Electron microscopic studies of the synaptic junctions have revealed, among other features, that the pre-and postsynaptic processes are separated by an extracellular space of 200-250 A (Palay, 1958; De Robertis, 1958). This gap may be slightly wider than that found separating neuronal and glial processes throughout the central nervous system and may contain a material of intermediate density to that of the pre-and postsynaptic membranes (cf. De Robertis, 1964; and Pappas, 1966). The synaptic cleft substance sometimes appears as a condensed line or band parallel to the pre-and postsynaptic membranes, or as a series of filaments or striations perpendicular to these membranes. It has been suggested that the synaptic cleft substance is made up of mucopolysaccharide material (Pappas & Purpura, 1966). Mucopolysaccharide material in the synaptic gap might be one reason for the strong cohesion between pre-and postsynaptic membranes, as well as play a role in diffusion and dissipation of transmitter substances. Electrical transmission has been demonstrated in the central nervous system of several species of fish (Bennett et al., 1966, a, b, c, d; Bennett & Pappas, 1965; Bennett et al., 1963, 1964). Where this mode of transmission occurs, neurons form distinctive morphological junctions (Pappas et al., 1965). For example, the spinal electromotor neurons of the Mormyrid electric fish are electrotonically coupled (Bennett et a]., 1963). In the light microscope, short thick dendrites connect the cells, and these appear to form cytoplasmic continuity of adjacent neurons (at arrows in FIGURE 1). However, using the higher resolving power of the electron microscope, a definite septum between the two dendritic processes can be demonstrated (FIGURE 2). The junction of the two apposing dendrites is formed by the fusion of the apposing plasma membranes, obliterating the extracellular space at these sites. The overall measurement of the fused membranes is about 140 A (inset, FIGURE 2). At high magnification, the fused membranes can be identified as being similar to the “external compound membrane” described by Robertson originally in the myelin sheath and later at club endings on Mauthner cells (see Robertson, 1966). These neuronal junctions formed by fusion of adjacent plasma membranes are not found in adjacent areas of the spinal cord where there is no special reason to believe that neurons are electrotonically coupled.