The field potentials (FPs) and neuronal responses evoked by stimulation of subcortical white matter were recorded in the sensorimotor cortex of intact brain in alert rabbits and in cortical slices. After a few short (500 msec) tetanizations (200 Hz) delivered via the stimulating electrodes, a long-term (several hours) increase of short-latency (1-3 msec) FP amplitude occurred in the brain as well as an increase of the same FP components in cortical slices. The probability of short-latency (up to 1.8 msec) unit discharges increased after the tetanization. Repetition of tetanic stimulation in the same experiment increased and prolonged potentiation in vivo though it had an opposite effect in vitro. The long-term potentiation of neocortical FP seems to be due to the potentiation in monosynaptic pathways.

Synaptic connections of 26 pairs of hippocampal neurons were studied in nonanesthetized rabbits by spike-triggered averaging of intracellularly recorded activity. Synchronized activity was detected in 5 pairs and considered to represent common inputs to the neurons recorded. Hyperpolarizing or depolarizing potentials with 3--4 ms latency were revealed in 3 additional pairs. These potentials are considered to be individual postsynaptic potentials (PSPs) evoked in the target neuron by spikes of the adjacent (0.5--2 mm apart) neuron. A quantum analysis of the individual PSPs was performed. The mean quantum content (0.27--0.65) and quantum size (35--200 mkV) were found to be of the same order as those of the excitatory PSPs previously recorded after intracerebral stimulation. It is concluded that most hippocampal synapses are of low efficacy.

It is suggested that in equations usually used to calculate curves of theoretical distributions of PSP amplitudes in the course of quantum analysis the effect of variations in the quantum values themselves is overestimated and variations in noise level are not taken into account. An equation allowing for both these parameters is suggested. Statistical analysis of "elementary" PSPs of the parietal ganglion ofHelix pomatia and of "minimal" PSPs of hippocampal neurons showed the necessity of allowing for noise variations. By means of the suggested equation theoretical distributions agreeing better with the experimental according to thex2 criterion can be found and variation in the quantum itself can also be determined indirectly.

It is suggested that formulas commonly used to calculate theoretical amplitude distributions of postsynaptic potentials (PSPs) for their further quantum analysis lead to the overestimation of the quantum value variations (sigma nu) and neglect the variations of a noise level (sigma 2n). A proper formula taking into account both parameters in proposed. The necessity to consider sigma 2n is shown by statistical analysis of "individual" and "minimal" PSPs recorded in the neurons of the central ganglion of the snail and hippocampal neurons of the rabbit. Taking into account sigma 2n resulted in a better agreement between theoretical and experimental distributions and provided an indirect method for sigma nu estimation.