CHAKACTERS IN AN UPLAND-EGYPTIAN COTTON HYBRID. 
leaves and other organs in as nearly as possible the same stage of 
development. But with two such different types of cotton as upland 
and Egyptian, and with the very great diversity which characterizes 
the hybrid between them, it is difficult to select for measurement 
organs which are contemporaneous. 
The investigation was carried on at the Cooperative Testing Station 
at Sacaton, Ariz. Valuable service was rendered by Walton G. Wells, 
who had supervision of the held plantings and did much of the work 
of hybridization, measurement, and computation in 1917, 1918, and 
1919, and by Walter F. Gilpin, who assisted in the work during 1919 
and took the place of Mr. Wells in 1920. Others who assisted in 
the investigation were F. Ben Clark, George J. Harrison, R. H. Man- 
they, and Max Willett. To G. N. Collins and J. H. Kempton the writer 
is indebted for much aid in the computation of the coefficients of 
correlation' and for valuable suggestions throughout the course of the 
investigation. 2 
ANCESTRY OF THE HYBRID. 
The cross determined upon was between the Holdon variety of 
upland cotton and the Pima variety of American Egyptian cotton. 
The Pima varietv was described and illustrated in an earlier publica- 
tion {25, pp. 292-294, pis. 18, 23, and 24) . The Holdon variety, which 
is an extreme representative of the Texas big-boiled type of upland 
cotton, was developed by D. A. Saunders from a plant selected in 1905 
in a field of the Jackson variety near Smithville, Tex. 3 
Plants of the Holdon variety were grown at Sacaton, Ariz., in 1916 
and flowers were cross-pollinated with Pima pollen, giving rise to an 
¥\ grown in 1917 and the latter to an F 2 grown in 1918. Owing to 
vvar conditions, however, an intensive study of these populations was 
not practicable. A population of Holdon was grown in 1917 from seed 
produced by self -pollinated flowers on the 1916 plants, and on a half 
dozen plants in this population flowers were cross-pollinated with pol- 
len from two plants of family P1A40 of the Pima variety. The latter, 
as indicated by the pedigree given in another publication {26, fig. 1, 
p. 229), was in the third strictly self-fertilized generation and was 
doubtless much more nearly homozygous than the Holdon family to 
which the female parent of the hybrid belonged. It would have been 
desirable, of course, to have started with a strain of upland cotton 
which had been as closely inbred as the Egyptian parental strain, but 
in view of the specific magnitude of the difference between the two 
varieties and considering the fact that none of the Holdon pi; 
showed the slightest evidence of other than upland characters, it is 
not believed that the results of the investigation are seriously 
prejudiced by this possible difference in the genetic purity of the 
parents. 4 
Very few bolls matured from the cross-pollinated flowers on each 
of the Holdon parent plants in 1917, and for this reason the ( i 
pollinated seed from all except one of the mother plants was mixed 
together and was planted as a single F x population in 1918. Progenies 
»The illustrations were prepared from photographs by Walter V. Gilpin, with the exception of 
Plates X and XT, which were pre; ared from photographs by lint ei I I Taj lor, 
3 A description of the Holdon varietv is given in "Distribution of Cotton Seed in I! Dept. 
Agr., Bureau of Plant Industry Do-. 1 163, }>■ !-' 13. I 
* The preponderance of self-fertilization in cotton, even when the flowers are naturally open pollio 
makes the chance; srood that most of the progeny of any individual in a well-bred sunk ^ 111 be 
approximately homozvRous (27, p. 11). 
