376 ROYAL SOCIETY OF CANADA 
ing fairly large canals. Proceeding towards the base, the canals became 
more and more defined, and at 2 em. below the tip thyloses were observed 
in some of the bundles. The thyloses here appeared mostly singly 
(Fig. 2), and were well-formed, bladder-like cells with cytoplasm and 
nucleus standing out clearly. In no case were the thyloses observed to 
fill up much of the canal, the greater part of the cavity remaining un- 
blocked. As will be shown below, this is just what ought to be expected 
under conditions of slow growth. 
In sections of stipes collected Sept. 23rd, 1906, which measured 
about 20 cm. in length and 3 cm. in diameter, the following was ob- 
served. 
As was to be expected in stipes which had ceased growth in length, 
all the tissues, from the tip to the base, were fully formed. The vascular 
bundles, however, did not show any canals whatsoever. In the region 
of the protoxylem, patches of cells which stand out from the adjacent 
parenchyma and phloem by the comparatively large size of their lumens, 
and by their irregular outlines, were invariably seen. The walls of these 
cells are thinner than those of the surrounding tissues, and there is com- 
paratively little protoplasm in their lumens. Whatever cytoplasm is 
present seems to be in a state of disintegration, the muclei standing out 
prominently. These groups of cells are the thyloses which have now 
come to fill up the whole cavity of the canal. 
These patches were also noticeable from the fact that spirals and 
rings of disorganised protoxylem elements seemed to be embedded in 
them, and this is explainable in view of what was observed in the younger 
stipes. In those it was seen that the protoxylem elements split away 
trom each other and floated off into the cavity caused by the growth of 
the surrounding tissue. When the thyloses began to make their appear- 
ance these came to lie in the. spaces between the thylosal cells. As the 
thyloses extended more and more into the canal they took up all the 
available space, and thus the remains of the disorganised protoxylem 
appear to be enclosed in the lumens of these cells. This, however, is 
impossible, for the cells could not absorb them through their walls in the 
manner of amœbæ. Hence the only reasonable conclusion to be arrived 
at is that they lie in the spaces between the walls of the thyloses, these 
spaces being the surviving portions of the original canal, the remainder 
of the canal being occupied by the thylosal cells. These spaces, however, 
are so small, that the protoxylem remains appear to be enclosed in the 
thyloses themselves. 
The thylose cells vary greatly in size and form, and frequently there 
is a large, irregularly circular cell in the centre of the patch, with smaller 
