50 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I40 



several rows of cells in the xylem between the outer margin of 1939 

 and the frost injury of 1940 are larger, thicker walled, and more 

 deeply lignified than the cells of the 1939 densewood. Farther out- 

 ward, immediately inward from the frost injury, these xylem cells 

 become still larger but are thinner walled and less heavily lignified. 



It may well be that growth in the Conservatory location caused the 

 diffuse margin of 1939 because we do know that some xylem was 

 formed while the tree was kept in the Conservatory. The lateness of 

 the season is to be noted, and this is an item in support of the thesis 

 that trees grow whenever conditions are favorable and that their 

 growth is not necessarily confined to one interval per year. Further, 

 it is to be noted that the Conservatory habitat does not directly explain 

 the diffuseness and variability of the 1940 margin. 



Another feature having to do with the transition from densewood 

 outward to lightwood is the alternation radially of sharp and diffuse 

 portions of contiguous growth layers. In TTC 5-7 all the contacts for 

 six years have diffuse margins over short arcs of circuit. The outer 

 margin of 1938 in TTC 5-7-b, for instance, is diffuse for more than 

 one-half its circuit ; it is diffuse on that part of its circuit where, 

 radially, neighboring contacts are sharp. In TTC 5-8-b and 33-9-a, 

 the increments for 1939 possess three growth layers each. The two 

 intra-annuals in each case can be described as partly sharp, but other- 

 wise complete, entire growth layers. Where one intra-annual is dif- 

 fuse, its neighbor radially adjacent to it is sharp. If one were re- 

 stricted to two radii either 90° or 180° apart on these sections, he 

 would have difficulty not only in correlating the sequences but also in 

 making sound interpretations based on growth-layer thicknessess. 



The matter of annual contacts showing diffuseness bears so directly 

 upon all growth-layer problems that at least two clear-cut examples 

 must be cited. In TTC 12-io-a, the outer margin of 1939 varies from 

 sharp through indefinite to diffuse. Were it not for methods of ab- 

 solute dating, one would judge 1939 to be intra-annual if he were 

 committed to the criterion of sharpness as belonging to annual in- 

 crements alone. The case is more decided even in XSC 11-2-a (pi. 30, 

 fig. 2) where the outer borders of 1937 and 1939 are weak contacts — 

 they are indefinite. Hence, there are fewer sharp growth layers than 

 there are years; the 5-year interval, 1937 to 1941, contained three 

 sharp growth layers. The import of this dearth of sharp growth 

 layers is quite obvious. Our notes for XSC 12-i-a (pi. 32) record 

 eight completely sharp growth layers for 5 years and then continue: 

 "This is the first specimen in more than a dozen that has a surplus of 



