394 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9. 
is effected through the agency of parallel fusions, but by far the greatest 
amount of it is brought about in other ways. 
As the nucleus advances into prophase, the amount of parallelism does 
not increase materially, but the reticulum loses the even and uniform 
appearance that has been so characteristic of it (fig. 4). The meshes 
begin to show a marked difference in size, the threads are no longer so 
evenly distributed, and are becoming quite unequal in diameter. The 
difference in size of the meshes seems to be due mainly to the fact that 
some of the threads are contracting and becoming thereby shorter and 
thicker. This process serves to bring threads which have been lying far 
apart nearer together, so that they occasionally parallel one another quite 
closely. It also helps to bring about a variation in the diameter of the 
threads. Those which contract will become thicker thereby. Further- 
more, contraction on the part of some threads will result in the stretching 
and attenuation of others. Another process seems to be going on at the 
same time, however, which is doing much to make some threads thicker 
than others. There are indications of a flow of material from one thread 
to another, or from one part of a thread to another part, resulting in a 
marked thickening of some parts and the absorption and disappearance 
of others. One may observe everywhere threads whose contents seem to 
have passed to adjacent knots or threads, causing these to enlarge, while 
the former have themselves become very thin and are either on the point 
of breaking or have already broken (figs. 4-7). It is evident that certain 
threads are being sacrificed, and that the material of which they are com- 
posed is being transferred to the threads which are to survive, or is being 
temporarily stored in the ''knots," formed where the threads join. A 
process of this kind, involving a certain amount of contraction, will neces- 
sarily result in occasional parallelisms. They are not sufficiently numerous 
in this material, however, to warrant the belief that they are of any far- 
reaching significance. They are what might be termed "chance" paral- 
lelisms. 
Synizesis. The condition of synizesis is apparently brought about by 
the continued contraction of the threads of the network. There is so 
much pull exercised upon the peripheral threads because of this shrinkage 
that they become torn away from the nuclear membrane on one side, 
and the contracting mass of threads shrinks into a bundle against the 
opposite side of the nucleus. Either before this breaking away occurs or 
while the bundle is being formed, the nucleolus leaves its central position 
and comes to lie against the membrane in the midst of the reticulum (figs. 
4» 5» 8, 9). Here it becomes somewhat flattened and cushion-shaped. 
The position of the synizetic knot is of interest. One cannot help 
being struck by the fact that during synizesis the threadwork lies, in the 
vast majority of cases, on that side of the nucleus which is nearest the 
periphery of the cell. Just what significance should be attached to this 
