VIEW OF THE GENUS CUPRESSUS. 823 
layer of cells. Within the hypoderm, we find palissade-parenchyma 
more or less well developed, and this is followed by layers of 
irregularly-shaped cells, which when more or less lengthened are 
horizontal, or at right-angles to the palissade-cells. It is rare 
that the section shows clearly the vascular bundle of the leaf, for 
if the section is thick it is obscure, if thin the tissues disintegrate 
and fall away, leaving a space; but the structure of the 
branchlet or axis from which the leaves proceed is usually well- 
defined, consisting of two, or of a ring of four, wedge-shaped (in 
section) masses of phloem and xylem, the latter directed towards 
the centre of the axis.  Resin-canals are often found in the 
parenchyma of the leaf; but their number and relative position 
are not so constant as in the species of ies and Pinus. 
The structure of the flat cotyledons, of which there are two, is 
the same as that of the adult leaves, with the exception that there 
is no hypoderm. The vascular cylinder of the cotyledon has a 
well-marked endodermal layer, and an undivided vascular bundle. 
The stomata are mostly on the upper surface, next the axis. 
The arrangement, size, and form of the cones offer good 
characters in many instances, but the large amount of inherent 
variability which these trees possess must never be lost sight of 
by the systematist. A similar remark applies to the form of the 
male catkins. In fact no one character can be relied on to 
determine a species of Cupressus, and the inferences to be drawn 
from the aggregate or sum of available characteristics require 
always to be carefully controlled by a consideration of the varia- 
tion incidental to the growth of each individual plant, as well as 
of that peculiar to the species and constituting a variety. 
How far these variable characters are physiological or the 
result of adaptation, and how far phylogenetic, it is impossible to 
say with accuracy. The difficulty is enhanced by the fact that 
in the earliest representatives of the genus with which we are 
acquainted the same variations existed. (See Renault, ‘Cours de 
Botanique Fossile: Cupressinew,’ 1885, p. 121; J. Starkie 
Gardner, ‘ Mon. British Eocene Flora,’ vol. ii. part 2, tab. 19.) 
The anatomical structure of the cone-scale shows very dis- 
tinctly its double nature: the vascular system of the bract and 
that of the scale being well developed, the latter having an 
inverse orientation of its elements. A vertical section of the 
axis shows that the vascular bundle for the bract, and that for 
the cone-scale are given off separately and independently. It is 
