GENE FOR FOURTH CHROMOSOME OF DROSOPHILA 



329 



Supposing, however, that mutations are equally likely to occur at 

 any locus in the chromatin, it could be explained as a result of 

 pure chance that no mutations had as yet happened to lie in the 

 restricted space of the small autosome. It was expected, neverthe- 

 less, that mutations in a fourth group would eventually be found, 

 and such a mutation has now in fact arisen. 



The new character is a recessive wing and leg abnormality, the 

 wings being held out from the body but bent backwards near the 

 base, and the metatarsal joint of the legs being frequently greatly 

 shortened and thickened. The wing is also apt to be curved, with 

 the dorsal surface convex, and shortened. The character varies 

 somewhat, but there is very rarely any difficulty in distinguishing 

 it from the normal form, unless the flies have been raised in very 

 dry bottles. Drought therefore hinders the manifestation of this 

 character, as it may also, and to a greater degree, in some way 

 hinder the development of the character 'abnormal abdomen,' 

 a case reported by Morgan. 



I found the mutation 'bent wing' in a race with bifid wings and 

 vermilion barred eyes, all three of these characters being in Chro- 

 mosome I (sex-linked). A cross of bifid vermilion barred bent 

 male by a pink-eyed female (pink is in Chromosome III) gave in 

 Fi all the males normal, and the females also normal except in 

 respect to barred, which is dominant; in the next generation, F2, 

 all combinations of the factors concerned appeared. If bent had 

 been in Chromosome III, no crossing-over would have occurred 

 between bent and pink in the Fi male, consequently, no pink bent 

 spermatozoa could have been formed, and thus (since pink and 

 bent are both recessive) no pink bent F2 individuals could have 

 been produced. Since these were produced, bent did not lie 

 in the third chromosome. Similarly, if bent had been in the first 

 (X) chromosome or in the Y, no crossing-over between it and the 

 sex-factor would have taken place in the Pi or Fi male, and conse- 

 quently no bent females could have been produced in F2. As bent 

 females were produced, bent did not lie in the first (X) chromo- 

 some, or in the Y. 



A bent barred eyed F2 male was then crossed to a female con- 

 taining three mutant genes in the second chromosome, namely, 



