GENETICS: R. GOLDSCHMIDT 
57 
h. The grade of intersexuality in one culture is subject to a regular and 
continuous variation around a mean. If we divide the distance between fe- 
maleness and maleness in 100 grades a given culture would show a symmetric 
variation of a certain range, say 20 grades, around a mean of 50 or 60, etc. 
What causes certain differences in the position of the modal class in sister cul- 
tures is not yet clear, but will be very important for the theoretical side of 
the question. It is to be expected that at the two end points of the series the 
individuals overlap on the normal, i.e., respectively normal females and males. 
This was actually found in the border cases. If all individuals which are ge- 
netically females are changed into males some minus individuals may be 
recognizable by their somewhat female shape of the wings, proving that the 
variability extends into the normal, too. It is to be hoped that the statistical 
treatment of these facts, together with the F2 results and the data regarding 
the male intersexes will open a way to an exact calculation of the relative val- 
ues of potency. 
c. In the crosses producing nothing but males, occasionally there appears 
a single normal female, hatching as the last individual of a given culture. It 
is very probable that we face here another case of non-disjunction (Bridges). 
A spermatozoon without X-chromosome ought to give a normal female with 
every egg (assuming naturally that the male factors are carried in the X- 
chromosomes) . So far I have succeeded only in finding a single spermatocyte 
II with 30 instead of 31 chromosomes. 
So far only the facts regarding the female intersexes have been re- 
corded, and they form a strong support for the views of my earlier papers 
in regard to the sex-problem. But the new results about the male inter- 
sexes show that in one important point the hypotheses have to be changed. 
I stated there that the male intersexes appear in F2 from the reciprocal 
cross, not giving female intersexes in Fi. The fact that exactly one- 
eighth of the males were intersexual corroborated strongly the Mende- 
lian formula of sex-inheritance used by me, working with two pairs of 
sex factors. The new results give a different aspect to the facts: 
1. The appearance of the male intersexes in the said crosses is a single event 
conditioned by the two races involved in the cross. The ratio \ is typical, 
too, only for the special combination. In other crosses any ratio between 0 
and 50 percent may be typical. 
2. A new fact of decisive importance is that male intersexes may be pro- 
duced in Fi. So there were a few individuals in the crosses female K X male 
S and female O X male H. And in the cross: weak Japanese female K X 
weak Japanese male H, nothing but intersexual males appear m ri. 6 These 
facts are of greatest importance for the whole question. Together with some 
other F2 results and the insight in some former errors of interpretation they 
make it very probable that the female part of the formulae F F M m = 9 
