Chromosomes and Genes 109 



premise, conclusions on the basis of the target theory would be worth- 

 less. In other words, the proof of the existence of the gene molecule 

 by calculating its size is based upon the premise that there is a gene 

 molecule the size of which can be calculated and that the target size 

 is its size. Here is a good example of this type of reasoning. Kurnick 

 and Herskowitz ( 1952 ) measured the DNA content of salivary nuclei 

 in Drosophila and came to the conclusion, by comparison with other 

 somatic cells, that the salivary chromosomes have about 1,000 polytene 

 strands. (The proportionality of the nuclear volume to the number of 

 chromonemata also agrees with this; see fig. Sb.) They argue that the 

 molecular weight of DNA polymerized is about a million, and thus 

 the mass of a single molecule is 1.6 X 10 ~^^ mg. In the 1,000-tene 

 nucleus, 710 X 10~^" mg. were found. A haploid chromosome set 

 would contain 0.71 X 10" ^^ mg. of DNA, that is, 44,000 such mole- 

 cules. The largest estimate of gene number in Drosophila — arrived 

 at by the Treffer theory by Timofeeff — is 10,000. Therefore, either the 

 gene must consist of a number of molecules, or the DNA in the gene 

 has a much greater weight. A number of 44,000 genes would be im- 

 probable, because a mean distance from one to the other of 17 m/x 

 would result. "It is difiicult to conceive of how resolvable bands could 

 be derived from the unhanded state with units so closely spaced." I 

 may be excused if I say that this and many former calculations are 

 exactly of the type of the somewhat facetious example of counting the 

 inhabitants of Mars. 



The Treffer theory is the major unproved premise of all these 

 calculations. Muller (1950a) has assembled powerful arguments 

 against the theory based upon genetic facts as well as upon better 

 knowledge of ionization effects. (See our discussion of the Treffer 

 theory in I 3 B a. ) He insists that there are abundant data which dem- 

 onstrate the importance of intermediate chemical reactions in muta- 

 genesis. It is difficult to estimate, in terms of atomic distances, the 

 spatial range of the reactions from which mutations result. We do not 

 even know whether they originate within or outside the chromosome, 

 especially the genetic material, and it is by no means certain that the 

 mutagenic effect resides within the particle that is hit. Some other 

 arguments of Muller, shared by many workers in the field, are as 

 follows: 



In radiation experiments, made under different conditions, results 

 are obtained which, if calculated, lead to different volumes of the 

 target. Stadler has found eight times more mutations in sprouting 

 seedHngs than in dry ones with the same dose, in spite of strict 



