36 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I40 



designation. The only serious ambiguity arises in case of multiple 

 growth layers in a year. Should we use springwood and summerwood 

 in relation to an outer growth layer formed, say in July, or late in the 

 growing season, whatever portion of the calendar this may span? At 

 first sight, earlywood and latewood, based upon mutual relationship in 

 time, seem to be more accurate and more justifiably descriptive than 

 previous terms. We have, however, found latewood present at the 

 beginning of a growth layer, and we have seen it within the body of 

 an annual increment as fragments, intermittent circles, and complete 

 circles preceding the latest earlywood of the increment. Surely, late- 

 wood should be later than earlywood. 



Terminology based upon position in the grozvth layer has not been 

 proposed, apparently, in a formal manner. "Inner part" and "outer 

 part" have been used descriptively (Holman and Robbins, 1939, p. 

 117), These are rather weak and subject to the same indefiniteness as 

 those previously mentioned. In the same category are the terms "red 

 ring" and "red autumn ring" (Douglass, 1928, p. 32). 



The third basis for discriminating between parts of a growth layer, 

 structure of the tissues, is committed neither to time of formation nor 

 to relative position within the growth layer or annual increment. 

 Lightwood and densewood are thus terms descriptive of the cellular 

 nature of the tissues. It seems worthy of note that, although botanical 

 workers have stressed xylem structure, they have commonly made use 

 of a nomenclature based on time. However that may be, a structural 

 term is eminently descriptive and can be applied to any type of growth 

 layer no matter when formed or how completely formed. The terms 

 for the two general types of xylem in growth layers, "light" and 

 "dense," have only structural and textural implications. 



The structural characteristics involved, it is generally agreed, are 

 radial dimension of the cell, thickness of the cell wall, amount of ligni- 

 fication, and deposition of dark materials such as tannins, gums, and 

 resins. The first tw^o are rather easily determined under moderate 

 magnification. In contrast, the third can be safely determined only if 

 staining has been done properly or if microchemical tests have been 

 performed. Lignin is concentrated in, but not confined to, the dense- 

 wood. The dark material mentioned as the fourth feature is visible on 

 unstained sections under a hand lens or naked eye. Of all the features 

 mentioned, this is probably the least diagnostic. 



The ordinary conception of a growth layer pictures lightwood fol- 

 lowed outward radially by densewood (pi. i, fig. i). In the lightwood, 

 the cells are wide radially and possess thin walls which are but slightly 

 lignified. In the densewood, these features are essentially the oppo- 



