e other by Drucker
(1937). Of all these the treatment by Drucker is the most complete
since he had the advantage of a knowledge of the earlier work. Although
he may have missed settlements in the interior, for present purposes we
must accept his list as a working basis.

Drucker mentions 23 villages, all located on the coast or along the
lower reaches of the Smith River. Kroeber gives 10 sites from which he
computes the population, at the Yurok rate of 45 inhabitants per
village, as 450. Waterman gives 14 places, which, at the same rate,
would yield 630. Drucker has house counts for 13 of his villages, with
a total of 88 houses or 6.76 houses per village. At the Yurok count of
7.5 persons per house, which Kroeber says applies to the Tolowa, the
average population per village would be 51. Kroeber's estimate of 45 is
thus quite close. There is no good reason to suppose, in view of the
lack of any good evidence to the contrary that the other 10 villages of
Drucker were smaller than those for which he gave house counts. Thus we
may add 68 houses, making a total of 156 and a population of 1,186.
Kroeber would of course reduce by one-third but the reasons for so
doing are no more compelling with this than with any other tribe.

Drucker (p. 226) states that his house counts are as of 40 to 50 years
ago. This means, first, that the houses were described to him by
informants as known to them in their youth to be inhabited (hence no
reduction necessary) and, second, that the counts represent the
situation during the period of 1885 to 1895.

Now the counts published for all the tribes hitherto considered were
based upon the conditions obtaining at approximately 1850, 35 to 45
years earlier. In other words, Drucker's figures cannot in any sense
represent the aboriginal state, for there must have been a marked
decline in population and in number of houses among the Tolowa between
1850 and 1890. The implication is, startling as it may seem, that the
population estimate given above is much too low.

Some idea of what may have happened can be secured by a brief
reconsideration of Waterman's Yurok data. It will be remembered that
Waterman shows detailed maps of 19 villages, including not only houses
once standing but also houses standing and inhabited when he saw them
in 1909. The ratio of the former to the latter is 189 to 38. There were
of course many more houses standing in 1890 than in 1909, although the
population did not decli