We may at this juncture have recourse to river mileage 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 persons 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 mentioned. Hence it is reasonable to apply the factor found for the Chowchilla, 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 peripheral region would then be approximately 2,500 and that for the southern end of the valley as a whole 6,920, or in round numbers 6,900. SOUTHERN SAN JOAQUIN VALLEY ... 6,900 _____________________________________ THE NORTHERN SAN JOAQUIN VALLEY