16 
W. E. AGAR. 
by these figures we are concerned with four separate pro- 
cesses, which do not always keep exact pace with one another. 
These processes are : the increase in intensity of synizesis up 
to a maximum, and then its gradual loosening out; the 
shortening and thickening of the chromosomes ; the complete 
separation of the ex-conjugants already mentioned ; and the 
development of a transverse constriction across each (uni- 
valent) chromosome. 
In PI. 2, fig. 17, synizesis is very pronounced, and the 
chromosomes have contracted considerably since the stage 
shown in PI. 2, fig. 16. Most of them are probably still in the 
form of rings, though the dense mass in the centre does not 
allow of any of them being followed out completely. There is, 
however, one free end which must have been formed by the 
breaking open of a ring. At s is seen another ring in the 
act of breaking through at the point of junction of the conju- 
gants. 
The completion of the process of separation of the conjugants 
can be followed in figs. 18-24. The^e are all in thick 
celloidin sections (with the exception of fig. 24), and untouched 
by the razor (except figs. 21 and 24). The dissociation of 
the conjugants, therefore, cannot possibly be ascribed to 
dislocation by the razor. 
In PI. 2, fig. 18, owing to the still considerable length of the 
chromosomes, and to the fact that synizesis is not quite 
resolved, it is not possible to determine the limits of all 
the chromosomes. Among those which can clearly be made 
out there is one complete bivalent ring’ (bl), three bivalents 
in which the constituents are united by their ends only 
(&2-64), two pairs of univalents, still, however, united by 
a thin thread (si, s 2), and a number of free univalents. Thus 
we get here every stage in the breaking apart of the rings. 
In the still complete ring bl it can be seen that the conju- 
gants are in the act of tearing apart at one junction, while 
the attachment at the other one is not intimate. 
In PI. 2, fig. 19, the synizesis is still at its height, and only 
a few of the chromosomes can be determined. Among the free 
