464 
Journal o f Agricultural Research 
Vol. XXIX, No. 9 
only locally in the cortex, but are 
frequently distributed throughout its 
radial section. The fact that the tissue 
is not immediately destroyed indicates 
that a high degree of adaptation has 
been established. 
DESCRIPTION OF THE FUNGUS 
The mycorrhizal fungus in roots of 
legumes has been described and illus¬ 
will be called haustoria in this paper. 
The vesicles have been regarded by 
those using the term as fruiting struc¬ 
tures, probably sporanges, a supposi¬ 
tion which is supported by Peyronel’s 
tentative account of their germination. 
The mycelium is nonseptate, at least 
within the host. In diameter it varies, 
depending somewhat upon the char¬ 
acter of the tissue in which it grows. 
The first strands which enter the root 
are 6 to 11 n in diameter, 
whereas those which develop 
later sometimes measure 12 to 
13 fx. The hyphal walls are 
conspicuously thick whether 
the strands are inside or out¬ 
side the host tissue. Within 
the tissue intercellular strands 
often show small pointed pro¬ 
jections extending between 
the rounded corners of the 
cells at their junction, suggest¬ 
ing a plasticity of the fungus 
wall in early development (fig. 
2). The comparative thick¬ 
ness of frhe wall, and the some¬ 
what angular molding of con¬ 
tour at intervals are the more 
distinctive mycelial charac¬ 
ters. The septa, if such they 
may be called, which are found 
in mycelium outside the host 
are usually curved, and appear 
to be derived from an interior 
lining of the wall rather than 
from the thick outer wall. 
The haustoria present a 
complete series of forms from 
the simplest protuberance of 
the mycelial wall to branched 
structures of great complex¬ 
ity. They rarely develop in 
the outer cells of the cortex in 
which the mycelium passes 
readily through cells. In 
deeper cells haustoria may ap¬ 
pear as small protuberances 
or as thick, blunt projections 
on intercellular strands. In 
deeper tissue, which appears to 
have a richer protoplasmic 
content, haustoria branch 
abundantly, producing at the 
ends of thin hyphal branches a much- 
lobed, very delicate structure, seemingly 
without wall or membrane, and often fill¬ 
ing a large part of the lumen of the cell 
(fig. 3). This terminal structure has 
been interpreted as the product of the di¬ 
gestion of the fungus by the host, and be¬ 
yond doubt it does disintegrate very soon 
into an amorphous mass. Whether 
the haustoria are digested by the tissue 
which they enter or whether they dis¬ 
integrate after a brief period of service 
to the fungus requires further study. 
trated so well by Janse, Magrou, and 
Peyronel that only the more important 
morphological characters will be 
described here. The characteristic 
structures of the fungus have been 
designated mycelium, arbuscles, and 
vesicles. The arbuscles are, as Pey¬ 
ronel recognized, haustorial structures, 
comparable to those produced by some 
of the Peronosporaceae, though some¬ 
what more elaborate in structure. 
Since there is no apparent advantage 
in the use of a distinctive term, they 
