De Assis in 1939 (a,b) described strain 648, which was related in cultural properties to Shigella alkalescens, although it, failed to ferment dulcitol. Antigenically it was entirely different from the ordinary S. alkalescens.\_ At that time this form was called alkalescens II, a designation that was adopted by Neter (1944). Approximately 20 strains of this type were isolated from cases of dysentery in Rio de Janeiro and Sao Paulo (De Assis et al., 1946; De Assis, 1947b). Two of the strains slowly utilized lactose (De Assis et al., 1946). In a recent publication, De Assis (1947b) designated these strains as Shigella tiet.1 Among the strains employed by Braun and Unat (1942a,b, 1943a,b) in their investigation of the inagglutinability of certain shigellae, a strain Clark was mentioned which, when received in this laboratory, proved to be culturally and serologically indistinguishable from S. tiete. The isolation of strains of this type at two points as distant as Brazil and Turkey suggested that S. tiete is more than a local Brazilian variant. For this reason, and because the confusion regarding this type reflects what may occur in the classification of shigellae in general, we feel justified in reporting our experiences with S. tiete. The strain Clark was obtained through the courtesy of Professor Braun both as "Ocl" and "Ol" variant. Dr. de Assis kindly sent us four of his strains including the lactose-fermenting cultures, "Gen" and "Cav." We are also indebted to Drs. Ewing, Francis, Neter, and Mollari for strains used in this work. S. tiete is a gram-negative rod that is nonmotile in fluid media and semisolid agar at 20 C and 37 C. It does not liquefy gelatin nor attack urea. It is anaerogenic, indole-positive, and it reduces trimethylamine oxide (Wood and Baird, 1943; Weil and Black, 1944). It forms acid rapidly from glucose, mannitol, maltose, and arabinose, and may or may not slowly utilize salicin. Dulcitol and rhamnose are not fermented. Lactose was not acidified by two of the strains of Dr. de Assis, but was slowly fermented by strains "Gen" and "Cav" as well as by Braun's strain Clark. On agar plates, the strains form both clear and opaque colonies, as observed in S. alkalescens (see figure 1 and references in Weil, 1947). All strains formed round smooth colonies. They grow evenly throughout fluid media, and suspensions are salt-stable. The clear variant is agglutinable. The opaque variant is inagglutinable when tested in the living state, but after boiling for 1 hour the bacteria are