NERVOUS SYSTEM AND BEHAVIOUR 459 



following experiments show. When obstructions of various kinds are 

 placed on the arms of starfish, they are removed by the efforts of the 

 animal. In the brittle-star Ophiura (Ophiodermd) brevispina, various reac- 

 tions are tried successively to get rid of the foreign body, and as each 

 effort fails to remove the irritant it is abandoned and another commenced. 

 Finally, as a last resort, the arm is amputated. When confronted with such 

 situations the animal shows no improvement of performance with repeated 

 trials. Other experiments suggest that starfish possess limited powers of 

 learning. Starfish (Asterias) can right themselves with any arms, but by 

 restraining certain of them it is possible to train animals to use preferen- 

 tially one or two particular arms for turning over. Such an induced habit 

 persists for several days. Again, by applying nocuous stimuli as reinforce- 

 ment, it is possible to train animals to reduce the number of fruitless efforts 

 made in trying to shake off an obstruction (61). 



Learning and Conditioning. Most animals are capable of learning or 

 profiting by experience. In contrast to mere sensory adaptation, learning 

 involves some lasting change in the central nervous system of the animal 

 whereby its behaviour in a given situation becomes dependent on previous 

 experience of the same or a similar situation. True learning has been 

 demonstrated only in animals with central nervous organization, and 

 appears to be an emergent condition of ganglionic arrangement or organi- 

 zation of grey matter. 



Animals differ enormously in their capacity for learning. Differences 

 depend not only on neural complexity, but also on the sensory avenues 

 open to the animal and on its modes of effector responses. Simple levels of 

 learning have been discovered in the more primitive phyla, including flat- 

 worms and annelids. Experiments have been designed so as to bring about 

 a reversal of response to a stimulus after combining with another, more 

 potent, stimulus. In Leptoplana the normal response to illumination is an 

 increase in locomotory rate, whereas mechanical stimulation of the head 

 causes the worm to stop. By touching the worm each time it is illuminated, 

 it has proved possible to train it to remain motionless in the light. The new 

 association is weakened and disappears when the mechanical stimulation 

 is omitted, but learning takes place more rapidly on a second occasion. 



Similar reversals in sign of response have been obtained with poly- 

 chaetes. Neanthes virens normally shows a negative response to illumina- 

 tion and mechanical stimulation. The animal emerges from its tube when 

 mussel juice is presented; when this stimulus is presented at the same time 

 as a photic or tactile stimulus, the response to the latter becomes positive 

 after some eighty trials. Again, the normal negative response of Hydroides 

 dianthus to shadows is withdrawal. In some animals the response is poorly 

 developed but can be strengthened by associating a tactile stimulus 

 with it. 



Marine animals with most highly-developed central nervous systems are 

 decapod crustaceans, cephalopods and fishes, and it is among them that 

 we may expect to find the most complex and varied behaviour patterns. 



