98 Nature of the Genetic Material 



physiologically as one. In practice, we would conclude nothing in 

 such a case until we knew the behavior of the physiological unit 

 following an exact reversal of the rearrangement to the original con- 

 dition, or following a recovery of that unit in its original arrangement 

 as the result of crossing over. Actually it has never been possible to 

 effect either of these reversals (except in the case of position effects 

 involving the heterochromatic regions, where the issue has been one 

 of deciding whether the gene has changed by virtue of its being next 

 to heterochromatin as opposed to euchromatin ) . Crossing over, on the 

 other hand, is the unique process which results in recombinations of 

 chromosomal parts without altering the physical distance between 

 genes. It is therefore capable of leading and has led in the examples 

 cited in this paper, to the recognition of smaller units within a por- 

 tion of the chromosome acting as a physiological unit. These smaller 

 units we still call genes." 



We shall discuss below in detail the factual material underlying 

 this discussion and its possible explanations. Here I point out only 

 that Lewis' discussion starts with "the definition of any particulate 

 unit," which to me means that the classic dogma of the particulate 

 gene is accepted a priori in its totality. Hence all further conclusions 

 are only a reflection of this basic dogma, which I shall try to criticize 

 and to destroy. 



Actually, the idea of the particulate gene makes sense only in 

 what I may call its naive formulation. There is no way of talking 

 about a gene if no mutant is known which can be distinguished (see 

 Whiting's happy formulation that the mutation creates the gene). If 

 this mutant locus formed with the non-mutated locus a pair of 

 Mendelizing alleles, the conclusion was drawn that at the original 

 locus of the chromosome a gene was located which had changed by 

 mutation into its allele. It is surprising how many geneticists speak 

 about genes, finding new genes, and so on, without realizing that 

 they mean a mutant locus, from the existence of which they extra- 

 polate upon the normal gene without knowing anything of its reality. 



It is remarkable how many geneticists are unable to concur in 

 this simple conclusion, stating, for example, that if something mutates, 

 the original something is the normal gene, though we know only 

 that a change has occurred at a point of the chromosome which 

 altered its function. The change at a point does not mean that this 

 point contains a separate corpuscle of unitary action. However this 

 may be, we cannot define a gene, even assuming the correctness of 

 the classic idea of the particulate gene, by the breakability of the 



