The study of sections, freeze-cleaved, and whole-cell preparations of Deinococcus radiodurans supported the contention that septa close assymmetrically and originate from discrete opposing locations on the cell surface. Tetrads and the larger associations (sheets) of cells in some strains were formed by alternate and synchronized divisions in two planes. The polarity initiating the second division in cells of the Sark strain, in particular, was often expressed in slower growing cells before completion of the first division so that the advancing margins of the first septum were diverted towards the nearest new pole; the resulting gap was closed later on, and consequently, the cell compartments of this coccus were in communication for some time after two rounds of nuclear segregation. Freeze cleaving showed that the initial generation of septa involved a short sulcus in the plasma membrane and not a circumferential infolding. The shape and form of the developing septum was inferred from sections but was displayed in whole-cell preparations by a technique which selectively and positively stained a septal component. Positive staining of the septum with uranyl salts was appreciable when the relative stainability of the peripheral wall (mainly peptidoglycan) was reduced by pretreatment with salts of low atomic weight metals (0.01-1.0%, w/v) such as cobalt, copper, iron, or zinc. Examination of these whole-cell preparations by stereoscopy showed that the septal diaphragm closes as a slit or long oval, and the advancing margin shows curvature towards the next axis of division. The mechanism and exact site of this positive staining was not elucidated; vancomycin blocking of the uncross-linked peptides of peptidoglycan was almost as effective as the transition metal salts as a foretreatment for staining septa.