loo Journal of Agricultural Research voi. xii, no. 2 



ether-extracted kernels. Their results may be due to the influence of 

 previously heating the kernels to 120° C, thus possibly decomposing 

 some of the gossypol, and to incomplete extraction so that the remain- 

 der was left in the three fractions, the extracts and the residue. They 

 also assume that the ethyl-ether extract of petroleum-extracted kernels 

 is always 2 per cent of the weight of the kernels. This is the case only 

 with long-continued extractions. 



SUMMARY 



Raw cottonseed kernels contain about 0.6 per cent of gossypol and are 

 highly toxic to rats. Ether extraction renders the material nontoxic 

 and gives a highly toxic extract containing about 2 per cent of gossypol. 

 Gossypol fed in milk diets in amounts equivalent to those contained in 

 the raw cottonseed diets has proved as toxic as raw cottonseed. Gossy- 

 pol may be quantitatively removed from the ether extract by precipita- 

 tion as its insoluble anilin compound. The extract is thus rendered non- 

 toxic to rats. The insoluble anilin compound of gossypol is not toxic 

 because of its insolubility. Gossypol prepared from this compound 

 possesses its original toxic properties. 



Cottonseed meal is much less toxic than raw cottonseed, owing mainly 

 to the oxidation of gossypol during cooking. 



Cottonseed meal, ether-extracted cottonseed, and gossypol have been 

 fed to small pigs in pens under comparable conditions. Cottonseed 

 meal has been found definitely injurious, while the ether-extracted raw 

 seed does not appear to cause cottonseed-meal poisoning. Gossypol has 

 been found toxic to pigs. 



If the presence of an injurious substance in the meal is disregarded, a 

 diet of cottonseed meal and corn meal has nutritive limitations which 

 may, under restricted conditions of living, lead to failure of pigs to thrive. 

 Such failure is a phenomenon distinct from cottonseed-meal poisoning. 



Outdoor exercise, access to forage and soil, and improved diets tend 

 to postpone or avert cottonseed-meal poisoning of swine. The defi- 

 ciency hypothesis that cottonseed-meal poisoning of swine is similar to 

 beriberi is untenable. 



LITERATURE CITED 



(i) Forbes, E. B., Beegle, F. M., and Mensching, J. E. 



1913. MINERAL AND ORGANIC ANALYSES OF POODS. Ohio Agr. Exp. Sta. Bul. 



255, p. 2II-23I. 



(2) Hart, E. B., and McCollum, E. V. 



1914. INFLUENCE ON GROWTH OF RATIONS RESTRICTED TO THE CORN OR 



WHEAT GRAIN. In Jour. Biol. Chem., v. 19, no. 3, p. 373-395, 11 

 charts, i pi. 



(3) HOGAN, A. G. 



1916. THE NUTRITIVE PROPERTIES OF CORN. In Jour. Biol. Chem., v. 27, 

 no. I, p. 193-208. Bibliography, p. 208. 



