39 
panics and surrounds the vessels in their lower course, where, in fact, 
they form the bulk of the haustorium (ac in all the figures). The 
transformation of these cells into vessels by the deposition of the 
reticulated thickening on the cell-wall, and the subsequent union of 
several cells into one duct, can be plainly traced. 
The space in the central portion not occupied by the tissues men- 
tioned is filled with a ground tissue of parenchymatous cells, not 
unlike those of the bark in form, but, for the greater part, of smaller 
size (gf in all the figures). In the lower half, especially in close 
proximity to the foster-root, these cells become also narrow and 
elongated. But at the two lateral extremities, where they curve out- 
ward to insert themselves longitudinally between and into the cells 
of the foster-root, many of them assume an inflated, club-shaped 
form (at x in Fig. 3, ^/in Figs. 4, 8 and 10), 
In all the sections we notice peculiar, striate bands dividing the 
parenchymatous tissue into shell-like layers (ss). Along these bands, 
cavities like those between the bark and the zone of prismatic cells 
are frequently met with, as in Figs. 3, 6, and 8. Similar structures 
were called '' stripes of separation " by Solras-Laubach in his excel- 
lent description of the haustorium of T/ies/um, the Old World near 
relative of our Comandra!^ Solms-Laubach has shown that they con- 
sist of parenchymatous cells, crowded together and crushed by the 
multiplying neighboring cells into a compact mass, in which the com- 
ponent individual cells are recognizable only with difficulty or not at 
all. As these notes are to furnish merely an anatomical description, 
I will reserve- my opinion in regard to the origin of these ** stripes," 
and win simply state that in Comaiidra they seem to be of a more 
complex nature than in Tkesium j for, in the former, we find them 
not only in more or less concentric zones, as in Thcsium, but every 
larger group of vessels with meristem tissue appears partly surrounded 
by such a shell of compressed cell-membranes. 
^ The difference in the manner in which the cells of these two par- 
asites are attached to those of their foster-plants deserves special 
attention. Solms-Laubach says (/. c, p. 545) that the terminal cells 
pf the haustorium are separated from those of the foster-plant by an 
irregularly developed layer of a homogeneous, yellow mass, possessing 
high refractive power; and (p. 547) that this mass is evidently 
intended to isolate the haustorium from its foster-root, and that 
wherever this mass is but poorly developed or entirely wanting, the 
foster-root endeavors to replace it by producing a more or less mas- 
sive corky layer. In the haustorium of Cofnandra I could detect no 
such mass nor any corky layer. Figs, 4, 5, 7, 9 and^ 10 show the 
points of contact in sections carried out in the three principal dimen- 
sions. 
. Fig. 4 is a longitudinal section, showing a group of cells from the 
vicinity of the point x in Fig. 3. Three cells, gt, belonging to what 
i have called the ground-tissue of the haustorium, enter some phloem 
of the Asier root ; the walls of all the cells can be plainly distin- 
Hermann Graf zu Solms-Laubach. iiber den Ban und die Entwickelung par- 
asitischer Phanerogamen, in Pringsheim's Jahrb. fur wissensch. Bot.. Vol vi. 
