1909] BAILEY—STRUCTURE OF WOOD IN PINEAE 53 
Pinus taeda Linn., and I have myself observed them in Pinus atten- 
uata Lemm. and several other species. Furthermore, spiral thicken- 
ings occur in the ray tracheids of Pinus Balfouriana A. Murr.; they 
are strongly developed in certain specimens and only feebly in others. 
Pinus strobiformis Engelm. and other pines from the southwestern 
United States show traces of their occurrence. 
From the development of the spirals of the tracheids in the first 
year’s growth and in areas of injury, it would appear that the spirals 
were an early development, yet the fact that they are not always 
present in the first and second annual rings, and are absent from the 
axis of the cone, would seem to show that the character cannot be 
primitive. It certainly would not be safe to assume that the presence 
of these spirals is an indication of close relationship with the Taxaceae, 
for spiral thickenings of the vessels in the dicotyledons are found in 
such widely separated families as the Betulaceae and Tiliaceae. 
INSTABILITY OF DIAGNOSTIC CHARACTERS 
From the preceding remarks on the development of wood-paren- 
chyma and spiral thickenings, and the sporadic and uncertain charac- 
ter of their appearance, one realizes the difficulty in finding any defi- 
nite basis upon which to separate Picea, Larix, and Pseudotsuga. As 
has been shown, one can no longer depend upon the presence or 
absence of wood-parenchyma and spirals to acomplish this end. 
Let us now turn to consider the stability of the other elements of 
the wood. The size, form, and location of the resin canals vary so 
greatly in material from different sources that no rule can be formu- 
lated to cover every case. It is true that Picea approaches nearer to 
the condition found in the nut and foxtail pines than the other genera. 
Thus certain spruces have considerable thin-walled epithelium and 
tyloses, yet thin-walled epithelium occurs in Larix and Pseudotsuga. 
Further, the same variability is characteristic of the pitting of the 
rays and tracheids, and of the weight and color of the wood. We are 
thus forced to the conclusion that in order to distinguish the woods 
of Picea, Larix, and Pseudotsuga, a careful study must be made of 
all the anatomical and gross characters, and comparisons made with 
type sections. In other words, one must apply here the same methods 
in differentiating genera which one usually uses in distinguishing 
species. 
