eras_. They are regarded as places of
refuge of former inhabitants of the country, contemporaneous with
ancient pueblos and cliff houses.]

[Footnote 25: This pinnacle is visible for miles, and is one of many
prominences in the surrounding country. Unfortunately this region is
so imperfectly surveyed that only approximations of distances are
possible in this account, and the maps known to me are too meager in
detail to fairly illustrate the distribution of these buttes.]

[Footnote 26: In certain cavate houses on Oak creek we find these
caverns in two tiers, one above the other, and the hill above is
capped by a well-preserved building. In one of these we find the
entrance to the cavern walled in, with the exception of a T-shape
doorway and a small window. This chamber shows a connecting link
between the type of true cavate dwellings and that of cliff-houses.]

[Footnote 27: The absence of kivas in the ruins of the Verde has been
commented on by Mindeleff, and has likewise been found to be
characteristic of the cliff houses on the upper courses of the other
tributaries of Gila and Salado rivers. The round kiva appears to be
confined to the middle and eastern ruins of the pueblo area, and are
very numerous in the ruins of San Juan valley.]

[Footnote 28: See "Tusayan Totemic Signatures," _American
Anthropologist_, Washington, January, 1897.]

[Footnote 29: An exhaustive report on the ruins near Winslow, at the
Sunset Crossing of the Little Colorado, will later be published. These
ruins were the sites of my operations in the summer of 1896, and from
them a very large collection of prehistoric objects was taken. The
report will consider also the ruins at Chaves Pass, on the trail of
migration used by the Hopi in prehistoric times in their visits, for
barter and other purposes, to the Gila-Salado watershed.]

[Footnote 30: Possibly the Shoshonean elements in Hopi linguistics are
due to the Snake peoples, the early colonists who came from the north,
where they may have been in contact with Paiute or other divisions of
the Shoshonean stock. The consanguinity of this phratry may have been
close to that of the Shoshonean tribes, as that of the Patki was to
the Piman, or the Asa to the Tanoan. The present Hopi are a composite
people, and it is yet to be demonstrated which stock predominates in
them.]

[Footnote 31: A Study of Pueblo Architecture: Tusayan and Cibola;
Eighth Annual Report of the Bureau of Ethnology, 1886-8