It has been shown (Kovacs and Bruckner, 1952; Bruckner et al., 1953a, 1953b, 1953c) that certain bacteria produce glutamyl polypeptides in which the y-glutamyl bond predominates; that produced by Bacillus anthracis (Ivainovics and Erd6s, 1937; Ivanovics and Bruckner, 1937; Hanby and Rydon, 1946) and by Bacillus mesentericus (Bruckner and Ivdnovics, 1937) contains the D-isomer alone, whereas that formed by Bacillus subtilis (Bovarnick, 1942) contains both D-and L-isomers in varying quantities (Thorne et al., 1954). Confirmatory evidence for the presence of the 'y-glutamyl bond has been provided recently by work with synthetic 'y-poly-L-(Waley, 1955; Bruckner et al., 1955a) and-y-poly-D-glutamic acid (Bruckner et al., 1955b). Other natural products in which y-glutamyl residues occur include pteroylglutamic acid conjugates and p-aminobenzoylpolyglutamic acid. These substances are important growth factors and are split by conjugases (Stokstad, 1954). It was thought that an organism that can grow on the bacterial polypeptide and can specifically split the 'y-glutamyl linkages might serve as the source of an enzyme to demonstrate the existence of 'y-glutamyl bonds in proteins.2 Indeed, it has been reported (Haurowitz and Bursa, 1949) that 'y-linked glutamic acid occurs in proteins, but this has not been conclusively demonstrated since there is as yet no reliable method for distinguishing between a-and-y-bonds in protein; an improved chemical method was recently described by Haurowitz and Horowitz (1956).