26 
SPERMATOGENESIS OP NORMAL 
to certain exaggerations which frequently occur, especially in 
hybrids from very distinct species, some very significant facts 
come to light. As we shall see later, in many hybrid pigeons, 
the pairing necessary to the formation of bivalent chromo¬ 
somes comes about with more or less difficulty or not at all. 
The chromosomes from the respective maternal and paternal 
species seemingly strive to maintain their own individuality in 
preference to fusing with a foreign plasma. The result is that 
they remain apart on separate spindles in some cases or unite 
feebly, or possibly sometimes fuse with members of their own 
species when two spindles are present instead of pairing with 
the chromosome from the opposite sex as they would do or¬ 
dinarily. 
Where the two kinds of chromosomes refuse to unite, but 
divide in such a way that the paternal and maternal elements are 
set apart in separate cells, the effect is that of a qualitative 
reduction. In such exaggerated cases, then, what amounts to 
a reduction occurs in the primary spermatocytes, but this does 
not prove that the reduction ordinarily occurs here. If certain 
of the chromosomes can overcome their incompatabilities suf¬ 
ficiently to form bivalent chromosomes, as is true in the off¬ 
spring from closely related forms, then their mode of pro¬ 
cedure could be that of the ordinary bivalent chromosomes. 
To explain reversion and variation in the offspring of fertile 
hybrids, we are forced to the conclusion that a segregation of 
the paternal and maternal plasmas occurs in the germ cells, and 
the only chance for such a segregation, if the cells are of the 
normal type, is through the reduction division. It Avill be re¬ 
called that eight chromosomes were given to each secondary 
spermatocyte through the division of the primary spermatocyte, 
but only four appear when the cell is again ready for division; 
that is, the chromosomes have again united in pairs. In matur¬ 
ation the first division is ordinarily the equation division and 
the second the reduction division. If this be true in pigeons, 
then each of the eight chromosomes received by the secondary 
spermatocyte is bivalent. When these bivalent chromosomes 
fuse to form the four chromosomes of the secondary sperma¬ 
tocyte, then the latter must each be quadrivalent. A transverse 
division of such chromosomes could manifestly result in more 
