98 



CLIMATIC CYCLES AND TREE GROWTH 



that when the underlining is compared with the amplitude and the A.D. 

 ratio, both the percentage amplitude and the A.D. ratio increase with the 

 stronger emphasis of each additional underline. In every class, however, 

 there is a large amount of scatter about the mean relations between these 

 three quantities. The source of the scatter lies in two factors: First, each 

 of the four or five underline classes covers a considerable variation in em- 

 phasis and hence each estimate of underlines has a personal element. Second, 

 many cycles which were quite strong in the cyclogram yield on summation 

 small amplitudes and A.D. ratios due to interfering cycles; in other words 

 the form of the cycle, the relation of the interfering cycles, and the brevity 

 of the data, have not permitted the satisfactory cancellation of the interfer- 

 ing effects on which the summation process is based. It is evident that the 

 cycle underlining is perhaps fully as important as the percentage amplitude 

 or the A.D. ratio. 



There is evident within each emphasis class a systematic decrease in per- 

 centage amplitude with increase in the number of rows; this arises partly 

 because the number of cycle repetitions is one of the factors considered in 

 assigning underlines. 



The preceding discussion is based on the following tables: 



Cycle emphasis class 



2 underlines, ES 

 1 underline, ES 

 No underlines, ES 



Addit. underl., AED 



Ave. Duration 

 Yrs. 



151 

 165 



169 (for 195 

 cycles) 



Cycle emphasis class 



2 underlines, ES 

 1 underline, ES 

 No underlines, ES 



Frequency 1 



15 

 17 

 13 



Ave. A. D. 

 ratio 



1.43 

 1.30 

 1.18 



Stand. Dev. 



±.23 

 ±.29 

 ±.20 



A. D. Ratio 



Stand. Dev. 



6.2 

 4.5 

 5.9 



Mean Amp. 

 p. cU 



20.3 



13.2 



Percentage refers to mean ring thickness. 



2 This table includes only cases in which the A. D. ratio had been computed. 



RECORDS OF PAST CLIMATES 



The preceding lists have contained the analyses of many tree-ring records 

 that originated in modern times, chiefly since 1750. Those records came 

 from thousands of square miles of area and showed similarity in the general 

 characters of climatic change for that region. Now we deal with the cog- 

 nate idea of extension in time. Appropriate questions at once occur to us: 

 If two areas have agreed for the last 150 years, did they agree for the last 

 1500 years or more? If the discontinuous periods in climate show resem- 

 blance to the sun in the last hundred years while the sunspot cycle was stable, 

 did they show resemblance during the preceding hundred years when the 



