398 BRAIN MECHANISMS AND LEARNING 



bipolar electrodes were permanently implanted in the lateral column of 

 the upper thoracic segments of the spinal cord. In order to prevent dis- 

 placements of the spinal cord during movements of the vertebral column, 

 the vertebra to be perforated was tightly fixed to the adjacent vertebrae by 

 means of a vanadium plate. The electrode was made up of two insulated 

 stainless steel wires, the tips of which were bared and i mm. apart. 

 Single electrical square pulses of o.i msec, duration and of intensities 

 which produced only tactile sensations when applied to the experimenter's 

 skin, were delivered to the cat's skin. The cutaneous area whose stimula- 

 tion evoked the potentials recorded by the spinal electrode was always 

 small and very localized. 



Fig. 4 

 Habituatit)!! of spinal potentials (recorded from the lateral column) evoked by 

 mild cutaneous shocks regularly repeated at intervals of lO seconds in a freely 

 moving cat. Whereas the late waves were greatly reduced or abolished with 

 twenty-nine trials, the primary short-latency component remained unaffected. 



Repetition of those tactile stimuli led to a progressive and oscillating 

 decline of the spinal evoked potentials (Fig. 3) which eventually could be 

 completely extinguished. As shown in Fig. 4, in contrast with the stability 

 of short latency components, the late waves of the evoked responses were 

 more susceptible to habituation. Usually the rate of repetition of stimuli 

 was one every 10 seconds, but decrement of response was also observed 

 with longer intervals between successive stimuli. Stimulation at faster 

 rates (once every i or 2 seconds) made the evoked potentials decline more 

 rapidly than at slower rates. As in afferent neuronal habituation in other 

 sensory pathways, great individual variation was observed in the number 

 of stimuli required to extinguish the spinal evoked potentials. That number 

 ranged between dozens and several hundreds. 



