20 BUREAU OF AMERICAN ETHNOLOGY [bull. 41 



and in one or more of the Hopi kivas. The phistering of the walls 

 was placed directly on the masonry. 



It is probable that the kiva walls were painted with various devices 

 before their roofs fell in and other mutilation of the walls took place. 

 Among these designs parallel lines in white were common. Similar 

 lines are still made with meal on kiva walls in Hopi ceremonies, as 

 the author has often described. One of the pedestals of kiva A is 

 decorated with a triangular figure on the margin of the dado, to 

 wdiich reference will be made later. 



The author has found no conclusive answer to the question why 

 the kivas are built under ground and are circular in form. He be- 

 lieves both conditions to be survivals of ancient " pit-houses," or 

 subterranean dwellings of an antecedent people. In this explana- 

 tion the kiva is regarded as the oldest form of building in the cliff- 

 dwellings. We have the authority of observation bearing on this 

 point. Pit- dwellings are recorded from several ruins. In a recent 

 work Dr. Walter Hough figures and describes certain dwellings of 

 subterranean character that are sometimes found in clusters," while 

 the present author has observed subterranean rooms so situated as to 

 leave no doubt of their great antiquity.^ 



The form of the kiva is characteristic and may be used as a basis 

 of classification of pueblo culture. The people whose kivas are cir- 

 cular inhabited villages now ruins in the valley of the San Juan 

 and its tributaries, in Chelly canyon, Chaco canyon, and on the west- 

 ern plateau of the Rio Grande. 



The rectangular kiva is a structure altogether different from a 

 round kiva, morphologically, genetically, and geographically. It 

 is peculiar to the southern and w^estern pueblo area, and while of later 

 growth, should not be regarded as an evolution from the circular 

 kiva. Several authors have found in circular kivas survivals of 

 nomadic architectural conditions, while the position of these rooms, 

 in nearly every instance in front of the other rooms of the cliff- 

 dwelling, has led others to accept the theory that they were later 

 additions to the village, which should be ascribed to a different race. 

 It w^ould seem that this hypothesis hardly conforms to facts, as some 

 kivas have secular rooms in front of them which show evidences of 

 later construction. The strongest objection to the theory that kivas 

 are modified houses of nomads is the style of roof construction. 



Kiva A 



This room (pi. 13), which is the most northerly of all of the 

 ceremonial rooms of Spruce-tree House, is, the author believes, the 



" Bulletin 35 of the Bureau of American Ethnology, Antiquities of the Upper Gila and 

 Salt River Valleys in Arizona and New Mexico. 



* In some cases the walls of the later rectangular rooms are built across and above them, 

 as in compound B in the Casa Grande group of ruins. 



