While the work of excavation and repair of Cliff Palace described in the preceding pages adds nothing distinctly new to existing knowledge of cliff-dweller culture, it renders a more comprehensive idea of the conditions of life in one of the largest of these interesting ancient, settlements in our Southwest. Of all the questions that present themselves after a work of this kind, perhaps the most important, from a scientific point of view, is, What relation exists between the culture of Cliff Palace and that of the neighboring pueblos? Directly across the canyon, in full view of Cliff Palace, there is a typical pueblo ruin, almost identical in character with many others scattered throughout the Southwest, some of which are known to have been inhabited in historic times by ancestors of Pueblo peoples still living. The contribution here made to the knowledge of cliff-dwelling culture will, it is hoped, shed light on the question, In what way are the cliff-dwellers and the Pueblos related?

The relationship in culture of the former people of Cliff Palace to those of the large pueblo ruin on the mesa across the canyon is most instructive. How were the inhabitants of these two settlements related; and were the two sites inhabited simultaneously, or is the pueblo ruin older than Cliff Palace? So far as the culture of the inhabitants of the two is known (and knowledge of the pueblo is scant), the two settlements were synchronously inhabited, but nothing in them gives indication of the period of their occupancy. These questions can be settled only by the excavation of this pueblo or of some similar ruin on the plateau.a Nordenskiöld, with the data possessed by him, did not hesitate to express decided views on this point:

We are forced to conclude that they [cliff-houses] were abandoned later than the villages on the mesa. Some features, for example, the superposition of walls constructed with the greatest proficiency on others built in a more primitive fashion (see plate XIII) indicate that the cliff-dwellings have been inhabited at two different periods. They were first abandoned, and had partly fallen into ruin, but were subsequently repeopled, new walls being now erected on the ruins of the old. The best explanation hereof seems to be the following: On the plateau, and in the valleys the Pueblo tribes attained their widest distribution and their highest development. The numerous villages at no great distance from each other were strong enough to defy their hostile neighbors. But afterwards, from causes difficult of elucidation, a period of decay set in, the number and population of the villages gradually decreased, and the inhabitants were again compelled to take refuge in the remote fastnesses. Here the people of the Mesa Verde finally succumbed to their enemies. The memory of their last struggle is preserved by the numerous human bones found in many places, strewn among the ruined cliff-dwellings. These human remains occur in situations where it is impossible to assume that they have been interred.

---

aA true comparison of the mesa habitation and the cliff-dwelling can be made only by renewed work on the former, which is now little more than a huge pile of fallen walls. Present indications show a greater antiquity of the mesa ruin, the site of which afforded more adequate protection. On this supposition the mesa ruins would be considered older than the cliff ruins, and those of the valley the most ancient. If the ruins in Montezuma valley are the oldest, we can not suppose that the culture originated in the cliffs and spread to the valley. The circular subterranean kiva bears indication of having originated in valleys rather than in caverns. Nordenskiöld does not mention the large ruin on the bluff west of Cliff Palace.

---

Closely connected with the relative age and the identity of the Mesa Verde cliff-house and pueblo culture are the age and relationship of different cliff-houses of the same region, for example, Cliff Palace and Spruce-tree House. The relative number of kivas may shed light on this point.

The relative proportion of the number of kivas to secular houses varies in Cliff Palace and Spruce-tree House. In the former there are about 7 secular rooms to every kiva; in the latter about 15. Long House has a still more marked difference, there being here only a few secular houses and a maximum number of kivas. Whether this variation has any meaning it is impossible to say definitely; theoretically, as compared with modern pueblos, the proportionately larger number of kivas points to a sociological condition in Cliff Palace characteristic of more primitive times. The larger the number of kivas relatively to secular rooms the older the ruin. Long House would be regarded as older than Cliff Palace, and Cliff Palace older than Spruce-tree House, Balcony House being the most modern and the last of the four to be deserted. A cliff-dwelling with a kiva but without secular rooms is rare, and one with secular rooms but without kivas is likewise unusual. Where the latter exists it is so situated as to indicate that it was subordinated to neighboring large cliff-dwellings.

The relative number of circular kivas in ruins and in modern inhabited pueblos where the circular form of room is found is larger in the ruins than in the inhabited pueblos. The proportionate number of circular rooms to secular rooms in cliff-dwellings of the Mesa Verde is also larger than in pueblo ruins like those of the Chaco. Apparently the older the pueblo the greater the relative number of kivas. If, as is suspected, a larger number of kivas indicates relatively greater age, the explanation may be sought in the amalgamation of clans and the development of religious fraternities. Hypothetically, in early days each clan had its own men's room, or kiva, but when clans were united by marriage and secret ceremonies were no longer limited to individual Clans, the participants belonging to several clans, a religious fraternity was developed and several clan kivas consolidated or were enlarged into fraternity kivas such as we find among the Hopi and other Pueblos.

From a study of kivas the conclusion is that Spruce-tree House is more modern than Cliff Palace. This conclusion is borne out also by the fact that the water supply at Spruce-tree House is more abundant than that at Cliff Palace.

In one or two architectural features Cliff Palace is unique, although sharing with other cliff-houses of the Mesa Verde National Park many minor characters. The first difference between Cliff Palace and Spruce-tree House, outside of the disparity in their size and the relatively large proportion of secular to ceremonial rooms in the latter, is the existence in the former of terraces and retaining walls. Spruce-tree House is built on one level, above which rise the secular houses while below are the ceremonial rooms or kivas. The contrast of this simple condition with that of Cliff Palace, with its three terraces and the complicated front wall at several levels thereby necessitated, is apparent.

There are several other ruins in the Mesa Verde Park in which the configuration of the rear of the cave led to the construction of the cliff-house in terrace form. This is well exemplified in the Spring House, where buildings on an upper level occupy much the same relation to those below as the ledge houses to the main ruin, and in ruins in the Canyon de Chelly, like those in Mummy Cave, where this relation of the buildings on the ledge to those on top of the talus is even more pronounced. Architectural features in cliff-houses are due to the geological structure of the cave in which they are situated rather than to cultural differences.

Nothing was found to indicate that Cliff Palace was inhabited during the historic period. The inhabitants were not acquainted with metals brought by white men to the Southwest. The absence of glass and of glazed pottery is also significant. No sheep, horses, or other beasts of burden paid them tribute. In fact, there is no evidence that they had ever heard of white men. These ruins belong to the stone age in America and show no evidence of white man's culture.

Except that it is prehistoric, the period at which Cliff Palace was inhabited is therefore largely a matter for archeological investigation to determine, and thus far no decisive evidence bearing on that point has been produced. It has been held that Cliff Palace is five hundred years old, and some writers have added five centuries to this guess; but the nature of the evidence on which this extreme antiquity is ascribed to the ruin is not warranted by the evidence available.

No additional information was obtained bearing on current theories of the causes that led the ancient occupants of the Mesa Verde cliff-dwellings to adopt this inhospitable and inconvenient habitat. It is probable that one and the same cause led to the abandonment of Spruce-tree House, Cliff Palace, and other Mesa Verde cliff-houses. The inhabitants of these buildings struggled to gain a livelihood against their unfavorable environment until a too-exacting nature finally overcame them. There are no indications that the abandonment of Cliff Palace was cataclysmic in nature: it seems to have been a gradual desertion by one clan after another. One of the primary reasons was change of climate, which caused the water supply to diminish and the crops to fail; but long before its final desertion many clans abandoned the place, and drifting from point to point sought home-sites where water was more abundant. All available data lend weight to a belief that the cliff-houses of Mesa Verde were not abandoned simultaneously, but were deserted one by one. Possibly the inhabitants retired to the river valleys, where water was constant, and later gave up life on the mesa. But even then the culture was not allowed to continue unmodified by outside influences. Where the descendants of Cliff Palace now dwell, or whether they are now extinct, can be determined only by additional research.

Evidence is rapidly accumulating in support of the theory that the "cliff-dweller culture" of our Southwest was preceded by a "pit-house culture," the most prominent feature of which is the small circular or rectangular rooms, artificially excavated laterally in cliffs or vertical in the ground, which served this ancient people either as dwellings or for storage. The side walls of these rooms were supported in some instances by upright logs, and commonly clay was plastered directly on the walls of the excavations. The architectural survival of subterranean rooms exists among the cliff-dwellings in circular underground kivas, the variations of which are so well illustrated in Cliff Palace.

In connection with these "pit rooms," which are never large, may be mentioned the large subterranean artificial excavations found scattered over the Pueblo area of the Southwest. Such occur in the Gila valley, and have been reported from the San Juan drainage; they have been identified as reservoirs and also as kivas. Some of these subterranean rooms are rightly identified as kivas, but others have architectural features that render this interpretation improbable. What their function was and how they are connected with the people who built the smaller subterranean rooms of the Southwest can be determined only by excavations and a study of the features of both types.

The most important step that remains to be taken in the scientific study of the ruins of the Mesa Verde National Park is to discover the relation of the culture of Cliff Palace to that of the neighboring pueblo. This will necessitate the scientific excavation and repair of the latter ruin and a comparison of its major and minor antiquities with those of Cliff Palace. The age of cliff-dwellings in different parts of the Southwest undoubtedly varies. Certain Pueblo ruins are older than some cliff-dwellings, and there are cliff-houses more ancient than Pueblo ruins. Continued research in the Mesa Verde region will doubtless shed light on the relative age of Cliff Palace and the great pueblo ruin opposite it.