380 



THE A ME IIK A X XATCLW L Z.s' T [Vol. L V 



TABLE IV 



Pi $ Wild Type X Fi c? Reduced (from Salmox x Eeduced 2) 



8,347.... 133 ! 112 60 76 72 j 4 43 1 



8,348.... 124 ; 119 46 57 67 40 3 



8,354... .i 88 i 105 | 47 43 80 I 4 47 | 



Totals^ . I 345"" I 33 6 I 153 17 6 219 I 8 1 1 30 I 4 



.sive and in an autosome all the females should be wild type. 

 Such is the case. Of the males, four should be red-eyed, 

 three salmon, and one orange. If the gene were dominant, the 

 ratios among the males should be four red, one salmon, and three 

 orange. The actual results are given in Table I. We tind neither 

 of the former ratios, but 1435 red-eyed, 800 orange, and 592 

 salmon. The fact that there is a linkage between the genes for 

 salmon modifier and reduced is further evidence that the gene 

 for salmon modifier is in the X chromosome. 



Orange eye color, then, is due to two sex-linked genes, one of 

 which when alone . produces salmon ; the other called salmon 

 modifier, when alone produces no visible effect, but when added 

 to salmon produces or^inge. The discovery of such a modifying 

 -nil' is not new as Bridges has demonstrated seven of them which 

 itiiHlity cos in eye color. It does lend support, however, to the 

 lircsi'ncr and behavior of such genes. 



These same orange reduced males when nialcd to riMlnced 

 frniales give a very different result from the cio^s to \hr wiM. 

 When cro.ssed to reduced, orange behaves as a >iii'jlc sfx-lliikcil 

 character (Table II). Why this difference ? Ai lii M it was our 

 impression that the reduced strain carried a non-crossover factor. 

 This was disproved, however, by mating reduced to other sex- 

 linked characters. The other alternative and the correct inter- 

 pretation wo think, is tliat the reduced strain is homozygous for 



