154 CLIMATIC CYCLES AND TREE GROWTH 



the cycles in lot-drawn curves. As a comparison, the periodogram numbers 

 of the analyses of 22 dated Coast Redwood specimens were correlated with the 

 numbers for 19 undated specimens. The correlation coefficient was 0.56 

 ± 0.08, strong evidence then that the cycles found are present in the trees 

 concerned and not accidental in origin. The correlation coefficient for a 

 similar comparison of the more homogeneous Central Pueblo Area curves 

 was 0.76 ± 0.05. As a check, the correlation coefficient was obtained be- 

 tween the periodogram numbers representing analysis of two additional 

 random groups. The value was 0.05 ± 0.11, again indicating random dis- 

 tribution of cycles in lot drawings. 



THE PRESENTATION OF CYCLE DATA 



In mass presentation of cycle data, as given in the lists of Chapter V, it 

 is impossible to represent all the peculiarities of the individual cycles. The 

 summation curves following this section supplement the lists but often do 

 not reveal significant detail seen in the cyclogram. It is hoped that it will 

 eventually be possible to publish cyclograms to accompany all cycles men- 

 tioned. The column titles in the tabular matter referred to are explained 

 below. 



Number — The numbers in this column refer to the summation curves 

 (pp. 156). The latter are placed in order of increasing cycle length, which is 

 the arrangement of the long chronology lists. 



Cycle — This column gives the exact cycle length used in the amplitude 

 computation. Since values of the cycle observed are subject to the inaccura- 

 cies already discussed, differences between two cycles of 2 to 3 per cent may 

 not mean that they are different ones. For example, 13.6 and 13.8 may really 

 represent the same cycle. Often a change in cycle length of 2 per cent or 

 more will be found when the assigned duration is changed somewhat. 



Occasionally a cycle is seen as two-crested, that is, with one set of subor- 

 dinate intermediate maxima; the notation in the lists, oc\, follows the cycle 

 length. Two subordinate crests are noted as oc\, and so on. If every other 

 maximum in an alignment is especially strong, indicating the presence of 

 the double-length cycle, it is noted in the lists as x2 (times 2). This para- 

 graph refers particularly to the cycle tables of the monthly sunspot numbers 

 and the geological material. 



Underlines — The significance of cycle underlines is discussed on page 146 

 and on page 147. 



Duration — The dates just preceding the first maximum and following the 

 last one were obtained at time of analysis. Dates were subsequently altered 

 by as much as three or four years at beginning or end, to permit an integral 

 number of rows in summation for amplitude. It is important to remember 

 also that many cycles do not appear or disappear suddenly; hence caution 

 is necessary in considering these dates as precise. 



First Maximum — The position of first maximum as given is that of the 

 highest part of the summation curve. Many cycles have two or more crests, 



