294 
Journal of Agricultural Research 
Vol. XXV, No. 7 
It is probably true that this apparent relation is due to the close relation¬ 
ship of each of these to the oil. 
Before basing generalizations on the gossypol content of cottonseed 
upon these analyses, it is to be noted that the nitrogen (protein) and the 
oil content of the seeds fall fairly well within the generalization of Bid- 
well concerning the interrelation of the quantities of these two constitu¬ 
ents in seeds from different parts of the Cotton Belt. This conformity 
indicates that the small number of samples of seeds analyzed form a 
representative series. The seeds from the Southwest have a tendency 
to be low in oil, those from the Southeast to be somewhat higher, and 
those from the Pacific coast to be still higher. The nitrogen has the 
reverse relation. The few exceptions which are evident are to be ex¬ 
pected. 
The analyses show that the seeds from the Southwest tend to be low 
in gossypol, those from the Southeast somewhat higher, and those from 
the Pacific coast regions still higher. Even more significant than this is 
the tendency of the gossypol to follow what may be termed the “rule of 
the oil." Seeds which are somewhat atypical of the region in which 
they are grown, as indicated by their oil content, vary correspondingly 
in their gossypol content. This shows that the gossypol content is 
closely related to the oil content, and only in a general way to the place 
of production. The seeds lowest in oil (Lone Star, 1917 or 1918, and 
Trice, 1918) have the smallest gossypol content, while the seeds highest 
in oil (Acala from Bakersfield, Calif.) are only 0.2 per cent below the 
highest in percentage of gossypol (Egyptian seeds). 
These results are of interest to plant physiologists. The correlations 
and variations herein recorded should prove useful in attacking problems 
dealing with the causes which underlie variation in chemical composition. 
The possibility of developing a gossypol-free variety of cotton with the 
retention of the attribute to develop oil, which is at present correlated 
with the development of gossypol, should be borne in mind. The state¬ 
ment herein made as to the “rule of the oil" should be interpreted to 
mean the simultaneous correlated appearance of gossypol and oil, and 
not a cause and effect phenomenon. 
SUMMARY 
(1) Gossypol was found in the kernels of every sample of cottonseed 
examined, in Ingenhouzia (Arizona wild cotton), and in a sample of 
commercial cotton-root bark. 
(2) The optical crystallographic properties of gossypol “acetate" 
are described. 
(3) The proportion of gossypol varies in raw cottonseed kernels from 
about 0.4 to 1.2 per cent, a variation of 300 per cent. 
(4) The gossypol content appears to depend upon factors other than 
varietal factors. If a varietal influence exists, practically it is masked. 
A variation of 200 per cent was found in samples of one variety from the 
same plantation, but from crops of different years. 
(5) The variation in the gossypol content was fairly regular in that it 
tended to vary directly with and bore a true relationship to the oil con¬ 
tent. This was true for all seeds from any one region, regardless of the 
regional tendency. 
