NO. l8 



TREE GROWTH AND RAINFALL CLOCK 



39 



Table 19, as stated heretofore, shows a higher correlation for 1898- 

 1941 than for 1850-1897. This is true for all month-intervals except 

 for March-April which has lower correlations in groups 4, 7, and 9 

 for 1898- 1941. Apparently March- April rainfall had greater influence 

 on tree growth during the earlier period than during the later. Group 

 5 did not conform except in the ratio of opposed trends. An exami- 

 nation of the temperature records readily available gives table 21. 



Obviously, a thermochemical or thermophysiological approach to 

 temperature problems via direct experimental evidence in conjunction 



CORRELATION 

 HOLMAN PASS TREE GROWTH AND RAINFALL 



RAINFALL INTERVAL 



MAR -JUL 



JAN -AUG 



MAR-JUN 



MAY - JUL 



MAY -AUG 



MAY-JUN 



JAN -MAY 



NOV-MAY 



MAR-APR 



.6 .5 .9 .8 .7 .6 .5 



TREND COEFFICIENT 



Fig. 6. — Charted correlations between tree growth of group 7 and rainfall of 

 various month intervals. 



with exact knowledge of growth initiation would give a much better 

 idea of relationships among temperature, soil moisture, and growth. 

 However, table 21 shows that for March, April, and May the average 

 temperatures and the average maxima were somewhat higher for the 

 period 1874- 1897 and progressively lower thereafter except in the 

 case of May when they were slightly higher again during 1910-1930. 

 June follows the same pattern but in a less decided fashion. It is 

 perhaps not illogical to speculate that with higher temperatures during 

 1 874- 1 897 growth began earlier in the spring and was therefore influ- 

 enced by March-April rainfall to a greater extent than after 1897. 

 This may be linked up with the discussion, under the preceding section, 

 of the parallel increase of average variation in dry-site trees and 



