664 



Journal of Agricultural Research 



Vol. Xri. No. 10 



In every other instance in which irregularities in sex ratios occurred, a 

 male or hermaphroditic parent was used in which the condition for nar- 

 row leaves was linked with either a determiner for maleness, femaleness, 

 or hermaphroditeness. If the irregularities were relatively slight, as 

 was the case when normal broad-leaved males and females were crossed, 

 maleness and narrow leaf were linked. A partial elimination of these 

 male gametes would produce the actual results obtained. Hermaphro- 

 dites of Lychnis dioica acted in the same manner as the males. Hermaph- 

 rodites of Melandrium album, which we may assume to bear the narrow- 

 leaf and hermaphrodite determiners linked, as they have undoubtedly 

 been derived from males, acted in the same manner as the narrow- 

 leaved males — that is, they produced only females when crossed to 

 normal broad-leaved females, in place of a i to i ratio. These results 

 can be explained on the assumtion of complete elimination of the male 

 gametes of the M. album hermaphrodite, which carry the hermaphrodite 

 mutant and its linked factor, narrow leaf, and in the case of the narrow- 

 leaved L. dioica males of the nearly complete elimination of the male 

 gametes bearing the mutant factor for femaleness and narrow leaf. 



In all cases it seems that the factor for narrow leaves has an inhibiting 

 action on the formation of the male gametes and results in the partial 

 or complete elimination of them. Elimination in the normal males is 

 not complete; otherwise this line would long ago have disappeared. In 

 the mutants in which narrow leaf is linked with femaleness, elimination 

 of male gametes bearing this mutant factor is nearly complete. Shull 

 has also given much evidence which shows that there is also some elim- 

 ination of female gametes bearing this mutant factor but to a less extent 

 than in the males. Such partial elimination of the female gametes was 

 shown in the cross heterozygous broad-leaved female (of the formula 

 FBFb) by a narrow-leaved male Fbfb, which produced two heterozygous 

 broad-leaved females, no homozygous narrow-leave'd females (the nearly 

 complete absence of both of these classes evidently being due to the 

 elimination of male gametes bearing the factor for narrow leaves and 

 femaleness), 630 broad-leaved males, and 463 narrow-leaved males. In 

 this cross all four types should have appeared in equal numbers.* The 



1 The effect of the linked factors "narrow leaves" and "femaleness" on the productioa of male and 

 female gametes can be most readily seen by the use of'this simple diagram 



in which the male gametes are placed on the upper side of the square and the female gametes on the left 

 side, while the number of each of the types of progeny are placed within the small squares, with their 

 respective gametic combiaations. 



