160 BOTANICAL GAZETTE [MARCH 
very satisfactory. By this stain the nucleoli of the fungus and 
host are stained red, the chromatin blue, and the ordinary proto- 
plasm orange. Fungus nuclei in the process of division take the 
violet stain readily, but in a resting condition they possess 
chromatin in very minute masses. The most of the nuclei shown 
in the accompanying figures had to be under-stained with violet 
in order to prevent the over-staining of the host-tissues. The 
nuclei are represented as containing conspicuous chromatin 
granules, because in properly stained nuclei the granules are 
differentiated. 
ERYSIPHE COMMUNIS. 
Erysiphe communis proved very favorable for study, and since 
Erysiphe, Spherotheca, Microsphera, and Podosphzra appear to 
agree closely in respect to their haustoria, the fungus on 
Geranium maculatum will be described as a basis for comparison. 
Fig. 8 represents such a case as De Bary described. The hypha 
from which the haustorium originated is not shown, as the mycel- 
ium is very easily so far separated from the leaves in the process 
of fixing and imbedding as to destroy the connection between 
haustoria and hyphae, a fact upon which De Bary (12, p. 27) 
and Harper (17) comment. The absorbing organ is seen to 
consist of a slender, proximal portion, the neck, penetrating the 
epidermal wall of the cell, within which it enlarges into a ves- 
icular, distal portion with a thin wall. On the interior, the 
vesicle is filled with a delicate, spongy protoplasm differing in 
no visible particulars from the protoplasm in the mycelium. A 
mature haustorium always contains one normal nucleus, a fact to 
which De Bary does not refer. He probably saw it as the 
“thick ball” surrounded by transparent protoplasm, to which 
reference has been made. Harper (17) for the first time men- 
tions the nucleus in his paper. ig. rz represents the only — 
I have seen of a haustorium with two nuclei. This haustorium 
is also septate and such septa occasionally present themselves 
(fig. 13). Fags. 8, 79, etc. show in longitudinal section that the 
ingrowth from the inner surface of the wall of the host-cell 
