22 
W. E. AGAR. 
the spindles already rotating in preparation for the second 
division. The details of this mitosis are not so clear as those 
of the first, owing to the fact that the chromosomes are apt 
to lose their regular outlines, and to adhere together to a 
certain extent. A pair of metaphases are shown in PI. 5, 
fig. 32. Owing to the rotating of the spindles one is seen 
from the pole, the other from the side. In the side view 
we see many of the chromosomes of the same “ tetrad” 
shapes as those appearing in metaphase I. This has pre- 
sumably arisen by a longitudinal division of the anaphasic 
chromosomes of the first division, with partial separation of 
the halves. 
The shapes of the chromosomes in the anaphase II cannot 
be satisfactorily made out, as they always appear to stick 
together and separate as two solid plates in which the out- 
lines of the individual chromosomes are only rarely plainly 
visible. Here and there one can often be seen to be dumb- 
bell-shaped, but they lose their transverse constrictions in the 
later anaphase (PI. 5, fig. 33). 
In spite of the obscurity of the details of this division it is 
plain enough that the chromosomes divide longitudinally, and 
not by the transverse constriction which appears in the pro- 
phase of the first division. It is by no means clear what is 
the meaning of this constriction, but it doubtless corre- 
sponds with the apex of the V's of the somatic or 
spermatogonial mitoses. PI. 1, fig. 7 shows a, 
daughter-plate of one of the smallest spermatogonia, in which, 
therefore, the chromosomes are comparatively small. An 
examination of the smaller elements shows that when the limbs 
of the V J s are very short we get an appearance strongly 
resembling the bipartite chromosomes of the meiotic prophase. 
The fact that the transverse constriction of the univa- 
lents, i.e. one of the two constrictions of the metaphase 
“ tetrads,” does not play any part in the second division will 
not come as a surprise to anyone who has followed the trend 
of recent cytology. Even the Copepod tetrads — -these classical 
examples of reduction by means of one longitudinal and one 
