AveusT 16, 1918] 
relations, into consideration. It is evident to 
any zoologist who has studied large series of 
specimens of a wide-ranging and plastic spe- 
cies that it is, of course, easy to mistake false 
for true intergradation, as, for instance, if 
adults in comparable age and condition are not 
used in comparison. It is likewise easy to 
overlook evidence of intergradation, since the 
latter is sometimes obscured by other cireum- 
stances. For example, two forms may meet on 
the edges of their ranges and intermingle on 
common ground, remaining perfectly separate 
and be apparently distinct species, yet else- 
where directly or indirectly through other 
forms completely intergrade. This is exempli- 
fied in our case of Aphelocoma by Aphelocoma 
californica woodhouseti, which remains an ap- 
parently distinct species where its range meets 
that of Aphelocoma californica immanis at the 
foot of the Sierra Nevada, but which eastward 
passes directly by continuous geographic inter- 
gradation into Aphelocoma californica texana, 
and thence through forms in eastern Mexico 
into Aphelocoma californica sumichrasti, which 
in turn intergrades individually with Aphelo- 
coma californica hypoleuca and Aphelocoma 
californica immanis. A parallel case is found 
in the mice of the genus Peromyscus, as shown 
by Mr. W. H. Osgood,® in Peromyscus mani- 
culatus austerus (Baird) and Peromyscus ma- 
niculatus oreas Bangs; and also in Peromyscus 
maniculatus gambelti (Baird) and Peromyscus 
maniculatus rubidus Osgood. To consider 
them species, because at some point where their 
ranges meet they remain distinct while at 
another similar place they intergrade, would 
clearly not best represent the facts. 
The use, therefore, of individual variation as 
one of the chief criterions of intergradation 
seems not only not illogical but necessary. 
Harry C. OBERHOLSER 
COTTON AS A SEED CROP 
Utinization of cotton seed as a source of oil 
and other valuable products has brought for- 
ward two questions for cotton breeders; first, 
the possibility of increasing the oil-content 
5 North Amer. Fauna, No. 28, April 17, 1909, pp. 
52, 53, 66, 69 and 70. 
SCIENCE 
167 
in the seeds of lint-bearing varieties; and 
second, the breeding of a lintless cotton, to be 
grown strictly as a seed-ecrop. In asking the 
first question it is assumed that the oil might 
be increased without reducing the lint, while 
the second is prompted by the idea that lint- 
less varieties could be harvested by machinery, 
thus avoiding the chief difficulty and expense 
in the production of fiber, the labor of picking 
the cotton by hand. 
Increasing the oil in cotton seed was under- 
taken several years ago in connection with the 
breeding of the Trice variety. A large amount 
of careful work was done by Professor S. M. 
Bain, of the University of Tennessee, as- 
sisted by the late Mr. Albert T. Anders, for- 
merly of the Bureau of Plant Industry, but 
without finding the definite differences that 
were sought as the basis of selection for oil- 
content. The fluctuations induced by condi- 
tions of growth or associated with various de- 
grees of maturity attained by the seeds were so 
large as to conceal inherent differences of in- 
dividual plants or progenies. As might be ex- 
pected from the greater proportion of kernel 
to shell, the oil-content seems to be higher 
in varieties with large seeds, more than 24 per 
cent. of oil, or 64 gallons per ton of seed, be- 
ing reported for the Meade cotton in Georgia, 
but large-seeded varieties are unpopular be- 
cause they do not have the highest percentages 
of lint. 
The breeding of lintless varieties might not 
prove difficult, since individual plants with 
entirely naked seeds have been found as chance 
variations in lint-bearing stocks. Degenerate 
“ slick-seeded ” plants with little or no fuzz on 
the seeds and only a sparse covering of lint 
are of rather common occurrence in the South- 
eastern States in ordinary short staple fields 
raised from “ gin-run” seed. Some of the in- 
ferior “Hindi” variations of the Egyptian 
type of cotton have nearly naked seeds. Fail- 
ure to eliminate the Hindi admixture in Egypt 
damages the crop to the extent of several 
million dollars every year. A lintless cotton 
would need to be excluded rigorously from any 
region where other varieties are grown. The 
seed must become much more valuable than 
