46 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I40 



A parenthetical statement about compression wood should be made 

 at this point. Ordinarily it would arouse no comment so far as this 

 study is concerned. However, the three growth layers of 1939 in TTC 

 33-13-b show an interesting similarity between intra-annuals and an- 

 nuals (see pi. 13). The densewood of the inner growth layer on one 

 radius has a sharp contact under low power whereas under high power 

 the densewood cells show themselves to be merely compression wood. 

 On the middle growth layer the outer contact is sharp except where 

 followed outward by compression wood. High power shows the 

 densewood to be present and sharply bounded completely around the 

 circuit. The outer growth layer has a sharp outer contact except 

 where followed by compression wood. Where this is true of the outer 

 growth layer, the middle growth layer just to the interior is sharp; 

 where the intra-annual is followed by compression wood, the outer 

 contact of outer 1939 is sharp and followed by a normal sequence of 

 lightwood. 



Atypical lightwood, whether entirely absent or merely poorly de- 

 veloped, not only impairs marginal definition but also, and what is 

 even more important, reveals information concerning cambial activity. 

 The densewood of one growth layer may merge over an arc of five 

 degrees or less with the densewood to the interior, as in the case XSC 

 2- 1 -a where the densewood of the outer growth layer of 1932 "dips 

 inward" to make contact with the densewood of the middle growth 

 layer of 1932 at a sinus (see pis. 14, figs, i and 2 ; 19, fig. i ; 29; 31, 

 fig. I ) .^ Or again, the lightwood may be absent over a much greater 

 portion of the circuit, as in TTP 23-i-a (see pis. 17, fig. i ; 25) 

 where the densewood of 1941 lies back against that of 1940. For a 

 short distance the densewoods of both 1940 and 1941 are flush against 

 that of 1939, the microscope showing that only the lightwoods are ab- 

 sent. The sections of TTP 24-14-a (pi. 19, figs. 2 and 3) show a more 

 extreme case in which the densewoods of two growth layers are in- 

 distinguishably in contact except for a distance of 10° to 15° where 

 lightwood and an injury are inserted (contrast "added on" lens of pi. 

 24, fig. 2). Theoretically, the final step should exist wherein light- 

 wood is totally absent around the circuit of a cross section. A method 

 of internal tagging or serial sections would be required to distinguish 

 between two diameter flushes totally lacking in lightwood or to trace 

 them longitudinally into a region where lightwood is inserted. How- 

 ever, longitudinally detached cross sections which show a progressive 

 decrease and final disappearance of lightwood indicate to us that the 

 theoretical condition actually exists. 



3 Clock, 1937, pi. 2A, shows an identical instance of merging growth layers. 



