[OG bub]

Increasing evidence that the dominance hypothesis, at least in its simplest form, was inadequate to account for the heterosis actually observed in various species led to increased interest in the nature of "heterotic loci" their relationship to the selective advantage of heterozygotes, the production of balanced phenotypes, and the development of homeostatic systems (Lerner 1954a 1954b; Dobzhansky and Wallace 1953). The earliest view concerning the nature of heterotic loci was formulated by East and Shull in 1908, as we saw earlier, in physiological language. Shull's concept was that "hybridity itself-the union of unlike elements, the state of being heterozygous-has a stimulating effect on the physiological activities of the organism." This idea, restated in genetic terminology by East in 1936, attaches to heterozygotes a versatility of development that is expressed in terms of alternative pathways in metabolism. These presumably arise in biochemical versatility stemming from existence of different alleles at the same locus. Because such versatility is not open to homozygotes, it must be associated with particular loci, which can be called heterotic loci. A necessary corollary is that heterozygosity is essential for vigor. The "homeostatic system" hypothesis (Lerner 1954a) attached no physiological virtue to the heterozygous state in itself; it reversed the causal sequence and made the physiological virtues observed in many heterozygotes consequences of selection under an outbreeding system. According to this second hypothesis "relationally balanced" (Mather 1943) or "coadapted chromosomes" (Dobzhansky and Wallace 1953) are produced as a result of long-term evolutionary development of balanced complexes of genes that have been integrated into the population gene pool.

Precisely what a heterotic locus represents in terms of physical structure remained a matter of spirited debate. Studies of pseudoallelism (e.g., Green and Green 1949; Lewis 195 1) suggested that the unit of specific physiological interaction is probably not a single Mendelian locus in the classical sense but a complex of more than one, and perhaps several to many, tightly linked genes. This had been implied by Jones (1917) when he formulated his hypothesis of dominant linked factors: if linkages are close, crossing over between loci will necessarily be rare, with the result that small chromosome blocks rather than individual Mendelian loci act, at least in the short term, as units of hereditary transmission. This proposition has appeal because chromosome blocks vary in size with the closeness of linkage and the number of sexual generations, thus providing opportunities for study of evolutionary change within small but potentially complex hereditary units. We now turn to the much studied effects of inbreeding in corn as an approach to understanding both inbreeding depression and combining ability.