64 AiNNUAL REPORT SMITHSONIAN INSTITUTION, 195 



The Texarkana reconnaissance resulted in the location of 50 arche- 

 ological sites, all of which are open occupational areas. At three of 

 them there are small artificial mounds of the variety which has been 

 called "Capped Ridge." Ten of the sites appear to belong to a non- 

 pottery horizon, probably the Balcones phase. Seventeen are large 

 village areas characterized by potsherds and appear to range in time 

 from Early Gibson Aspect to Middle Fulton Aspect. At least two 

 sites are related to the Coles Creek culture. The remainder are small 

 sites of indeterminate afiiliation. Of the total, 16 sites have been 

 recommended for extensive excavation and analysis. 



Excavations were carried on at the Whitney Reservoir from March 

 6 to June 18. During that period five Indian sites — three rock- 

 shelter and two open sites — were extensively excavated and two 

 historic sites were studied and recorded. One shelter called Picto- 

 graph Cave contained material from two different periods, the first 

 probably dating before A. D. 1200 and the second sometime subse- 

 quent to that date but pre-Columbian. The early occupation is com- 

 parable in many respects to the Round Rock focus in Texas, while 

 the second has not yet been correlated with other remains. The 

 data obtained from the shelter give interesting information pertaining 

 to changes in diet and population density during the two periods of 

 occupation. The second, known as Buzzard Shelter, is not far from 

 the first, and also gave evidence of an early occupation in the lower 

 depths of the fill. The later occupation in the shelter suggests certain 

 similarities to that of the Toyah focus. While there is considerable 

 similarity between the cultural sequence found in the two shelters, 

 there are specific differences in artifact types and stratigraphic pro- 

 portions. The third shelter, known locally as Sheep Cave, is the 

 largest of the three, and the material from it agrees in the main with 

 that from the other two. Five flexed burials were found there, how- 

 ever, and study of the physical type represented should throw some 

 light on the relationships of the people. 



Three weeks were spent in the excavation of a small occupational 

 area on the second terrace of the Brazos River at the Steele site. 

 The evidence of occupation on the surface covers about an acre in 

 extent and it is underlain by an unlmown number of occupational 

 levels of considerably greater extent. Traces of occupation extend 

 to a depth of at least 15 feet, and it will be necessary to use mechanical 

 equipment to excavate a deep trench in order to make stratigraphic 

 studies. The site appears to be a significant one in that the most 

 recent occupation was prior to the advent of pottery and the bow and 

 arrow in that area. 



The Stansbury site, the location of a historic Indian village, was the 

 fifth area excavated. Material from it includes trade items of French, 



