168 Nature of the Genetic Material 



separate "genes" with similar but clearly distinguishable effects. In 



some cases, each has its own series of multiple alleles. The important 



and unexpected fact is that these mutants behave, nevertheless, like 



alleles; that is, if one chromosome contains one mutant and the 



homologous chromosome the other (the same for three mutants), 



an effect like a homozygous mutant is produced. Since it is assumed 



that two different "genes" in different position are involved, which, 



therefore, should not be allelic, the phenomenon has been called 



"position pseudoallelism" (Lewis). In terms of genes, these facts 



a + a b 

 could be written — 7- = mutant type by pseudoallelic action, = 



normal. Thus strictly adhering to the classic notion, we could also 

 say (Lewis) that the two genes in the one chromosome have no 

 effect, but, if located in two chromosomes, they act like alleles. 

 This was likened to Sturtevant's original description of the position 

 effect of two Bar alleles in such arrangements, a position effect which 

 since has found a different explanation, as we have seen. Therefore, 

 the term position alleles was proposed by Lewis. 



It is desirable at this point to realize the logic behind these 

 conclusions. The fact is that there are two different but adjacent 

 mutants, different because they can be separated by crossover, and, 

 in some cases, show shghtly different effects and different series 

 of multiple alleles. Both are recessive and therefore normal if hetero- 

 zygous; but their distinction as two loci breaks down when they are 

 present in a compound and act as alleles. Therefore some special 

 condition must be present which changes the classic relationship of 

 allelism. This special condition may mean that the classic concept of 

 allelism does not suffice. This conclusion is not drawn, but in order 

 to preserve the classic concept it is assumed that it makes a difference, 

 here, whether the two genes are located in the same or in different 

 chromosomes, that is, the old Sturtevant effect for Bar which turned 

 out to be something quite different. 



This leads to a further step in the explanation within the classic 

 theory: the two genes (or more of them) have originated by tandem 

 duplication of one. After the dupHcation, one of the duplicates has 

 assumed a new function (mutated to a new allele), but the old 

 allelism has been retained, thus becoming pseudoallelism. If this 

 effect should disappear, we would face the origin of a new gene, 

 though by definition the new gene would act just like a multiple allele 

 of the old gene, but without allelism. This would require the change 

 of the gene, released from the fetters of allelism, into one with 



