44 



The Journal of Heredity 



obtained. The segregation of the F2 

 generations with respect to the pre- 

 sence or absence of a hgule is shown in 

 Table II. In these crosses there are 

 apparently three factor differences 

 operating in the inheritance of the 

 ligule character. In a total of 520 

 F2 plants only 10 were without ligules. 

 Assuming that three independently 

 transmitted factors are involved, the 

 presence of any one of which gives rise 

 to a ligule, the deviation from the ex- 

 pected ratio is 1 . times its probable 

 error. In some other crosses (Nilsson- 

 Ehle 1909; Love and Craig 1918) the 

 presence or absence of a ligule ap- 

 parently involves one, two, three, or 

 even four factor difTerences. 



THE F2 AND F3 GENERATIONS OF A 

 VICTORY-GARTON 784 CROSS 



The parents of the Victory-Garton 

 784 cross differ with respect to panicle 

 type, seed color, and ligules. Segrega- 

 tion with respect to these characters 

 appeared in the Fa generation. Ap- 

 proximately three-fourths of the F2 

 plants had black and one-fourth had 

 white seeds. The segregation with 

 respect to ligules was 489 with ligules 

 to 38 without ligules. The 88 families 

 of this cross which were grown in F3 

 verified the F2 segregation. Of these 

 F3 families, 36 bred true to the liguled 

 condition; 20 segregated in the ap- 

 proximate ratio of 3 liguled to 1 non- 

 liguled plant; 23 segregated in the 

 approximate ratio of 15 liguled to 1 

 non-liguled plant; and 9 bred true to the 

 non-liguled condition. The respective 

 number of families expected in each 

 category on a two-factor basis was 

 38.5, 22, 22, and 5.5 respectively. 

 Theory and observation agree very well 

 in this case, the value of P being 

 . 4609. Deviations as great as this, due 

 solely to the errors of random sampling, 

 would be expected in about one-half of 

 the cases. 



DISCUSSION 



From the facts brought out above 

 and an examination of the figures, it is 

 obvious that the segregation observed 



in the Fj generation of fatua-Victory 

 crosses and fatua-Garton 748 crosses is 

 very different from that among the 

 progeny of heterozygous false wild oats. 

 In the latter there is apparently but a 

 single factor difference involved where- 

 as in the fatua crosses there are un- 

 doubtedly several factor differences 

 involved. Zade (1918) accounts for 

 this difference by assuming that the 

 natural cross between fatua and culti- 

 vated forms which gave rise to a par- 

 ticular false wild oat occurred many 

 years previous ("Kreuzung jahrelang 

 zuriickliegt . . . ") to the time when 

 the false wild form was actually ob- 

 served. If one accepts Zade's hypothe- 

 sis as an explanation of the origin of the 

 false wild oats reported in this paper 

 one must account for the selective 

 elimination of all phenotypes except 

 the three types, cultivated, interme- 

 diate, arid true false wild oats. In each 

 case the intermediate and true false 

 wild oats were apparently identical 

 with the variety in which they were 

 found except with respect to awn 

 development, articulation of the seeds, 

 and pubescence around the articula- 

 tions. These three characters collec- 

 tively depend on a single factor differ- 

 ence for their particular development 

 in each of the three categories, homo- 

 zygous cultivated, heterozygous false, 

 and homozygous false forms. How 

 may one account for the elimination of 

 all the liguled plants which result from 

 a cross between fatua and non-liguled 

 orientalis forms or the elimination of 

 all open-panicled plants? The white 

 color of the false wild Victory seed 

 and the yellow color of the false wild 

 Aurora seed are other instances of 

 recessive characters whose exclusive 

 presence must be explained either on 

 the basis of selective elimination or the 

 possibility that the progeny of the 

 particular fatua-sativa natural crosses 

 showed no segregation with respect to 

 color of seed. Similarly in the case 

 of Garton 784 in which false wild oats 

 were found the exclusive occurrence of 

 black seeds could only be explained on 

 the basis of selective elimination, or 

 that in this case false wild oats ap- 



