When the Snake clans lived at Tokónabi, and later at Wukóki (on the Little Colorado), so far as known they had no subterranean rooms isolated from the others for ceremonial purposes, but used rooms so closely resembling other apartments that they may be called “living rooms.” Even when they came to the Hopi mesas they may not have had at first a specialized ceremonial chamber. A study of Arizona ruins reveals no rooms identified as ceremonial that are isolated from the house masses. This is true of cliff-dwellings and pueblos, and it is probable that the differentiation and separation of kivas from secular houses, found in modern Hopi pueblos, are an introduced feature of comparatively late date. At Zuñi a rectangular room, not separated from the house mass, serves as a kiva, the custom in this respect approaching more closely that found among their kindred, the ancient people of the Little Colorado river, than among the more modified Hopi of the present time. While some of the rooms identified as ceremonial in preceding pages are rectangular in shape and not isolated from secular rooms, the circular type seems also to have been found in Utah, and at Kitsiel and ruins near it. South of Marsh pass circular kivas are less abundant, and it appears that somewhere in this region is a line of demarcation between ruins with circular kivas and those with rectangular kivas. In prehistoric ruins from Marsh pass southward[43] to the Gila valley no rooms have ever been identified as kivas, although in the cavate ruins called Old Caves, near Flagstaff, are subterranean rooms entered from the floor of a room above, which may have served for the performance of religious rites.[44] From a comparison of some features of the kivas in the cliff-dwellings of the San Juan and its tributaries with those of the Navaho Monument it would appear that while the ceremonial rooms of the latter in certain details are like those of the former, in some cases their form and position are different. So far as this resemblance goes, it may be reasoned that the San Juan ancients influenced by their culture the northern Arizona cliff-dwellers, but there is scant evidence of the reverse, that is, that the San Juan pueblos borrowed from the culture of the northern Arizonians any architectural features, especially in the form and construction of their kivas. The theory would be logical that the prehistoric migration of culture was down rather than up the river, and the symbolism of the pottery contributes interesting data supporting this conclusion. FOOTNOTES: [21] This ruin may be that called Tecolote, which appears on many old maps. [22] Among other names cut on the walls of this ruin is that of Lieutenant Bell, 1859. [23] A few broken-down walls of rooms stand at the side of the trail just before one reaches the dangerous part.