NO. I GROWTH LAYERS IN TREE BRANCHES — CLOCK ET AL. 5 1 



sharp growth layers rather than an equahty or dearth for the number 

 of years involved." 



The special features that impart sharpness to a contact are consid- 

 ered, perhaps unconsciously, to reside in the densewood alone. Not 

 infrequently the worker is brought up short by contacts showing ap- 

 parently a transition from densewood out into lightwood only to real- 

 ize by close study that the lack of contrast is due solely to atypical 

 lightwood. The preceding densewood is as it should be ; the succeed- 

 ing lightwood fails to give the expected contrast. If the cells do not 

 enlarge, if the walls thicken somewhat, or if lignification is excessive, 

 the sharply distinctive contrast with the densewood fails and, espe- 

 cially under lower powers of magnification, the one growth layer 

 seems to merge into the other. This is illustrated by the outer contact 

 of 1939 in TTP 21-3-b where the margin lacks contrast because in the 

 lightwood of 1940 the cells did not become large. The growth layers 

 of TRSp i-i-a appear to be indefinite over portions of their circuits 

 because the lightwood fails to contrast with the preceding densewood. 



The lack of contrast in lightwood can become somewhat com- 

 plicated, as in 1939 of TTP 21-2-a. Over three- fourths of the circuit 

 an exterior lens is immersed in what one would call normal densewood 

 under low power, A radial sequence passing outward through the lens 

 is as follows: (i) about seven cells of decreasing widths; (2) from 

 one to three cells exactly like the previous densewood except that their 

 radial widths are somewhat greater; and (3) from one to four nar- 

 row cells forming the outer part of the densewood of 1939. The in- 

 definiteness at the inner margin of the lens is not so much a matter of 

 the outer margin of number ( i ) above — this outer margin is actually 

 quite definite — as it is rather a lack of contrast between the narrow 

 cells of (i) and the only slightly wider cells of number (2) above. 



So-called "curtains" give an appearance of diffuseness to the growth- 

 layer margin outside of which they lie (pi. 9, figs, i and 2). In XSC 

 2-3-b, the outer contact of 1938 is highly indefinite; it does not bear 

 those decisive characteristics to be expected in an annual increment. 

 The growth layer for 1939 (see also pi. 19, fig. i) is in general more 

 heavily lignified than that of 1938 and, in fact, over a great share of 

 its extent it seems to be merely a curtain added on to 1938. Therefore, 

 in XSC 2-3-b we have an example of an annual increment subsiding 

 from the rigid characteristics considered normal for such an incre- 

 ment in contrast with intra-annuals which rise to, and maintain the 

 level of, annual increment characteristics. 



Perhaps it is pertinent at this point to mention how transitional 

 margins can vary in their intensity and thus throw doubt on the best 



