me) BOTANICAL GAZETTE [JANUARY 
probably divide simultaneously, since no 3-celled proembryos were 
found. No mitoses in the proembryo have been observed, nor any 
trace of evidence that the nuclei divide amitotically. I have 
already, in former papers (3a, 3b), called attention to the very 
curious fact that almost no mitoses in the critical stages of devel- 
opment of Araucaria have been observed. It is a very curious 
and puzzling fact, not to say a very annoying one. SAXTON has 
recently called attention to a similar state of affairs in another 
southern hemisphere form, Actinostrobus pyramidalis (18a). 
The four free nuclei may occupy almost any position with 
reference to one another. It has already been mentioned that the 
position of the fusion nucleus appears to depend on how much it 
is displaced through the violence of the contact between egg and 
sperm. It may lie near the middle of the archegonium, as it does 
in A gathis (8), or more generally near the bottom. The succeeding 
divisions take place wherever the fusion nucleus has been left. 
This same displacement would probably tend to conceal any 
polarity that the fertilized egg might possess. The commonest 
appearances of the proembryos are shown in figs. 20 and at. 
Sometimes the four nuclei may all lie at the bottom of the pro- 
embryo, as in figs. 23 and 25. ‘The subsequent divisions do not 
appear to follow any definite order nor are they simultaneous. 
Whether the 4 nuclei were tetrahedrally placed (figs. 20, 21), placed 
in a single vertical plane (fig. 23), or in a curved line around the 
bottom (fig. 23), or in any other position, seems not to affect the 
ultimate result. Irregular division continues for two weeks or 
more before the final arrangement of the cells in tiers. The number 
of cells or free nuclei at this time varies considerably. No counts 
of less than 32 nor more than 45 were obtained from an examination 
of a considerable number of embryos of about this stage. The 
number of proembryos showing the beginning of wall formation 
was.so small that it cannot be certainly said that some of those with 
32 free nuclei might not have had more at the time of wall forma- 
tion. Many of these seemed as large and as definitely arranged 
as the ones with a greater number of nuclei. It seems to me, 
therefore, that the number of nuclei at the time of wall formation 
is probably variable. 
