498 



SCIENCE 



[N. S. Vol. XXXIII. No. ! 



peared with wings shorter than the abdomen. 

 The square, short wings were broad and often 

 blistered; in individuals obtained later their 

 ends were crenated at times as though incom- 

 pletely unfolded, but at other times the ends 

 were quite even. This fly, bred to his sisters, 

 reproduced the same condition in a few of his 

 male offspring, and in closely related stock a 

 few males of this kind appeared in successive 

 generations. By repeated intercrossing a 

 large number of these males have been pro- 

 duced and some females. It is important to 

 note that although I have paired numbers of 

 these males and females together I have ob- 

 tained offspring in only a very few eases and 

 these were all normal flies. Inbred these nor- 

 mal flies have produced 964 normal males and 

 females; 6 short winged males and 2 females. 

 Whether the original short winged females 

 were virgins may appear doubtful from these 

 facts. On the other hand the males and fe- 

 males with rudimentary wings are fertile, both 

 with wild flies and with the mutations to be 

 next described. The frequent failure to obtain 

 fertile eggs from pairs of flies with rudimentary 

 wings may seem therefore to indicate that this 

 combination is generally sterile, but whether 

 because the eggs are not fertilized or being 

 fertilized do not develop is uncertain. This 

 point is being further investigated. The 

 extraordinary deficiency of individuals of this 

 class in the second filial generation (see be- 

 low), may be due to the same causes that ac- 

 count for the deficiency when inbred. Similar 

 deficiencies, though not so extreme (see the 

 second generation of white-red eyed and 

 black-yellow crosses), run through nearly all 

 of the results with my new races. 



If a male with rudimentary wings is bred 

 to normal females of the original stocJc all of 

 the offspring have long wings. These inbred 

 produce variable percentages of long and rudi- 

 mentary wings. A partial census of the re- 

 sults gave 5,850 normal wings to 83 males 

 with rudimentary wings. No intermediate 

 types appeared. Later several males with 

 rudimentary wings were bred to a new wild 

 stoch. The offspring had long wings. These 

 flies inbred produced in the next generation 



8,459 normals, males and females, and only 

 32 males with rudimentary wings. The ex- 

 periments show that the condition of rudi- 

 mentary wings is sex-limited. By suitable 

 combinations this sex-limited character has 

 been combined with another sex-limited char- 

 acter, viz., white eyes. The theoretical ques- 

 tions involved in such a combination have 

 been discussed elsewhere.'' 



Miniature Wings. — In the seventh genera- 

 tion of the beaded wing stock a fly appeared 

 with wings like the normal in form, but ex- 

 tending no further than the end of the ab- 

 domen. Flies with wings of this kind are 

 much more viable than those with rudimen- 

 tary wings. Pure cultures of thousands of in- 

 dividuals are now on hand. The character 

 also shows itself to be sex-limited. Many 

 combinations between these flies and those 

 with other modifications have been made. 

 Only a few of these can be now mentioned. 

 Miniature wings have been combined with the 

 white eyes, also a sex-limited character. 



If a male with miniature wings is crossed 

 with a wild female all of the offspring have 

 long wings. These inbred have produced 409 

 flies with normal wings, males and females, 

 and 178 males with miniature wings. No 

 females with miniature wings appeared in the 

 second generation. Miniature wings are, there- 

 fore, sex-limited in inheritance. The recipro- 

 cal cross, viz., females with miniature wings 

 by wild males, gave in the first generation 50 

 females with normal wings and 51 males with 

 miniature wings. Evidently the male-pro- 

 ducing spei-m of the wild fly does not carry 

 the character essential for the formation of 

 long wings. This same spermatozoon also 

 lacks, as I have tried to show, one of the fac- 

 tors essential for the formation of red eyes, 

 likewise a sex-limited character. The Fi flies, 

 inbred, produced 785 normal males and fe- 

 males (in approximately equal numbers) and 

 827 flies with miniature wings, of which 430 

 were females and 397 were males. This result 

 gives the expectation for this combination. 



'Proc. Soc. Experimental Biology and Medicine, 

 Vol. 8, October, 1910, and American Naturalist, 

 February, 1911. 



