490 BOTANICAL GAZETTE [JUNE 
by bulbous swellings from one cell fitting into corresponding depres- 
sions of the next. Sometimes the fungus thread branches dichotom- 
ously and forms a rest for another hypha in the fork (fig. 4), and 
again the two often intertwine and bend around each other (fig. 3). 
Contrary to expectation, the walls of the three types of hyphae 
are uniform in thickness, and although irregular thickenings of 
cellulose in the walls of the stroma-forming hyphae make them 
uneven, they are not really thicker. Pits (fig. 1a) through the 
walls are abundant, and although in a few rare cases these seem 
to be open, in the great majority a definite separating membrane is 
very clear. The cells of the smaller hyphae are binucleate, but the 
nuclear condition in the larger is difficult to distinguish, because 
in the few places where the contents are visible they contain dark- 
staining granules. As definitely as could be distinguished, how- 
ever, they have many multinucleate cells which have probably 
developed from the binucleate condition by division of the 
nuclei. 
In the compact zone above the central region the stroma appears 
to be uniformly composed of the same large empty-celled hyphae 
already mentioned, which radiate out in parallel rows from the 
center to the periphery; but if the filaments are spread apart or 
a cross-section is examined the same differentiation ‘of the three 
sizes is seen. The walls show the same thickenings and pits already 
described, and the cells fit into each other in the same way, making 
a very firm dense pseudoparenchyma. In some places, beside the 
incurving of cells to each other, they hold together in the corners, 
much like xylem cells. One peculiarity of this part of the stroma 
seen in longitudinal section is the large number of holes bordered 
by walls (fig. 5). These are cut ends of protuberances from the 
cells (figs. 5, 6), and are one of the methods of mechanical support 
not found in the previous zone. They are not the technical clamp 
connections of Dr Bary’s description, because they occur not 
only near the septa, but from any place along the cell to another in 
the same hypha, and because they cannot be so named until their 
method of development has been established. Here they are 
blind tubes that reach from one cell to fit flat against the wall of 
an adjacent hypha and clamp the two together. 
