INTERACTION BETWEEN ADAPTATIONS 18/6 



the normal organism natural selection has been sufficient to 

 co-ordinate them all. 



If now from the principle of natural selection we remove all 

 reference to its cytological details, there remains a process strik- 

 ingly similar in the abstract to that impelled by the principle of 

 ultrastability. Thus natural selection co-ordinates the reflexes 

 by repeated application of the two operations : 



(1) test the organism against the environment ; if harmful 



interactions occur, remove that organism ; 



(2) replace it by new organisms, differing randomly from the 



old. 

 And it is known that in general these rules are sufficient, given 

 time, to achieve the co-ordination. Similarly, the principle of 

 ultrastability leads to the repeated application of the two 

 operations : 



(1) test the organism against the environment ; if a harmful 



interaction occurs, change the values of the step-functions 

 responsible for it ; 



(2) let the new values provide new forms of behaviour, differing 



randomly from the old. 

 The analogy between genes and step-functions is most interesting 

 and could be developed further ; but it must not distract us now 

 The point at issue is : if natural selection's method of action is 

 sufficient to account for the co-ordination between spinal reflexes, 

 may not ultrastability's method of action also be sufficient to 

 account for the co-ordination between cerebral responses, consider- 

 ing that the two processes are abstractly almost identical ? 



It may be objected that the spinal reflexes do not have to fear 

 disorganisation by new learning ; but the objection will not stand. 

 We are comparing the ontogenetic progress of the cerebral re- 

 sponses with the phylogenetic progress of the spinal reflexes. As 

 the species evolves, its environment changes and new reflexes have 

 to be developed to suit the new conditions. Each new reflex, 

 though suitable in itself, may cause difficulties if it compounds 

 badly with the pre-existing reflexes. Thus, a bird that developed 

 a new reflex for pecking at a new type of white, round, edible 

 fungus might be in danger of using the same reflex on its eggs. 

 Evolution has thus often had to face the difficulty that a harmoni- 

 ous set of reflexes will be disorganised if an extra reflex is added. 

 Again, consider the biochemical systems. As most, if not all, 



197 o 



