492 BOTANICAL GAZETTE [JUNE 
Xylaria. From the evidence there is no basis for believing in any 
earlier differentiation of Woronin hyphae as initiating the coiling 
process; but they do become differentiated in the center of the coil 
by an increase in size very soon after the perithecial primordia are 
well started (fig. 10). Growth in the size of the perithecium is 
accomplished by increase in the length of the wall hyphae, and also 
by the addition of other hyphae around the outside. With this - 
increase in size the wall layers become thinner and compact, 
and some of the inner hyphae decrease in size and become | 
absorbed, probably furnishing nourishment for the fertile branches 
(figs. 10, 11d). A large number of perithecia start but few mature, 
and these, logically in relation to their food supply, are mostly 
toward the inner line of the perithecial zone. The others remain 
intact and are scattered throughout the fruiting region, apparently 
inhibited from further growth at any stage in their development, 
and remaining without change at that stage. As an exception to 
this, some of the larger perithecia that have reached the point 
where they contain ascogonia and then become abortive show signs 
of deliquescing and disintegrating. As a rule the older perithecia 
are toward the top of the stroma and the younger stages are down 
toward the substratum. 
Within the perithecium the filament in the center develops into 
the Woronin hypha (fig. 10). It increases in length, and in size 
and number of cells; and after this some of the cells round out, 
increase in size, and eventually separate from each other to form 
the ascogonia. The enlargement is not uniform as to the size and 
shape attained, for these ascogonia (fig. 115) are many of them 
uneven and contorted in outline, and many of them retain narrow 
stalklike connections with cells from which they have not become 
completely separated. 
The nuclear condition in the Woronin hyphae and ascogonia 
is the critical point, and is hard to determine because of the extreme 
variations in the size of the nuclei and the difficulty of determining 
successive stages. As said before, the cells of the Woronin hyphae 
are binucleate (fig. 9), and as they enlarge (fig. 10) they show 4 
steady and marked increase in the size of the nuclei from the 
time the knot is well formed. There are one or two divisions, after 
