No abstract

327and the parts of the genotype the genes of a nucleus must be related in action, and the characters must be related in development." Without such a correspondence genes cannot be either "completely independent " or essentially additive ".Williams' attempt to partition the genotype in order to justify his arbitrary separation of genetics into genetics and somatics has led to an unnecessary attack on gene interaction which only confuses the whole problem of hybrid vigour.

No abstract

1. The problem of inbreeding and the fixation of apparent heterosis in the cultivated tomato and in inbreeding crop species generally is discussed. The case is argued that ‘heterosis’ as normally understood in out-breeders is not likely to exist as a general phenomenon in inbreeding species.2. Selections were made among segregating progenies of two successful, commercial F1 hybrids, and the performance of the selected lines, up to the F4 generation, shows that the superiority of the hybrids over their parents can be fixed in purebreeding lines.3. Owing to selection effects in inbreeding species of crops the best parents will differ in respect of fewer quantitative loci than is to be expected from general experiments on quantitative inheritance. In the present material, desirable recombinants were isolated at a frequency of 1 in every 1000–1500 F2 individuals.4. The utility of F1 hybrids in the tomato thus depends on the economy of effort involved in their development as compared with the isolation of pure lines. It is concluded that the production of hybrids is to be favoured only when reliable methods of prediction of their performance are available, and then as an interim measure while pure-line forms become available.

No abstract