664 
Notes. 
to one another, and in some places these parallel portions present the 
appearance of a single thread which has undergone longitudinal 
fission. That this is not the case, however, appears probable from 
the sudden divarications of these portions of the thread and the way 
in which they often lie across each other, and also from the fact that in 
later stages the divarications are not so conspicuous. As the thread 
thickens the parallel portions become more regular in their disposition, 
and finally transverse fission divides it into a number of chromosomes, 
each composed of two portions lying more or less exactly parallel 
to one another. Sometimes the two portions of a chromosome form 
a loop which is possibly derived from a loop in the original thread, 
and sometimes they are twisted round one another. Thus it appears 
that the double form of the chromosomes before the formation of the 
nuclear plate is in this case not due to longitudinal fission of the nuclear 
thread, but to the lateral approximation of portions of it in pairs. 
As the chromosomes arrange themselves at the equator, they 
become simultaneously shorter and thicker, and they dispose them- 
selves so that the plane of division between their two short rod-like 
parts is vertical and is not in the equatorial plane. These two parts 
are in close contact with one another and seem fused together at 
their inner extremities, while their outer or peripheral ends are often 
slightly parted. So that when seen from the poles at this stage they 
appear triangular, quadrate, or ring-shaped. Viewed from the equa- 
torial plane each chromosome has the characteristic humped form. 
In the nuclear plate each undergoes horizontal longitudinal fission 
beginning at the inner end (so that seen in profile each chromosome 
now appears T-shaped), and as this proceeds two daughter-chromo- 
somes are formed. As these latter separate from one another, the 
rod-like portions, which form them, diverge from one another, so that 
a diamond-shaped space is enclosed between the two daughter-chro- 
mosomes. When the latter are approaching the poles their two rod-like 
portions part asunder from one another, so that at the poles there appear 
twice as many short, straight chromosomes as there were at the equator. 
From the process described it appears probable that each chromo- 
some in this karyokinesis represents two of previous nuclear divisions, 
which have become more or less completely united end to end. 
Their double nature is shown in their mode of origin of the two 
parallel portions of the chromosomes which exist prior to the forma- 
tion of the nuclear plate and in the separation of the two parts of the 
