AND OF HYBRID PIGEONS. 
25 
the previous generation, although only half the number is 
present. There may exist also a residual nucleolar-like mass 
of linin at some point in the nucleus. The spindle is a trifle 
shorter and slenderer than that of the previous division. The 
number of fibers appears to be four, thus again corresponding 
with the number of chromosomes. 
The manner of dividing is like that of the parent cell; tlie 
chromosomes each break in such a way that a stringing out of 
the sheaths gives rise likewise to a system of interzonal libers 
which, as division proceeds, construct at the equator to form an 
intermediate body. The cytoplasm throughout, seems to re¬ 
main in a comparatively quiescent condition, exhibiting no per¬ 
ceptible changes. Chromatin fragments may be present in 
some instances. 
The reconstruction of the daughter nuclei differs in some 
respects from that of the preceding generation. The chromo¬ 
somes do not fuse so readily, but lie side by side for some time 
in the form of four hollow vesicles. These usually fuse later 
v/ith the result that one large hollow sphere of chromatin is 
formed near the center of the nucleus, giving the spermatid 
a very characteristic appearance (Fig. 1, st.). Numerous fine 
fragments of chromatin migrate to the nuclear membrane which 
has appeared in the meantime, and form a thin shell along its 
inner surface. Fibers of linin radiate out from the central mass 
to the periphery of the nucleus. The centrosome persists and 
together with the tip of the spindle moves out into the cyto¬ 
plasm to one side of the nucleus after the nuclear membrane is 
formed (Fig. 16, i.). The remnant of the spindle which sur¬ 
rounds the centrosome seemingly becomes reconverted into 
sphere substance. The new cell thus formed from the division 
of the secondary spermatocyte is the spermatid and is the cell 
which will ultimately be transformed into the spermatozoon. 
REDUCTION. 
In the division of the spermatogonia there is very clearly . 
a longitudinal splitting of the chromatic loops. In the sperma¬ 
tocytes of normal pigeons there is no satisfactory evidence other 
than the remarkable decrease in the number of chromosomes to 
indicate the exact nature of the division but in hybrids, owing 
