388 GENES MODIFYING NOTCH. 



not appear to have been avoided by Castle, hence his appeal to con- 

 tamination of genes to help him out of an apparent contradiction. In 

 the present case of Drosophila the experiment is of a kind to demon- 

 strate clearly whether contamination had occurred or not, and the 

 results clearly show that it did not occur, even under the unusually 

 favorable opportunities that heterozygosis for 24 generations offered. 



(4) A modification of the Notch character appeared several times in 

 the course of the work. This variation, called short Notch (fig. 94, 6,), 

 is in the opposite direction from the selected type "produced by 

 selection." By proper tests it is shown that this variation is due to 

 another modifying gene situated in the X chromosome itself. When in 

 homozygous condition the gene shortens and broadens the Notch wing, 

 producing a greater amount of curvature at the end. This variant, 

 too, can be brought back at any time to the original or atavistic type 

 by breeding to wild flies. 



(5) In the course of the work a number of other mutations occurred, 

 some of which modified the wing in somewhat the same way as the 

 Notch gene itself (nick and cut), others modified the wing as domi- 

 nants (truncate), or in the homozygous condition (deformed eyes, etc.). 

 Other modifications causing serrations or notchings on the end of the 

 wings are known in Drosophila; the location of these genes in other 

 chromosomes or at other levels than Notch in the X chromosome 

 shows that they are different from Notch. Were it not possible, as it 

 is hi this case, to check up such modifications that resemble somewhat 

 the character under selection, one might easily be led to entirely 

 erroneous deductions. 



(6) In the course of the experiment two females appeared with 

 exceptional sex ratios, viz, 76 9 to 1 cf and 119 9 to 10 cf. Their 

 occurrence is undoubtedly due to the appearance of a lethal in the 

 "normal" X chromosome of the Notch mother, because hi several 

 cases such changes in the sex-ratio in Drosophila have been shown to 

 be due to such a situation. In consequence of two lethals in the 

 mother, one in each X chromosome, every son will die, except for an 

 occasipnal cross-over that will give rise to a normal son. That this 

 result is not due to the production of a homozygous Notch female by 

 non-disjunction is demonstrated by the kind of daughters produced, 

 which were half normal, half Notch. All must have been Notch if the 

 mother had been a double lethal Notch female (XX Y). 



