300 BOTANICAL GAZETTE [MAY 
appearance of the coenocentrum. The phenomenon is remark- 
able. It consists of an inward movement of the protoplasm 
toward the center of the oogonium, and a consequent floating 
out of the larger vacuoles to the periphery. The position of the 
nuclei is also affected so that they pass from the interior to vari- 
able distances between the coenocentrum and the wall of the 
oogonium. At a certain stage the nuclei may be found in a hol- 
low sphere of protoplasm, at or just outside of the position 
where the wall of the oospore will finally be laid down. This is 
the very interesting and striking condition to which Mr. Stevens 
has applied the term ‘‘ zonation.” 
Zonation isa very prominent stage of oogenesis in A. candida, 
' although not as conspicuous as in A. d4#. It is similar in all 
important particulars, as may be seen by comparing fig. 2 of 
this paper with figs. 6z, 62, 64, 65 of Mr. Stevens’ article. The 
agreement is made more striking by the fact that in A. candida 
also the nuclei are usually in mitosis at this time. At the com- 
pletion of zonation the ooplasm may be said to be differentiated 
from the periplasm, and only lacks the necessary female sexual 
nucleus to become the oosphere ready for fertilization. 
At this point in oogenesis A. candida and A. biti depart 
widely from one another. The oosphere of A. candida is uni- 
nucleate. One of the nuclei from a point near the periphery of 
the ooplasm slips back to the center and takes a position near 
the coenocentrum (figs. 3, 4, 5). One may find stages where 
such a solitary nucleus lies midway between the coenocentrum 
and the periplasm, but it is exceedingly difficult to determine in 
such cases whether the nucleus is returning to the center of the 
ooplasm or passing to the exterior. In the former case it 
would always be a resting nucleus, in the latter it might be in 
mitosis. 
There can be no question but that the oosphere of A. candida 
is quite generally if not universally uninucleate. There are 
theoretical reasons why in the light of Mr. Steven’s results with 
A, bliti one may expect some variation in this particular, and 
that A. candida may at least occasionally have a multinucleate 
