THE NATURE OF THE GENE 297 



because as keen a thinker as Muller still believes there may be an 

 clement of truth in it. The newer discoveries regarding the scute 

 locus, however, shift the discussion to a very different field (see 

 page 306, position effect). 



It should be pointed out also that it is always dangerous to draw 

 conclusions from exceptional cases without considering whether 

 the complicated phenotypic behavior is not more likely to be 

 caused by developmental conditions than by an assumed property 

 or structure of a gene. Glass (1933a), for example, has described 

 a case that he likens to the scute case and from which he draws 

 definite negative conclusions about the gene. He finds in 

 Drosophila a mutant facet notch which is an allele of facet but 

 which produces wing notches, without an eye effect. Facet-facet 

 notch heterozygous is normal, except for 0.26 per cent notched 

 flies, which resembles the behavior of some of the scute com- 

 pounds. Glass uses these facts to criticize the quantitative 

 nature of multiple alleles. But, looking at the case from the 

 standpoint of development, it is easy to conceive that the hetero- 

 zygote is actually intermediate when different thresholds for the 

 two processes (eye and wings) are assumed, an interpretation that 

 actually goes back to Wright's earliest work. 



C. The Hypothesis of Thompson 



Thompson (1925, 1931) derived a theory of the gene from the 

 experimental studies on the Bar locus in Drosophila, especially 

 those of the Zeleny school. The general idea is that the gene 

 consists of a main particle anchored in the chromosome, the 

 protosome, to which a varying number of one or more particles, 

 the episomes, are attached. Mutation is due to the loss of one or 

 more episomes and sometimes also to their addition. Two or 

 more episomes of the same kind are attached to each other and 

 form a side chain, and different episomes are attached separately 

 to the protosome. Varying numbers of the same kind of episomes 

 produce quantitative series of multiple allelomorphs. Qualita- 

 tive series are based upon numerical variation of more than one 

 kind of episome. 



It is obvious that this theory contains the combined elements 

 of most of the other theories mentioned. Quantity of active 

 episomes explains the same facts as quantity of gene molecules, 

 and the calculations that Thompson makes from the Bar data 



