428 
Journal of Agricultural Research 
Vol. XIV, No. 10 
this change has not been ascertained, but it seems plausible to 
suppose it is an oxidation process. 1 The gossypol appears to be 
converted into D-gossypol, which is very similar in color reactions to 
gossypol, but which is much less soluble and has a higher melting 
point. In the hot-pressing process the highly toxic gossypol, under 
favorable conditions, is quickly changed to this considerablv less 
toxic substance. 
COOKING PROCESS 
The two meals which we used so extensively may be designated as 
the “short-cooked'’ meal and the “long-cooked” meal. The short- 
cooked meal was obtained from an oil mill using single “cookers/’ or 
steam-jacketed drums. The raw kernels after decortication and 
crushing are conveyed to the drum, where they are stirred by large 
revolving paddles. The steam pressure in the jackets was about 40 
pounds. The kernels themselves are not subjected to steam pressure. 
In this drum the kernels are cooked for 28 minutes; then the drum is 
emptied; and the kernels are ready for the “cake former.” In the 
meantime they are kept hot. The cake former allows a definite 
amount to pass out onto haircloth mats. The mat is folded and 
molded by light pressure and then carried to the hydraulic press. 
When enough cakes have been molded, pressure is exerted and main¬ 
tained until the oil ceases to drip. The hard cake is then removed, 
usually allowed to cool, and then ground into the meal of commerce. 
From this particular mill sufficient meal was obtained for the experi¬ 
ments with rats, rabbits, fowls, and swine. In order to follow the 
change in toxicity on cooking, we took samples from the cooking 
drum at 5-, 10-, and 20-minute periods and also a sample of the kernels 
cooked the full period—28 minutes in this case. These products 
were all fed to rats as described on pages 431-431 (see fig. 1). A 
sample of oil from these fully cooked kernels was also obtained; how¬ 
ever, as it contained practically no gossypol, it was not fed, but was 
assumed to be nontoxic, as was a similar sample secured from this 
mill at another time. 
The long-cooked meal was obtained from a mill using a 5-“stack” (or 
drum) “French continuous cooker.” In this type the steam-jacketed 
drums are placed one over another. The kernels enter the top drum and 
pass down as the drum below is emptied. In this mill the cooking opera¬ 
tion consumed about two hours. The kernels and meal were considerably 
darker than the short-cooked meal and would be considered as off color, 
the meal being almost reddish brown instead of the usually highly 
regarded “bright-yellow” meal. Samples of fully cooked kernels and oil 
1 It is possible that this change may be otherwise effected in the unknown conditions ol the colloidal 
material. Our reason for believing that it is not due to heat alone is that B-gossypol obtained by heating 
gossypol to 180 0 C. differs in properties from D-gossypol. This temperature is about 8o° higher than the 
temperature of cooking cottonseed in the manufacture of cottonseed meal. 
