12 THE AMER1CAX XATL RALIST [Vol. LII 



The F 2 Generation 5 

 A glance at Pig. 4 shows that the range of variation of 

 the F 2 group is greater than that of the F a ; that the 12- 

 feathered tail has reappeared in considerable numbers ; 

 that the " curve" is at least bimodal with one apex in the 

 14, 15, 16 rows, and the other in the 12 row; that there 

 are a few individuals that approach the lower range of 

 variation of the fantail, viz., those with 24, 25 and 26 

 tail feathers. 



There is a distinct return of one of the grandparental 

 types, viz., the 12 class. The 13—16 groups clearly corre- 

 spond to a large part of the heterozygous group seen in 

 Fj. Whether the range to the right of this middle group 

 in the F 2 's is significantly different from that in the F t 

 can not be determined by inspection, as the number of in- 

 dividuals is too small. If the F\ and the F 2 groups are 

 made into curves the results show that it is' doubtful if 

 the wider range in F 2 is significant, although the large 

 12-feathered class in F 2 makes the F 2 variability much 

 more marked than the variability in F ± . 



Back Cross 



Some of the F x birds, both males and females, were 

 back-crossed to fantails. Twenty-three offspring were 

 obtained which differed strikingly as a group from the 

 F, and F 2 lots. The number of tail feathers (Fig. 5) was 

 greater; no 12-feathered birds appeared (the lowest num- 



ber was 14) ; while the highest number included birds with 

 30 and 31 tail feathers. The latter would undoubtedly 

 pass-for-fantail, so far as the number of tail feathers was 



