1840. Hence he could scarcely be expected to remember population figures from a date much earlier than his childhood. The names and location of the villages themselves were at least semipermanent and could have been derived from the informant's parents even if not from his own memory. Hence it is probable that the figure furnished to Gifford more nearly represents the number of inhabitants in 1850 than in 1840. The average value for all 49 villages is 20.8 persons. Yet 7 villages are stated to have held 15 persons, 11 villages 10 persons, and 3 villages 5 or less persons. Such a condition argues a rapidly declining population, for no normal aboriginal settlement is likely to have contained less than 20 inhabitants. Gifford's average of 21 persons per village must, however, be accepted as representing the closest we can get to the value for the period of 1850. This means a population of 588 for the Stanislaus and 882 for the Tuolumne. The total is 1,470 for the foothill region. Between 300 and 400 may be added to account for scattered remnants along the lower courses of these rivers and on the San Joaquin itself, or 1,800 for the entire area under consideration.

To summarize, we have the following estimates for the Stanislaus-Tuolumne watershed at or about the year 1851:

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, Mariposa 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 pertaining 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 Tuolumne and the upper San Joaquin. Ryer, in a bill submitted July 31, 1851, claimed to have vaccinated 695 persons 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 Wozencraft 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 number 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 Mariposa 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 between 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 Mariposa 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, Kroeber 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 located villages.

Of the 35 village sites in Yosemite and below El Portal, 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. To this may be added, according to Ryer and to Johnston, 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. Merriam'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 territorial comparisons. There are under consideration, including those previously discussed, five small river systems, those of the Stanislaus, Tuolumne, Merced, Mariposa, and Chowchilla. Physiographically and ecologically 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 differences. It is of interest, therefore, to estimate the village 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, Gifford, and Merriam.