426 
Journal of Agricultural Research 
Vol. XIV, No. 10 
diets designed to reveal whether the failure of their animals was due to 
inadequate diets or to the presence of a mildly toxic substance. We 
believe we have accomplished this by comparing ether-extracted cotton¬ 
seed kernels with cottonseed meal. 
After the appearance of Rommel and Vedder’s ( ij ) preliminary note, 
we conducted some preliminary experiments with pigs (j6) in which we 
found gossypol isolated from raw cottonseed to be markedly toxic to 
these animals and also demonstrated that while animals on cottonseed 
meal quickly sickened and died, pigs which were fed on the ether- 
extracted cottonseed kernels were not affected. This fact seems to us 
to outweigh the hypothesis advanced regarding deficient diets as the 
cause of cottonseed-meal injury, and to confirm our previous view. 
With reference to the presence of a deleterious substance in cottonseed 
meal, we would emphasize at the outset the fact that pigs, rats, and 
rabbits grow better on a diet containing ether-extracted kernels than on 
a diet containing cottonseed meal. Pigs and rabbits are so quickly 
affected by cottonseed meal that we were led to state (15) that— 
generally speaking, the meal and the kernels are toxic to the same degree. 
Many subsequent experiments by us with four species of animals 
have shown that meal is far less toxic than kernels for rats and hens, but 
the fact remains that cottonseed meal, even a thoroughly cooked meal, 
is highly injurious to rabbits and pigs. 
Following the paper by Richardson and Green (10) we fed rats 
on cottonseed products. The vast difference between the highly toxic 
raw seed and the very slightly toxic cooked product was immediately 
evident. This indicated to us that in the manufacture of cottonseed 
meal the gossypol undergoes some change whereby the meal is rendered 
much less toxic than the original kernels. 
In this paper we report some of the experiments conducted to ascer¬ 
tain to what extent the change in toxicity takes place under oil-mill 
conditions. These experiments led to the conclusion that there still 
remained a toxic factor in all the samples of cottonseed meal and cotton¬ 
seed flour which we fed. Rats and hens are less affected by this factor 
than rabbits and swine. In fact, in diets well supplemented with milk 
powder the toxic factor for rats may remain entirely masked. 
OIL-MILL TREATMENT OF COTTONSEED PRODUCTS 
In order to compare the effects of cooking on the toxicity of cot¬ 
tonseed, we have used products obtained from cottonseed-oil mills. 
For comparing the effect on different species, two meals were exten¬ 
sively fed. These, together with other samples, were collected by 
one of us (Carruth) personally at the oil mills. Information was 
also obtained regarding the length of time, steam pressure, and 
other details of the cooking process. It is evident to us that the 
