272 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I40 



The last phase of anatomical problems has to do with what we have 

 called lenses. Such a growth layer has been called "locally absent" 

 (Douglass, 1939, p. 19), referring specifically to that part of the 

 growth layer which is not there. "Locally absent in places" (Douglass, 

 1939, p. 19) has been used in the same way. Elsewhere, Douglass 

 (1938, p. 8) has used "often absent," "sometimes absent," and "oc- 

 casionally absent" in reference to the total absence of the growth layer 

 from the core or cross section in hand. Of course, these may be syn- 

 onymous with "locally absent" insofar as the remainder of the tree is 

 concerned. In contrast, Clock (1937, pp. 8-10, 48-51) referred to a 

 lens as a "locally present" growth layer. His use of "missing" and 

 "absent" rings refers only to the specimens in hand. 



A high percentage of the lenses examined in the present work 

 proved to be lenses only insofar as the lightwood is concerned, the 

 densewood being continuous around the circuit in immediate contact 

 with the densewood of the previous growth layer. The same thing had 

 been noted previously by Douglass (1928, p. 32) and by Clock (1937, 

 p. 73), no doubt under lower magnifications. 



In the present work, out of some thousands of examples, all lenses 

 were proved to be intra-annual with the exception of less than half a 

 dozen. The lenses of the exceptions were in all cases the outermost 

 growth layers lying just under the cambium in the position which 

 would be occupied by temporary lenses. No case is known where 

 xylem lay outside the annual lenses. In at least some of the exceptions 

 the densewood was entire around the circuit. 



Sampling problems. — The different sequences on different radii of 

 a single section as well as along the longitudinal extent of a branch, so 

 abundantly illustrated heretofore, make it quite clear that sampling 

 problems have a real existence. An increment core is essentially a 

 single radius, and a cross section is a group of radii emerging in the 

 same plane from a single point on the axis of branch or stem. In an- 

 atomical work having to do with growth layers such as arcs, half- 

 lenses, lenses, and complete, entire growth layers with variable mar- 

 ginal definition, all grown at the lower forest border, a single radius or 

 even an entire cross section gives an inadequate picture of what the in- 

 dividual growth layers do throughout the tree. A sequence of sharply 

 bordered growth layers on one radius may be duplicated in numbers 

 on a second radius but not by the same growth layers which were 

 sharply bordered on the first. Variability, its nature and amount, has 

 been amply illustrated previously. How misleading one radius can be 

 or how much in error the matching of two radii, growth layer to 

 growth layer, can be, is well shown by a comparison of different radii. 



