REPORT ON THE BUREAU OF AMERICAN ETHNOLOGY. 45 



included more or less comparatively recent pottery in the lower 

 levels, the earthenware of the earliest inhabitants of the site is of 

 finer quality and of finer decoration than that manufactured by the 

 historic Hawikuh people not long before the abandonment of their 

 settlement. 



Although the study of the archeology of Hawikuh has been barely 

 commenced, the results of last season's work give promise of a 

 material addition to our knowledge of an important phase of Pueblo 

 culture, and it is hoped will ultimately open the way to the solution 

 of related problems in southwestern archeology. 



Besides the routine work of his desk Mr. Hodge gave what spare 

 time he could while in Washington to continuing his work on the 

 bibliography of the Pueblo Indians. 



During July and August Dr. J. Walter Fewkes, ethnologist, com- 

 pleted his report on the Heye collection of West Indian antiquities 

 and in the autumn made a brief archeological reconnoissance in south- 

 western Colorado, returning to Washington the middle of November. 

 His plan of operations was to visit the ruins in the McElmo district 

 and determine their architectural features in order to define with 

 greater exactness the characteristics they share with the cliff dwell- 

 ings and pueblos of the Mesa Verde National Park. The object was 

 to gather material that would enable him to construct a classification 

 of the prehistoric buildings of the Southwest from structural data. 

 The Mesa Verde cliff dwellings and pueblos belong to a type or group 

 of ruins distinguished by the structure of the roof and other feat- 

 ures of the ceremonial room or kiva. The aim of the field work in 

 1917 was to investigate the distribution of this form of kiva and to 

 discover other peculiarities of the Mesa Verde type or group at points 

 remote from the plateau and thus enlarge our knowledge of the 

 geographical distribution of the types. 



It was found that the ruins in Montezuma Valley and the McElmo 

 and its tributaries show extensions westward of the Mesa Verde type, 

 and as the field work progressed much was added to our knowledge of 

 the characteristics of great houses and towers, the examples of which 

 on the Mesa Verde have been little investigated. 



The most noteworthy group of the ancient ruins visited in the 

 course of his field work were three clusters of great houses, castles, 

 and towers situated a short distance over the State line on the north- 

 ern tributaries of the canyons of the McElmo. 



The most important result of the field work in 1917 is the conclu- 

 sion that the ruins of the McElmo region indicate a people allied to 

 those of Mesa Verde, who reached a high degree of architectural 

 technique, surpassing any in America north of Mexico. Evidence was 

 gathered that it was preceded by a stage indicated by one-house con- 

 struction, and the suggestion is made that it antedated pueblos, on 



