The figures, considering physiographic differences and varying coverage by ethnographers, are quite consistent. 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 Chowchilla. 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 probably 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 population must have been very small in the period of the early 1850's owing to extreme attrition suffered from the Spanish and particularly from the gold miners. Kroeber gives only 20 villages on all three streams, most of them on the Mokelumne. Merriam adds another 3, making 23 in all. At Gifford's population value this means 480 persons. The official sources are of little help since none of the agents or commissioners reported specifically on the area. Evidently there were too few survivors among the natives to warrant the trouble of placing them under the 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 the days before the Gold Rush. F. T. Gilbert (1879, p. 113) says that the Mokelkos, by which he means all the Indians between the Mokelumne and the Cosumnes in the hills and as far as Stockton on the plain, had 12 rancherias of 200 to 300 each and numbered about 3,000 in all. He, however, 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 towns, 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 numbered then about 2,000. In 1852, however, there were only 4 rancherias left, with 390 inhabitants. Gilbert was referring explicitly to the lower course of the rivers, whereas the villages cited by Kroeber were definitely above this region in the foothills. We may accept Gilbert's figure of 390 on the lower Mokelumne, to which may be added 110 for the lower Cosumnes and Calaveras and 480 for the upper villages, making a total of 980 or, let us say, 1,000.