FHWKES] RELATION OF COMPOUNDS TO PUEBLOS 151 
Culturally, all northern! and central Arizona ruins, ancient and mod- 
ern,seem to show a dual composition, having connections on the one 
side with the Rio Grande pueblos and on the other with the habitations 
of the Gila Valley. Whether the pueblos of New Mexico in the Rio 
Grande drainage were derived from the compounds of the Gila or 
vice versa, is an open question, but there seem to have been two foci 
of cultural distribution in the Southwest. 
Hopi traditions support the theory that the ruins in the Verde and 
Tonto Valleys were settled by offshoots from the ‘great house” 
builders of the Gila and Salt Valleys in prehistoric times, and that the 
ruins along the Little Colorado were peopled in part by clans from 
the same river valleys. There appears to be no way of ascertaining 
the sources or the relative age of the Rio Grande culture, whether 
derivative or autochthonous.? 
The geographic limits of the ruins called “‘compounds”’ appear to 
be the plains of the Gila-Salt Basin. Following up the tributaries of 
the Gila-Salt, these ruins give place to pueblos and cliff-houses, and 
even where there are extensive plains, as in the Little Colorado Basin, 
the construction of ‘great houses’’ like Casa Grande does not appear 
to have been undertaken. The so-called Casas Grandes ruins in 
Chihuahua, however, belong to the same type as the compounds in 
the Gila-Salt Valley of Arizona, although larger and apparently more 
ancient. The environmental conditions of the deserts of southern 
Arizona and northern Mexico, like the ruins, are quite similar. 
Although without kivas, the cliff-houses in the Sierra Madre of 
northwestern Mexico resemble in many features those of Arizona, 
Colorado, New Mexico, and Utah, and stand in somewhat the same 
relation to the casas grandes along the river of the same name as do 
the cliff-dwellings near Roosevelt Dam to Casa Grande. This in- 
dicates the existence of a homogeneous culture, and shows that, 
where similar environmental conditions existed, the inhabitants con- 
structed similar dwellings. By this method of reasoning, the conclu- 
sion is reached that the Sierra Madre and the Arizona cliff-dwellings 
were not derived one from the other, but arose as independent modi- 
fications of a similar culture.’ 
Thus, it would appear that while architecturally there is con- 
siderable difference between the compounds and the pueblos, some 
of the latter may have housed descendants of the inhabitants of the 
former. None of these pueblos, however, are found in the neighbor- 
1 Except possibly those on the San Juan and its tributaries. 
2 There is evidence that some of the oldest Hopi villages were settled by clans from this region. 
3 The circular subterranean kiva, so constant a feature of the cliff-houses of northern Arizona, southern 
Colorado, and the Rio Grande and San Juan drainage, does not exist in the cliff-houses of southern Ari- 
zona, nor in the Sierra Madre in Mexico. This form is not found in the cliff-houses of the Red Rocks, on 
the Verde, or on any tributary of the Gila; it originated in the eastern part of the Pueblo area, and its influ- 
ence was not suflicient to be felt in any prehistoric pueblo on the Colorado River. 
