Studies in the Morphology of Torreya californica.207 
later, the first stage in the germination of the megaspore had been 
reached, and the embryo-sac was bi-nucleate (Fig. 1.) Four- 
nucleate and eight-nucleate embryo-sacs have also been found, the 
latter occurring on July 6th. Later on numerous free nuclei are 
seen in the protoplasm lining the embryo-sac. Wall formation 
seems to proceed by in-growing tubes as described by Mile. Sokolowa 1 
for so many Gymnosperms. In two cases I have seen the alveoli 
from opposite sides of the embryo-sac approaching the centre, but 
not yet meeting. In one of these cases archegonia were already 
present. Mile. Sokolowa says that Cephalotaxus is exceptional in 
that the partitioning of the alveoli begins before they have met in 
the middle region, and in this point Torreya agrees with Cephalo¬ 
taxus. 
Jager mentions that in Taxus there is an axial core of cells 
which are smaller and richer in protoplasm than the rest, and have 
conspicuous nuclei. He also describes the cells at the base of the 
endosperm as small, rich in protoplasm, and meristematic, but he 
does not speak of them or figure them as forming a distinct layer. 
In Torreya californica there is no special axial core, but about the 
middle of August differentiation begins to occur in the basal region. 
The outermost layer of the endosperm gradually becomes distin¬ 
guishable from the rest by its conspicuous nuclei and contents and 
its meristematic character (Fig. 2). The nucleus is generally 
centrally placed and suspended by protoplasmic filaments (Fig. 3). 
When sections of an ovule gathered at the end of August were 
treated with iodine it was found that this layer was rich in starch 
while the rest of the endosperm was almost or quite starch-free. In 
sections of fresh material collected at the end of September the layer 
very conspicuous. It was one cell thick through most of its extent, 
and reached rather less than half-way up the endosperm, enclosing 
th& rest of the tissue in a kind of cup (Fig. 28). During thegrowth 
of the prothallus which takes place chiefly in the downward direction 
(cf. Figs. 4 and 28) the surrounding nucellar tissues become crushed 
and apparently absorbed (Fig. 2). It seems possible that the outer 
layer of the base of the prothallus may be of an epithelial nature,, 
and secrete a ferment. In that case the growth of the prothallus 
through the nucellus would be comparable with the growth of an 
endogenous root enveloped in its “digestive sac.” 
1 C. Sokolowa. “ Naissance de l’endosperme dans le sac 
embryonnaire de quelques Gymnospermes.” Bull, de la Soc. 
Imp. des Nat. deMoscou., 1890, p. 446. 
* L. Jiiger. “ Beitriige zur Kcnntniss der Endospermbildung und 
zur Embryologie von Taxus baccata.” Flora, 1899, p. 241. 
