296 PHYSIOLOGICAL GENETIC8 



presented on page 230, and their interpretation in terms of 

 development was discussed. The main point was that the 

 different scute allelomorphs could be arranged in a definite 

 xrics according to the pattern of bristles that each allele removed. 

 Each allele acted independently in the compound in such a way 

 that only those bristles were missing in a compound which both 

 alleles affected in common. Let ABCD be four different bristles, 

 and abed their suppression; then the allele sc x may produce the 

 pattern AbcD, and the allele sc v the pattern abCD. The com- 

 pound sc x /sc v is phenotypically AbcD/abCD and therefore normal 

 in ACD but with the bristle b missing. If we leave out of this 

 formula the symbols for the unchanged bristles, we have bc/ab; 

 i.e., the areas affected are steplike, and the parts common to both 

 steps (6) are affected. Serebrowsky, Dubinin, and their col- 

 laborators Gerschenson, Agol, and Lewit (1929-1931) conclude 

 that this special behavior can be explained only if the steplike 

 arrangement of the bristle pattern as controlled by the different 

 alleles has its counterpart in a similar arrangement of parts of the 

 gene. A projection of the different patterns in the order derived 

 for each two alleles from the common parts in the stepladder 

 would then represent a picture of the structure of the gene, i.e., a 

 structure of a definite length, made up of a series of smaller 

 elements arranged like steps or as a bundle of sticks the ends of 

 which do not coincide. 



We have already discussed some of the aspects of the case 

 and mentioned criticism and countercriticism (page 237). 

 A priori, there is no clear reason why a pattern of phenotypic 

 effects ought to be projected as a similar pattern into the gene. 

 Such a projection would help in no way to an understanding, 

 since it is impossible to see how a pattern in a small part of a 

 chromosome could control the formation of a parallel pattern 

 upon the thorax of a fly. If this were imaginable, one ought 

 to expect that, in general, the organism ought to be a replica of 

 its chromosome pattern. The old idea of the manikin in the 

 sperm head would be revived. The idea was therefore rejected 

 by most geneticists, some of whom tried to find an embryological 

 explanation for the facts (Sturtevant and Schultz, 1931; Gold- 

 schmidt, 19316), whereas others assumed that the basic descrip- 

 tion was erroneous (Child, 1936). Though the originators of the 

 theory seem to have deserted it now, it ought to be mentioned 



