SHULL: DUPLICATION OF A LEAF-LOBE FACTOR 433 
interest in certain peculiarities of inflorescence and capsule characters, 
which will be discussed elsewhere. (b) The second fact which sug- 
gested the probable occurrence of duplication of the B factor in this 
strain was the frequent preponderance of rhomboidea-like plants in 
the pure-bred families. A large portion of the plants which, because 
of precocious development, did not reach the full typical adult con- 
dition, showed, nevertheless, highly developed rhomboidea characters, 
even when no trace of the elongated A lobes was discernible. Other 
plants in the same pure-bred families showed conspicuous elongation 
of the primary lobes, thus making certain that the A factor was 
present. If two B factors were present, namely B and B’, associated 
with only one A factor, the greater ease with which the A character, 
as compared with the B character, was suppressed by unfavorable 
conditions, would be readily comprehensible. 
With this clue to a possible interpretation of the rosette characters 
in the Tucson strain, especial note was taken of the F.2 families from 
crosses between the same Tucson biotype and B. Heegeri simplex, 
but in every case the number of individuals which fruited with rel- 
atively undifferentiated “juvenile’’ condition of the rosettes was so 
great that the F. ratios gave no fully convincing proof of the correct- 
ness of the hypothesis. The results of such crosses are given in 
Table 2. 
TABLE 2 
The Composition of the Fz Progenies from a Cross between Bursa bursa-pastoris heteris 
from Tucson, Arizona, and B. Heegert simplex, grown at the Station for 
Experimental Evolution 
Fecigrec Dydmbers heteris rhomboidea | tenuzs | simplex | Ratio 
Pi F Pp | AB ab Ab | ab B:6 
11505 13202 | 87 44 16 | 4 6:55 y-0 
x 12353 13203 | 189 66 18 I2 S50 211 
11425 | 13204 | 7 59 13 9 copay een 
13205 18 13h 7 (6) V ALASE= I 
Seon oie 13206 45 2a 4 Bs jh 72h eur 
Ahi 354 13207 35 18 6 4 5.30: 1 
aoe 13208 142 122 | 33 14 5-62 : I 
Wl Gtal se eeqa lho rouge const | 592 | = 364 | 97 51 | 6.46 : I 
Fiepected....s\.....- +. Veet ema be cape! We aan # i her 17 15.00: 1 
The results in the several families were quite consistent, but the 
deviation from the expected 15 : 1 ratio were in all cases very con- 
siderable. If it is taken into account, however, that the demon- 
strated difficulty in genetical studies with this Tucson strain arises 
