290 
Journal of Agricultural Research 
Vol. II, No. 4 
While little is accurately known of the origin of most of these varieties, 
the evidence seems to Prof. Balls (1912, p. 105) to justify the conclusion 
that “the majority probably arose as single-plant 'selections.' In the 
case of Yannovitch this is definitely known." 
The impression prevails in Egypt that these varieties can scarcely be 
distinguished from one another except in yield and lint percentage, in 
adaptability to particular soils and climatic conditions, and in the length, 
color, and fineness of the fiber. Nevertheless, when grown in Arizona 
from imported seed, most of them could readily be distinguished even in 
the blossoming period, or at the latest after the first bolls were set. In 
characterizing them, however, it was necessary to ignore the numerous 
hybrids with Hindi cotton which appeared in most of the plantings. 
As compared with the Mit Afifi, the Ashmuni produced lower and more 
bushy plants. The Abassi variety had remarkably long and pointed, 
relatively slender bolls, very different from the plump, short-pointed 
bolls of the Mit Afifi (PI. XXV, fig. 1). The Yannovitch plants averaged 
taller than the Mit Afifi, but there was so much variation among the plants 
grown from imported seed of this variety that a close comparison was 
impracticable. The Nubari variety differed from the Mit Afifi in its more 
compact habit of growth, in its larger, more frequently 5-lobed leaves, 
in having the bracts of the involucre (PI. XXII, fig. 1) more grown 
together at the base, and in the much longer, more tapering bolls. The 
Sakellaridis differed from the Mit Afifi variety in many of its characters, 
notably in the much larger proportion of deeply 5-lobed leaves, involucral 
bracts with long teeth which extended nearly to the base of the bracts, 
and conical, very abruptly and very sharply pointed bolls (PI. XXV, 
fig. 4). In the habit of growth, in the shape and small size of the leaves, 
and in the shape of the bolls the Sakellaridis variety showed some resem¬ 
blance to Sea Island cotton. 
The conclusion that these varieties originated by mutation is supported 
by the following facts: (1) The derivation of each from a single plant 
discovered in a field of very different cotton; (2) the distinctness of their 
botanical characters, especially in the recently developed Nubari and 
Sakellaridis varieties; and (3) their tendency to remain uniform, which 
is, however, finally nullified by the ample opportunities afforded in Egypt 
for cross-pollination with other types and for the mixing of seeds at the 
gins. 
ORIGIN OF NEW VARIETIES IN ARIZONA 
Twelve years ago seed of the Mit Afifi variety, imported from Egypt, 
was planted at Yuma, Ariz. The resulting plants were generally unpro¬ 
ductive, late in ripening, and produced fiber of poor quality. Selection 
carried on for several years resulted in some improvement in these 
respects, but the progress was not very encouraging. Although the 
plants showed considerable fluctuation, until 1908 there was no clear 
evidence that any of them had exceeded the limits of the characters of 
the Mit Afifi variety. 
