Lnternal Factors of Sex Determination 403 
of spermatozoa, presumably both functional, there may be a 
connection between the two conditions. This tacitly assumes 
that all of the eggs are alike, which was apparently taken for 
granted by McClung, but not examined. Since the accessory 
is found in one half of the spermatozoa, McClung conjectured 
that those spermatozoa that contained the accessory produce 
males, presumably because of the seeming necessity for the new 
males to have an additional chromosome, the accessory. This 
argument proved invalid in the light of the observation of 
Wilson, that the female and not the male contains an additional 
chromosome, and of Stevens that the female of Tenebrio molitor 
contains the large chromosome (the homologue of the acces- 
sory) and the male the small chromosome. 
In order to show how this process works itself out, let us take 
the case of Anasa tristis, in which there are 21 chromosomes in 
the male somatic cells and 22 in the female. There will be 
rr chromosomes in half of the spermatozoa and ro in the other 
half. All of the eggs may be supposed to contain 11 chromo- 
somes. When a 1o-chromosome spermatozoén fertilizes an egg, 
there is formed a male with 21 chromosomes, whose sperm will 
again contain 10 and 11 chromosomes; when an 11-chromosome 
spermatozoén fertilizes an egg, there will be 22 chromosomes — 
the new individual will contain 22, and its eggs will later all 
contain 11 chromosomes. The accompanying scheme shows 
this result graphically : — 
Spermatozoén Egg Male Germ-cells 
Io ————> II = 21 ———>_ roorrr 
or or Female 
II ————-, III = 22 —_+>  rrorir 
The evidence that the eggs are all alike and contain the larger 
number (z.e. 11 in the case just cited) is indirect, but apparently 
conclusive. The chromosomes of the polar spindles have not 
been counted, but the somatic cells contain 24 chromosomes, 
as well as the early germ-cells before the pairing of the chromo- 
somes. Moreover, in those cases in which the accessory of the 
