The structure of the cell wall of Streptococcus faecalis was studied in thin sections and freeze fractures of whole cells and partially purified wall fractions. Also, the structures of wall preparations treated with hot trichloroacetic acid to remove non-peptidoglycan wall polymers were compared with wall preparations that possess a full complement of accessory polymers. The appearance of the wall varied with the degree of hydration of preparations and physical removal of the cell membrane from the wall before study. Seen in freeze fractures of whole cells, the fully hydrated wall seemed to be a thick, largely amorphic layer. Breaking cells with beads caused the cell membrane to separate from the wall and transforned the wall from a predominantly. amorphic layer to a structure seemingly made up of two rows of "cobblestones" enclosing a central channel of lower density. Dehydration of walls seemingly caused the cobblestones to be transformed into two bands which continued to be separated by a channel. This channel was also observed in isolated wall preparations treated with hot trichloroacetic acid to remove non-peptidoglycan polymers. These observations are consistent with the interpretation that both peptidoglycan and non-peptidoglycan polymers are concentrated at the outer and inner surfaces of cell walls. These observations are discussed in relation to possible models of wall structure and assembly.Walls of gram-positive bacteria are composed of peptidoglycan to which is covalently linked one or more non-peptidoglycan polymers, such as anionic, teichoic, or teichuronic acids, and neutral polysaccharides. Currently, little, is known about the exact topological localization of the various wall polymers. The question of exclusive or enriched domains within walls remains open.Several kinds of evidence have been used to promote the idea that various polymers in walls are enriched in specific wall areas. First is the appearance of walls of a variety of gram-positive species in thin sections. Frequently, two electron-dense bands enclosing a less dense central zone are seen (16). This appearance has led to the idea that the outer and inner dense bands might be enriched in anionic polymers which would bind larger amounts of the heavy metal stains used to improve the contrast of these preparations. Although phosphate groups in the teichoic acid of Streptococcus faecalis have a high affinity for uranyl acetate, this affinity was greatly reduced after exposure of walls to osmium tetroxide (14). Since fixation with osmium almost invariably precedes uranyl acetate staining of thin sections, it seems unlikely that the tribanded appearance is due solely to selective binding of uranyl ions by phosphate groups.However, in support of the idea that anions are concentrated at wall surfaces are studies of sectiots of Bacillus subtilis in which tribanded walls were visualized in sections that had never been exposed to either heavy metal stains or osmium tetroxide (29). The tribanded image observed was interpreted as indicating an ac...