176 
Journal of Agricultural Research 
Vol. XXVIII, No. 2 
elude that any loose compound that may have been formed in butter would not 
materially influence the results of these experiments. Whatever the form in 
which the gossypol eventually may have been found and whatever its source, care 
was taken that it should be distributed evenly throughout the medium and that 
the state of subdivision should be as fine as the condition of the preparation would 
permit. 
Whenever the gossypol was not administered per os it was given dissolved in 
peanut oil. Published data on the chemistry of gossypol, as well as the authors’ 
experience, indicate that the best way to prepare this substance for intraperitoneal 
administration is to dissolve it in oil. The rate of oxidation as judged by color 
changes is much less in this medium than in aqueous alkali or in alcohol. Where 
the object was to determine the toxicity of a given sample of cottonseed rather 
than the toxicity of gossypol itself in the pure state, the ether extracts of the seeds 
to be tested were used. Their method of preparation has been described. It is 
obvious that for these experiments the proper procedure was to use the whole 
ether extract rather than pure gossypol prepared from a given sample of seed. 
The oil solution of gossypol and the oily residue obtained in the ether extraction 
of cottonseed kernels were always administered intraperitoneally, the most 
suitable way to inject an oil solution. 
ANIMALS USED 
Under the conditions obtaining in these experiments, rats appear somewhat 
unsusceptible to gossypol intoxication and do not show some of the typical 
symptoms of cottonseed poisoning found in certain farm animals. Because of the 
convenience of performing a large number of experiments simultaneously, how¬ 
ever, they were used. The results establish the relation of the toxicity of differ¬ 
ent samples of seed to one another and to pure gossypol. All animals were of 
the same sturdy stock, reared in the laboratory. Variability was thus reduced 
to a minimum. 
A few grown mice and a number of grown rats were used. In the early feeding 
experiments, however, growing rats proved best for the tests, so that the discus¬ 
sion is confined chiefly to experiments with them. 
RESULTS OF EXPERIMENTS 
INTRAPERITONEAL INJECTION OF PURE GOSSYPOL AND OF THE ETHER 
EXTRACT OF DIFFERENT VARIETIES OF COTTONSEED KERNELS 
Gossypol was injected intraperitoneally because it is readily absorbed from an 
oil in the peritoneal cavity. 4 Gossypol is more potent by this route than by 
oral or subcutaneous administration. Apparently the injected gossypol leaves 
the oil and some of it is dissolved in serum exudate. The oral administration 
of gossypol dissolved in oil is followed by diarrhea and also by paralysis of the 
intestine and stomach musculature. From these phenomena it follows that 
the dosage varies with the degree to which the local phenomena are superimposed 
upon one another. In one case gossypol was recovered in the oil found in the 
stomach, the oil having been given by a stomach tube seven days before death. 
Absorption from the subcutaneous tissue is very poor. The intravenous admin¬ 
istration of an emulsion of an oil solution of gossypol was not attempted because 
with the relatively small quantities of gossypol available no means were at hand 
for the preparation of an emulsion sufficiently fine to preclude fat embolism. 
4 Later experiments have shown that cottonseed oil becomes emulsified slowly, whereas peanut oil does 
not. No deleterious symptoms from the injection of pure cottonseed oil or peanut oil were observed during 
the lifetime of the animals. 
