1908-9. ] THE STELE OF OSMUNDA CINNAMOMEA. 523 
ceeding nodal pocket. A short distance up it opens into the pith and 
becomes as large as the lower one. It happens sometimes, however, 
that the tracheids so multiply that such a connection is not effected 
and even the pith may be obliterated. 
Fig. 17 is a drawing of a young stele that has acquired a permanent 
pith, though even here the tendency of the xylem to encroach on the 
pith is evident. In the lower portion of the figure a leaf-trace is repre- 
sented as about to take its departure. The pith in this region is broader 
than elsewhere, but it is continuous downward into a pocket like the one 
to be next pointed out. It will be noticed that this leaf-trace contains 
a considerable segment of the cylinder of xylem, and that the foliar gap 
or medullary ray is comparatively wide. At the upper left of the pith 
is represented the next pocket, a narrow bay of from four to six paren- 
chyma cells and continuous with the pith. 
At many other nodes, just as in Fig. 16, the pocket is not continuous 
at first with the ‘‘pith,’’ but opens out into it later on. Indeed this 
becomes, even before the seedling stage is passed, by far the most usual 
mode of origin. Thus in Fig. 4 a rather large pocket is shown at the 
upper left—a pocket that farther up connects with the tissues enclosed 
by the xylem. A pocket in an adult stem is figured in Fig. rr in the 
lowermost entire bundle. The one tracheid centrad of it shortly dis- 
appears so that it joins with the parenchyma just within. 
The nodal pocket is associated with some interesting variations 
that throw light on its homologies. If the pocket be large as in Fig. 9, 
an isolated group of endodermal cells may make their appearance in it, 
which, when the pocket opens inward, in most cases, perhaps in all (I 
regret not being able to make a more definite statement on this point) 
connects with the internal endodermis. Occasionally the nodal pocket 
does not open into the tissues surrounded by the xylem, just as at the left 
in Fig. 7, and in Fig. 8. If Fig. 7 be studied, it will be seen that the 
connection may at times be made very late, that is some distance above 
the level at which the leaf-trace separates. Thus in the upper middle a 
bay or medullary ray (into which the nodal pocket has opened) almost 
penetrates the xylem, and in some instances actually does so farther up 
the stem, just as has happened in the lower middle of the same figure. 
In conclusion the phenomenon of cladosiphony, whether in the seedling 
or the adult, is to be looked for in stems or portions of stems where the 
xylem is comparatively thick, and where it apparently shows a centripetal 
proliferation. 
