337 
Actinostrobus pyramidalis , Miq. 
evidently before the splitting of the chromosomes, and sixteen can be 
counted on the equator of the spindle (Fig. 36 a). At this time no segrega- 
tion of the chromosomes into two groups (representing o* and $ ) could be 
seen, but this would hardly be expected except at an earlier stage of division. 
The spindle may be oblique to the long axis (Fig. 36) or parallel to it 
(PL XXVII, Fig. 37). It may be that it starts parallel to the plane of 
fusion of the maternal and paternal nuclei, and subsequently becomes parallel 
to the long axis, but the number of preparations of these stages is not 
sufficient to substantiate the suggestion. Fig. 37 shows the two daughter 
nuclei reorganizing, with the spindle still quite evident between them. No 
cell-plate is laid down and the spindle disappears completely. 
The basal nucleus next divides to form two daughter nuclei, placed one 
above the other, or somewhat obliquely (Figs. 38 and 39), and is very 
quickly followed by the upper nucleus, which divides in a transverse plane. 
We have thus the arrangement represented in longitudinal section in Fig. 40, 
in which it is most clearly shown that cell-plates have been formed from the 
spindles, traces of which can still be seen between the pairs of daughter 
nuclei. The segmentation is then completed by the formation of a cleavage 
plane between the two pairs of daughter nuclei, giving the arrangement 
shown in Fig. 41. 
The later stages are sometimes difficult to interpret, because it is often 
impossible to say whether a group of embryonic cells is formed from one 
fertilized archegonium, or from two adjacent archegonia, the wall between 
adjacent archegonia being exactly like that between adjacent proembryo 
cells. In any case, it is certain that some cell divisions follow the stage of 
Fig. 41, probably only occurring in the two basal cells, but it is unlikely 
that their sequence or number is constant. Parts of proembryos are shown 
in section in Figs. 42 and 43, in which the divisions in progress are probably 
the last which occur in the formation of the mature proembryo. 
The mitotic figures of Fig. 42, from both the sections in which they are 
seen, are drawn on a larger scale in Figs. 42 a and 42 b. The stage of 
division is slightly later than that shown in PI. XXVI, Fig. 36, after the 
splitting of the chromosomes, and thirty-two chromosomes, i. e. sixteen 
for each daughter nucleus, can be counted in each mitosis. 
Each cell of the mature proembryo, except the two apical cells, which 
appear to take no further part in the development, gives rise by a mitotic 
division to two cells of very unequal size, the larger of which becomes the 
suspensor, and the smaller the embryo initial. This mitosis always occurs 
towards that side of the cell nearest the basal end of the prothallus 
(Figs. 44 and 45). In this division the sporophytic number of chromosomes 
(16) has again been counted (Fig. 45#)* The two resulting cells (sus- 
pensor and embryo initial) are shown very shortly after their formation 
in Fig. 46, and somewhat later in Fig. 47, when the difference in size of the 
