270 T. H. Morgan 
embryos. These embryos become female- and male-producers— 
the winged migrants—whose eggs are large or small. ‘The former 
contain six chromosomes in the polar spindle of nearly equal size 
which seem to represent the same combinations existing in the 
preceding generation. I[n the male spindle a difference appears, 
there being one largest, four intermediate and one smallest chromo- 
some. The only way in which such a combination can be accounted 
for, on the assumption that here too the original number 1s still 
retained—as in P. fallax—is that the two smallest of the eight have 
shifted over from their larger partners and have united with each 
other; the larger partners have also combined to produce the very 
large chromosome. ‘The “purpose” of the shifting 1s a preparation 
for the polar division in which one of the larger chromosomes and one 
of the smallest passes entirely out of the male egg to leavesix chro- 
mosomes in the egg.2_ These six often appear as five, owing to the 
subsequent union of the largest and the smallest sex chromosomes. 
Hence we find in the spermatogenesis two pairs of ordinary chro- 
mosomes that unite during synapsis and two sex chromosomes that 
have no pairs. ‘These, in consequence, pass over to one poleat the 
first spermatocyte division that represents the “reduction division.” 
The theoretical questions connected with these results will be dis- 
cussed later. 
PRECOCIOUS SEXUAL DIFFERENCES CORRELATED WITH MALE AND 
FEMALE EGGS 
The relations described in the preceding section throw an impor- 
tant light on the differences in size of the male and female eggs. 
It is obvious that the difference is connected with the precocious 
development of the sexual individuals and not withthe condition of 
maleness or femaleness of the egg, as such. In other animals size 
differences may also exist, the female being much larger than the 
male, yet the eggs are all of the same size. The differences in 
adult size in the latter case result from differences in the rate or 
time of growth after hatching. It is therefore of unusual interest 
2Tn P. fallax the same result is probably shown when ten chromosomes appear in the polar spindle, 
two being of greater size. 
