88 T. H. MORGAN 



many females as males shows that the male was non-lethal; 

 for, had he been lethal, half the daughters of this cross would 

 have been homozygous for lethal and would not have appeared. 

 This would have resulted in equality of red females and white 

 males. The test of this male was continued into F2 by breeding 

 three Fi pairs, which gave: 



red 9 9 white 9 9 red d^ white cf 



L 242 AA 25 12 



L242AB 34 20 



L 242 AC 12 11 7 19 



This result is in conformity with the view that the grand- 

 father was non-lethal, and impossible on the view that he was 

 lethal. On the first view there should be two kinds of daughters 

 (exclusive of cross-overs), viz., those heterozygous for lethal and 

 free from white which would give a 2:1 sex-ratio, and those free 

 from lethal but heterozygous for white (type B) which should 

 give a 1:1:1:1 proportion. Fortunately both types were met 

 with among the three daughters selected at random. 



Theoretically, none of the white F2 females in the cross ana- 

 lyzed should be heterozygous for lethal, except those rare cases 

 due to crossing-over. These should be as frequent as the cross- 

 over males, viz., one in about two hundred. Of the ten grand- 

 daughters tested, nine gave (table 5) 1 :1 ratios, but one (L 9 D) 

 gave apparently a 2:1 sex-ratio, although it was not seen in time 

 to test further. Such a female mated to a red male in Pi and 

 the daughters again mated to red males should give results in 

 many ways the converse of those obtained in tables 3 and 4, for, 

 now, the red males would live and the white males die. 



The locus of Lethal I 



In table 3 the number of cross-overs (red males) was two, and 

 of non-cross-overs (white males) was 370. Table 4 gives 4 cross- 

 overs to 589 non-cross-overs. The total of 6 cross-overs to 959 

 non cross-overs gives the percentage of crossing-over between the 

 loci lethal and white as 0.6 per cent. 



