OF MUTANT CHARACTERS. 281 



There was another possible method of distint:;uishing between these 

 views, which was tried. It had been found that the occurrence of the 

 forked-bar deficiency had distributed the Hnkaj^c relations in the first 

 chromosome in a definite way. All crossing-over in the rcf^ion between 

 forked and bar was eliminated, as proved by direct tests with forked 

 and bar, and likewise by tests of the decrease in the amount of crossin^;- 

 over from that which nominally occurs between the nean^st lori on 

 either side, namely, rudimentary and fuseil. In the case of dachs- 

 lethal there was no other gene known to be included in the deficient 

 region itself, so that direct tests were impossible; and even worse, 

 there were no loci close enough to dachs to give a meiisure of the 

 decrease unless it were very marked. It was thought possible tlrnt 

 a rather extensive disturbance might be initiated by a reliitively short 

 deficiency, since the shortened chromosome might well prevent perfect 

 synapsis for a much longer region because of the "pucker." 



The only practical but unsatisfactory test that could be made w;is 

 through black, which was the nearest workable locus to the right, and 

 star, which was the only locus to the left that could be used without 

 inaccuracy. 



If the dachs locus were deficient it could still be controlled by means 

 of the haploid dachs flies produced by testing with dachs. Thus the 

 proposed experiment involved a female carrying a star dachs-defi- 

 cient second and a black second chromosome, to be tested by means 

 of a dachs black male. Star and dachs were put in the same chromo- 

 some because that method was far easier, and also because the recip- 



rocal back-cross ( ^ cf X6 9 ) offered a means of carrying 



on the dachs-lethal stock with no opportunity for crossing-over, since 

 the only heterozygote was the male, in which sex no crossing-over 

 occurs. The stock had been run with star and dachs-lethal in opposite 

 second chromosomes. To get them into the same chromosome a feiruile 



( j was out-crossed to a dachs male. The star dachs cross- 

 over offspring contained a chromosome of the desired composition 

 (- — -^—j. To be sure of retaining this chromosome, and not 



getting a plain star dachs chromosome by crossing-over, a male of this 

 type was used in the next step, which was an out-cross to black (culture 

 2613). All of the star offspring of this cross were of the desired com- 

 position (51A_ j . Stock was started by mating such males to black 



females and repeating each generation (table 128). No sjiecial pains 

 need be taken to see that the females are virgin in this stock, which is 

 an advantage not possessed by the similar selecteil stocks of the sex- 

 linked mutants where the heterozygote has to be the female and virgin. 



