114 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL, I40 



Point two of the introductory paragraph has to do with the form of 

 growth layers in three dimensions. Entire growth layers may become 

 partial, and vice versa. All various forms of partial growth layers 

 merge the one into the other. Lenses may be discrete bodies, like a 

 "patch" on the cylinder of a branch or stem. Or, again, a lens may be 

 the longitudinal extension of a growth layer (text figs. 34, 35). A 

 cross section at one place shows the growth layer to be entire; at 

 another, it appears as a concurrent lens; at a third place, it appears 

 as a single lens ; and last, if the section be taken at exactly the correct 

 place, it appears as divided densewood. From a three-dimensional 

 standpoint, lenses testify to cambial activity local in time and space. 



Half-lenses represent the visible portions of lenses and as such par- 

 take of all variations characteristic of them. Theoretically perhaps, a 

 half -lens could exist independent of a lens, its densewood, sharp for 

 a distance, weakening, diffusing, and giving way tangentially to light- 

 wood. The concept of an independent half -lens is difficult to visualize 

 through physiological activity. In three dimensions, local failure of 

 densewood to form in connection with a lens may produce a half -lens. 



Arcs physiologically involve the failure of densewood to form on 

 opposite edges of a lens or lens "finger," or the failure locally of the 

 densewood of an entire growth layer which in cross section gives a 

 long arc. Whether or not an arc can exist independently merits the 

 same remarks as those given in connection with half-lenses. 



No matter how we employ cross sections to illustrate growth layers, 

 the three-dimensional aspect cannot be neglected. To do so, is to miss 

 the significance of cambial activity, the important link between growth 

 factors and the anatomy of the xylem. 



ECOLOGIC SIGNIFICANCE 



The importance of cambial activity has been constantly emphasized 

 both directly and indirectly. Description, classification, and attempted 

 interpretation of growth layers are tasks superficial and misleading 

 without a proper understanding of the physiology of growth. To such 

 a restricted approach, a strictly two-dimensional habit of thought 

 only adds confusion. Growth layers mirror cambial activity and 

 growth processes, and without a knowledge of such activities no solid 

 basis exists for the study of growth layers as a record of the nature 

 and variations of growth factors. 



All the work, of which the present paper is a partial report, demon- 

 strates clearly that cambial activity and growth processes within the 

 body of a tree grown under decided lower forest-border conditions 

 are intermittent in time and space both locally and regionally. If 



