NO. I GROWTH LAYERS IN TREE BRANCHES — CLOCK ET AL. 43 



Perhaps it is pertinent to emphasize the comparison between the 

 feebleness of the outer densewood of the annual and the strength of 

 the densewood of the intra-annuals (see 1939 of TTC 34-6-a, pi. 14, 

 fig, 2). This emphasis is all the more pertinent because the situation 

 has been encountered repeatedly throughout the work. Failure to 

 lignify normally is well shown in TTP 24-3-b (pi. 19, fig. i) in which 

 lignification has been fairly normal in the densewood of the inner 

 growth layers but has decreased in the outer growth layers until it is 

 present in such small amounts that the densewoods of the outer 

 growth layers are extremely faint and only visible with great dif- 

 ficulty. In some instances there is not only a failure of lignification 

 but also a failure of v/all thickening. The outer cells of the densewood 

 of 1939 and 1941 in Con T 2-0-base bi (pi. 2, fig. i) are light colored 

 and thin walled, but narrow in radial direction. The case of Con T 

 2-7-a has especial interest because the densewood of 1941 is distinct 

 from the earlywood of 1942 over most of the circuit by reason of 

 narrowness of cells only. In different parts of the circuit the cell walls 

 of 1941 densewood and 1942 earlywood may be of equal thickness and 

 equal lignification, or the walls of the earlywood of 1942 may be 

 thicker and more heavily lignified than the densewood cells of 1941, 

 or the cell walls of both may be normal in comparative thicknesses. 

 These comparisons hold true outside of the compression wood. In the 

 case of Con T i-15-a, the contact between 1941 and 1942 is weak and 

 the densewood of 1941 is atypically developed. The outer three or 

 four rows of cells of this densewood are thinner walled, less lignified, 

 but narrower than those cells immediately to their interior. The same 

 set of features is present in the densewoods of the first and second 

 flushes of 1942 in Con T 2-7-b. 



From the above examples, which could be multiplied many times 

 over, it is clear that densewood is not universally a simple constituent 

 of a growth layer, that it has complexities of various types, and that 

 these complexities can detract from the sharpness of contact between 

 the densewood and the following growth layer. 



Features that are chiefly exterior to the contact and that detract 

 from marginal definition are varied and intermingled to such an extent 

 that a simple classification becomes difficult. Therefore, no attempt 

 is made to hold a single example to one category. A seemingly ir- 

 regular rate of cambial activity imparts a ragged outline to a growth- 

 layer margin as seen in cross section. This irregularity is most strik- 

 ingly shown on the outer margin of the xylem just under the cambium. 

 In fact, a rugose outer face to the xylem is nearly, if not quite, as 

 common as a smooth one. Most certainly the cambium, cell for cell, 

 does not cease activity simultaneously throughout its extent. 



