Cates .— The Tnsomic Mutations of Oenothera. 557 
Oe. nanella x rubrinervis. 
Lamarckiana 
I 
Lamarckiana 
I 
Lamarckiana 
subrobusta 
rubrinervis sub>obusta dwarfs 
rubrinervis rubr. subr. dwarfs dwarfs 
It may be worth while determining how many of these complicated facts 
of genetic behaviour can be explained by the simple hypothesis that in the 
origin of nanella and of rubrinervis there has been what we may call double 
non-disjunction in one or both parental germ cells, so that we have for 
A AACDEFG ^ chromosome being absent), while 
nanella the formula 
for rubrinervis we have 
BBCDEFG 
give an F x 
ABODEFG 
AACDEFG,AACDEFG 
+ 
Oe. nanella x rubrinervis will then 
The former combination would be 
BBCDEFG ABCDEFG 
L amarckiana subrobusta. 
expected to produce Lamarckiana , and in its pollen formation synapsis 
might be expected to take place between the A and B chromosomes instead 
of between the two A's and the two B’s. It should therefore breed true, as 
it actually does. The other F x combination {subrobusta) might be said to 
be heterozygous for an AB pair of chromosomes. When selfed it would 
therefore be expected to give the following: 
A ACDEFG AACDEFG ABCDEFG 
AACDEFG + ABCDEFG + ABCDEFG 
nanella 1 subrobusta Lamarckiana. 
But according to de Vries, we get in fact rubrinervis where we should 
expect to get Lamarckiana. In view of the fact that there is some con¬ 
fusion over the rubrinervis forms and that subrobusta was only recognized 
as a separate type in 1913, we might suppose that there was something in¬ 
complete in the analysis. But the hypothesis of double non-disjunction 
meets with other difficulties. It will explain the result of Oe. rubrinervis x 
Lamarckiana thus: 
BBCDEFG v ABCDEFG _ BBCDEFG ABCDEFG 
ABCDEFG X ABCDEFG ABCDEFG ABCDEFG 
rubrinervis Lamarckiana rubrinervis Lamarckiana. 
But, as Miss Lutz pointed out, nanella x Lamarckiana introduces a fatal 
, AACDEFG ABCDEFG AACDEFG , . 
1 cu ty, or we lave AACD£fG x AliC DEFG ~ ABCDEFG ’ W ' C 1S 
subrobusta , whereas we are supposed to obtain the two parent forms in the 
F 1 of this cross. It therefore appears to be impossible to explain the 
