16 BUREAU OF AMERICAN ETHNOLOGY [bull. 35 



intervals posts rising above them into the superstructure. Most of the 

 ruins show their extent and plans by the lines of these bowlders pro- 

 truding from the soil. There is sufficient basis, however, for the 

 belief that the bowlder-earth wall belonged to a lower story and that 

 above it were a structure of earth and sticks of lighter character and a 

 flat roof resting upon the posts. So far as known, no adobe brick or 

 block construction, like that used in Mexico, was practised in any 

 part of this region. Wherever villages of this type occur (on lower 

 Gila and Salt and southward through Mexico), the ruins show a high 

 mound surrounded with lower masses of debris. These have been 

 designated temples by Gushing, and have generally been regarded as 

 council houses, sacerdotal or defensive structures, a surviving type of 

 which is Casa Grande, near Florence, Ariz. Irrigation works are 

 always associated with ruins of this kind, and on the rivers men- 

 tioned these works are the most important found in America/ 1 



CLIFF-DWELLINGS 



T\ nile structures of this class were not so numerous or large as 

 those of the San Juan watershed, yet their distribution was extensive 

 throughout the mountain region, where they exist whenever the 

 physiographic conditions are favorable. They have been found in 

 the canyons of the San Francisco, Tularosa, Blue, Eagle, Pueblo, 

 Salt and Gila, and Gila Bonita, and they reach westward to the 

 Tonto basin. Generally they occupy sites in narrow canyons near the 

 headwaters of rivers and consist of a row of small one or two story 

 houses placed under an overhanging cliff or in front of a cave. 

 Single houses were often placed in niches or fissures in the rocks. 

 Scarcely any pocket large enough to shelter a human being but 

 shows evidence of occupation, mostly temporary perhaps, though 

 frequently such places are blackened with smoke and contain much 

 refuse. 



The cliff-houses are in construction like those of the San Juan 

 region, and they preserve the details of building that have disap- 

 peared from the open-air pueblos. A comparison of artifacts found 

 in the pueblos and cliff-dw T ellings shows their similarity, and indi- 

 cates clearly that ,the two classes of structures are work of the same 

 people. 



A niche adjoining a cliff-house on the San Francisco river near the 

 mouth of the Blue was devoted to milling, as shown by the well-worn 

 cavities in the floor (see p. 47). On Pueblo creek and at the head 

 of the Tularosa similar milling rooms were excavated in the soft 

 rock. 



" Hodge, Prehistoric Irrigation in Arizona, American Anthropologist, VI, July, 1893. 



