342 T. H. Morgan 
product gives rise to female parts; the other sperm (or sperms) also 
developing, give rise to male parts. If in the bee, as in the phyl- 
loxerans, the sperms are allof one class, no such complications arise 
in these cases as might theoretically arise if gynandromorphism 
should appear in a species having two kinds of sperm, one with, 
the other without, an accessory. Here two categories of cases 
appear. If any egg can be fertilized by any sperm,1.e., no select- 
ive fertilization occurs, then, should the female-producing sperm 
enter and fuse with the egg pronucleus, the resulting parts would be 
female—if another suchsperm also entered at the same time, but did 
not fuse, it would give rise to male parts. Should this other sperm 
be a ‘“‘male-producing sperm,”’ so called, it might either produce 
male parts for all we know to the contrary, although it has no 
accessory, or else it might not produce normal development of 
any kind. 
On the theory of selective fertilization the following scheme 
would give the results: 
If a female egg carrying the female determinant be entered by 
more than one sperm carrying the male determinant and one com- 
bines with the nucleus, the parts that result from this combination 
will become female (since the female determinant dominates) the 
other sperms can produce male parts, hence an hermaphrodite 
results. But if a male egg (carrying the male determinant) be 
entered by more than one female sperm carrying no sex determi- 
nant, the derivatives of the combined nucleus will produce a female 
and that derived from the male sperm will produce male parts. 
There is still another way in which gynandromorphism may be 
regarded. If one sex is heterozygous the simultaneous but inde- 
pendent development of dominant and recessive characters would 
produce gynandromorphism; which could occur only in that sex 
that is heterozygous. Should both sexes produce gynandro- 
morphs, the result would mean either.that both sexes are heterozy- 
gous; or that abnormal fertilization, as explained above, is respon- 
sible for the effects. 
The following evidence on the effects of castration have a bear- 
ing on the problem of sex. 
It has been shown for the mammals and birds especially that 
