452 RADIATION BIOLOGY 



of the deficient region differed slightly in the different cases of any one 

 series of experiments in consequence of minute differences in the positions 

 of breakage. By noting the phenotypic effects produced in the different 

 cases, and by comparisons of the manner of grouping of these effects from 

 case to case, the linear sequence and the functions of the genes contained 

 in the affected portion of the chromosome could be ascertained with great 

 nicety of resolution, yet over a considerably greater length of chromosome 

 than in the work with the scute region of Drosophila. 



Returning now to the studies on the scute region, a deficiency that 

 according to genetic tests lacked exactly that portion of chromosome 

 lying between the leftmost and the rightmost of all the breaks which had 

 been dealt with, and thus was deficient for the whole group of four 

 adjacent genes comprised in the preceding analysis, was found on cy to- 

 logical examination to occupy a length of about half a micron on the 

 salivary chromosome and to constitute not more than a half of one double 

 band as seen in ordinary preparations. In this way it was shown that 

 the maximum length of the individual genes here dealt with was just 

 beyond the resolving power of visible light, even in the salivary chromo- 

 somes, and occupied less than one ordinary band. From the total length 

 of a complete set of sahvary chromosomes it could then be readily 

 reckoned that there would be room for some 8000 such genes, if they w^ere 

 similarly spaced throughout, while if they were contained only in the 

 chromatic regions there would be some 3000. These estimates of gene 

 number, divided into the size of one complete set of chromosomes when 

 it is most condensed (as at mitosis or in spermatozoa), gave maximum 

 estimates for gene volume (Muller, 1935a). The approximation figures 

 for gene number thus arrived at (and therefore also those for size) proved 

 to be in satisfactory agreement with estimates of gene number (and size) 

 obtained by two quite independent methods, which involved a larger risk 

 of error. One of these methods was based on the minimum "map dis- 

 tance" found between genes in representative portions of the genetic 

 linkage diagrams. The other was based on the frequencies with which 

 gene mutations recurred in the same locus, as compared with the fre- 

 quencies of gene mutations in different loci. In some of the applications 

 of the latter method, also, radiation had been used for the production of 

 the gene mutations studied. 



Radiation genetics has provided evidence not only concerning the 

 manner of subdivision of the genetic material along the chromosome, i.e., 

 in a longitudinal direction, but also concerning its possible compoundness 

 in a transverse direction. In fact, the obtaining of evidence on this ques- 

 tion, at a time when it was thought that the gene might be composed of 

 several or many identical units, termed "genomeres" (Muller, 1926b, 

 formed one of the principal motives for the present writer's first work 

 on the production of mutations by radiation (1927, 1928a, b, d). For if; 



