134 CLIMATIC CYCLES AND TREE-GROWTH 



maximum or minimum. Consider, for example, a yearly curve of 

 temperature, low in winter and high in summer. Impress upon this, 

 as we have in Arizona, a summer rainy season which lowers the daily 

 averages and produces a slight summer minimum. The maximum is 

 split and driven each way, but owing to the lag in effects the higher 

 maximum comes in June. If a cycle is split we need to know whether 

 it is the maximum or minimum that changes. If only one changes we 

 get a double-crested curve and if both maximum and minimum split 

 we get a three-crested curve. In the 120 or 130 analyses of western 

 groups, certain cycles, obviously the same in each case, were sometimes 

 found single, sometimes double, and very rarely triple. Hence, it is 

 evident that the comparison of dates of maxima and minima is a 

 complicated process. 



Fourth caution: Interference cycles — If some tree cycles arise, 

 as is possible, from an interference between some external short cycle, 

 say 10.5 months, and the annual seasons, then it is evident that the 

 time of maxima would not necessarily be the same in different geo- 

 graphical locations, for the time of favorable season is different. Com- 

 parison between the northern and southern hemispheres would be 

 needed to settle such cases, for similar conditions in the two hemi- 

 spheres would reverse the cycle. A single curve from Tasmania 

 suggests a split 35-year cycle, with major maximum about 1891 and 

 minor maximum in 1908. In the early Arizona curve the maximum 

 of the 35-year cycle was put about 1900, but in the recent study of 

 western groups this 35-year cycle is usually split into two 17-year 

 cycles whose maxima come in 1892 and 1909, thus agreeing with 

 Tasmania. 



Fifth caution: Cycle centers — In the western zones it was found 

 that each zone had a homogeneous central area with scattering varia- 

 tions about it and that intermediate points, such as the Charleston 

 Mountains, partook of the variations of each zone near it. It is not 

 impossible that we shall find several more central homogeneous areas 

 from which certain typical effects spread out. It is evident that in 

 such conditions many intermediate places will have badly mixed 

 conditions, so that prediction of any kind will become additionally 

 difficult. 



Flagstaff area synthetic curve — The mean curve covering the area 

 from the Grand Canyon to the Rim shows very excellent similarity 

 to the individual curves composing it, but many of the short periods 

 have disappeared and multiples of 7.0 years are left prominent, 21 

 years being by far the strongest. Residuals between the synthetic 

 curve and the real growth-curve show a 9.4-year cycle in the latter 

 part of the eighteenth century. Crests are too high (in the natural 

 curve) at 1793 and 1891 and the minima at 1847 and perhaps 1880 



