LOCK : 
IK PLANT BREEDING 
Only 25plants were raised in F a , viz., 11 (m.p.f.), 6 (p.f.), 
2 (fc), 4 (w). 
But in F 3 3 plants from seeds showing (m.p.f.) in F 2 had 
the following offspring:_ 
No. 2 ... 18 (m.p.f.) ... 3 (m.f.)... 
No. 7 ... 48 „ ... 8 ,, ... 
No. 8 .. 16 , o 
(pX).. 
14 (w.) 
Total. 
37 
A proportion which closely approaches that which is 
expected on the above hypothesis. 
The final test of the hypothesis will be to make artificial 
rosses between,^, (f) p i anta and (w) plantS- From certain 
these cinsses aH the original types should be obtained.* 
t yet been caried out. 
This experiment has i 
IV.—THE VIEWS OF DE VRIES. 
De Vries has extended the operation of Mendel’s Law to 
a large number of characters belonging to plants of widely 
separated natural orders. In his first paper upon the law of 
segregation of hybrid characters (67) de Vries propounded, 
as a general rule, that the hybrid always shows one only of 
apair of antagonistic characters, and this in all completeness. 
Dominance, he asserts, is always complete. But this is by 
no means the case, as Correns (11) has well pointed out, and 
as en e imself observed in many of the characters of 
peas. In the remainder of his discussion of this class of 
hybrids de Vries follows Mendel closely. 
In a second paper de Vries grouped together Mendel’s 
examples in Hieracinm, the false hybrids of Millardet and 
a series of cases discovered by himself in crossing different 
forms of Oenothera and other species, as « false hybrids ” ; 
