192 BOTANICAL GAZETTE [SEPTEMBEx 
Jongitudinal anticlines (jigs. 18, 19, 20). Then by the appearance 
of other longitudinal and transverse anticlines a very large number 
of spermatogenous cells are formed. The nuclear divisions in these 
spermatogenous cells occur simultaneously over larger or smaller 
blocks, commonly extending over one-tenth to one-fifth the area of 
a longitudinal section of the antheridium, but never over the whole 
of it at once. In several of the antheridia examined there were found 
to be from 35 to 50 of these cubical spermatogenous cells on a single 
diameter, and from 125 to 160 of them in the length of the antheridium. 
This means that there are from 100,000 to 250,000 of these cells in a 
well-developed antheridium. Each of these cubical cells divides 
later by a diagonal wall to form two triangular-prismatic sperma- 
tozoid mother-cells. There are thus formed from 200,000 to 500,000 
spermatozoids in each antheridium. 
The organization of the spermatozoid in the mother-cell begins, 
as in other described liverworts, by the elongation and coiling of the 
nucleus. The presence of a blepharoplast was not demonstrated. 
When mature the spermatozoid is coiled to about one and a half tums 
in a flat spiral, whose axis is perpendicular to the broader side of the 
triangular-prismatic mother-cell. 
The most striking peculiarity shown in the development of the 
spermatozoid is the fact that the individuality of the chromosomes 
is visibly persistent in the ripe spermatozoid. Careful study of the 
mitotic figures in spermatogenous cells at various stages of develop- 
ment showed the number of chromosomes to be eight or ten. In 
preparations of ripe antheridia, which had been fixed in Flemming’s 
solution and stained in Flemming’s triple stain, when washed s0 as 10 
show well the chromatin in the vegetative nuclei round about, the 
spermatozoids appeared as single dark blue coils. When wie 
the sections were washed out more completely, so that even the nu 
were of a faint blue, the color remaining in the spermatozoid was . 
fined to a number of fine threads of nearly the length of the sperm | 
tozoid. These threads were twisted about each other slightly 9° ” | 
each thread in its length made a complete turn about the n® 
cylindrical spermatozoid (fig. 8a). A careful count of these er 
which could best be made in optical transverse sections of iY ve 
of the spermatozoid, showed that the number is constant and dea 
