334 T. H. Morgan 
Doncaster, Raynor and Durham have brought forward some 
astonishing facts connected with the assumption of certain charac- 
teristics by one or the other sex and their transmission. The 
evidence is too complicated to go into here but the assumptions 
necessary to account for these results throw important light on sex 
as determined by one or by the other parent." Experiments along 
these lines bid fair to give a further insight into the problem. 
Doncaster based his first interpretation of the facts for Abraxas 
on the assumption that each sex gives off male and female bearing 
gametes, and on the further assumption that in the first generation 
there is a coupling so that the male ova bear the A. grossulariata 
character and the female ova the A. lacticular character; further 
that in the gametes of the male there is no coupling, that after 
fertilization dominance attaches to the sex brought in by the ovum. 
Bateson and Punnett have simplified this hypothesis and explained 
the facts by two (three) assumptions; (1) that the female is heter- 
ozygous for sex, the male homozygous; (2) that femaleness is domi- 
nant to maleness; (3) that in the first generation whenever the 
two dominants, femaleness and the grossulariata coexist there is a 
repulsion between them so that each cell gets one or the other of 
these factors, not both. Here the assumption made for the insects 
on the basis of primordia determining sex is reversed, the female is 
heterozygous in the moth according to Punnett and Bateson, and 
the male, according to the McClung-Stevens-Wilson hypothesis. 
Wilson asks, apropos of Bateson’s and Punnett’s view, if the male 
is homozygous what is the meanig of the formation of two kinds 
of spermatozoa in many insects? In the case of the bee the con- 
tradiction is even more inexplicable without a further assumption. 
One escape would be to deny that the accessory chromosome, 
which alone gives a difference in the two kinds of sperm, has any- 
thing to do with sex determination, yet a definite relation between 
that chromosome and sex has been demonstrated and it 1s difficult 
to believe that this constancy has no relation to sex determination. 
Another way out of the dilemma would be to assume that in some 
cases the discriminating factor lies in the male and in other cases 
in the female. Here the facts for Phylloxerans furnish the data to 
support such a view. ‘The parthenogenetic egg contains the possi- 
