302 BOTANICAL GAZETTE [MAY 
by a quantity of granular cytoplasm from the antheridium 
(fig. 5), which immediately mingles with that of the oogonium. 
Although there is no proof that more than one sperm nucleus 
ever enters the oosphere, it is well to note that other nuclei may 
be present in the upper portion of the antheridial tube. Such 
an example is shown in fg. 3, where one perfectly organized 
nucleus occupies the neck of the tube. It would not seem 
strange to the writer if material were sometimes found where 
several nuclei occasionally or frequently occupied the tip of 
the antheridial tube, and were introduced into the ooplasm at the 
time of fertilization. 
The sperm nucleus begins to move towards the center of the 
oosphere, and finally fuses with the female nucleus near if not 
actually adjacent to the coenocentrum. Figs. 6 and 7 show the 
two sexual nuclei close to one another, and fig. 8 is of a fusion 
nucleus, 
An important change usually comes over the coenocentrum 
before or during the fusion of the sexual nuclei. The body so 
clearly differentiated at the time when the antheridial tube first 
enters the ooplasm becomes less and less clearly defined, and at 
last completely disappears. 
The dissolution of the coenocentrum may be accomplished 
in two ways. The structure may increase in size, the outline 
becoming vague, until there is finally present an irregular mass 
of protoplasm usually surrounding the fusion nucleus. Or the 
coenocentrum may fragment into several portions, as is shown 
in fig. 7, and these later swell and merge into an unorganized 
granular protoplasmic mass. 
It seems quite certain that the coenocentrum is not a perma: 
nent structure in the protoplasm. Arising as an accumulation of 
protoplasm at the center of the oogonium, the coenocentrum 
becomes most pronounced when the oosphere is organized, but 
finally all trace of the body is lost in the ripening oospore. 
The suggestion of Swingle (98) that the coenocentrum is + 
organ of the oosphere seems less probable when we consider its 
temporary character. Yet it is possible that the coenocentrum 
