SUMMARY : Toluene-treated washed suspensions of rumen bacteria break down proteins largely to amino acids ; in the absence of toluene bacterial deaminases are active. Unlike the deaminases, the presence of proteases does not depend, to any great extent, on the presence of readily attacked protein in the diet of the host animal. Extracts of acetone-dried powders of the bacteria also show proteolytic activity. Rumen protozoa are also proteolytic, and ammonia appears to be the end product of their nitrogen metabolism. Ammonia production due to the protozoa is not as sensitive to toluene as is the case with bacteria. Much of the ammonia production in the rumen in the absence of substrate appears to be due to the endogenous metabolism of the protozoa. Extracts of acetone powders, and extracts prepared by simple freezing and thawing of the protozoa, contain active proteases.In an artificial rumen apparatus it was shown that when digestion was complete, about half the N and C of added casein could be recovered as ammonia and volatile fatty acids respectively. Most of the remainder could not be accounted for analytically, and was presumed to be used for microbial growth, which had occurred. When starch or some other polysaccharides were added to the artificial rumen apparatus as well as casein, the production of ammonia was lowered. This was shown not to be due to any effect on proteolysis or deamination, and was presumed to be due to the increased utilization for microbial growth of some breakdown product of casein.From a t least the time of Zuntz (1891), it has been suspected that microorganisms of the rumen split the proteins fed to the host animal; there has been, however, little unequivocal direct evidence for this proteolytic activity. Schlottke (1936) showed in vitro that glycerol extracts of rumen protozoa contained a protease active a t pH c. 6.1, and also a dipeptidase, but was unable to find much activity in glycerol extracts of rumen bacteria. However, Sym (1938) showed active proteolysis a t pH 6.3 by suspensions of rumen bacteria and of protozoa, and also by extracts of acetone powders of these microorganisms; he found peptidase activity weak in his preparations. Both these authors found little or no free protease in the supernatant rumen liquor. Nikitin (