No. 636] GAMETIC RATIOS IN DROSOPHILA 



55 



of the mutants are little if at all below normal in de^ee of 

 viability] that is, when such mutants are compared with the 

 wild t^-pe under identical conditions, the observed ratios show 

 little or no deviations from expectation l)eyond tlioso dno to 

 random sampling. As an example may be mentioned white- 

 ocelli, which is known to have maintained itself with practically 

 undiminished frequency through 175 generations of competi- 

 tion, under unfavorable conditions of culture,'' with the wild- 

 type. In the main, the mutant races that show normal viability 

 are those whose somatic effects are "slight". Thus, white-ocelli 

 affects the color of the tiny group of three ocelli on the top of 

 the head. The character, though involving so small an area, is 

 perfectly sharp and definite, and under proper conditions of 

 illumination and magnification is fairly easy to classify. The 

 same is true of many other "slight" mutations, such, for ex- 

 ample, as speck, cross-veinless, and hairy, which are among the 

 most valuable Drosophila mutations. On the other hand, 

 mutants that involve more extensive or manifold changes, such 

 as club, notch, rudimentary, and delta, are also among those 

 poorest in viability. Some of these changes in themselves inter- 

 fere with the success of the individuals possessing them: flies 

 with "spread" wings or "dachs" legs are liable to become 

 caught in the culture media and die. These changes are also 

 sometimes obviously accompanied by serious internal derange- 

 ments. In the ease of streak, for example, it can be seen that 

 the internal muscles of the thorax are largely replaced by bub- 

 bles ajid blood sinuses. The correlation between inviability and 

 the extent of the visible change is high, but is lessened by the 

 cases in w^hich the internal accompanying changes are of slight 

 disadvantage. Thus, the mutant "pads" resembles "club" 

 very much, and appears to be a greater change in the same 

 direction, but is nevertheless far freer from inviability. Con- 

 versely, certain mutants that are usually lethal occasionally do 

 produce offspring, which are then not as strikingly different 

 from the wild-type as some other mutants that have good 

 viability. Lethal-10 very occasionally survives, and the indi- 

 viduals are scarcely to be distinguished from dwarfs of a cer- 

 tain nuitant race (dwarfoid) that is little inferior to the wild: 

 type in viability. 



• Biol. Bull., 1920, pp. 231-23 



