120 Worsdell. — On the Comparative Anatomy of 
nuclei ; no sieve-tubes were ever seen. There is a well-marked 
cortical region in the haustorium of this species. 
Having entered a Bamboo-root, the haustorium may either 
grow straight through the cortex and penetrate directly, with 
its apex, the central cylinder (Fig. 6), or, having entered the 
cortex of the host, may branch, the branches bending at right 
angles to the cylinder or running parallel to it through the 
cortex (Fig. io). As a haustorial branch thus pursues its 
course through the inner region of the cortex, some of its 
lateral cortical cells in contact with the central cylinder 
acquire dark purple contents, and begin to dissolve the cells 
lying against it ; this fermentative action goes on until a cell, 
having opened out a cavity in the conducting-tissue of the 
host, grows out into it as a papilla or sucker ; this cell has 
fine granular contents, and dark lines in the convex portion 
of its wall, which look like pits ; by these lateral suckers 
a large portion of the cylinder becomes destroyed (Figs. 7, 
8, 9). In Fig. 6, which represents a haustorium directly 
penetrating the central cylinder of the host, the xylem- 
elements and their accompanying proteid-cells are seen to be 
in contact with and spreading among the conducting elements 
of the cylinder. 
The lateral roots are of endogenous origin, as indicated by 
the projecting flap of cortical cells, already referred to, nearly 
always found on either side of a lateral member. No young 
stages of lateral roots were seen in this plant. In some 
instances there is great difficulty in determining whether the 
branch is of endogenous or exogenous origin, the appearance 
being greatly in favour of the latter mode. There seems little 
doubt, however, that the endogeny becomes more or less 
obscured as the branch advances in age and thickness. 
What has been said of the root-apex in describing the last 
species, applies equally in this case (Fig. 11). I need only 
mention that some apices appeared to be destitute of a 
root-cap. 
The anastomosis of these roots is a curious phenomenon. 
At the point of contact of two roots the cells of its outermost 
