846 Comparative Animal Physiology 



In contrast to reptiles and birds, mammals have been conditioned in ways 

 which indicate the general site of persistent neural modification, if not its 

 nature. Reflexes which can occur without the cortex are susceptible of con- 

 ditioning, for example, eyelid reactions, pupillary response, psycho-galvanic 

 reflex and knee jerk. However it is not proved that the conditioning per 

 se is not in higher centers. A word association may be formed in man on a 

 single trial, but spinal reflexes (in intact animals, including man) are very 

 difficult to condition.-^^^- ^^^ Ability to learn appears to increase in the fol- 

 lowing order: rats (and some other rodents), cats, dogs, raccoons. Primates 

 excel in problem-solving and use of tools; cortical conditioning in them is 

 rapid and varied, and usually the RC diff^ers in detectable manner from the 

 RUnCS. 



There is no evidence for conditioning of sympathetic ganglia or various 

 sensory nuclei. It has been claimed that in acutely spinal dogs certain spinal 

 reflexes can be conditioned.'^^-- ^^"^ However, in chronic preparations, altera- 

 tions in leg response by conditioning methods do not occur according to a 

 typical learning curve; the response is variable in form and may be either 

 extension or flexion. -^^ In chronically spinal rats crossed extension, responses 

 of the rump to tail stimulation, and some other reflexes are gradually "ex- 

 tinguished" after many stimulations at the rate of one every fifteen seconds; 

 these responses return (disinhibition) after general vigorous excitation."*"^*^ 

 True conditioning of the spinal cord has not been demonstrated, but per- 

 sistent alterations in spinal excitability certainly do occur. 



Subcortical centers of the brain can be conditioned in certain mammals. 

 Decorticate dogs^''** and rabbits,^'''' as well as animals lacking the striate 

 or visual cortex (rabbits,'*^^ rats"^'^' -'*0> have been conditioned to light and 

 to sound. Even some responses acquired prior to decortication can be retained 

 after the operation. The kinds of response that can be conditioned in de- 

 corticate dogs are limited, and are of widespread activity rather than re- 

 stricted response. ^'^^ Numerous experiments after removal of the visual cortex 

 indicate conditioning of generalized responses to light but not detailed visual 

 discrimination. In cats auditory conditioning normally involves the auditory 

 cortex; if this is lacking some "learning" occurs at the level of the inferior 

 colliculus, and lacking this a modified, low sensitivity response occurs at 

 even lower levels.^*^ The dependence on specific cortex decreases in the 

 order: man, monkey, dog, rabbit. Startle responses to sound are made by 

 cats after transection at the level of the superior colliculus. In non-operated 

 rats startle responses to sound and light can be extinguished by repeating 

 stimuli at intervals of many seconds; the extinguished response can be dis- 

 inhibited, and the threshold for startle responses can be raised by simulta- 

 neous shocks to the feet. 



What are the probable loci of persistent modifications in response? In- 

 tervals between stimufi are so great. that sense organ responses cannot have 

 been altered. It is possible to substitute stimulation anywhere on the sensory 

 pathwav (spinal cord, sensory cortex) for the CS. However, no conditioning 

 occurs if the motor nerve or motor cortex is stimulated as the UnCS.-*^** The 

 motor (mechanical) response need not be made during conditioning; the 

 ventral roots to the leg of a dog were cut and the dog conditioned by a 

 sound plus shock to the leg, then after nerve regeneration the muscular 



