W. F. R. Weldon 
231 
those of their pure-bred yellow-seeded ancestors ; one kind, produced from green 
seeds, should give rise to apparently pure-bred green-seeded plants; while the 
third kind, being yellow seeds produced by the union of one gamete with 
"dominant" and one with "recessive" properties, should give rise to plants 
identical with the hybrids of the first generation. This result was actually 
obtained. Those seeds of the second hybrid generation, which had green coty- 
ledons, gave rise to plants which in turn produced only green seeds. From the 
yellow seeds of the second hybrid generation 519 plants were raised, of which 
166 produced only yellow seeds, behaving therefore like pure-bred plants of 
yellow race, and 353 had both yellow and green seeds, in the proportion of three 
yellow to one green, as in the first hybrid generation. The observations were 
continued through six generations; and the descendants of those plants of the 
second generation which produced only one kind of seed remained throughout 
apparently pure-bred, producing each its proper kind of seed as regularly as its 
pure-bred ancestors ; the hybrids of every generation behaved exactly like those 
of the first or second generation. The result of this process is that the percentage 
of hybrids diminishes in every generation : for suppose each plant in each gener- 
ation to produce four seeds only, the plants of the first hybrid generation will 
produce one apparently pure-bred yellow-seeded and one apparently pure-bred 
green-seeded plant, with two hybrids; in the tliird generation there will be four 
apparently pm-e-bred yellow plants from the apparently pure yellow-seeded plant 
of the second generation, and two others, one from each hybrid; there will in the 
same way be six apparently pure green-seeded plants, and four hybrids. Assuming 
this process to go on, the numbers in succes.sive generations will be : 
Generation 
Yellow 
Hybrid 
Green 
1 
1 
2 
1 
2 
1 
3 
6 
4. 
6 
4 
28 
8 
28 
5 
120 
16 
120 
6 
496 
32 
496 
etc. 
So that the number of apparently pure-bred forms of each kind increases, 
and in the {n + l)th hybrid generation, the three classes of plants are in the 
proportions of 2" - 1 : 2 : 2" - 1. 
Results closely similar to these were obtained with seven sets of differential 
characters ; namely 
1. The shape of the seed (round and feebly wrinkled, dominant ; irregular and 
deeply wrinkled, recessive). 
2. The colour of the cotyledons, already dealt with. 
