526 THE AMERICAN NATURALIST. | [Vor. XXXIX. 
(8) The process of evolution is less definitely expressed in 
the genus Populus which shows but little advance since the 
Middle Cretaceous, and it may possibly be regarded as having 
attained the culmination of its development. 
While it is impossible at present to discuss the anatomy of 
these plants in the light of their ancestral forms, the study of 
existing species may serve an important purpose as a working 
basis for future paleontological research in this direction, and it 
may eventually prove to be the key which shall unlock the door 
now concealing important records of the past. In this sense it 
is necessary to formulate a theory of possible descent as a 
working hypothesis, and this can be obtained only through a 
detailed study of. the anatomy of the Salicacez in cemparison 
with what has already been ascertained to represent certain laws 
of development in the gymnosperms. This hypothesis may be 
stated briefly before proceeding to a discussion of those anatom- 
ical details which may lend it support. 
A comparison of the ferns, arborescent pteridophytes such 
as Calamites and Lepidodendron, and the arborescent types of 
gymnosperms and angiosperms, shows, in general terms, that 
the so called wood increases in proportion to the requirements 
of mechanical support. This latter is usually met by the dis- 
position of the tissue in such a manner as to afford the most 
effective resistance to external stress, and it is therefore dis- 
posed in monostelic stems, in the form of a definite cylinder 
which constitutes the secondary xylem structure. It is also an 
essential feature of the same law, that the secondary xylem 
should be in all cases radially external to the protoxylem, and 
that its augmentation in secondary growth of the stem must 
always arise in radial succession. 
In the polystelic ferns where the secondary xylem consists 
almost wholly of vessels and the wood elements are numerically 
subordinate, these latter are distributed among the other ele- 
ments of the xylem and do not form a specialized mechanical 
region. Such strengthening as these plants require, is accom- 
plished in the first instance by the sclerenchymatous bundle 
sheath, and secondarily by the more general conversion of the 
fundamental tissue into hard sclerenchyma. It is similarly 
