NO. 19 CYCLES IN TREE-RING WIDTHS ABBOT 5 



mean form of the even-numbered 46-year cycles in the upper curve. 

 It will be observed at once that they dififer in that the upper curve 

 shows a strong 23-year cycle additional to that of 46 years, while in 

 the lower curve this feature is missing. This, of course, is but to 

 emphasize the fact already brought out above, namely, there is a 

 strong 92-year cycle besides those of 23 and 46 years. 



I suspect that meteorologists are rather inclined to distrust the 

 origin and existence of the 23-year cycle which was indicated in 

 my former paper as arising from the variability of solar radiation 

 and expressing itself in weather and many phenomena depending 

 thereon. Their hesitation is no doubt due to the smallness of the 

 percentage and the absence of a theory of the solar variations, and 

 to the reversals of phase which occur in the terrestrial manifestations 

 of the subordinate solar periodicities, as strikingly illustrated in 

 figures 16 and 17 of my paper "Solar Radiation and Weather 

 Studies ". I have still nothing to offer but statistical proofs of the 

 important influences of the many periodicities integrally related to 

 23 years, l^it T feel somewhat supported in my conclusions by the 

 recent paper of E. A. Cornish ' on the precipitation at Adelaide. 

 Australia, and the unreserved acceptance of this period as real by 

 Dr. F. J. W. Whipple while he criticized the method of Mr. 

 Cornish in presenting it. 



In closing, I may draw attention to an important inference from 

 this study of Professor Lyon's measurements of the tree-ring widths 

 at Fairlee, Vermont. The 46-year cycle seems to be clearly in evi- 

 dence since the year 1544, and covering eight recurrences of it. 

 Counting back from the year 1911 when the eighth cycle ends (see 

 curve D, fig. i ) , we find that the low phase, corresponding no doubt 

 to deficient water supply, ended about 1896, Now referring to figure 

 26 B of my paper " Solar Radiation and Weather Studies ", the low 

 l)hases of Lakes Huron and Erie also ended about 1896, and also 

 46 years previously thereto, at the ends of the great droughts of the 

 forties and the nineties of the nineteenth century in our Northwestern 

 States. Does not this long-continued series of 46-year precipitation 

 cycles at Fairlee plainly warn us of the probable recurrence of a major 

 drought beginning soon after the year 1975, and lasting about 10 

 years ? 



"" Quart. Journ. Roy. Met. Soc, vol. 62, p. 481, 1936. 



