Canyon de Chelly was occupied because it was the best place in that vicinity for the practice of horticulture. The cliff ruins there grew out of the natural conditions, as they have in other places. It is not meant that a type of house structure developed here and was transferred subsequently to other places. When the geological and topographical environment favored their construction, cliff outlooks were built; from a different geological structure in certain regions cavate lodges resulted; in other places there were "watch towers;" in still others single rooms were built, either lone or in clusters, and these results obtained quite as often if not oftener within the historic period as in prehistoric times. Notwithstanding the possible division of the De Chelly ruins into four well defined types, there is no warrant for the assumption of a large population. The types are interrelated and to a large extent were inhabited not contemporaneously but conjointly. There are about 140 ruins in Canyon de Chelly and its branches, but few of them could accommodate more than a very small population. Settlements large enough to furnish homes for 50 or 60 people were rare. As not all of the sites were occupied at one time, the maximum population of the canyon could hardly have exceeded 400; it is more likely to have been 300. The character of the site occupied is one of the most important elements to be studied in the examination of ruins in the pueblo country. In De Chelly whatever defensive value the settlements had was due to the character of the sites selected. It is believed, however, that other considerations dictated the selection of the sites, and that the defensive motive, if present at all, exercised very little influence in this region. The sites here are always selected with a view to an outlook over some adjacent area of cultivable land, and the structures erected on them were industrial or horticultural, rather than military or defensive. The masonry of the ruins and the constructive expedients employed by the builders are an insurmountable obstacle in the way of the hypothesis that the cliff ruins represent a primitive or intermediate stage in the growth of pueblo architecture. The builders were well acquainted with the principles and methods of construction employed in the best work found in other regions; the inferiority of their work is due to special conditions and to the locality. The presence of a number of extraneous features, both in methods and principles employed, is further evidence in the same line. These features are certainly foreign to this region, some of them suggest even Spanish or Mexican origin, which implies comparatively recent occupancy.