SOUTHWESTERN ARCHEOLOGY — ROBERTS 513 



in type. Cosmos Mindeleff commented on this difference in 1896 and 

 suggested that it was too marked to be attributed wholly to a question 

 of environment.^ Kidder, in 1915, separated southwestern culture 

 into two major divisions on the strength of the dissimilarities,'' and 

 again pointed them out in 1924. In the latter publication, however, 

 with pottery as a criterion, he concluded that in some respects these 

 aberrant sites were allied to the Pueblo ruins.^° Nelson had recog- 

 nized the distinction, and in 1919 indicated it on his diagrammatic 

 chart, although he did not give a detailed discussion of the problem. 

 The situation was not accorded the attention which it merited — actu- 

 ally was overlooked at the first conference — until Gladwin and others 

 working in the district, beginning in 1927 and continuing through 

 subsequent years, obtained definite evidence that the types were dif- 

 ferent. The full import of this did not crystallize at Pecos but at 

 Gila Pueblo, Globe, Ariz., in April 1931, when a classification was 

 drawn up for that division by workers interested in its problems. The 

 results of the Gila Pueblo conference were presented to a larger group 

 of southwestern students at the Laboratory of Anthropology at 

 Santa Fe, N. Mex., in September of that year. The Santa Fe session, 

 which took the place of the biennial Pecos conference in 1931, dis- 

 cussed and adopted the Globe recommendations. There have been no 

 general meetings of that nature since. 



From the knowledge amassed during the many years of inves- 

 tigations and on the basis of understandings reached in various con- 

 ferences, most southwestern archeologists today synthesize the data 

 broadly and briefly as follows: Scattered over the area are the 

 remains of a basic sedentary, agricultural, pottery-making culture 

 which has two major provinces comprising the plateau and desert 

 patterns (fig. 1). The plateau division, which falls under the 

 Pecos Classification, includes the regions of the San Juan, the Rio 

 Grande, the Upper Gila and Salt, the Little Colorado, most of Utah, 

 and a portion of eastern Nevada. The desert domain, summed up 

 by the Globe Classification, occupies the territory extending from 

 the Colorado River on the west to approximately the New Mexico 

 line on the east, from Flagstaff, Ariz., on the north to northern 

 Sonora on the south, with its center lying in the middle Gila Basin. 

 The northern boundary follows roughly the thirty-fifth parallel from 

 the Colorado, swings slightly north to include the Flagstaff section, 

 thence southeastward across Anzona, conforming for the most part 

 to the great diagonal ridge sometimes called the " Mogollon Rim " 

 or the " Verde Breaks ", and continues along the Gila Mountains as 

 far as Safford, Ariz. The eastern boundary extends from Safford 



8 MindeleCf, C, 1896, pp. 186-187. 



» Kidder, 1917. 



»» Kidder, 1924, pp. 105-106, 107. 



