The small northern cluster was in all probability inhabited by five or six families only, as contrasted with the main cluster, which had sixteen or seventeen, while the smaller clusters had each only two or three families. The strong presumption of the later building and occupancy of the smaller clusters, previously commented on, is supported by three other facts of importance, viz, the amount and height of the standing wall, the character of the sites occupied, and the extraordinary size of the rooms.

Although as a rule external appearance is an unsatisfactory criterion of age, still, other things equal, a large amount and good height of standing wall may be taken to indicate in a general way a more recent period of occupancy than wall lines much obliterated and merged into the surrounding ground level. The character of the site occupied is, however, a very good criterion of age. It was a rule of the ancient pueblo builder, a rule still adhered to with a certain degree of persistence, that enlargement of a village for the purpose of obtaining more space must be by the addition of rooms to those already built, and not by the construction of detached rooms. So well was this rule observed that attached rooms were often built on sites not at all adapted to them, when much better sites were available but a short

distance away; and, although detached rooms were built in certain cases, there was always a strong reason for such exceptions to the general rule. At a late period in the history of the Pueblos this rule was not so much adhered to as before, and detached houses were often built at such points as the fancy or convenience of the builder might dictate. As the traditions are broken down the tendency to depart from the old rule becomes more decided, and at the present day several of the older Pueblo villages are being gradually abandoned for the more convenient detached dwellings, while nearly all of them have suffered more or less from this cause.

The tendency to cluster rooms in one large compact group was undoubtedly due primarily to hostile pressure from outside, and as this pressure decreased the inherent inconveniences of the plan would assert themselves and the rule would be less and less closely adhered to. It therefore follows that, in the absence of other sufficient cause, the presence of detached rooms or small clusters may be taken in a general way to indicate a more recent occupancy than a ground plan of a compact, closely built village.

The size of rooms is closely connected with the character of the site occupied. When, owing to hostile pressure, villages were built on sites difficult of access, and when the rooms were crowded together into clusters in order to produce an easily defended structure, the rooms themselves were necessarily small; but when hostile pressure from surrounding or outside tribes became less pronounced, the pueblo-builders consulted convenience more, and larger rooms were built. This has occurred in many of the pueblos and in the ruins, and in a general way a ruin consisting of large rooms is apt to be more modern than one consisting of small rooms; and where large and small rooms occur together there is a fair presumption that the occupancy of the village extended over a period when hostile pressure was pronounced and when it became less strong. It has already been shown that, owing to the social system of the pueblo-builders, there is almost always growth in a village, although the population may remain stationary in numbers or even decrease; so that, until a village is abandoned it will follow the general rule of development sketched above.

Fig. 285.—Ground plan of ruin on southern side of Clear creek.

Along the southern side of Clear creek, which discharges into the Rio Verde from the east, about 4 miles below Verde, there is a flat terrace from 30 to 40 feet above the creek and some 2 or 3 miles in length. Scattered over almost the whole of this terrace are remains of houses and horticultural works, which will be described later. Near the western end of the terrace a low hill with flat top and rounded sides rises, and on the top of this occurs the ruin whose ground plan is shown in figure 285.

This ruin commands an outlook over the whole extent of the terrace and seems to have been the home pueblo with which were connected the numerous single houses whose remains cover the terrace. The