ENVIRONMENT 97 



inches to 16 or 20 feet. The upper trees have rings of modern type, 

 while the lower ones show enormous rings up to a centimeter in size. 

 They exhibit two characteristics which go with larger water-supply 

 than noted to-day in Arizona. The centers of the white parts of the 

 youth-rings show sometimes a softening that gives an effect almost of 

 an abnormal ring. And when the tree is old the red part of the rings 

 is very massive and wide in proportion to the rest, and the ring 

 sequence is subject to characteristic " surges" which are common in 

 European and other wet-climate trees. In this surging there is con- 

 siderable difference between largest and smallest rings, but the change 

 from large to small or the reverse is gradual, so that the mean sensi- 

 tivity is low, though the rings show strong variations. This sort of 

 thing is very different from the habit of the living Arizona trees. 



Certain small white needle-shaped crystals discovered in these 

 ancient stumps were identified by Dr. F. N. Guild (1920, 1921) as the 

 first observed occurrence of terpin hydrate as a natural mineral. 

 On account of the location, it was named " Flagstaffite." 



RAINFALL CORRELATIONS 



If successive years were exactly alike, the rings would all be of the 

 same size, with some alteration with age or injury. But successive 

 years are not alike, and in their differences there are climatic factors 

 which appeal strongly to the tree. In northern Arizona, with its 

 limited moisture and great freedom from pests and with no dense 

 vegetable population, and with the seasonal correlations above de- 

 scribed, this controlling factor is unquestionably rainfall. This is 

 entirely in accord with the rainfall comparisons given below. 



Prescott growth and rainfall — This was worked out to 1908 in 

 Volume I. Its insertion here is to call attention to figure 7, which 

 gives tree-growth and rainfall at Prescott extended to 1923, with a 

 new calculation of rainfall from growth, using the method described 

 in the previous volume. The discrepancies in the last few years 

 probably arises from the error of boring trees too near the roads, as 

 was the case with the recent collection. The calculations and plotting 

 for these curves were done by Mr. D. A. Hawkins. 



Flagstaff tree-records and rainfall — The official Weather Bureau 

 records at Flagstaff began in September 1898. Hence, there are very 

 few years for comparison with tree-growth. A gain has been made by 

 using fragmentary records beginning in 1888 and filling in the deficient 

 months by estimation, using for comparison various records in other 

 localities of northern Arizona, such as Holbrook, Fort Defiance, 

 Prescott, and so forth. Practically all the precipitation after November 

 1 falls as snow, and hence that date is used as the beginning of the 

 year in reckoning rainfall. But even so the total rain does not show a 



