The uuclear componeuts of the sex cells of four species of cockroaches. 495 
in the rings which result.from median attachment, these lugs appear 
for a time as projections in the plane of the cell’s equator. Finally 
when the daughter chromosomes have pulled out completely from 
one another (fig. 31), the lugs are lost. 
It will be seen that the rings lie vertically in the cell and 
tangentially to the spindle (fig. 30). Before the two daughter chro- 
mosomes break away completely from one another (text-figure III, f), 
they remain in contact at their ends while the ring continues to draw 
out, so that the result is the closing up of the interior of the ring 
(fig. 31) until, ultimately, there appears a very narrow space running 
longitudinally through the daughter chromosome (fig. 32). It is to be 
remembered that this space is not a true longitudinal split, but simply 
the area between the arms of the V-shaped chromosomes. A true 
longitudinal split does, however, appear under favorable conditions 
in these chromosomes, running parallel with the cleft just described 
(text-figure III, f) and raore conspicuously in those chromosomes 
which have been pulled out from subterminal attachment. This split 
is the line along which the chromosomes of the second spermatoc.yte 
division will divide and it becomes more evident during the prophase 
stages of that division. 
Figure 33 represents an optical section through the anaphase 
(fig. 32) of the primary spermatocyte. Seventeen bodies may be 
counted. Sixteen of these lying centrally, are the ordinary chromo- 
somes. In the telophase, they become massed together (fig. 34) and 
it is impossible to distinguish them as separate chromosomes. Con- 
spicuous interzonal fibres extend for a time from one nucleus to that 
of the other daughter cell (fig. 34). — We shall now return to 
the odd chromosome. 
b) The odd chromosome during the first mitosis. The odd chro- 
mosome may be traced through the primary spermatocyte divison 
stages as a homogeneous, pear shaped body, conspicuously placed 
at the^ periphery of the spindle and readily distinguishable from the 
ordinary chromosomes (figs. 25 to 34). In nearly all cases it advances 
towards one pole before any of the other chromosomes (fig. 27). There 
is no constancy with respect to which end is directed towards the 
pole (figs. 28 and 32). The longitudinal split is conspicuous at all 
times. In an optical section, such as is given in figure 33, one may, 
by focusing, assure himself of the presence of the odd chromosome, 
sixteen of the bodies being clearly seen to be composed of two 
portions in vertical direction while a sqventeenth chromosome (lying in 
Archiv f. Zellforschung. III. 33 
