124 
S. IKENO : 
All cases mentioned in the above Table (with some exceptions) may be 
explained according to one of several ways discussed below, either alone or 
combined. 1 
1. Reversion during the Formation of Gametes —Nos. 1, 2, 3, 4, 5, 6, 7, 
8, 9, 10, 12, 13, 21, 22, 25, 26, 27, 28 and 30 in the above Table may be 
included here. All of them are explainable when we will assume a certain 
reversional change of allelomorphs in cells concerned in the formation of gametes, 
either male or female. The question, in what stage of development such a 
change will take place cannot be yet exactly answered, but its occurrence 
might be perhaps sought in the reducing division leading to their formation, 
especially in synapsis stage. In some cases the reversion of one single 
allelomorph into its corresponding suffices to explain the phenomenon, but in 
others that of two or even three allelomorphs must be assumed. 
To begin with the simplest case : the white of the constitution ccrrbb 
produces normally the gametes crb ; suppose that the reverse mutation of the 
allelomorph c into C takes place in some cells concerned in the reducing 
division, then the gametes Crb will be produced besides the normal ones crb. 
Since the number of the mutated gametes is certainly very small as compared 
to that of normal ones the gametes Crb will meet in fertilisation most 
commonly with crb, though very exceptionally the meeting of the mutated 
male and female gametes Crb might occur ; the resulting zygotes are phæno- 
typically the same in both cases, viz. orange, but genotypically different, viz. 
Ccrrbb in the first and CCrrbb in the. second case. The production of 
oranges from whites, as seen in Nos. 6 (F 3 ), 7 (Ff 12, and 13 may be due 
to such a reverse mutation. For instance, in No. 6 one orange has been 
derived in F, from the seed taken on a Ft white plant, and that orange 
was found to segregate in F 4 into 3 oranges : 3 wlrites, proving thus itself to 
be of the constitution Ccrrbb, though here the actual numbers of the two 
kinds of segregates in F 4 do not very well agree with the calculated (4‘5 : 1*5), 
evidently on account of the small number of individuals. No. 12 might 
1 What are given below are, as above stated, mere trials to explain the appearance of 
unexpected individuals, and would contain necessarily some defects and mistakes, especially as 
the very poor germination of seeds makes the explanation difficult, and in certain cases even 
almost impossible. It would be quite possible that in future some of them might be replaced 
by much better ones. 
