> i a 
MEMOIRS FROM THE DEPARTMENT OF BOTANY OF COLUMBIA COLLEGE. 177 
present in every instance. Polygonum scandens and P. arifolium have a comparatively 
weak development of hypoderma and starch-ring, but these tissues are clearly distin- 
guishable. P. Californicum is the most puzzling case; I have been unable to distinguish 
with the material at hand between a stereome cylinder, as characteristic of the other sub- 
genera, and the woody fibres of the xylem, but it would not be safe to conclude that 
either is wanting; we can regard the woody fibres as having become strongly sclerotic for 
reasons to be mentioned below. 
Third: If the stereome is comparatively weak, is it supported by a strong develop- 
ment of xylem or otherwise? In P. bistortoides and P. alpinum we find a strong cylinder 
of stereome elements and a comparatively strong development of xylem. P. Pennsyl- 
vanicum and P. Virginianwm have a rather weak and more or less interrupted stereome 
cylinder, but it is supported by a very heavy development of xylem, and in the case of 
the former species, where the stereome cylinder is broken more than usual, we find fasci- 
cles of stereome elements at the inner sides of the collateral bundles. 
The two species just considered require rigidity, hence their peculiar structure. On 
the other hand, such plants as P. aviculare, which prefer to grow in localities where they 
are continually buffeted about in one way or another by nature, need an axis whose 
characteristic qualities are wiryness and toughness instead of rigidity. This is accom- 
plished by the development of small fascicles of stereids in place of a stereome cylinder. 
This principle is carried still further, in the example chosen for this study, by the 
development of fascicles in the hypoderma, thus producing the quality of wiryness and 
toughness which could not otherwise be obtained. 
Fourth: In cases where the normal development of stereome is a cylinder, if this 
cylinder is broken by growth, are the breaks placed economically? In the case of fasci- 
cles, are the latter placed to mechanical advantage? The fact that they are is answered, 
for the first question, in P. Pennsylvanicum and P. Virginiaum. In the latter species we 
find one class of breaks in the stereome cylinder, namely, those made by the penetration 
of the starch-ring; these breaks come between the collateral bundles, or what we may 
term interradially, so that the break is made with the least possible loss of strength. We 
find two kinds of breaks in the stereome of P. Pennsylvanicum, the one as in the case just 
cited, the other radial or radially with the bundles, thus causing quite a serious loss of 
rigidity to the axis. However, a strong fascicle of stereids at the inner side of the bundle 
supports this radial interruption and probably gives at least equal strength. As to the 
second question: In the cases examined, the stereome fascicles of the cortex were radial 
with those of the pleromatic cylinder as well as the collateral bundles, thus producing 
the tough and wiry character of the plants of AvIcULARIA. 
