Journal of Agricultural Research voi. xxvn. no. 7 
5 io 
means. Finally, the percentage of 4-lock bolls was determined in 1923 
for the homozygous F 3 progenies 21-21 and 21-22, the former dominant 
and the latter recessive for petal spot. Soil heterogeneity was not a 
factor in this comparison, the two progenies having been grown in adja¬ 
cent and conterminous rows. The percentages were 5.11 ±0.33 and 
5.33 ±0.34, the difference having been much smaller than its probable 
error. 
The two plants of 1917 which became the progenitors of the two spot¬ 
less families were selected late in the fall in the course of a search for 
plants having exceptionally numerous 4-lock bolls. The unusual char¬ 
acter of their flowers was first noticed in their progenies. The spotless 
character is extremely rare in the Pima variety and, indeed, has not 
been observed except in the descendants of these individuals. Con¬ 
sidering this fact and the absence of linkage between spotless petal and 
a high percentage of 4-lock bolls, the coincidence would be an extraor¬ 
dinary one if the original spotless plants had been unrelated. It.is 
much more probable that, although discovered almost by accident in a 
field grown from “bulk” seed, they were both descended from an indi¬ 
vidual presenting a chance combination of the two characters. 
DISCUSSION 
Comparatively few characters of the cotton plant have been proved 
conclusively to segregate in definite Mendelian fashion. It has seemed 
worth while, therefore, to present rather fully the evidence that the 
character petal spot affords an example of such behavior. The sharp¬ 
ness of the segregation in this case is interesting because -the recessive 
allelomorph usually is expressed in a faintly spotted rather than an 
absolutely spotless condition. It is also of interest that the character 
proves to be decidedly variable in the homozygous populations, whether 
recessive or dominant, variation being shown not only from plant to 
plant and in response to environic influences, but among flowers pro¬ 
duced simultaneously on the same individual plant. 
Petal spot, unlike such characters as fruitfulness, size of the boll and 
length and abundance of the fiber, is of no direct agricultural or com¬ 
mercial importance. Nevertheless, any character which may serve as 
the “hall mark'' of an agriculturally desirable strain or variety is likely 
to prove useful. This is particularly the case in a crop like cotton, since 
the commercial value of the product depends largely upon its uniformity 
and this, in turn, upon the genetic purity of the stock. Cotton is sub¬ 
ject to cross-pollination and, consequently, in order to maintain pure 
planting seed of a variety, it is often necessary to rogue seed-increase 
fields which have been grown in proximity to another variety or stock, 
in order to eliminate accidental hybrids. It is obvious that roguing can 
be done thoroughly only if the stock to be rogued possesses characters 
which make it easy to distinguish the hybrids resulting from cross¬ 
pollination with other stocks. It is also important that the distinguishing 
characters be recognizable early in the summer, before there has been 
opportunity for extensive cross-pollination of the parent stock by the 
hybrids present in the same field. 
The Pima variety is so uniform morphologically that it is usually im¬ 
possible to detect, in time for roguing, the offspring of accidental crosses 
between strains which have been selected on the basis of fruitfulness, 
earliness and properties of the fiber and seeds. Such a character as spot¬ 
less petal meets all the requirements of a serviceable “hall mark” for its 
