1900] THE HAUSTORIA OF THE ERYSIPHEA 161 
surrounds and accompanies the neck for a distance, as De Bary 
reported. It seldom extends down to the body of the hausto- 
rium, however, and sometimes is absent altogether (fig. 76). The 
collar from the cell-wall is somewhat different from the wall 
from which it takes its origin, and usually stains little, while the 
remaining portion stains with safranin intensely. The outer sur- 
face of the neck adheres closely to this collar, the outer boundary 
of which is represented by the middle circle in fig. 77. 
But fig. 8 does not represent the usual conditions in one par- 
ticular. Almost always a mature haustorium is surrounded by 
a thick, sheath-like layer which, De Bary says, belongs to the 
protoplasm of the host-cell. It is clear that, by the use of hand- 
cut sections of fresh or alcoholic material and by a study of the 
haustoria from the leaf surface, he was not able to make out the 
structure of this sheath fully. De Bary describes it as an irregu- 
lar mass; but it is not extremely irregular and is bounded by a 
very thin membrane of about the delicacy of the plasmic mem- 
brane of the host-cell. In fig. 9, for example, the middle cell 
contains an optical section of a haustorium. Within the center 
lies the relatively large nucleus with its red-staining nucleolus and 
fine chromatin granules. Surrounding the nucleus is the spongy 
protoplasm staining orange. The haustorium-wall surrounds 
this protoplasm. Exterior to the haustorium are the contents 
of the sheath described by De Bary as belonging to the proto- 
plasm of the cell, while bounding the whole, in contact with the 
contents of the host-cell, is the delicate limiting membrane of 
the sheath itself. 
The substance of this sheath looks like protoplasm at first 
glance and stains with orange, but it is not vacuolated like pro- 
toplasm. It consists of a dense, homogeneous, finely-granular 
mass, most frequently gathered into lumps of varying outline, 
which appear very slightly granular (figs. 9, 77, 16). Rosen 
(26, p- 258) observed in Puccinia asarina, growing in the intercel- 
lular spaces of Asarum, that the branched haustorium was con- 
nected with the nucleus of the host-cell in the majority of cases, 
and either adhered closely to it or entered it with disorganizing 
