[111] These are then a mixture of pure dominants and cross-bred dominants, and are now inextricably confused. This time the round seeds may have been all on particular plants—showing recessive seed-shape as a maternal character. It seems just possible that this fact suggested the idea of “round” seeds on the coloured plants in the last generation. Till that result is confirmed it should be regarded as very doubtful on the evidence. But we cannot at the present time be sure how much difference there was between these round seeds and the normal maples in point of shape; and on the whole it seems most probable that the roundness was a mere fluctuation, such as commonly occurs among the peas with large indented seeds.
[112] Is this really evidence of segregation of characters, the flower being the unit? In any case the possibility makes the experiment well worth repeating, especially as Correns has seen a phenomenon conceivably similar.
[113] Being a mixture of heterozygotes (probably involving several pairs of allelomorphs) and homozygotes.
[114] This looks as if the violet colour was merely due to irregularity of xenia.
[115] Pure recessives.
[116] Pure recessives in coats showing maternal dominant character.
[117] Now recognized as pure homozygotes.
[118] This seems almost certainly segregation by flower-units, and is as yet inexplicable on any other hypothesis. Especially paradoxical is the presence of “white” seeds on these plants. The impression is scarcely resistible that some remarkable phenomenon of segregation was really seen here.
[119] Being now homozygotes.
[120] Being heterozygotes exclusively.