296 
Journal of Agricultural Research 
Vol. II, No. 4 
A point of interest in connection with the origin of varieties in this 
type of cotton is the independent appearance of forms which are nearly 
identical in several of the characters by which they differ from the parent 
stock. As was pointed out on a preceding page, the Yuma variety is 
almost the exact counterpart, in foliage, involucres, and bolls, of the 
Nubari variety, which appeared in Egypt three years before the Yuma 
variety appeared in Arizona. Both varieties are derived from the Mit 
Afifi, which has very different characters. It has also been shown that 
the Pima variety, a descendant of the Yuma, bears a close resemblance in 
some of the characters by which it differs from its parent to the Somer- 
ton variety, which had appeared simultaneously with the Yuma three 
years previous to the appearance of the Pima variety. This independent 
appearance of the same or very similar new characters is paralleled in 
the case of Oenothera Lamarckiana , which has repeatedly given rise to 
identical or nearly identical mutants. 
While mutation in these two plants appears to be essentially the same 
phenomenon, it is very much more active in Oenothera Lamarckiana . In 
De Vries's cultures during the first seven generations 1.5 per cent of the 
total number of individuals were mutants. On the other hand, in the 
Egyptian cotton grown in Arizona one conspicuously mutating indi¬ 
vidual among many thousand appears to be the limit of expectation. 
Even the heterogeneous stock of the Mit Afifi variety, which was intro- 
duced into Arizona 12 years ago, has produced only three or four mutants 
of a striking character, and in the more closely selected Yuma variety, 
which is now 7 years old, only one noteworthy mutant, the Pima, has 
been detected among the several thousand plants grown in progeny 
rows in each generation. 1 
The evidence at hand indicates that Egyptian cotton is a mutable 
group and that the mutability is of a type very similar to that occurring 
in Oenothera Lamarckiana . In seeking an explanation of the occurrence 
of mutation in Egyptian cotton it is therefore in order to consider cer¬ 
tain theories which have been advanced to account for the mutability 
of Oenothera. 
Shortly after the publication of De Vries's work it was suggested by 
Bateson and Saunders (1902) that the appearance of mutants in Oeno¬ 
thera Lamarckiana is due to hybridization. Other biologists have since 
adopted this idea. Thus, Tower (1910, p. 315-316), discussing the 
results of his experiments in crossing different species of Eeptinotarsa, 
a genus of beetles in which, beginning with the sixth hybrid generation, 
the hybrid bred true except for the occasional appearance of mutant-like 
individuals, states: “These strains . . . gave results which strongly 
suggested that the interpretation of a mutative period, as described by 
De Vries in Oenothera Lamarckiana , may well be the variability which 
follows complex processes of hybridization." 
1 It does not follow that numerous minor or undesirable variants, eliminated each year in the process of 
roguing, might not, if subjected to the test of line breeding, prove to be mutants. 
