128 



CLIMATIC CYCLES AND TREE-GROWTH 



accord again till 1617, and from there on it decreases its accuracy, and 

 the variations typical up to that point disappear from the curve. 

 The curve from 1700 on shows much less of the sunspot variations, 

 but in the Grand Canyon group, one of its components, and others 

 also, the half cycle shows well from 1850 to the time of collecting and 

 with almost the early regularity. Brief calculations show that the 



1800 



1900 



1950 



60 



80 



Fia. 18 — (1) Flagstaff area mean curve, FAM; average of eight groups, 

 standardized and smoothed; (2) synthetic curve; (3) residuals 



long variation in the earlier curve agrees exactly in phase with the 

 recent years, and so we find through practically 600 years a mean 

 value of the sunspot cycle of 11.30 ± 0.02 years. From the correlation 

 diagram already referred to on page 104 of Volume I, we see that the 

 most direct relation between the double-crested growth-curve of the 

 above stated length and the single-crested sunspot curve is that a 

 growth maximum occurs at the time of a sunspot minimum. On 



