290 BOTANICAL GAZETTE [APRIL 
P. albicaulis. The ninth, P. reflexa, though in the climatically dry 
region, occupies moist sheltered spots. The remaining three species 
are in climatically moister regions, and one of them rather belongs 
to the Atlantic forest flora. The evidence, therefore, favors the 
view that narrowness of the spring tracheids is linked with special 
need for economizing water, and is thus encountered in xerophilous 
species. 
But another interpretation is possible. The first six species, 
having the narrowest tracheids, all show wood of characteristic 
_ Structure, differing from that of the others in that their ray paren- 
chyma has “‘piceoid”’ pitting, and this conforms with characteristic 
cone structure, which agrees with that of diploxylic species. Thus 
these species belong to the peculiar subsection PARA-CEMBRA. 
Again, the first four of the species belong to one group of this sub- 
section, namely Parrya, while the remaining two belong to the 
other group BALFOouURIA. 
The remaining species differ from the first six in structure of 
medullary rays and of cone (whose scales have a terminal umbo), 
so that they are members of the subsection Cempra. And of them 
the first two species, with narrowest spring tracheids, belong to the 
subdivision of CEMBRA known as the group Eu-cEMBRA; while the 
last four, with widest spring tracheids, are included in the other 
subdivision STROBUS. 
Thus, in the whole series of species belonging to section I, 
diameter of spring tracheid is rigidly linked with systematic post 
tion. It is conceivable, therefore, on the one hand, that width of 
spring tracheid is a systematic or purely morphological character, 
not an epharmonic one; or, on the other hand, that these groups 
and subsections are not natural monophyletic groups, but are poly- 
phyletic collections of types whose likenesses are determined by 
ecological factors. Yet, on the whole, it seems simplest to suppose 
that the evolution of section I of Pinus has been determined by 
available water supply, and that group PAaRrRYA and group STROBUS 
represent extreme types, the former the most xerophilous and the 
latter the most hygrophilous. 
The suggestions above given harmonize with the facts relating '° 
East Indian pines. P. Gerardiana, belonging to the group PARRY 
