NATURE AND ORIGIN OF HETEROSIS 219 



environments. This deleterious character of most mutations seems surpris- 

 ing, especially because in modern biology the process of mutation is regarded 

 as the source of the raw materials from which evolutionary changes are con- 

 structed. 



A little consideration shows, however, that the ada])tively negative char- 

 acter of most mutations is by no means unexpected. Indeed, since every mu- 

 tation has a finite probability to occur in any generation, the mutants which 

 we observe in our fields and laboratories must have arisen many times in the 

 history of the species. The rare mutants which confer adaptive advantages 

 on their possessors in the environments in which the species normally lives 

 have had the chance to become established in the species populations as 

 components of the normal species genotype. In a more or less static environ- 

 ment, the genotypes of most species are close to the upper attainable level of 

 adaptedness. 



The above argument may seem to prove too much. In the absence of use- 

 ful mutants, evolution would come to a standstill. The paradox is resolved 

 if we recall that the environment is rarely static for any considerable periods 

 of time. Furthermore, most living species occur not in a single but in several 

 related environments. Genotypes which are adaptively valuable in a certain 

 environment may be ill adapted in other environments, and vice versa. It 

 should be possible then to observe the occurrence of useful mutations if we 

 place the experimental organisms in environments in which their ancestors 

 did not live. 



Progressive improvement of domesticated animals and plants in the hands 

 of breeders constitutes evidence that useful mutations do occur. The genetic 

 variants which are being made use of by breeders have arisen ultimately 

 through mutation. These mutations have been arising from time to time, be- 

 fore as well as after the domestication. But while they were deleterious in 

 the wild state, some of them happened to be suitable from the standpoint of 

 the breeders. They were useful in the man-made environment or they were 

 useful to man. Favorable mutations can be observed also in wild species, 

 provided that the latter are placed in unusual external or genetic environ- 

 ments. This has been demonstrated in experiments of Spassky and the writer 

 on Drosophila pseudoobscura. Several laboratory strains of this fly were sub- 

 jected to intense selection for fifty consecutive generations, and improve- 

 ments of the viability have been observed in most of them. 



Many, perhaps most, deleterious mutants are nearly or completely reces- 

 sive. Others are more or less dominant to the "normal," or ancestral, state. 

 The fate of the dominant deleterious mutants in jwpulations of sexually re- 

 producing and cross-fertilizing species is different from that of the recessives. 

 By definition, deleterious mutants in wild species lower the fitness of their 

 carriers to survive or to reproduce, and in cultivated species impair the 

 qualities considered desirable by the breeders. Natural and artificial selec- 



