THE NATURE OF THE GENE 377 



modifies the general picture of independent and particulate hereditary 

 units. Moreover, it is only conceivable if different regions of the chro- 

 mosome have different properties; it is meaningless to talk about the 

 rearrangement of a linear series of identical units. These differences at 

 different points in the chromosome must have arisen in the past, and 

 it is therefore unreasonable to deny that they may arise to-day; but the 

 process by which they arise is exactly what we mean by gene mutation. 



4. The Size of the Gene 



Various attempts have been made to estimate the size of the gene. It 

 is, of course, by no means clear that a gene has a definite and constant 

 size; if the gene is a compound structure, it might grow gradually 

 between two divisions. But the estimates of gene size are all rather 

 rough at present, and it is very unlikely that the size of the gene is 

 sufficiently variable to affect the order of magnitude. 



All estimates of the dimensions of the gene start from considerations 

 of the observed size of chromosomes or parts of chromosomes. Chromo- 

 somes certainly change in apparent dimensions during the processes of 

 growth and division. Many of these variations can be explained as 

 results of the changing degree of spiraHzation of the chromonema; and 

 the relation between the gene and the chromosome, and thus the 

 estimates of the size of the former, always depends on the structure 

 which we find or assume for the chromosome thread. The most securely 

 founded of all statements about the condition of the chromonema is 

 probably that it is completely straight and uncoiled in saHvary gland 

 chromosomes. An estimate of the length of the chromonema associated 

 with a gene can be made on this assumption. Muller and Prokofieva^ 

 studied seven chromosome breaks occurring in a short region at the 

 left end of the X chromosome in D. melanogaster . They found that the 

 breaks apparently occurred in only four different places; and they 

 concluded that these four places were the connections between five 

 successive genes in the linear sequence. Further, Muller^ obtained a case 

 of a minute deficiency which was viable in a homozygous condition and 

 showed a phenotype indicating the absence of only two genes, yellow 

 and achaete. By cytological examination of the salivary gland chromo- 

 somes in these cases they deduced the interval between successive genes 

 to be about 125 m/x; the yellow-achaete deficiency was just visible as 

 the loss of a part of a dark band. But it should be noted that the chromo- 

 somes are extensible, in the usual cytological preparations, by at least 



' Muller and Prokofieva 1935. ^ Muller I935«« 



