in Lilhim Mart agon: II. Spermatogenesis. 207 
include a complete chromosome. Sections about 7-5 \x thick 
from Hermann-material are the most useful, and it is well to 
compare them with others about 5 V thick. Figs. 16-22 are 
drawn from such preparations. Of these, Figs. 20 and 22 
give the general appearance of the nuclear section as faithfully 
as I can render it. The others are slightly diagrammatised 
by the omission of such chromosomes and fragments as would 
make the figure obscure. In doing this I have of course been 
influenced in retaining or omitting detail by my own inter- 
pretation of the structure. This interpretation is based on 
very careful study of every detail in the karyokinesis. 
Most of the chromosomes retain their characteristic shape 
throughout the process of division (Figs. 19 and 21). The 
angle of the V lies in the equatorial plane of the spindle, and 
the legs usually point outwards or towards one pole or the 
other (Fig. 19). Occasionally the legs point towards different 
poles, and then the V-shape is lost, though the chromosome 
retains traces of its original form (No. 8 in Fig. 18, and the 
middle chromosome in Fig. 19). One result of this arrange- 
ment is, that when the chromosomes of a whole nuclear plate 
are seen sideways there seem to be more than twelve, for it is 
easy to mistake each leg of the V for a separate chromosome. 
The whole of the spindle is not shown in the nucleus to the 
right of Fig. 20, yet fifteen free ends can be counted on it. 
Besides the V-shaped chromosomes, there are usually two or 
three in each spindle which are straight or irregularly twisted. 
One end of such a chromosome commonly lies in the equatorial 
plane, and longitudinal fission of the segments begins there, as 
is the case also in vegetative nuclei. The segments of V-shaped 
chromosomes are also formed by longitudinal fission, but they 
begin to separate at the angle (Fig. 21). Hence most of the 
diaster-segments are V-shaped, but a few are hooked (Fig. 22). 
The reconstruction of the daughter-nuclei from the diaster- 
segments goes on in the usual way, and is shortly followed 
by the formation of a cell-wall between them. Each of the 
four segments of the mother-cell rounds itself off and begins 
an independent existence as a young pollen-grain, but for 
