ii 
38 
endeavoring to reach, the surface of the foster- root (Fig. 3). This 
" bark " consists of large, rounded, parenchymatous cells, which are 
usually distorted or torn at the outer surface, and in the vicinity of 
certain cavities, which will be mentioned hereafter. It is often 
formed df two or more shell-like layers, which grow successively 
longer, overlap one another, and thus make the surface of the haus'- 
torium appear as in Fig. la. 
Under this outer covering there is a zone of oblong, prismatic 
cells, that are radially arranged in several close rows (pc in Figs. 3 
and 6). This tissue does not penetrate into the foster-plant ; we do 
not notice it in a tangential section taken as deep as the one in Fig. 8, 
but in sections higher up toward the top of the haustorium it is 
always seen encircling the central portion. Its peculiar cambium- 
like structure, and the manner in which the cells outside and inside 
of it seem to proceed from it, induce me to consider it as a kind of 
meristem from which, on one side, the bark (eventually with its 
successive layers) may originate, while on the other it may contribute 
to the development of the parenchymatous outer part of the central 
portion. 
Very often there are empty spaces, caused by the stretching 
and tearing of the tissues, between the zone of prismatic cells and 
the bark (Fig. 6, sp), and also between the successive layers of the 
bark. Besides, an empty space is usually to be found extending 
along the curved surface of the foster-root, where the bark is attached 
by a few rows of cells only (Figs. 6, sj> and 8, sj>). 
The central portion of the haustorium, which lies within the two 
zones described, and enters the body of the foster-root, consists of 
several distinct kinds of cells. Most conspicuous among them by 
their reticulated walls, their zigzag course and pecuHar arrangement, 
are the vascular cells {vc in all the figures). They arise singly 
(Fig. 8), or in groups (Figs. 8 and 3), either at the apex of the haus- 
torium, closely attached to the vessels of the foster-root, often even 
entering them, or at some distance from them. Continuing their 
course m zigzag lines they meet other vessels of the same kind, and 
unite with them m bundles, the butt ends of which form a rather 
compact ring in the centre of the haustorium. From the different 
groups of this ring, numerous single ducts proceed upward and con- 
verge into a large bundle, more or less distinctly divided longitudi- 
nally into two halves, and finally join the plerome of the Co7nandra 
rootlet (see Fig. 3, and for the upper half, Fig. 6). It usually hap- 
pens that a zone around the lower part of the vertical axis of the 
haustorium contains no vessels. (Figs. 6 and 8). In Fig 6, there- 
fore, we see no vascular cells in the lower half, except a small por- 
tion of an arc (joining two groups of vessels) which was sliced off by 
the knife. 
In the upper part of their course the vascular bundles are enclosed 
by a few rows of elongated, narrow prismatic cells (see Fig. 6) which 
are similar to the prismatic cells inside of the bark but are 
niuch narrower and with thinner walls. There is little doubt that 
they perfonn the functions of the cambium of stems and roots 
Anotlier kind of elongated (but not prismatic) active cells accom- 
