E. J. BUTLER. 9 
though in some of the distorted bundles, where it is difficult to dis- 
tinguish between the sheath and the outer layers of the xylem, 
cells may occasionally be found, apparently lying in the xylem, 
containing hyphe. These are really sheath cells, infolded into 
the xylem by the distortion of the whole. 
As in the case of several other fungi whose habitat is chiefly 
between the cell walls, some branches penetrate the latter and pass 
into or across the cell cavity. These intercellular hyphe are of 
two kinds. The one are similar to the haustoria of other Peronos- 
poraces, being simple or branched, thin-walled processes, arising 
from intercellular hyphe, and sometimes forming small fungal 
clumps within the cells (plate V, fig. 1). The others correspond to 
the “ Verbreitungshyphe’’ described by Guttenberg (1905) in Usti- 
lago Maydis, and like them are thick-walled filaments, often 
clothed with a distinctly two-layered cellulose sheath (plate V, 
fig. 3). This sheath, as in the maize parasite, consists of an outer 
portion deposited by the host cells around the penetrating hypha, 
in a defensive attempt to check its growth. The attempt often 
succeeds and the resulting thick-walled pegs projecting into the 
lumina are not uncommon (plate V, fig. 4). The inner part of 
the sheath is the wall of the fungus filament itself. 
The intercellular mycelium is sparse in the stem and attains 
its greatest development in the leaves. The hyphe frequently 
occupy the intercellular spaces at the angles of the mesophyll cells, 
but are also found in the lamelle between two cells. They branch 
freely, have thick, rather gelatinous, cell walls, and clear 
protoplasm. In the stem, haustoria are not common, and are 
chiefly small, often button-shaped,: processes, not unlike those of 
Cystopus. Their full development occurs in the leaves (plate V, 
fig. 1). Here, they are often extraordinarily numerous and 
usually simple or branched finger-shaped processes, sometimes 
much convoluted and occupying a fair proportion of the cell 
cavity. Their wall is thin while young, but older ones often have 
distinct cellulose caps (fig. 2) which are possibly provided by the 
host plant aiter the organ has ceased to be active, asin the case 
of Cystopus candidus (Guttenberg, 1905, pp. 8 and 9). 
