484 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1916. 



ruder walls. These considerations throw doubt on the theory that 

 the character of masonry or the amount of erosion are good criteria 

 by which to determine the relative age of pueblo buildings and 

 cliff dwellings. 



The second question, "How old are the cliff dwellings?" is impos- 

 sible to answer, but we know that cliff dwellings were not inhabited 

 in historic times. Comrades of Coronado, in 1540, left no records 

 of cliff dwellings in New Mexico, but Castaiieda mentioned many 

 inhabited pueblos, which is an argument in support of the theory that 

 the latter was a phase of architecture subsequent to the cliff house. 

 No light is thrown by the writer's excavations on relative chro- 

 nology, for no one can tell the age of either type of ruins in the 

 Mesa Verde; nor is it possible to say that the buildings at Mesa 

 Verde are older or younger than some of the Chaco, Animas, Mc- 

 Elmo, or Montezuma Valley ruins, although the walls of the pueblos 

 on top of Mesa Verde are as a rule worn down more by the elements 

 than are those in the valley and therefore of a greater antiquity. 

 The annual erosion of an artificially exposed wall on Mesa Verde 

 is probably about the same in extent as in case of walls in the valleys 

 mentioned. Since, as a rule, the walls of the latter stand higher 

 out of their mounds, the logical conclusion would be that the higher 

 walls are more modern than the lower, but other facts must also be 

 considered before this can be stated as a law. 



In a general way we can explain the supposed later construction 

 of pueblos in the San Juan or its tributaries by the theory that 

 their ancestors lived on the Mesa Verde, and that they left their 

 ancient homes and settled in the Mancos or Montezuma Valley, 

 whence they later spread down the river to distant points, as far 

 as evidences of their culture can now be traced. This would seem 

 9 more natural conclusion, considering all the facts, than the theory, 

 formerly advocated by the writer, that pueblos were developed in the 

 river valleys before the ancients went on the mesa. 



The likeness of the pueblo excavated to the open-air community 

 houses along the San Juan and its tributaries is close enough to 

 indicate identity of culture. As long as we were unacquainted with 

 the essential features of the pueblos on the plateau we were unable 

 to make close comparisons with adjacent cliff houses. The resem- 

 blances of community houses 100 miles distant from the cliff dwell- 

 ings were known to be close, but the pueblo excavated furnishes 

 us with a connecting link in our chain of similarities near at 

 hand and on that account is of preeminent importance in a cultural 

 comparison. 



The resemblance between the pueblo on Mesa Verdfe and those 

 of the McElmo and Montezuma Canyons, and the similarities to 



