1 I \ PHYSIOLOGICAL GENETICS 



heterozygote, however, follows closely the Bar thermophene, the 

 Bar gene thus being dominanl for the whole facet-temperature 

 relation. Figure 29 gives the curves constructed by Hersh 

 (1934) for these relations. It shows as one of the consequences 

 of the divergence of the temperature-facet curves for B { , B, and 

 B/B' thai at lower temperatures the heterozygotes assume more 

 and more the />' type and finally, just below 17°, reach it, so that 

 here B' appeals dominant over B. The additional fact regarding 

 B i action is then its different thermophene, which leads auto- 

 matically to dominance of Bar at higher temperatures and makes 

 a simple quantitative arrangement of the data impossible, when 

 B and B l arc simultaneously present. (Hersh, 1934, however, 

 points out that a single order for all temperatures is obtained if 

 not the end result — the number of facets — is used as a measure, 

 but the instantaneous increase in facets at all temperatures 

 measured by the algebraic value of the first derivative, see 

 page 105). 



The old discussion, whether the whole Bar series is quantita- 

 tive or qualitative, thus resolves itself into the following: the Bar 

 components are all quantitative, and the series of Infrabar is also 

 quantitative. But Infrabar has, in addition, a different quality, 

 which is expressed in its different thermophene and therefore 

 makes the phenotype not simply dependent upon the quantities 

 involved, when Bar and Infrabar are both present. We know 

 that B is a double dose of a small but differentiated chromosome 

 segment. What Infrabar is, is not yet known. If it is, as the 

 genetic data indicate, a duplication of only a part of this region, 

 the lack of the rest of the region would have to account for the 

 different qualities. At present, the analysis may not be pushed 

 further. There remains, of course, the question whether or not 

 one may draw a conclusion from this case concerning the action 

 of the gene, since usually such a duplication woidd not be called a 

 gene mutation. This question will be answered later when 

 discussing the theory of the gene. 



The Bar series has always served as a typical example for a 

 series of multiple allelomorphs and has therefore been cited too by 

 Goldschmidt when he tried to prove that such series are based 

 upon different quantities of the same gene. The proof of actual 

 dosage (with the necessary additions for B l ) in this case requires 

 a comparison with other cases in which both actual dosages and 



