270 Allen . — The Spermatogenesis of Poly trichum juniperinum. 
antherozoid it is destined, as will be seen, to function as a blepharoplast. 
When a pair of sister androcytes can be recognized as such (as in Figs. 1 
and 2, PL XV), the blepharoplast in each is usually to be found in a position 
corresponding to that of a central body in the division recently completed. 
The blepharoplast is sometimes (Figs. 1-3) the centre of a system consisting 
of a few radiations much like those that surrounded the single central body 
which was found previous to the initiation of mitosis in each androcyte 
mother- cell. The cytoplasm immediately about the blepharoplast is 
commonly somewhat more dense than that in other parts of the cell. The 
cytoplasmic contraction and rounding up, which began in the androcyte 
mother-cells, have now gone on to a considerable extent. The walls (except- 
ing those laid down in the course of the last division) are much thickened 
and show evidence of softening. The chromatic substances within the 
nucleus still exhibit more or less of the forms of the chromosomes ; in some 
cells (Fig. 2) there is often no body as yet recognizable as a nucleole. It was 
pointed out in the paper just cited that nucleoles make their appearance 
much later in the androcyte nuclei than they do in the daughter nuclei 
formed by any of the androgonial divisions. However, in some androcytes 
of about the same age as those shown in Fig. 2 (e.g. Figs. 1 and 4), bodies 
appear whose nucleolar nature is evidenced by their usually rounded shape 
and their affinity for safranin. Later (Figs. 13, 14, &c.), a single fairly large 
nucleole is always present in the androcyte nucleus. But this nucleole, 
though sometimes, especially in the earlier stages (Fig. 4), rather irregular 
in shape, is usually more or less rounded ; it is neither so irregular nor so 
large as the nucleolar masses of the preceding cell generations, which were 
found to consist of a nucleole about which more or less chromatin was 
aggregated. 
The blepharoplast, originally (Fig. 1) very small, early begins to increase 
in size. In the androcytes shown in Fig. 2, each blepharoplast is somewhat 
elongated. The blepharoplasts of the cells shown in Figs. 3 and 4 (which 
are in the same antheridium as are those shown in Fig. 2) seem to have 
grown more or less in all dimensions, although that of Fig. 3 is longest in 
a direction oblique to the plane of the section. Whatever be the differences 
in this respect at this early period, the growth tendency very soon shows 
itself exclusively in an elongation (Figs. 5, 6 , 7, &c.). Excepting that it lies 
in the cytoplasm, I cannot find that at this time there is any rule governing 
the position of the blepharoplast. It may lie in contact with the nucleus or 
with the plasma membrane, or it may be at any distance, or its different 
parts may be at varying distances, from the nucleus on the one side and 
from the plasma membrane on the other. In the early stages of its elonga- 
tion (Figs. 5, 10, 12), the blepharoplast is likely to be more or less uneven in 
thickness ; sometimes (Fig. 5) one end is noticeably thicker than the other. 
But as its growth goes on (Figs. 11, 13), it becomes a curved rod of quite 
