PUEBLO RUINS NEAR WINSLOW, ARIZONA. 539 



Onr excavations at old Cunopavi, interrupted in their inception, were 

 too small to betray much of ancient customs, but we found in the 

 graves such a close resemblance to those of Sikyatki that there must 

 have been a strong likeness between the two peoples. The decorations 

 on pottery are closer to those of Sikyatki than to that of Walpi or the 

 Middle Mesa. Cuiiopavi is no longer a pottery-making pueblo to the 

 same extent as Oraibi and the East Mesa, but in old times this condi- 

 tion does not seem to hold. 



CONCLUSION. 



In order to show the advance made in the interpretation of the 

 problem of the origin of the Hopi Indians by the fieldwork of 1890, it 

 may be well to close my report with a summary of theoretical results 

 obtained by the expedition. The objective material collected demon- 

 strates that ancestors of some Hopi clans lived in ancient times on the 

 Colorado Ohiquito and at Chaves Pass, over a hundred miles south of 

 Walpi, a few miles from the head waters of the Gila Salado drainage 

 area. The material from the most southern ruin examined is almost 

 identical with that from the Verde Valley. The step which remains to 

 be taken is a searching investigation of the ruins from the crest of the 

 Mogollones south to the great ruins near Tempe and Phoenix. When 

 that is done we will have what has never been done before in south- 

 western archaeology, the tracing of a migration legend of a pueblo 

 people, step by step, by archaeological methods. 1 



The material collected from the Chevlon ruin shows that some Zufii 

 clans probably formerly lived farther down the river than their descend- 

 ants do at present, and that their culture was almost identical with the 

 Hopi, who lived in the neighborhood or possibly in the same pueblo. 

 The arguments bearing on this conclusion can be satisfactorily stated 

 only by a technical discussion of the material. 



Studies of the habitations of the ancient people of the southwest 

 have shown me, as stated in my report for 1895, that the culture of the 

 j)eople who built cliff houses, cavate dwellings, and villages in the 

 plains or on mesa tops, was the same, and that these different architec- 

 tural structures are adaptations to environment. I am convinced, from 

 my studies in 1896, that the color and character of ancient pottery 

 follows a similar law, and varies according to the material employed. 

 In other words, that a classification of pottery by colors is purely 

 arbitrary and useless in separating different pueblo people. The 

 character of the symbolism is more important. 



1 It may be suggestive to show that the ruins at Chaves Pass are twice the distance 

 from Walpi that they are from Casa Montezuma in the Verde Valley, and almost mid- 

 way between the present Tusayan pueblos and those near Tempe, near the Salado 

 River, in southern Arizona. The ruin at Rattlesnake Tanks, which I have not stud- 

 ied, lies midway between the Verde ruins and those at Chaves Pass, on the trail 

 connecting them. 



