86 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I40 



increment of TTC 34-5, divided densewood is present in sections a 

 and b (50.5 cm. and 42.5 cm. from the tip) but not in sections c and 

 d (35.5 cm. and 27.5 cm. from the tip). In the 1939 increment of the 

 same branch a complete, entire growth layer in sections a becomes a 

 lens in b and c and divided densewood in d. In the 1941 increment 

 of TTJ 2-12, sections a (46 cm.) contain a complete, entire growth 

 layer, a lens, and concurrent divided densewood, whereas b, c, and d 

 (35 cm., 33 cm., 22 cm.) contain only a complete, entire growth layer. 

 In the 1939 increment of TTP 24-3, the complete, entire growth 

 layer of a, b, and c (42, 35.5, 30 cm.) becomes divided densewood in 

 d (20 cm.). In the 1940 increment of XSC 9-6, a complete, entire 

 growth layer of a (52 cm.) becomes divided densewood in b (46 cm.) 

 and remains so at least to within 29 cm. of the tip. In the 1938 in- 

 crement of XSJf i-i, divided densewood is present in a and b (65 

 and 50 cm.) but absent in c (36 cm.). In the 1938 increment of XSP 

 i-i, a lens of a (49 cm.) becomes divided densewood in b (33 cm.) 

 and has disappeared outward by the time c (19 cm.) is reached. 

 Equally vivid and numerous are the examples where complete, entire 

 growth layers, or lenses, dwindle inward on the branch and ultimately 

 disappear. 



Interesting situations exist in some branches because the increment 

 for one year has increasing complexity inward, whereas that of 

 another year has increasing simplicity inward as if the years in any 

 one branch were complementary to each other. 



Divided densewood, if it is detected, can give a clue to the presence 

 of a very weakly developed growth layer which otherwise would be 

 considered to be nonexistent. In TTP 24-3-a the intimate contact of 

 the densewoods of 1938 and 1939 and of the interior concurrent lenses 

 of 1 94 1 with the densewood of 1940 resemble closely what has been 

 called divided densewood. At first, such densewood was thought to be 

 of little importance. Later, it was found that divided densewood, and 

 even undivided densewood, could represent two distinct growing sea- 

 sons. The so-called invisible continuation of a lens around the circuit 

 is revealed at places by a slight separation of the densewood where a 

 row of wider tracheids breaks the normal outward sequence of pro- 

 gressively narrower cells. In certain instances, as TTP 24-3-a, the de- 

 tection of an annual increment depends upon such slight evidence 

 that it easily escapes notice. It is there, nonetheless. 



The outer densewood of 1940, in TTC 33-ii-a (text fig. 30 ; pi. 12, 

 fig. 2), is multiple by the insertion of one or more rows of lightwood 

 cells between the bands of densewood. In other words, the densewood 

 of 1940 is made up of three bands over half the circuit and two over 



