232 PLANT LIFE 
In this particular nuclear division we find 
that the chromosomes are not longitudinally 
divided and the moieties then distributed be- 
tween the two daughter nuclei, but that the 
whole process is carried through in another 
way. 
The earliest stages resemble those of an 
ordinary vegetative nucleus which is about 
to divide. The chromatin-containing §gela- 
tinous strands make their appearance, and the 
chromatin is arranged in parallel streaks. 
But instead of going on to differentiate and 
finally to divide, the chromosomes proceed 
to unite in pairs. We have very strong 
grounds for believing that in no case is this 
union a chance one, but that a chromosome 
descended from one contributed by the sperm 
unites with another corresponding to it but 
derived from the egg. In other words, each 
pair consists of a chromosome of maternal 
and a paternal origin. 
The net result, then, of the approximation 
and union of the paternally and maternally 
derived chromosomes to form the respective 
pairs is, of course, a reduction to one-half of 
the number apparently present. 
Each pair now behaves as if it were a single 
chromosome. They flock to the equator of 
the spindle, but when they divide there, what 
happens is simply a disjunction of the two 
members of each pair, one of the members re- 
treates to one pole, the other one to the other 
pole. Hence a real reduction is now effected, 
