ity to arsphenamine liver injury in protein-depleted dogs that could be prevented by prior feeding of protein or the sulfur-containing amino acids. As with chloroform liver injury, arsphenamine liver injury was not prevented by nonsulfur-containing amino acids. This led Whipple to emphasize the importance of body protein stores and their content of S-containing amino acids in protecting the liver against toxic agents, but the exact mechanism of the protective action was not established.

In 1940 Sydney C. Madden and Whipple reviewed eight years of their work and that of others on the plasma proteins; they emphasized the presumptive role of the liver as the site of plasma protein synthesis and were able to confirm for plasma protein synthesis in the dog the dietary essentiality of those amino acids found by William C. Rose to be indispensable for growth in the rat.

During World War II, with the collaboration of Merck & Co. Inc. in supplying pure amino acids, Madden et al. formulated a number of pure amino acid mixtures and tested them for their effectiveness in promoting synthesis of plasma proteins when given orally or parenterally. When given either orally or parenterally with an adequate intake of nonprotein calories several of the amino acid mixtures could completely satisfy all the metabolic requirements for maintenance of weight and nitrogen balance in the dog and at the same time support ample plasma protein and hemoglobin regeneration. This work was also important because it led to the demonstration at Rochester in a few human subjects that a mixture of the essential amino acids, or an enzymatic digest of casein containing all of the amino acids, when given parenterally along with adequate nonprotein calories, could maintain positive nitrogen balance for several days. These results were independently confirmed by Robert Elman et al. in St. Louis and were important be-