NATURE OF THE GENETIC EFFECTS 403 



In practice, the dominance of one allele over the other appears in most 

 cases to be virtually complete, if a superficial inspection only is made. In 

 such cases the heterozygote cannot be distinguished from the dominant 

 homozygote by ordinary means. Moreover, whether or not the domi- 

 nance is as strong as this, the normal allele is, with but rare exceptions, 

 the dominant one in the above sense of the term. For although the 

 mutant is sometimes loosely termed dominant in any case in which the 

 heterozygote is readily distinguishable from the homozygous normal, 

 nevertheless it is usually found even in such a case that the heterozygote 

 is still more different from the homozygous mutant. At least this is true 

 for hypomorphic and amorphic mutants. 



Despite the apparently complete dominance of the normal gene in the 

 great majority of cases, a more searching investigation, e.g., by measure- 

 ment of numerous individuals and statistical analysis of the data, or by 

 the use of biochemical techniques or studies on rates of survival, usually 

 shows that the heterozygote is after all not quite the same phenotypically 

 as the homozygous normal type. For example, studies on lethal and 

 sublethal mutations in Drosophila (Stern et al, 1948, 1951, 1952; Muller, 

 1950b, c) indicate a dominance of the normal gene of something in the 

 neighborhood of 96 per cent, leaving about 4 per cent of expression to the 

 mutant in the heterozygote, on the average; i.e., the viability of such 

 heterozygotes is only about 96 per cent as great as that of homozygous 

 normals. This is also expressed by saying (with a slight alteration in the 

 use of the term dominance from that of our former definition) that the 

 mutant in such cases has "4 per cent of dominance." This amount is so 

 small that it has often been assumed that, for practical purposes, the 

 mutant may be regarded as completely recessive. However, as will be 

 seen in Sect. 20, even this small amount of dominance turns out to be very 

 significant in deciding the way in w^hich such mutant genes affect the 

 population in which they occur. It must be understood, further, that 

 the amount of dominance varies greatly from one type of gene to another, 

 and that, for a given pair of alleles, it can in some cases be influenced to a 

 considerable extent by differences in genes in other loci than their own, 

 which modify the result. 



As a result of the near-saturation potency of even one dose of most 

 normal genes, hypermorphic mutations of normal genes would seldom 

 produce an effect recognizably different from the normal. This fact, as 

 well as the lesser probability of a change to greater than of one to lesser 

 effectiveness (see p. 398) explains w^hy such mutations of normal genes 

 have so seldom been found. It also explains why genes are sometimes 

 found which when homozygous give a normal phenotype, yet when 

 crossed to a mutant hypomorph or amorph produce a heterozygote mani- 

 festing the mutant character to a distinctly greater degree than the usual 

 normal does. These so-called isoalleles are hypomorphic enovigh, in com- 



