ELIMINATION OF SEX CHROMOSOMES 497 
to get polar body spindles in this type. The winged individuals 
produce only male eggs, so far as I have found. Therefore the 
fifteen equatorial plates of the winged migrants supposedly belong 
to male eggs. Now it is well recognized that the large number 
of chromosomes is more likely to represent the typical number, 
for, when one or two are lacking they may be in other sections, 
or be cut, or obscured by the other chromosomes. It appears 
then that twelve chromosomes are present in the male eggs. 
I found only two cases in wingless individuals in which the 
number of chromosomes could be clearly counted, fig. II a U and 
W. In both ten chromosomes are present, but two are of double 
size and probably represent the four X’s fused together in two 
pairs. There is no visible evidence therefore of the loss of one 
chromosome in the eggs that give rise to the male-producers as 
I have postulated for P. caryaecaulis. | 
After the polar bodies have been given off from the large and 
thessmall eggs—unfortunately I have never found a spindle in 
process of division—two kinds of embryos are found, namely, those 
containing twelve and those containing ten chromosomes. It 
appears therefore that two chromosomes are lost from the smaller 
egg. Morever only ten chromosomes appear in the sperma- 
togonial cells. The evidence therefore may seem to point to the 
conclusion that in this species there is usually no loss of chro- 
mosomes in any of the eggs (at least none that can be pointed 
out) of the stem-mother at the time when the polar bodies are 
formed. Hence the separation into the male-producers and the 
female-producers at this time can not be shown to be due to the 
loss of one of the twelve visible chromosomes. 
It may appear therefore that the evidence contradicts the 
hypothesis offered to account for the change in the other species, 
but it will be noticed that the comparison rests in the assumption 
that the two equal XY chromosomes in the spermatocytes of P. 
fallax correspond to the large X and the small x of P. caryaecaulis. 
But it may equally well be true that there are two large X’s 
with two small x’s attached (that do not become visible) in P. 
fallax. In other words not only are there twice as many ordi- 
nary chromosomes but twice as many sex chromosomes also. The 
