36 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 68 



constructed stone structures in the region where the pueblo origi- 

 nated. Where there were several clans there were several towers ; 

 when one clan, a single tower. In course of time rooms for habitation 

 or possibly for other purposes, clustered about these towers ; these 

 units consolidated with rooms and kivas of another type forming a 

 composite pueblo. In this form we find the towers rising above a 

 mass of secular rooms. The archaic form of ceremonial room or 

 tower survived in Cliff Palace and other Mesa Verde ruins.^ 



Several circular kivas and towers seen by the author have one or 

 more incised stones, bearing a coiled figure resembling a serpent. 

 One of the best of these has also peripheral lines like conventional 

 symbols of feathers. An obscure legend of the Hopi recounts that 

 the ancestral kivas of the Snake clan, when it lived at Tokonabi, or 

 along the San Juan were circular in form. While at present only a 

 suggestion, it is not improbable that towers and round kivas may have 

 been associated with Snake ceremonials, especially as this cult is 

 known to have survived among Keresan pueblos like Sia and Acoma. 

 The Snake clan of the Hopi according to traditions came from the 

 north or the region of circular kivas. 



From their similarity in external shape and distribution, circular 

 ruins and round towers have been regarded as in some way connected. 

 It by no means follows that rooms inside their external walls were 

 identical in use. For instance, the so-called Great Tower on the clififs 

 overlooking the San Juan, described and figured by Prof. Holmes, 

 is said by him to measure 140 feet in diameter, and to have double 

 walls connected by partitions, forming a series of encircling rooms. 

 This ruin may be classified not as a tower but a circular ruin, and the 

 same may be said of the so-called Triple-wall Tower, rising on the 



' A more extended discussion of towers is reserved for a monograph, now in 

 preparation, on " Prehistoric Towers of the Southwest." The author has 

 made several new observations on these structures some of which differ con- 

 siderably from those of his predecessors. 



Morgan, " Houses and House Life of the American Aborigines " (Contr. to 

 Amer. Ethnol., Vol. IV), has pointed out, page 191, that the round tower at the 

 base of Ute Mountain must have been entered through the roof, as no lateral 

 doorways were visible, and Montgomery's observations on towers in Nine 

 Mile Canyon point the same way. These facts tell in favor of the theory that 

 towers and kivas are morphologically identical, as Morgan indicates. An 

 absence of pilasters on the inner walls of towers indicates that the roof was not 

 vaulted, as in most Mesa Verde cliff dwellings and in the pueblo. Far View 

 House, of the Mummy Lake group. Towers belong to what I have designated 

 the second type of kivas, or those with flat roofs, and are less abundant in 

 the San Juan area. 



