1907] Brainerd,— Mendel’s Law of Dominance in Viola 215 
but in no instance is there a failure to comply with the laws of Mendel. 
The following points should be noted:— | 
1. Inthe sowing numbered 338 all five offspring are like the parent 
a.b,— have green capsules and buff seeds; a recessive character (in 
this case a double recessive) must always breed true. Here we have a 
new and stable form, which as a whole is unlike either of the original 
species. 
2. We may infer that the plant from which seeds numbered 333 
were obtained was probably pure, that is, was really aB and not 
a.Bb; had it been the latter, one or more of the six offspring would 
probably have had buff seeds, or been a.b. Perhaps as regards cap- 
sule-color 334 was also pure,— A.b and not Aa.b. If so, we should 
have here in color characters complete reversions to the respective 
grandparents, V. papilionacea and V. hirsutula. 
3. In sowing 339, though only one seed germinated, it tells the 
whole story regarding the form of the parent, — it must have been 
the di-hybrid Aa.Bb, and thus exactly like the immediate parent F,. 
A plant with purple capsules and brown seeds produces one with green 
capsules and buff seeds; a feat to be expected of a di-hybrid, though 
only once in every sixteen plants that it reproduces. 
4. The proof is equally clear that in sowing 337 we have the seeds 
of another di-hybrid; it throws both sorts of capsules and both sorts 
of seeds. 
5. In sowing 340 the result is somewhat ambiguous ;— the mother 
may be either Aa.Bb or Aa.B; it could not be A.Bb or A.B. 
6. 'Though three apparent A.B's were sown, none proved to be 
pure or stable like no. 1 of fig. 2. But in the offspring of a di-hybrid 
the chances are that only one in nine, that have the form A.B, will 
turn out to be that in reality. 
I much regret that the experiment was not conducted on a larger 
scale; but it should be borne in mind that the object was not to 
that is now quite unnecessary; but, 
verify the Laws of Mendel, 
conversely to prove that as these seedlings in their metamorphoses 
obey the laws of Mendel, their parent was of hybrid origin. Disap- 
pointed in my expectation that the artificial production of violet 
hybrids would be attempted at a certain ‘biological laboratory,’ it 
occurred to me over two years ago that equally valuable results might 
be gotten by observing the behavior of the offspring of the supposed 
