Continuing along the left bank of the river, from the row of cavate rooms, just described, on the first mesa, we round a promontory and enter a small canyon,[17] which is perforated on each side with numerous other cavate dwellings, large and small, all of the same general character as the type described. Here, likewise, are small external openings which evidently communicated with subterranean chambers, but many of them are so elevated that access to them from the floor of the canyon or from the cliff above is not possible. A marked feature of the whole series is the existence here and there of small, often inaccessible, stone cists of masonry plastered to the side of the rocky cliff like swallows' nests. All of these cists which are accessible had been opened and plundered before my visit, but there yet remain a few which are still intact and would repay examination and study. Similar walled-up cists are likewise found, as we shall see later, in the cliff-houses of the Red-rock country, hence are not confined to the Verde system of ruins. Cavate dwellings similar to those here described are reported to exist in the canyons of upper Salado, Gala, and Zuñi rivers, and we may with reason suspect that the distribution[18] of cavate dwellings is as wide as that of the pueblos themselves, the sole requisite being a soft tufaceous rock, capable of being easily worked by people with stone implements. In none of the different regions in which they exist is there any probability that these caves were made by people different in culture from pueblo or cliff dwellers. They are much more likely to have been permanent than temporary habitations of the same culture stock of Indians who availed themselves of rock shelters wherever the nature of the cliff permitted excavation in its walls. That the cavate lodges are simple "horticultural outlooks" is an important suggestion, but one might question whether they were conveniently placed for that purpose. So far as overlooking the opposite plain (which had undoubtedly been cultivated in ancient times) is concerned, the position of some of them may be regarded good for that purpose, but certainly not so commanding as that of the hill or mesa above, where well-marked ruins still exist.