352 BUREAU OF AMERICAN ETHNOLOGY [bull. 118 



cult than are the western yellow pines. Thus our third problem, 

 that of cross dating, requires application of new techniques. 



Dating by tree-ring analysis depends primarily upon the simi- 

 larity of climate and of weather effects in trees over a considerable 

 area, so that the cross identity of the specimens can be established 

 in every ring. As this cross dating must be developed to a certainty 

 beyond question before dates may be published for southeastern 

 mounds, the publication of these dates is necessarily delayed during 

 a few years of study. Dr. Douglass, working in the Southwest, pub- 

 lished his first date 25 years after the beginning of his study. In 

 undertaking the middle western and southeastern work it was esti- 

 mated that from 3 to 5 years of work would be required before dates 

 could be produced. Of that period, 2 years of part-time study have 

 been covered. Approximately 500 specimens of various woods have 

 been collected from modern trees and from the mounds of eastern 

 Tennessee. Of these, five modern red cedar specimens carry records 

 old enough to extend a master chart back to 1321 A. D. Onto this 

 record the sequences of the best mound specimens are to be matched. 

 Mound specimens come from Sites Nos. 19, 1, and 5, of which Site 

 No. 19 is best represented. 



The techniques to be employed in progressing toward the much 

 desired result, dating the mounds, will consist of measuring ring 

 areas along definite lengths of the ring in the circuit, measuring 

 ring widths, mathematical analysis of their curves, and analysis of 

 their cycles by means of the Douglass cyclograph, and comparisons 

 of all double rings and injury rings, such as may be attributable 

 to late spring frosts. The degree of cross dating of modern trees 

 must be expressed as a comparison for the degree of cross dating of 

 modern and of mound specimens, and hence as a measure of the 

 certainty of dates obtained from the latter. Dates on southeastern 

 mounds are not yet ready for publication, but their publication may 

 be expected within 2 or 3 years of further study. 



In connection with the analysis of tree rings for dating, another 

 aspect of tree-ring records has been studied. Since the modern 

 cedar growth of eastern Tennessee trees has been shown to have a 

 correlation with water year precipitation records of 69.5 per- 

 cent ±3.64, with a standard error of 5.40, and a correlation with 

 water year run-off of 58.7 percent ±1.68, with a standard error of 

 2.49, it is evident that the variations in annual precipitation of the 

 past may be computed with a fair degree of accuracy from the tree 

 growth of the past. The completed analysis of southeastern juniper 

 rings may thus be expected to produce not only the dates of Mound 

 Builder structures, but also material on past weather fluctuations of 

 importance in archaeological, meteorological, and ecological studies. 



