The nuclear couqionents of the sex cells of four species of coekroaches. 493 
fused. Whatever its nature, it is tbe line of Separation of the cliro- 
mosomes of the primary spermatocyte metaphase. 
The bipolar arrangement is of short duration, judging by the 
scarcity of such stages in the material examined and gives place tö 
the stage represented in tigure 24, where the chromosomes lose tlieir 
attachment to the poles and become distributed throughout the nucleus 
as short rods (fig. 25). The chromosomes are constantlv sliorteuing 
and becoming capable of deeper staining through these phases. Tbe 
longitudinal split is seen throughout. The spindle uow becomes evident 
(fig. 26) witli its fibres attached to the chromosomes in such a way 
that they approach the equator of the cell as rods, longitudinally 
split and of slight curvature. 
These stages may be observed in a single section of one cyst 
and therefore it is probable that they are rapidly passed through. 
The continuity is also assured by the fact that the cells lie close 
together in a single cyst. The spindle fibres do not become evident 
until a comparatively late stage (fig. 26) so that it is impossible to 
Orient, definitely, such a cell as is shown in tigure 25, for the centro- 
somes are indistinguishable from certain mitochondrial bodies which 
lie around the nuclear membrane during these stages. The chromo- 
somes are now ready to enter the metaphase, but a description of 
this process will be deferred until the beliavior of the odd chromo- 
some and plasmosome has beeil given. 
b) The odd chromosome. The peculiar pear shape of the odd 
chromosome, so characteristic throughout the earlier stages is retained 
through all of the phases just described for the ordinary chromosomes 
(figs. 19, 20, 22, 23, 24, 25 and 26). It retains also its longitudinal 
split. Inasmuch as it is impossible to determine which centrosome 
is migrating in such stages, as are represented in figures 22 and 23, 
it is not possible to discover whether the odd chromosome migrates 
or whether it invariably remains attached to the stationary pole. When 
the clh'omosomes begin to move away from tlieir respective poles 
(fig. 24), the attenuated end extending out in a thread is lost to view 
and the odd chromosome becomes apparently free from any attach- 
ment to the other chromatin elements (fig. 25). When the chromo- 
somes arrange themselves in the equatorial plate, the odd chromo- 
some lies invariably in the outer series of spindle fibres (fig. 27). It 
is easily distinguished from tbe ordinary chromosomes throughout 
these stages and it may be found in at least some of the cells of a 
given stage. It is not represented in tigure 21, for it lies beneath 
