caved in, was covered in the same manner. None of the lower story
rooms had an outlet to the apartments above, and the evidence tended
to prove that the second story houses were reached from the bottom of
the cave over the roofs of the front row of houses by means of ladders.

Most of the rooms were well supplied with apertures of the usual
conventional form; sometimes there were as many as three in one
room, each one large enough to serve as a door. But there were also
several small circular openings, which to civilised man might appear
to have served as exits for the smoke; but to the Indian the house, as
everything else, is alive, and must have openings through which it can
draw breath, as otherwise it would be choked. These holes were three or
four inches in diameter, and many of them were blocked up and plastered
over. A large number of what seemed to have been doorways were also
found to be blocked up, no doubt from some ulterior religious reason.

A peculiar feature of the architecture was a hall not less than forty
feet long, and from floor to rafters seven feet high. Six beams were
used in the roof, laid between the north and south walls. There were
rafters of two different lengths, being set in an angle of about
ten degrees to each other. The west wall contained twelve pockets,
doubtless the cavities in which the rafters had rested. They were,
on an average, three inches in diameter, and ran in some six inches,
slanting downward in the interior. The east wall was found to contain
upright poles and horizontal slats, forming a framework for the
building material. The interior was bare, with the exception of a ledge
running along the southern side and made from the same material as the
house walls. It was squared up in front and formed a convenient settee.

At the end of this hall, but in the upper story, was found a house