1. The isolated hemisected frog spinal cord has been used to examine the effects of changes in ionic composition of perfusing medium on the intracellularly recorded e.p.s.p.s produced in single motoneurones by direct stimulation of individual dorsal root fibres through a separate intracellular micro‐electrode. 2. The monosynaptic single‐fibre e.p.s.p.s usually reveal two distinct components, early and late. The early component is resistant to replacement of external Ca2+ by Mn2+ whereas the later is reversibly abolished. It is concluded that the junction between primary afferent fibre and motoneurone provides joint electrical and chemical transmission. 3. The average amplitude of the unitary electrical e.p.s.p. varies at different junctions from 25 to 430 μV, mean 124 ± 17 μV (n = 50). It is relatively stable and changes its amplitude only with changes in the height of presynaptic spike. The individual amplitudes observed in a given cell usually have a normal distribution suggesting that any variability in electrical response is entirely due to noise. 4. The average amplitude of the single‐fibre chemically mediated e.p.s.p. varies from less than 20 μV to 1·7 mV, mean 222 ± 33 μV (n = 71). A positive correlation was found between the amplitudes of chemical and electrical e.p.s.p.s recorded from different motoneurones. 5. Chemical e.p.s.p.s evoked by consecutive impulses in a single presynaptic fibre show statistical fluctuations in amplitude. The fluctuations occur in quantal steps in a manner described by binomial or Poisson statistics. Only in a few cases the deviation from stochastic distribution can be attributed to variable invasion of nerve impulses into the terminal region. 6. The amplitude of the unit e.p.s.p. varies between 33 and 104 μV, mean 66·4 ± 4·3 μV (n = 19). The mean quantum content (m) varies from less than 1 to more than 10. The m is reversibly reduced by Ca2+ lack and by Mn2+ until the chemically mediated response fluctuates according to a Poisson distribution with the unit e.p.s.p. equivalent to the single quantum of transmitter. 7. When paired or repetitive stimuli are applied to the presynaptic fibre the facilitation of the chemically mediated unitary e.p.s.p. can be observed as well as the post‐tetanic potentiation. The amplitude of the electrical e.p.s.p. remains unchanged during the period of post‐tetanic potentiation, suggesting that the latter is attributable to some change restricted to the specific presynaptic mechanism responsible for the transmitter release but not to changes in presynaptic spike height.

SUMMARY1. Monosynaptically connected primary afferent fibres and motoneurones of the isolated spinal cord of the frog were injected with horseradish peroxidase (HRP). Six labelled afferent fibre-motoneurone pairs were reconstructed-and. subjected to detailed analysis.2. Frog motoneurones possess eight to twelve dendritic arrays displaying some dorso-ventral asymmetry. Dorsal dendrites exhibit a rostro-caudal extent of 1-7-2-6 mm (average 2-2 mm).3. Primary afferent fibres bifurcate in the dorsal funiculus. First-order collaterals emanate from the main ascending and descending branches, at an average distance of 407 ,sm. The average number of boutons per collateral is 670.4. To reach a contacting bouton the presynaptic spike must pass on average five bifurcations and then zero to twelve boutons en passant, attached to a single terminal collateral branch. The structural equivalent of the axon cylinder of the collateral tree roughly preserves cross-sectional area. The branch power ranged between 1-15 and 3*35 (average 2 06). Primary afferent fibres usually form clusters of contacting boutons (contact regions).Connexions between an afferent fibre and a motoneurone comprise from five to twenty-three contact regions (average 12-5). Each contact region contains one to twelve contacting boutons (average 3 3).6. In two of three experiments contacting boutons were found to be significantly larger than non-contacting boutons. The average diameter of the former was 2-6 ,um (range 1-2-4-0).7. In five out of six cases more than one collateral belonging to the same fibre participated in the connexion with a given motoneurone.8. The average number of contacting boutons per motoneurone and collateral is 19-1. It was estimated that each collateral could supply not more than thirty-five motoneurones. This would be less than 8-5 % ofthe motoneurones with their dendrites which cross the termination space of a single collateral. The average number of contacting boutons forming one primary motoneurone connexion was 41-5 (range 21-72).

SUMMARY1. The sensory-motor synaptic connexions in the frog lumbar cord have been used to examine the relationship between the statistical characteristics of the unitary excitatory post-synaptic potential (e.p.s.p.) and the number and organization of synaptic contacts determined when the primary afferent fibre used in evoking the e.p.s.p., and a motoneurone in which it was recorded, were both labelled with horseradish peroxidase (HRP).2. A significant correlation is found between the number of contacting boutons and the amplitude of the chemical component of the unitary e.p.s.p.s generated at the same connexions.3. The amplitude fluctuation patterns of the single-fibre e.p.s.p.s could be fitted by both Poisson and binomial distribution. The number of presumed Poisson release sites as estimated from the ratio Vmax/v (where Vmax is the maximal amplitude of the chemical component of e.p.s.p. and v is quantal size) is always less than or equal to the total number of boutons observed histologically. In three connexions there was a close correspondence between the number of binomial release units, n, and the number of contact regions formed by the tight clusters of contacting boutons.4. The unit potential amplitude estimated from the Poisson distribution is found to be two to three times smaller than the quantal size calculated from binomial distribution. A similar numerical relationship was found between the number of contacting boutons and the number of contact regions. It is suggested that at a single bouton, transmission results in release of a single quantum of transmitter, whereas the binomial quantum probably reflects the multi-quantal release occurring simultaneously at boutons comprising a contact region.5. A significant correlation is found between the mean quantum content estimated either from Poisson or binomial distribution and the number of contacting boutons and contact regions respectively, indicating the dependence of quantal release on the magnitude of synaptic surface.6. No correlation is found between the motoneuronal soma diameter and the quantal size, although the former is significantly correlated with the number of contacting boutons.

No abstract

In experiments on the isolated frog spinal cord the relationship between the statistical properties of the unitary EPSP and the number of synaptic contacts was determined when the primary afferent fibre used in evoking the EPSP and the motoneuron in which it was recorded were both stained with HRP. The size of the chemical component of the EPSP corresponds to the number of presynaptic boutons. Less obvious numerical correlation exists between the number of contact zones and the number of binomial units.