Previous electron microscopic examinations of outer segments of photoreceptors suggest that many flattened saccules of cones are continuous with the cell membrane and that their lumina connect with the extracellular compartment but that most saccules in rods appear to lack these connections. The saccules probably contain photolabile pigment, and certain potentials appear to result from dipole formation during pigment bleaching. The detection of dipoles from rod saccules may require that the lumina of rod saccules connect with extracellular space, and questions have been raised whether the interpretation of micrographs is correct or the isolation of rod saccules is the result of artifact. Accordingly, lanthanum and barium precipitates were produced near fixed and unfixed frog photoreceptors. Lanthanum precipitates appeared to infiltrate the saccules of fixed cones and the few surviving cones exposed prior to fixation, but no rod saccules were infiltrated except occasional, most basal saccules or saccules within narrow zones of probable damage. Barium precipitates did not infiltrate saccules of either variety of unfixed photoreceptor, but they did occasionally infiltrate around the saccules at points of damage in rod outer segments. The results thus support the view of the patency of saccules of frog cones and are consistent with, but do not prove, the isolation of saccules of frog rods.