328 LOCK : STUDIES IN PLANT BREEDING 
the form of the endosperm is passively moulded. It appears 
possible, therefore, as Correns points out, that in the 
present case the dimpled form of the cotyledons may be 
impressed upon them by the testa. 
There are indeed certain difficulties in the way of accept¬ 
ing this view. The endosperm of a grain of maize is quite 
fluid at a time when the external form of the grain is already 
clearly distinguishable, but the same is by no means true of 
the cotyledons of a pea. Tschermak cites as a further 
difficulty the appearance of dimpled seeds upon plants of 
the first generation from a cross between a race of Pi sum 
arvense having smooth seeds and a smooth-seeded race of P. 
sativum ; an example of the appearance of a new (or atavistic) 
character on crossing. But it does not appear that this ob¬ 
servation has any very direct bearing upon the point at issue. 
Gregory (39) has shown that the starch grains of the 
cotyledons of a dimpled pea resemble those of a smooth 
pea, and differ widely from the smaller starch grains of a 
truly wrinkled variety; and this renders it likely that the 
shape of a wrinkled pea may depend on something other 
than the cotyledons, i.e. upon the testa. On the whole, there¬ 
fore, the view of Correns appears to be the more probable. 
The following example illustrates several interesting 
points with regard to seed characters. In speaking of these 
it is very necessary to bear in mind that a seed includes por¬ 
tions contributed from two separate generations. Thus in 
seeds bom on a first cross plant (F t ) the testa belongs to the 
maternal generation, but the cotyledons to the next genera¬ 
tion (F 2 ). So that on the hypothesis here adopted the 
shape of a dimpled seed is a character belonging to one 
generation, whilst that of a wrinkled seed belongs to another. 
Example VI. —Sutton’s French Sugar pea (“male” parent) 
was used to pollinate a native pea with smooth yellow seeds. 
The former plant had coloured flowers with a red standard 
and purple wings ; and there were purple spots in the axils of 
the leaves. The testa in freshly ripe seeds was almost always 
