Numerous dyes of the phenyl-methane, xanthene, azo, and quinone-imide groups received attention. Their possibilities were not exhausted, but most of them were discarded to permit concentration on a few of the most promising. Of the latter, rosaniline hydrochloride gave most favorable results. The 2 most objectionable contaminating forms can be eliminated by exposing crude cultures of the nitrite-oxidizing organism to the action of this dye in 1.0% concentrations for 5-30 min. The other contaminating form can be eliminated by use of mechanically operated, sterile, minute glass pipettes in fishing colonies from the surface of a solid medium. A detailed comparison was made of the nitrite-oxidizing organism and of the contaminating form which could not be eliminated by use of rosaniline hydrochlo ride. This contaminating form is capable of rapid growth in nitrite medium, but of only very scanty growth on nutrient agar. It is incapable of oxidizing NH3 to nitrite or of oxidizing nitrite to nitrate. It is of interest here because certain investigators have described an organism with similar morphological characteristics as a Nitrobacter. This organism stains well with cold carbol fuchsin and when thus stained many of the cells appeared to possess a flagellum-like attachment at one pole which in many instances was several times as long as the cell. In some preparations the organisms were clustered together with the flagellum-like attachments radiating away from the center of the cluster. In many instances the cells were unevenly stained and somewhat pointed at the ends. This organism forms a colony differentiated from the colony of the nitrite-oxidizing organism by its characteristic appearance and unevenness of edge. In stained preparations the nitrite-oxidizing organism appears singly, in pairs, and in irregular clusters; it stains evenly and varies from oval to spherical. The morphology of the nitrite-oxidizing organism is such that it is difficult to determine whether it should be classed as a coccus or as a rod form. A detailed study was made of the morphological, physiological, and cultural characteristics of the 2 contaminating forms capable of rapid growth on ordinary culture media; they are pigment-formers. One organism is a Gram-negative coccus, usually occurring in pairs and forming a small, circular, pink colony. The other is a small, Gram-negative, extremely motile bacillus forming a small, circular, yellow colony. Both were capable of rapid growth in an inorganic nitrite medium but neither was capable of oxidizing nitrite to nitrate. By eliminating these forms from cultures of the nitrite-oxidizing organisms the process of isolation is greatly simplified. Although actinomyces were encountered in this study, they did not interfere to any great extent with isolation of the nitrite-oxidizing organism or with the study of the contaminating bacterial forms.