164 Nature of the Genetic Material 



fabric but nothing about the associations as they actually function." 

 I expressed the latter point in genetical terms by saying that the 

 mutant locus is a reality, but not the normal gene. It may be added 

 that the recent experiments of Mazia (I 2 B a), which may lead to 

 a better understanding of the supermolecular associations in the 

 chromosome, will hardly change the situation described in the quo- 

 tation. 



Some of the ideas presented, or rather hinted at, by Watson and 

 Crick in discussing their new model of nucleic acid structure are 

 relevant at this point in our discussion. We quoted their statement 

 (I 2 B b bb) that the different possibilities for attachment of the 

 purine-pyrimidine rings to the molecular axis spell out the code for 

 different genie action. The idea, ff I understand their hint correctly, is 

 that the innumerable possibilities of permutation within a polymerized 

 molecule in regard to the many different monomeres possible, as well 

 as to the presence of different sets of the flat rings, may have different 

 genetic effects (assuming DNA to be the genetic material; but it 

 would not be diflBcult to transfer the same idea to the protein if DNA 

 is the template or scaffolding). It seems that Watson and Crick took 

 the step from the molecular to the chromosomal level rather lightly. If 

 we think in terms of a gene molecule, and assume it to be one 

 polymerized DNA, we do not gain much. We should expect innumer- 

 able possible permutations of the single gene, with very different 

 actions (the spelled-out code). These do not agree with the genetic 

 facts, which show only a few (20 or 30 are a few in this connection) 

 different actions with only small quantitative differences, the multi- 

 ple alleles. If, however, we could regard the genie part of the chromo- 

 some as a single giant hypermolecule in which the individual sub- 

 polymeres (perhaps identical with Mazia's particles?), the parts of 

 the whole, behave as assumed in regard to molecular permutations, 

 we could parallel such a structure with the ideas on segments, 

 patterns, fields, and hierarchy which we have developed. We hope 

 that the biochemists, who, as a rule, hear only of the classic concept 

 of the gene, will get acquainted with these newer ideas and see 

 whether they do not agree more closely with their fact findings. 



We have not, thus far, assembled all the theories of position 

 effect which have been proposed, though some of them were 

 mentioned. (My former discussions of this subject are found in 

 Goldschmidt, 1938, 1944, and 1946.) A brief enumeration may suffice: 



1. A gene near a chromosomal break mutates. Very few geneti- 



