CYCLOGRAM ANALYSIS 



25 



small and microscopic rings. Extra large rings and frost or other injury- 

 effects are noted also; they sometimes help. A standard time scale is pre- 

 pared along a horizontal axis, and the deficient rings are marked in their 

 proper relative places, with solid ordinates made longer and longer as the 

 rings are found more and more minute or absent. It is usually very easy 

 by comparing two such plots to test cross-identity in two trees and show 

 whether they were living trees at the same time. Of course, one of two plots 

 compared may represent a well-established chronology with its dates all 

 known, and the other may be from a log picked up in an ancient ruin. 



Mr. H. S. Gladwin, Director of the Gila Pueblo at Globe, Arizona, felt 

 that too many rings were omitted in this form of plot and has proposed an 

 increment plot that promises to be of important use. The increment from 

 each ring to the next is measured and inserted on the usual time scale in two 



60 



70 



80 



90 



1900 



10 



-J^^ 



Original measures 



Standardizing line 



2.0 



1.0 



Standardized measures 

 G5 Eberswalde, Germany, group 

 Fig. 12 — Standardizing. 



colors; pluses are one color and minuses are another, and such colored lines 

 are proportional in length to the increments. Thus nearly every ring is 

 represented. Averages may be taken and variable rings sometimes have 

 two colors on them. Cross-dating by this method becomes very strong. Its 

 successful application is even more dramatic to the student than the ordinary 

 skeleton plot, but its field use is curtailed by the need of measurement. 



Standardized curves are curves equalized or brought to a uniform mean 

 value so that one tree record with large average growth shall not dominate 

 other records of small mean growth. If, for example, we have ring records 

 that differ greatly in mean ring size and if we plot the record of each tree 

 separately and analyze it for cycles, we shall find the cycles existing both in 

 the small-growth records and those in the large-growth records as well. 

 But if we average the curves together and analyze the mean, we get the cycles 

 in the large-growth records and lose some or most of those in the small-growth 



