NO, I GROWTH LAYERS IN TREE BRANCHES — CLOCK ET AL, 261 



late in the season just before frost crushed it. Low magnifications 

 give such contacts a very indefinite appearance. 



The Hnkage and relations between the closing phases of regular- 

 season growth and postseasonal growth, where present, determine in 

 large measure the definition of the outer margin of the annual incre- 

 ment. There is, of course, the ideal margin whose appearance in cross 

 section is a smooth curve abruptly separating the narrow, thick-walled, 

 heavily lignified cells of densewood on one side from the large, thin- 

 walled, slightly lignified cells of lightwood on the other. The definite 

 cessation of growth was followed some time later by a decisive, pro- 

 longed, and even an explosive burst of cambial division and cell en- 

 largement. This is the ideal concept of a growth layer, or commonly 

 called "annual ring," and of a boundary between two of them. Un- 

 fortunately, all growth layers are not so well formed nor all contacts 

 so sharply defined. Where conditions fluctuate decisively for pro- 

 longed intervals between those promoting and those prohibiting 

 growth, the appearance and the margins of the growth layers approach 

 the ideal, but where the conditions fluctuate less decisively over rather 

 short intervals on both sides of the conditions separating growth from 

 nongrowth, the appearance of growth layers recedes from the ideal 

 concept and the margins become blurred for any one of several rea- 

 sons. Such indefinite contacts are due either to irregularity of margin 

 or to added growth, both of which are common features in the 

 branches of the trees under study at the extreme lower forest border. 



Investigation of the outer margin of the xylem just under the cam- 

 bium has yielded a mass of information which can be used for an 

 understanding of the contacts between two growth layers in xylem 

 away from the cambium. Briefly, the outer margins of densewood are 

 irregular because ( i ) alternate radial columns of xylem protrude into 

 the cambium which loops over the columns, (2) alternate groups of 

 columns protrude, (3) outer cells in alternate columns or groups of 

 columns enlarge rather than remain narrow as in typical densewood, 

 (4) outer walls of the outer cells remain thin and nonlignified singly, 

 in groups, or collectively. To what extent these features are postsea- 

 sonal is problematical. Undoubted additional, or postseasonal growth 

 takes several forms : ( i ) Isolated large cells here and there under the 

 cambium ; (2) isolated groups of cells in the form of incomplete, im- 

 mature lenses; (3) a zone of postseasonal growth entire over the cir- 

 cuit ; or (4) a typical compact zone of densewood grading outward 

 into cells increasingly larger, thinner walled, and less lignified. Every 

 exterior lens and "outer thin" growth layer is postseasonal growth in 

 a very definite sense. It should be added that the four types of irreg- 



