4 BOTANICAL GAZETTE [JANUARY 
it to have passed the corner without having altered its shape in the 
least. It was evidently creeping along the convex side of the tube 
and was killed just as it was rounding the corner. That the male 
cells actively change position and shape on their own initiative is 
shown also by the manner of entering the archegonium, to which 
attention will be called in a succeeding paragraph. 
In the account of the archegonium already published (3b), it 
was stated that no ventral canal cells or nuclei had so far been 
observed, but that it was unlikely that they were not formed. 
Archegonia have since been observed in which such a ventral canal 
nucleus (figs. 7 and 8) had been cut off. There still remains some 
doubt whether this nucleus is regularly cut off. All the cases 
observed were in gametophytes some of whose archegonia had 
already been fertilized. I have not yet found a gametophyte with 
archegonia unfertilized where ventral canal nuclei were present. 
Two explanations are tenable. Either this division is delayed 
almost up to the time of fertilization, and the canal nucleus degen- 
erates very quickly so as to leave no trace of itself in fertilized 
archegonia, or it does not occur normally, but only in those arche- 
gonia in which fertilization has been delayed beyond the usual 
time. ‘Though the latter appears the less probable supposition on 
general grounds, the evidence available is more in accord with it. 
In one case (fig. 8) two small nuclei were present. The egg nucleus 
in this case appears small. What would ordinarily be taken for 
the ventral canal nucleus is larger than in fig. 7. In the majority 
of cases the cytoplasm of the egg in which ventral canal nuclei 
were found appears to be undergoing degeneration. It clumps 
together and has very indefinite structure. It cuts with difficulty. 
In some archegonia, of normal appearance in other respects, there 
was present a zone of fibrillar cytoplasm surrounding the very 
large nucleus. These archegonia were also invariably found in 
gametophytes with one or more that had already been fertilized. 
Eames (8) reports the regular formation of the ventral canal 
nucleus in Agathis immediately before fertilization, and its rapid 
degeneration. CouLTER and LaNnp (6) were unable to demon- 
strate its formation in Torreya. The expectation, therefore, is 
strongly in favor of its being formed normally before fertilization. 
