INDUCED MUTATIONS IN PLANTS 1271 



On a priori considerations polyploidy would be expected to decrease the 

 frequency of occurrence of variants due to gene mutation, but it would 

 be expected to increase the frequency of those resulting from various 

 sorts of chromosomal aberration. 



The relation of polyploidy to both sources of germinal variation 

 is a consequence of gene reduplication in the polyploid genotype. The 

 chromosome complement of the polyploid species consists of two or more 

 groups of chromosomes, each group analogous to the entire chromosome 

 complement of related nonpolyploid species. Probably the genomes 

 combined are never wholly identical, but in general they must be identical 

 in a considerable proportion of the genes which they include. This 

 proportion doubtless will differ widely in different polyploid species, 

 depending on differences both in original constitution and in the extent 

 of evolutionary change subsequent to the establishment of the polyploid. 

 We have at present no sound basis for an estimate of the proportion of 

 genes reduplicated in any specific instance. 



Now, so far as the reduplicated genes are concerned, recessive muta- 

 tion of a gene in one chromosome group will in general be masked by the 

 unmutated duplicate gene in the other. Since almost all of the radiation- 

 induced mutations in the nonpolyploid species are wholly recessive, this 

 will result in the elimination of practically all detectable mutation of the 

 reduplicated genes. Mutation of those genes which are present as 

 dominants in only one of the chromosome groups of the polyploid will 

 be readily detectable. Thus the polyploid with two groups of chromo- 

 somes, though it may have twice as many genes as a related nonpolyploid 

 species, would ordinarily yield a smaller number of detectable mutations. 

 In the polyploid with three groups of chromosomes, since the probability 

 of reduplication of any given gene is even higher, the frequency of 

 detectable mutations should be still lower. In the comparison of 

 existing species it is, of course, impossible to compare the polyploid 

 with its unchanged parent species, or to make accurate allowance for 

 the changes which have occurred in the polyploid itself. However, 

 a general relation of the type described may be expected to apply in 

 the comparison of any species of the same polyploid series, and in inter- 

 preting the results of irradiation in polyploid species it must be assumed 

 that a large proportion of the gene changes actually induced may be 

 phenotypically undetectable. 



On the other hand, the frequency of heritable variations due to 

 chromosomal aberrations is increased rather than decreased in polyploids. 

 In nonpolyploid species the loss of a part of the chromosome complement 

 is usually — perhaps always — lethal to the gametophyte. Such losses, 

 or deficiencies, are a very frequent result of irradiation. If the deficiency 

 is induced in a mature parental germ cell or in somatic cells it is readily 

 detectable, and in many instances the chromosomal segment affected 



