94 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I40 



sequence, the lightwood being in the center and densewood lying both 

 interior and exterior to it. No curtain recognizable as such exists in 

 sections a, i"] cm. inward, or in section c, 15 cm. outward from h. 

 Either it is entirely absent from these sections or it is represented by 

 some other type of growth layer. Perhaps the most interesting situa- 

 tion exists in the increments for 1934-1935 of XSC 6-2. The sections 

 for XSC 6-2-a, 117 cm. from the tip of the branch, show two curtains, 

 one of which could actually, and perhaps more logically, be interpreted 

 as two thin, sharp, complete, entire growth layers, and the other of 

 which changes laterally into a sharp lens (as does the curtain at start 

 of 1939 (pi. I, fig. 2). These curtain effects are even more conspicuous 

 in sections &, 100 cm. from the tip, and are excellently developed in c, 

 85 cm. from the tip. In fact, XSC 6-2 is remarkable not alone for its 

 curtains but also for its multiplicity ; for its concurrent lenses ; for a 

 narrow, weak, "outer" growth layer which constitutes the outermost 

 growth layer of a multiple annual increment; for thin densewood 

 bands one cell in thickness; and for multiple divided densewoods. 

 The increment for 1935 in XSC 6-3-a contains two curtain growth 

 layers. One of these changes laterally into a growth layer with mul- 

 tiple densewood which, in turn, becomes a sharp lens. 



The conditions present when a curtain is formed must resemble 

 those present when densewood is formed — inhibited cell enlargement, 

 accentuated cell maturation, and increased deposition of tannins, res- 

 ins, and lignin. During the formation of a curtain, then, these condi- 

 tions are present at the start of growth-layer formation as well as at 

 the close. One wonders if the rate of cambial activity is less than for 

 normal lightwood. In XSC 2-2-c, the curtain in 1938 is definitely 

 related to late-spring frost injury which indicates restricted cambial 

 activity. Not the frost effects, particularly, but the restricted cambial 

 activity may be highly suggestive as to the reason for the existence of 

 curtains. 



The presence of curtains, of course, detracts from the sharpness of 

 the growth-layer boundary against which it lies. This is increasingly 

 true, the lower the power of magnification used. 



In some instances, curtains could be more accurately classed as com- 

 plete growth layers than as partial growth layers. 



Postseasonal growth as a term is applied to any growth which pro- 

 duces an incomplete growth layer, however restricted, after the major 

 growth flush (or flushes) has been completed. The formation of post- 

 seasonal growth may occur, theoretically and practically, at any time 

 during the calendar year except for that part of the growing season 

 which is the chief period when diameter growth occurs. Of course, it 



