The nuclear couipoueucs ot' the sex cells ot' four species of cockroaches. 491 
and inasmuch as it staius with the same degree of intensity as that 
body, it is iudistinguishable from it. However, it is later clearly 
marked off from the chromosome-nucleolus (fig. 11 as the polarization 
stages are entered. As a rule, it is regulär in outline, but sometimes 
it appears roughened ,fig. 14). Throughout the remaiuder of the 
primary spermatocyte stages, it maintains its position near the chro- 
mosome-nucleolus. 
We shall now returj to the point where the history of the 
ordinary chromosomes was left and trace tlieir development through 
the ensuing stages up to the formatiou of the metaphase figures. 
3. The synapsis stages. 
a) The ordinary chromosomes. A decided change in the ap- 
pearance of the loops becoues evident in that some of the threads 
show a thickening which bejins at the looped or anti-polar portion 
fig. 14) and proceeds polewad, ultimately affecting all parts of the 
loops. At first only a few 'oops exhibit this sudden thickening 
fig. 16) as may readily be setu in optical sections (fig. 15). Düring 
this process, such au optical ection shows clearly that there is at 
the same time a reduction in mmber of the threads (compare fig. 15 
with fig. 12 '. When, finally, tlis thickening has extended to all of 
the loops (fig. 17), an optical setion (fig. 18) shows that the number 
of loops is sixteen — a fact whih, under favorable conditions, it is 
possible to determine in lateral \ew. In optical sections, such as 
the one given in figure 18, the pints where the arms of the loops 
intersect the optical plane are so lefinite that one may count tliem 
with certainty. Some of the pointsappear dumb-bell in shape, thus 
leading to the assumption that the treads are longitudinally doubled. 
Such a longitudinal doubling is mot evident at a little later stage 
fig. 19). The writer has attemptedto observe a side-by-side con- 
jugation of the threads of the earlie stages to form the thickened 
looj\s of the later series, but in thi. he was not successful. The 
writer believes, however, that such aninterpretation of the behavior 
of the ordinary chromosomes during iese stages is warranted bv 
the evidence of reduction in the numbe of threads accompanying the 
thickening and that the process is ontof parasynapsis. A further 
treatment of the evidence for this conclröon will be given in a later 
section of the paper. 
b) The odd chromosome. During thcsynaptic stages, the chro- 
mosome-nucleolus remains on the outer b«-der of the loops, with its 
