36 BOTANICAL GAZETTE [jyuLy 
are, without doubt, nothing but the cut ends of fibers. The two 
centrospheres in fig. 77 are practically alike, but the one at the 
upper pole is represented in median section and the other in 
surface view, the fibers in vertical view appearing as dots. How 
ever, it must be admitted that occasionally the deeply staining 
points are really granules (fig. g), but the cases are so rare that 
we have not regarded such granules as a functional part of the 
mitotic mechanism. 
After a study of the germinating spore had failed to show 
any centrosomes, the nuclear figures were examined in other 
phases of the life-history, particularly in the apical cell and its 
younger segments, and in the developing antheridia. The 
apical cell and the rapidly dividing cells near it are quite favor- 
able for study. The character of the mitoses in this region is 
represented in figs.gand zo. The lower pole in fig. 9 shows 
the more usual condition, although the rays are frequently as 
strongly developed as those shown at the upper pole, a consid- 
erable number of the rays reaching to the Hautschicht. A care- 
ful examination of this figure will show that there is no definite 
centrosphere like those in figs. r2 and 73. In later stages (jig. 
10) the spindle becomes sharply bipolar and the radiations dis- 
appear. 
The antheridia were examined with particular interest because 
Schottlander (30) had reported centrosomes during all stages in 
the development of the antheridium of Marchantia, and Belajeff 
(2) had found blepharoplasts throughout the development of the 
spermatogenous cells of Marsilea. However, nothing which 
could be interpreted as a centrosome was found in our material, 
which furnished a series from the initial cell up to stages in 
which more than thirty cells appear in a transverse section of 
the antheridium. Unfortunately, the material showing the last 
two or three divisions preceding the formation of the spermato- 
zoid mother-cells was not satisfactory, and, consequently, 0 
positive statement can be made in regard to blepharoplasts, 
although we should assume them to be present during the last 
one or two mitoses. 
In the germinating spore a differentiated area, already 
