1902] GAMETOPHYTES AND EMBRYO OF PODOCARPUS Iot 
(Arnoldi, 1900, 6). The two nuclei here lie side by side, and 
are not surrounded by starch or furnished with so distinct a 
__ protoplasmic sheath as in the Cupresseae and Taxodieae. A 
denser and more granular area can be observed in immediate 
contact with the nuclei, but it is not sharply defined and fades 
__ imperceptibly in the general protoplasm of the archegonium. 
In fig. 33 the second division has just occurred, and the nuclei, 
| only two of which are shown, have not yet reached their full 
| size. When the sixteen-celled stage is reached, the protoplasm 
: of the embryo has become separated from the disorganizing 
| area above (fig. 34). There is probably no further division 
| 
before cell walls are formed, but the absence of important stages 
here leaves this point in doubt. 
Fig. 38 shows the proembryo just as it is breaking through 
o base of the archegonium. It consists here, as invariably, of 
_ three tiers. First is a rosette of generally fourteen nuclei, which 
ni oo from one another by cell walls, but are in open 
S emication with the archegonium above; next come the 
Wace. aay fourteen in number, and below there 
a fo ree with two nuclei not yet separated by a wall. 
8 oo noticed that the walls between the rosette nuclei 
he eatnee those separating the suspensors below. This 
evidence that the rosette and the suspensors are estab- 
: after cell walls are formed, through the division of an 
€ upper tier. It is in this way that they are — 
cium. Between rosette and suspensors a thick 
soon formed, a novelty, I believe, in the gymno- 
| (Figs. 38, 39, 41). As the suspensors elongate, 
Sourtary. to the rule, remain for some time at the 
wee (he. 39). This is probably connected with the 
= 
as 
