4. Habitat areas 5A-6B: the Yokuts, part of the Mono, and the Southern Miwok. [76] 0 5 10 20 30 MILES 7A 4C HOLKOMA 4B 14 0 5 10 20 30 MILES 1~~~~RU ioi Bt AKISMNESESR .. N&ap 5. Habitat areas 7A-14: ~~~~~~~theNrhrnYkt, eta adNrhenMwk [77] U~~~~~~~~~~ul. ju IO A V~~~~~~~~~~~~~~~~~~~~tc to,,S - ; 9 \ BOZENAT5OYB\ MILES~~~~~~~~~~~~~~~ L AK\M [78] Map 6. The Lower San Joaquin River and Delta areas (particularly areas 8 and 13).