The village distribution of the Karok was treated briefly by Kroeber in the Handbook, pages 99 to 102, and far more exhaustively in a later paper (1936). For the latter he secured the services of two good informants, a very elderly Indian man named Ned and a woman, Mary Jacops, with whom he examined the area carefully. The list set forth on pages 30 to 34 of his publication must be regarded as definitive. It is true that Merriam has a very complete list of Karok villages but his names vary linguistically from those of Kroeber to such an extent that, save in a few instances, it is extremely difficult to reconcile them. However, since Merriam's total is 115 for the same territory where Kroeber finds 108 and since Merriam does not give house counts the Kroeber list may be used exclusively. Ned gave house counts but Mrs. Jacops did not. Kroeber amplified wherever possible with data from Curtis (cited by Kroeber, p. 30, as The North American Indian, 13:222). Ned's counts were very cautious since he distinguished frequently between the number of houses he had seen at a given site and the number he had heard were there. On the basis of such distinctions Kroeber reduces the total count by a factor of one-sixth. He states (p. 35): Among the Yurok ... two occupied houses may be reckoned for each three house sites recognized when full detailed data are at hand. They are obviously not detailed for the Karok. I must take issue with two points. With the Karok the counts were not based upon house sites recognized but on the memory of _inhabited houses_ by informants. Hence the house site or pit theory cannot apply. In the second place, a reasonably thorough examination of the _published_ material on the Yurok, Wiyot, and the Karok shows that the data for the Karok presented by Kroeber represents the fullest detail of all with respect to the number of houses. Apropos of the same question it is of interest to point out the house counts given by Ned for the fifteen villages also provided with counts by Curtis for 1860. Kroeber has tabulated these himself and shows that, despite variation in individual detail, the total for Ned is 60 and that for Curtis is 57-60. The identity is remarkable. Commenting on this situation, Kroeber makes the following very significant statement (p. 35. fn.): It is apparent that, for any particular settlement, no precise figure, even by a good informant, is very reliable unless based on an enumeration of named houses. But for a larger series of settlements the particular variations, resulting from changes of residence or difference of times referred to, tend to cancel each other out and to yield _comparable and fairly reliable totals_. (Emphasis mine.) The present writer, consequently, can see no necessity for a gross reduction of one-sixth of the computed population.