THE MUTATED GENE 133 



We have mentioned Ephrussi and Beadle's transplantations 

 of Drosophila eyes, to which we shall return later. When sepia 

 eye disks were implanted into vermilion hosts, a sex difference 

 in pigment development appeared (1937), the males being 

 lighter. Transplants performed with disks of different age 

 showed that the surroundings cannot be made responsible for 

 the effect, which must be caused by conditions in the disk itself. 

 This does not contradict the former conclusions, as the disk is 

 itself a developmental system in which the pigment-forming 

 reactions are only a part of all reactions. 



B. Differences of Dosage by Deficiencies 



Next to the instances of sex-linked genes, mutant genes lying 

 opposite deficiencies, i.e., deletions of a part of the chromosome, 

 provide the most illuminating material on the action of genes in 

 single doses as compared with double doses and the heterozygote. 

 Though deficiencies are known for many organisms used in 

 genetic experimentation, most of the information pertaining to 

 the present problem comes from work in Drosophila. Mohr 

 (1923a) discovered that most mutant genes contained in a section 

 of a chromosome opposite a deficiency in the sister chromosome 

 had in this their simplex condition an exaggerated effect; e.g., 

 a light eye color was still lighter, a forked bristle extremely 

 forked. This applies to the majority of mutant genes thus 

 studied by Mohr and his followers, with the exception of mutant 

 characters which seem to be already at the extremest, physiologi- 

 cally possible level, like white eyes, prune, purple, and yellow. 

 A first explanation for this phenomenon was given by Bridges 

 (1922), who assumed that the deleted piece of the chromosome 

 contained a number of modifiers the removal of which upset 

 the balance of the genes in favor of a more extreme action of the 

 mutant gene. This interpretation was not very probable at the 

 outset, as it ought to lead in different cases to results of opposite 

 character, since it can hardly be supposed that the deleted pieces 

 always contain only one type of modifiers. Mohr, moreover, 

 could show that the phenomenon is independent of the length 

 of the deletions in one region, which also rules out the balance 

 interpretation. Goldschmidt (1927c) proposed a different expla- 

 nation, based upon the general idea that the activity of the 

 genes is proportional to their quantity. If, for example, the 



