IN THE TROPICS. 
347 
De Vries’ account is a little obscure, and he speaks in one 
place of “ Constant Flesh + Delila x White ” as a term in 
the series. He appears to mean FD WD., which can only be 
spoken of as Flesh + Delila x constant Delila or more 
properly Red x constant Delila. But I believe the above 
to be a fair account of his explanation. 
In F 3 de Vries actually obtained evidence of the existence 
of all the terms (a)-(i) in F 2 . 
Bateson (3) has criticized this account and suggested as a 
more likely hypothesis that the gametes of the heterozygote 
R x W consist of R, D, F, and W in equal numbers. R being 
dominant to the other three members and W recessive to the 
first three. This would give the following terms in F<>:— 
9 R, namely, IRR, 2RF, 2RD, 2RW, 2FD 
3 F, namely, IFF, 2F W 
3 D, namely, 1DD, 2DW 
1 W, namely, 1WW. 
This scheme plainly fits the facts equally well ; and it 
will only be possible to learn which of the two methods of 
gamete-formation really took place, by a very extensive 
study of subsequent generations, or, as Mr. Bateson has 
himself pointed out to me, by crossing Fj with the recessive 
on a large scale. 
The above scheme also illustrates very clearly Bateson’s 
view of a compound allelomorph , which is essentially opposed 
to the above view of a combination of colours in the flower 
held by Mendel and de Vries. 
Bateson and Saunders (6) regard a compound character, 
when crossed with an allelomorph of simple recessive nature, 
as breaking up into simpler and possibly component elements. 
When similar elements of this kind meet in fertilization 
there is no further resolution, and a permanent new character 
thus arises by a process of analytical variation. The term 
hyp-allelomorph is used for the constituent elements of a 
compound allelomorph. 
