Distinguishing between Two Types of Gene Action in Quantitative Inheritance.

MONG the widely divergent results obtained from “quantitative A character” crosses one common type is characterized by (a) F1 mean approaching that of the smaller parent strain and (b) positive skewness in the frequency distribution of Fz measurements. Among many examples might be cited crosses involving differences of corolla tube length in tobacco (EAST 1913; SMITH 1937), fruit size in squash (SINNOTT 1937) in peppers (DALE 1929; KAISER 1935) and in tomatoes (MCARTHUR and BUTLER 1938), weight in chickens ( JULL and QUINN 193 I). Results of this kind have long been recognized as incompatible with the early hypothesis that quantitative characters might be determined in general by genes having arithmetic effects without dominance or interaction. This hypothesis had been proposed (EAST 1910) as a reasonably simple scheme which might and did accord with the main features of size segregation in certain crosses. But other crosses, giving the sort of result under discussion here, were shown by EAST (1913) to be better accounted for if the relevant genes were assumed to have a multiplicative, or geometric, action. Many subsequent workers have adopted the same interpretation for comparable cases (DALE, SINNOTT and SMITH, among others). LINDSTROM (1935) has proposed a partial return to the original hypothesis, in dealing with skewed distributions of fruit size in the progeny of hybrids between a small-fruited wild tomato and two large-fruited domesticated varieties, namely: that the factors have arithmetic effects, without interaction, but with partial dominance of the small size genes. In discussing his results LINDSTROM pointed out that ‘‘ . . . it does not help matters by assuming dominance as a necessity in explaining the numerous facts of heterosis and prepotency, and then conveniently discarding dominance for most other forms of quantitative inheritance” and that “ . . . the mere fact that quantitative character data seem to fit a logarithmic curve does not necessarily rule out dominance.” Perhaps neither scheme of gene effects will turn out ultimately to be strictly applicable. It seems likely, on the whole, that the genes determin-