160 Nature of the Genetic Material 



brought heterochromatin to the neighborhood of the section of 

 position effect. Here the position effects were produced by breaks 

 even more than twenty-five bands distant. (Still more extreme actions 

 were found for the Notch effect. ) In the euchromatic case, sections of 

 the chromosome are delineated containing about half a dozen 

 bands within which the assumed locus of the "gene" is located and 

 within which any rearrangement break produces the position effect 

 (though in some cases certain points of breakage seem to be preferred 

 as agents of position effect). Another very remarkable feature is that 

 these sections may overlap, as can be seen in yellow-scute, white- 

 Notch, Notch-diminutive. Hence a break at the same point may 

 produce either effect. We do not know what this means, but we may 

 think of an influence of the location of the second break. 



What do these facts mean in terms of the theory of the genetic 

 material? Let us first consider the normal locus, the location of which 

 is established by the deficiency test. Can this test still be regarded as 

 valid? I do not believe so. Since we know that many homozygous 

 deficiencies produce the mutant effect, without any locus existing, 

 since we know further that deficiencies located at different points 

 ( Sutton ) can give the same deficiency effects, the localizing of a point 

 mutant by the deficiency test has become obsolete. If, as it has 

 happened, three distinguished researchers (Demerec, Muller, and 

 Prokofyeva; see Muller and Prokofyeva, 1935) find the same locus in 

 completely different bands, using the deficiency method, this is no 

 longer a cause for quarreling, because we know that both may be 

 right. This does not mean that there are two loci; it means that the 

 concept of the locus in its strict form is wrong and with it that of the 

 gene. What has actually been proved by all these facts is that chro- 

 mosomal segments exist, probably of different lengths, which may vary 

 from two to many salivary bands, within which any change of order 

 of the elements by removal of parts (deficiency) or rearrangement 

 (duplication, inversion, translocation) produces approximately the 

 same phenotypical effect; and that all these rearrangement effects 

 behave as multiple alleles. There is no normal locus within a segment 

 and no point mutation of such a locus; an invisible rearrangement is 

 prone to be localized in or next to any band of a segment. This 

 conclusion, which tallies with all the facts analyzed in former chapters, 

 is, in my opinion, unavoidable, though I realize how difficult it is for 

 classic genetics to free itself of ingrained notions which have served 

 so well. It is easily realized that this new point of view may be for- 

 gotten when dealing with elementary genetics, which can be described 



