32 BULLETIN 1164, U. S. DEPARTMENT OF AGRICULTURE. 
BOLL SHAPE. 
Shape of the boll, as expressed by the boll index (diameter relative to length), 
gave a frequency distribution in F 2 of the Holdon-Pima hybrid (fig. 35) which indi- 
cated that this character involves several factors. 
Balls (6, pp. 162, 163) records similar behavior of the boll index character in 
upland-Egyptian hybrids studied by him. 
McLendon (34, pp. 200-205), working with sea-island-upland hybrids, concluded 
that the short thick type of boll is recessive, although he did not find the long, 
slender type to be fully dominant. But his evidence of a monohybrid ratio of 3 
intermediate to 1 short and thick is not convincing, especially when it is considered 
that his classification was made by inspection rather than by measurement. 
BOLL SURFACE 
Satisfactory grading with respect to boll surface, if it can be regarded as a single 
character, was not found to be practicable in the Holdon-Pima hybrid, although 
the parent varieties differed conspicuously. 
Balls (4, pp. 59, 60), while recognizing the probable complexity of the character, 
dealt with it in studying an upland-Egyptian hybrid, reporting an intermediate Fj 
and an approximately 1:2:1 segregation in F 2 into upland type (smooth), Fj type 
(intermediate), and Egyptian type (pitted). He states also that F 3 progenies of two 
smooth and of one pitted F, plant bred true. In a later publication (6, p. 142), 
however, he states that this apparently simple type of inheritance did not hold good 
upon further study of the F 3 data. 
McLendon (34, pp. 205-210) found that in sea-island-upland hybrids F x was inter- 
mediate, while the data for numerous F 2 populations show ratios varying from 1 : 1 to 
29:0 for pittedness as contrasted with smoothness. 
FIBER LENGTH. 
The first generation of the Holdon-Pima hybrid had a mean fiber length which 
approached the mean of the longer fibered (Pima) parent while F 2 gave a unimodal 
and nearlv svmmetrical frequency curve, indicating that several factors are involved 
(fig. 38). > 
Balls (4. pp. 6, 7, 77-80) obtained data from an upland-Egyptian hybrid from 
which he deduced a 3 to 1 ratio for length of fiber, long being dominant. He reports 
cases of intensification in crosses of Mit Afifi Egyptian with a short-staple cotton, Y x 
having considerably longer fiber than the Egyptian parent. 
McLendon (34, pp. 211-217), studying length of fibe'r in hybrids of sea-island 
with upland varieties, obtained an intermediate and rather variable F x . Several of 
his F 2 frequency curves are strongly bimodal, with long fiber in excess, while other 
F 2 populations gave plurimodal curves. 
FIBER COLOR. 
In the Holdon-Pima hybrid the F t was approximately intermediate in color of 
fiber, and the frequency curve of F 2 was unimodal and fairly symmetrical, although 
flatter than the curves for most of the characters (fig. 39). 
Balls (4, pp. 4, 80, 81), crossing a brown-fibered with a white-fibered cotton, 
reports a distribution in F 2 of 30 brown, 66 cream, and 32 white. In a hybrid of Mit 
Afifi Egyptian, which has much browner fiber than the Pima variety, with a white- 
fibered upland cotton, F x was intermediate (cream colored), and an F 2 segregation 
into 12 brown, 21 cream, and 11 white was recorded. It is evident, however, that 
the segregation was not so sharp as this classification would indicate, for in discussing 
the results in F 3 one of the F 2 parents is described as "creamish" and another as 
"whitish." An F 3 progeny of 30 individuals from a brown-fibered F 2 plant showed 
only brown fiber, but with more variation in the intensity of the shade than occurred 
in the Mit Afifi parent of the hybrid. In a later publication (6, p. 146) it is implied 
that the inheritance of this character is more complex than was supposed at first to 
be the case. 
SEED FUZZINESS. 
The entire fuzziness of the upland parent was almost completely dominant over 
the partial fuzziness of the Egyptian parent in the first generation of the Holdon-Pima 
hybrid. The frequency curve of the second generation (fig. 41) was unimodal, and 
the mode coincided with the Fj mean. Several factors appear to be involved. 
Fletcher (18) records dominance of the fuzz in a cross of a fuzzy-seeded with a 
naked-seeded cotton. On the other hand, Fyson (20), crossing a smooth-seeded tree 
