350 BUREAU OF AMERICAN ETHNOLOGY [»u^. H8 



and Archaeology of the University of Kentucky and in charge of 

 the archaeological excavations carried on for the Tennessee Valley 

 Authority in the Norris Basin, wrote to suggest that the work be 

 initiated in the area to be covered by the Norris Lake, as all the 

 standing timber was being cut. This would permit the collection of 

 as many specimens as might be desired from a number of different 

 species of trees. Moreover, settlements in this district were known 

 to have been earlier than in most other regions of the South and the 

 Middle West, and the log cabins of the pioneers still stood, for the 

 most part, inhabited. These cabins which stood below the level of 

 the future lake were to be razed, and we were welcome to sections 

 from the logs. As this program of land clearing offered an unusual 

 opportunity for opening our program, we accepted the invitation 

 and left for La Follette, Tenn. Ten days were spent in La Follette 

 in the collection and examination of specimens. Instructions were 

 given in preservation of the partially decayed wood being taken 

 from the post holes and from the steps of ancient mound structures 

 and grave coverings. Arrangements were made for the collection of 

 several hundred specimens of modern woods, the work to be done 

 by a field man in collaboration with Mr. Holley and his crews, who 

 sent in reports of large trees being felled. Our remaining 4 weeks 

 were spent collecting and studying oaks in an area in which the 

 University of Chicago was interested in dating the mounds they 

 excavated. Part-time work was carried on for the Tennessee Valley 

 Authority during the rest of the year, and the summer of 1935 was 

 spent with the Authority in Knoxville. 



The immediate result of the initial work on eastern Tennessee 

 trees has been a partial answer to several problems. Our principal 

 question regarding tree-ring work in the Mound Builder area was in 

 regard to the humidity of the area. In the arid Southwest, pines, 

 pifion, and Douglas fir carry records sensitive to variations in annual 

 precipitations. The curve of their growth records gives a high corre- 

 lation with precipitation records for the same area. Within an area 

 the records of all normal specimens of a species cross date, or agree 

 with each other in their pattern of successive large and small rings. 

 While the actual width of 1 year's growth will vary from tree to 

 tree, dependent upon the drainage and soil conditions, the relative 

 ring widths in the pattern of one tree have been found to agree with 

 those of the patterns of other trees through the same period. The 

 patterns of ring growth have been tested at various heights upon a 

 tree and at various radii on the cross section of a tree. All tests 

 indicate uniformity of relative annual ring width in the growth 

 patterns of normal uninjured trees. 



Trees growing too near water to be dependent upon seasonal 

 precipitation do not show sensitive records, however; with a con- 



