234 
Journal of Agricultural Research 
Vol. XXI, No. 4 
A more careful examination was made at Sacaton in 1914 of 124 indi¬ 
viduals taken at random in a field planted with ordinary “gin run” seed 
of Yuma cotton. Of these, 117 were classified as true Yuma, while 7, 
or 5.5 per cent of the total, were classified as distinctly off-type. Pro¬ 
genies grown in 1916 from selfed seed of the off-types and of plants 
regarded as typical Yuma proved to be, in most cases, remarkably 
uniform in the stem, leaf, and boll characters which had distinguished 
the parent individuals. A form which appeared to be analogous to the 
so-called “bull-stalk” variation in Sea Island cotton, characterized by a 
tall stem, long internodes of the axis and fruiting branches and rather 
stiff, semi-erect leaves, was so frequent as to be included within the 
normal range of variation of Yuma cotton, although this type also 
proved to be heritable. 
Cooperative roguing was carried on during and after the substitution 
of the Pima for the Yuma variety. In 1916, 1917, 1919, and 1920 an 
estimated total of 3,600,000 plants were examined; and, by actual count, 
10,221 plants, or 0.28 per cent of the total number, were removed from 
the seed-increase fields of the Pima variety. There was little variation 
in the amount of elimination from year to year, the figures having been 
0.21 per cent in 1916, 0.42 per cent in 1917, 0.20 per cent in 1919, and 
0.20 per cent in 1920. Although the roguing was much more rigorous 
than that practiced with the Yuma variety, the percentage of the total 
plants which were removed from the Pima fields in all four years was 
less than one-third as great. Most of the Pima “rogues” departed so 
little from the type that they would have been disregarded in roguing 
a field of the much more variable Yuma. A large majority of the plants 
removed were merely sickly or more or less sterile. Indisputable evi¬ 
dence was therefore obtained that the closer line-breeding which had 
been practiced from the beginning with the Pima variety had resulted 
in the development of a much more uniform stock. 
The most distinct off-types in the Pima fields, in addition to a few very 
aberrant individuals of which the characters indicated hybridization 
with some other variety of cotton, were either tall, vigorous individuals 
having exceptionally long intemodes (the “bull-stalk” type) or stiff, 
slender, grayish plants having semierect branches and leaves, the leaves 
being narrower, more deeply lobed, and with more incurved and undulate 
margins than in typical Pima. Both of these forms had their counter¬ 
parts in the older Yuma variety, and it seems not improbable that they 
arise by recurrent mutation or recombination in all varieties of Egyptian 
cotton. Variation in the shape of the bolls was much less frequent and 
extreme than in the Yuma variety, but a few of the plants had shorter 
and less pointed, or longer and more taper pointed, bolls than in typical 
Pima. 
In a bulk field of Pima cotton at Sacaton in 1917 certain individuals 
were selected as slightly off-type. Flowers were bagged on these plants, 
