KIDDER-GUERNSEY] ARCHEOLOGICAL EXPLORATIONS IN ARIZONA 201 
also that toward the end of the period of occupancy of this district 
the population withdrew to the vicinity of Marsh Pass, where the 
culmination of the culture, so to speak, was reached in the pueblos of 
the pass and the great cliff-houses of its tributary Sagi Canyon.' 
Regarding the culture as a unit, it may be assigned a position as a 
subgroup of the great, northeastern Kiva-culture.? That branch of 
southwestern civilization has not yet been clearly delineated, but it 
appears to have comprised all the true cliff-dwellings and pueblos of 
the San Juan drainage, with outposts running down and across the 
Colorado and, in somewhat later times, down the Rio Grande. At 
present, in addition to the division under discussion (which we may 
term the Kayenta), we can recognize in the San Juan district two 
definite subgroups: Chaco Canyon and Mesa Verde-McElmo. The 
position of two other groups is less certain: The Montezuma Creek, 
which should perhaps be classed with Mesa Verde; and the Aztec— 
Bloomfield, whose ruins, so far as we can tell from surface indica- 
tions, are allied architecturally to Chaco Canyon and ceramically to 
Mesa Verde. 
This very general classification of San Juan sites leaves unac- 
counted for the great and important mass of remains in the Canyon 
de Chelly and the lower Chinlee. Of these the authors have no per- 
sonal knowledge beyond their very brief examination of the Nockito 
cliff-house. 
The Kayenta group differs from the others most strikingly in 
pottery and in kiva construction. Its redwares are much more 
abundant than are those of any other San Juan region, the poly- 
chrome redware being, so to speak, its trade-mark. Black-and-white 
also differs in its shapes (handled bowls, colanders; lack of mugs and 
pitchers) and decorations (“heavy ” designs, etc.) from that of the 
other groups. 
The kivas are characterized, to permit ourselves a paradox, by 
their lack of character; they do not show the strict orthodoxy of 
form displayed by the six-pilastered Mesa Verde and Montezuma 
Creek type,* or of the low-benched style which appears to belong to 
the larger Chaco Canyon ruins.* As was brought out in the text, 
they vary greatly in size and interior arrangement, some having small 
recesses, some very large ones, many none at all. 
1This process—i. e., early diffusion in small sites, later concentration in large centers 
with high cultural specialization, and lastly more or less abrupt abandonment of whole 
regions—is a common phenomenon in southwestern archeology. Examples are: Chaco 
Canyon, Mesa Verde, Lower Gila, Casas Grandes. It has not yet been satisfactorily 
explained, though an attempt to account for it on the basis of climatic change has 
been made by Huntington (1914). 
2 Kidder, 1917. The arrangement there given is somewhat different from the present 
one; final classification is not yet possible. 
3 See Fewkes, 1908, 1909, 1911, a; Morley, 1908; Kidder, 1910. 
4Pepper, 1899. 
