1026 
perhaps been the most spectacularly successful one to 
date, and has called forth very great cooperation from 
archaeologists in supplying specimens for analysis. 
Though tree-ring dating of ancient ruins in the South- 
west is continuing and may well be extended to entirely 
new regions, it seems probable that the greatest poten- 
tial contributions of dendrochronology will be in the 
fields of meteorology and hydrology. 
TREE HISTORIES 
Rainfall and River Flow in Western North America. 
Series which contain significant indications as to past 
rainfall and runoff have now been derived for a number 
of localities; for three of these, the Colorado River 
Basin as a whole [25], southern California [27], and 
eastern Oregon [20], the ring data are sufficiently sensi- 
tive and numerous to permit quantitative analysis in 
terms of climate. It must be recognized that no ring 
indices are yet available, nor is it likely that any will be 
derived, which are perfectly representative of the rain- 
fall of any particular locality. Even the best of the 
present indices contain enough “random”’ variation to 
require caution in such quantitative conclusions. Yet 
these indices do show a fidelity to the march of rainfall 
which, in view of the potential refinements, leads to the 
expectation that a vast body of significant climatic 
data will eventually be obtained from selected trees. 
The more comprehensive drought chronologies now 
published are noted in Table I, which also lists the 
longer arctic chronologies of temperature. 
TasLe I. Tree-Rine INnpicEs* 
Number Rings Maximum 
Location and reference number ° ei length 
trees (000) (yr) 
Rainfall Chronologies 
Central California [5-IJ........... 15} 22 3219 
Central Pueblo area (estimated) (to A.D. 
EPA eae dene bee edie er AIS Statler CNY GAAS 150) 
Colorado River Basin [25]........ 240} 60 696 
Eastern Oregon [20].............. 340) +35 668 
Northern Arizona [5-I]........... 86} 14 519 
Oregon-California [3]............. 25 6 476 
Pacific Slope, U. S: [27]....-...... 121) 22 593 
Southwestern Canada [26]........ 66) 15 525 
Southwestern U.S. ruins [28]..... 168} 32 1551 
Western Nevada [16].... ........ +200} +13 685 
Western United States [5-II]..... 305} 52 371 
Temperature Chronologies 
American Arctic [13, 14].......... +250) +30 970 
amilanculiltS| haere. ccs herein 214, 11 55 
Birallevngl [Lio pcs ceoosesncesvontoos 2641 50 
INGER, Il, Zell oooocs eqgascaconor +300} +50 537 
Norway WO) ck: fcassede aye ie 91) 21 477 
Sweden [2] ers ces eit esis ea +300} +50 468 
* Tabulations, in some cases of a number of series, are to be 
found in almost all cited references, supplemented by plotted 
curves. In some reports the total number of rings studied was 
several times the number used in the indices. 
Throughout the West, sensitive timber conifers are 
found to provide histories primarily of winter precipi- 
tation, usually of the October—June interval. For several 
decades it has been generally accepted that the spotty 
convection storms of summer in the Rocky Mountains 
had no apparent systematic influence on ring-width 
CLIMATOLOGY 
fluctuations in these trees, though they were recognized 
as of occasional importance in producing false rings.! 
Most fortunately for chronology building, it was 
found that the drought conifers show a tendency to 
extraordinary longevity, high sensitivity, and slow 
growth on particularly adverse sites. The extremely 
long unbroken records derivable from these trees pro- 
vide the means for estimating the noncyclical trend 
in rainfall, if any, from one century to the next. 
In developing a 658-year index for the Colorado River 
Basin the chronologies in a number of smaller drainage 
basins throughout the West were analyzed, the growth 
relationships in some of them being displayed in Fig. 
2. Panel 1 of this figure is of special interest here, as it 
1890 1900 910 1920 1930 ‘1940 1890 1900 1910 
| ANIMAS |; Ht 
i 
4.SOUTH PLATTE 
a 
1890 1900 1910 1920 1930 1940] 1890 1900 1910 1920 1930 
Fig. 2.—October—June rainfall and water-year runoff (in 
acre-feet) are compared with preliminary growth indices in 
six drainage basins of the Rocky Mountains. Superposed 
smooth curves emphasize the longer fluctuations. 
compares the Mesa Verde growth index, plotted in Fig. 
1 over its entire length of more than 1200 years, with the 
October—June rainfall at nearby Durango and the water- 
year runoff of the Animas River. The degree of fidelity 
and the limitations in the rainfall and runoff record in 
the best tree indices thus far derived are perhaps best 
1. An extensive quantitative analysis now (June, 1951) in 
progress indicates that a small but significant additive influence 
of July-September rainfall on the growth of the following 
year is present in many ring series in this region. The reinter- 
pretation of the numerous centuries-long indices of rainfall, 
now developed for the West, in terms of the total annual rain- 
fall of July—-June, rather than the more limited winter interval, 
may prove far more important climatically than the fact of a 
few per cent improvement in the tree-growth-rainfall correla- 
tions. 
