ZOOLOGY: A. H. STURTEVANT 
557 
is present, as is shown in column 3 of table 1 and in figure 3. When 
the other piece of the Nova Scotia chromosome, lying to the right of 
purple (and, in many experiments, that part of it that is also to the 
left of speck) is tested, a still different result is obtained, as shown in 
column 4 and figure 4. 
TABLE 1 
PERCENTAGE OF CROSSING OV^R 
GENES 
Normal 
r 1 r 
*^nl *-iir 
'-III 
^iir 
^iir 
Star black 
39.3 
0.0 
0.0 
42.4 
38.2 
0.3* 
Star purple 
40.2 
0.0 
X 
45.4 
40.6 
0.3* 
46.5 
X 
25.9 
47.3 
X 
X 
Star speck 
48.3 
0.4 
49.0 
46.0 
49.0 
47.4 
Black purple 
5.9 
0.2 
0.4 
5.8 
3.6 
0.1 
Black vestigial 
16.9 
1.2 
X 
8.9 
X 
X 
Black curved 
22.7 
1.3 
25.6 
7.5 
X 
X 
49.3 
1.1 
48.8 
9.3 
43.5 
47.0 
Purple vestigial 
11.1 
1.2 
X 
0.6 
X 
X 
19.0 
1.1 
20.5 
1.7 
X 
X 
Purple speck 
46.5 
1.2 
47.4 
2.8 
42.2 
47.4 
36.7 
0.2 
X 
0.0 
X 
X 
Curved speck 
29.8 
0.1 
35.2 
0.2 
X 
X 
fTotal length 
94.0 
1.4 
56.1 
50.1 
84.0 
47.8 
* These few (6) crossovers are doubtful. None of them were tested; and there is apt to be 
a small percentage of error in classifying Star flies. 
t Calculated as Star black + black purple + purple curved + purple speck in all but the last 
two columns, where purple speck is used. In no case further corrected for double crossing 
over. 
It follows from these results that the original Nova Scotia chromosome 
contained at least two factors causing reduced crossing over, and each 
affecting chiefly the region in which it lies. We may call these two fac- 
tors 'Ciii' and ^Ciir' (C for crossing over, after Muller,^ II to dis- 
tinguish them from similar factors affecting other chromosomes, / and 
r for left hand and right hand) . 
By appropriate matings it has been possible to obtain females homo- 
zygous for Ciir, and a surprising result has been observed: the effect 
produced by Cur when heterozygous disappears when it becomes 
homozygous (see columns 5 and 6, figures 5 and 6). There can be no 
doubt of this result, as it has been obtained repeatedly, and has often 
been checked by tests of the offspring. It has also been paralleled more 
recently by results obtained by Muller^ (1916) and by me'^ with Cm. 
Similar tests of homozygous Cui have not been possible since no chromo- 
