798 THE AMERICAN NATURALIST. (VoL. XXXIX. 
growth. This case is therefore to be regarded as lying outside 
the normal course of development for the family as a whole, 
while at the same time it serves to give emphasis to the specific 
effects of external influences in modifying the character of the 
growth. 
It is a general feature of the Salicaceze to which there are no 
well defined exceptions of a noteworthy character, that there is 
substantially no differentiation of spring and summer wood. 
Partial exceptions seem to be shown by Sa/ix longifolia in which 
the last six rows of cells sometimes become radially compressed, 
or by S. cordata and S. /evigata in which three rows are some- 
times so modified. The fact that in all these cases the modifi- ` 
cation is not a constant feature, but that it arises sporadically, 
indicates clearly that it is due to some exceptional and transitory 
influence, and that it does not in any way express the usual 
characteristics of growth. On the contrary, it is not an uncom- 
mon feature in S. zzgra for the first half of the growth ring to 
present the most dense structure, this being due, not to a 
diminution of the vessels with a relative preponderance of the 
fibrous mechanical cells, but to the fact that the latter are char- 
acterized by thicker walls than elsewhere, with a corresponding 
diminution of the cell cavity. Inthe investigated species of 
poplar, there seems to be no deviation from the general rule 
thus expressed. In both genera there is substantially no altera- 
tion in the form and size of the wood cells as between those of 
the inner and those of the outer face of the growth ring (Figs. 
6, 7), and the demarcation between the growths of successive 
seasons is determined by other factors of which the occurrence 
of a limiting zone of resinous wood parenchyma is the most im- 
portant. From these considerations it appears that the charac- 
od of the growth ring, in its more general aspects, possesses no 
diagnostic value for specific purposes, or even for generic differ- 
entiations, but that it may be regarded as of value for ordinal 
purposes with respect to their great breadth and the absence of 
any marked differentiation between the spring and summer 
woods. Since the radial compression which is generally so dis- 
"dos a feature of the summer wood, is recognized to result 
rom increasing cortical pressure (De Bary, '84, p. 501) it must 
