gto} CHAMBERLAIN—DIOON EDULE 423 
same mass, shows the same number. The much smaller eggs of 
Coniferales have comparatively short free nuclear periods, showing 
only 32, 16, 8, 4, and 2 free nuclei; while in the small egg of Sequoia 
the first mitosis is followed by a wall. 
While it might be tempting to regard this as a reduction series, 
and to regard forms with a large number of free nuclei as primitive, 
it seems probable that the number of free nuclei is correlated merely 
with the size of the egg. That the large number does not prove 
antiquity is shown by the fact that Dioon has 1000 free nuclei, 
while Pinus has only 8 (i.e., walls appear in connection with the 
8-nucleate stage), and yet Pinus is much more ancient than Dioon. 
We should expect to find the critical forms among the lower Cycado- 
filicales as they began to be differentiated from some heterosporous 
fern ancestry. These forms, which are still to be discovered, 
should have small eggs with complete segmentation not preceded 
by any free nuclear period, the free nuclear feature appearing later 
in connection with an increased size of the egg. Although this is 
highly theoretical, we believe that it suggests an explanation of 
the structures as we find them in living gymnosperms. 
DIOON; THE CELLULAR EMBRYO 
At the last mitosis of the free nuclear or evanescent segmenta- 
tion period, permanent walls are developed at the base of the 
proembryo, the walls in the upper part being evanescent and 
disappearing as before. The nuclei in this upper region do not 
divide any more and many of them disorganize. In the basal 
region nuclear division, always accompanied by cell division, 
progresses rapidly; and while one can still estimate with some 
degree of exactness the extent to which the free nuclear stage has 
Progressed, a small-celled tissue is formed (fig. 7). At the stage 
shown in this figure, there is practically no indication of regions 
corresponding to the “rosette,” suspensor, and embryo proper 
regions of many conifers. At this stage the archegonium jacket 
is still conspicuous, and at the base of the proembryo the jacket has 
become extended into a tissue several cells deep, contrasting sharply 
With the adjacent starchy cells of the endosperm. Cellulose walls 
have formed on the peripheral surfaces of the cells of this proembryo 
and the egg membrane is still intact. 
