guration of the rock, and not conversely as we do. It was,
therefore, a large joint-tenement structure, harboring, perhaps, when
fully occupied, several hundreds of families.

In regard to ingress and egress, not only have I found no doors in any
fragments of exterior walls, but the many persons I have asked have
always assured me that there had been none, that the house was entered
by means of ladders, ascending to the top of each story in succession,
and descending into the rooms also by ladders and through trap-doors in
the roofs. They have also assured me that each room of each story
communicated with the one above and below, also by means of trap-doors
and ladders. It is quite certain that there are no staircases nor steps,
and that consequently ladders were used, in the same manner as they are
still used by the Indians of the pueblos of Zuni, Moqui, Acoma, Taos,
and others. Ingress and egress, therefore, must have taken place, not
horizontally "in and out," but vertically "up and down." I have not been
able to identify any one of the trap-doors referred to, but I should not
be surprised to hear that they have been subsequently found in the
north-west corner of each room. By referring to the diagram of the floor
(Pl. III., Fig. 4), it will be seen that the rectangular spaces between
the beams and overlying poles are almost everywhere large enough, if
the superstructure of splinters (or brush) and clay is removed, to give
passage to any man. The ladders themselves have completely disappeared.

On one and the same floor, I found in the side walls at a few places,
the remains of low and narrow openings through which a man might pass in
a stooping position and "sidling." Nowhere could I see the full height
of these small doorways, so that I do not know whether there was a
lintel, or whether they terminated in an open angle, like the doorways
of Yucatan. I have seen openings showing the peculiar so-called
"aboriginal arch" of Yucatan on a small scale, and I also have seen that
an accidental "knocking-out" of one or two stones from the walls
produced a hole or gap very similar in shape to the doorways at Uxmal
and other pueblos of Southern Mexico, though of course on a small scale.
It is self-evident that, the coincidence being accidental, I do not
place any stress upon it in view of "tracing relationships." The
coincidence is of ethnological, and not of ethnographical, value. As far
as I could ascertain, they were certa