TREE RINGS AND HISTORY — SCHULMAN 461 



the year's ring, is commonly found to be essentially over by the end 

 of July. Since the tree was sampled June 6, 1950, the ring for 1949 

 is complete ; that for 1950 had not yet started on this core. Similarly 

 the Tucson core (loAver), taken May 22, 1951, ends with the 1950 ring. 

 The Mesa Verde core (center), on the other hand, was cut later in the 

 season, October 2, 1951, and therefore has the 1951 ring complete. 



That the same general pattern of thick and thin rings may be 

 found in selected trees over a fairly large area is also illustrated 

 in the photographs. These ring series represent not only the suc- 

 cession of good and poor growth years, but also, as will be pointed 

 out below, a significant record of excessive rainfall or drought for 

 the entire year preceding the end of the growth season. Anticipat- 

 ing these conclusions, we note some interesting details of climatic 

 history in plate 1. Although the ring growth in individual trees 

 must, even at best, be subject, to a considerable extent, to purely 

 local or random influences, certain outstanding climatic events in 

 the Southwest are well shown. The unusually snowy "haylift" winter 

 of 1948-49 in northern Utah-Nevada resulted in the very thick 1949 

 ring in the Nine Mile tree (upper) ; the rainfall excess was less pro- 

 nounced to the south, as the Tucson core suggests (lower) . The wide- 

 spread drought of the "dustbowl" year 1934 is represented by a 

 very narrow ring in all three areas. The general deficiency in growth 

 in the outer two decades or so in the more northerly areas of the 

 Colorado River basin, and since 1920 in the Gila River tributary 

 basin, represents a rainfall fluctuation of major economic importance. 



Further, 1904 is generally represented by a narrower ring than 

 1902 in southern Arizona, in correct relation to the rainfall; but 

 in northeastern Utah the trees, like the rain-gage records, show the 

 July 1901-Junc 1902 interval to be the drier year. The flow of the 

 Colorado River at the Grand Canyon is largely subject to the rain- 

 fall in northern Colorado and Utah; the water year 1902 (ending 

 September 30) had the minimum flow of record preceding 1931. 



Occasionally, us m 1931, ring growth in a year of over-all drought 

 may not be correspondingly small, even in the most sensitive trees. 

 Further, the low rainfall of the 1898-99 season and the correspondingly 

 narrow ring which is observed over most of the Colorado River basin 

 are in great contrast to the very heavy main-stream runoff for 1899, 

 the result of an unusual concentration of heavy vSnows in the head- 

 waters area. The importance of a sufficiently dense and widespread 

 sampling network is thus obvious. 



DENDROARCHEOLOCY 



The method of overlapj^ing patterns by means of which prehistoric 

 beams may be dated is illustrated in highly idealized form in figure 1. 



