4. THE DYNAMICS OF THE EPIGENETIC SYSTEM 



49 



which did not depend upon the use of integrals of the motion and the construc- 

 tion of a statistical mechanics. One possibihty in this direction has been 

 presented in prehminary form by Sugita (1961), who uses a representation of 

 biochemical systems which is inspired by the structure of McCuUoch-Pitts 

 type neural nets (McCulloch and Pitts, 1943) and the formalisms of auto- 

 mata theory. Such a formalism is applicable to a very wide class of complex 

 systems whose "microscopic" structure is not well defined, and there is every 



My. (/?,) My. 



Figure 7. 



Figure 8. 



reason to believe that its application to certain types of biochemical system 

 may be of considerable value and lead to significant results. However, it is 

 clear that before automata theory can be applied to biochemical systems it 

 must be expressed in terms of many-valued logic, or even better, in terms of 

 continuous rather than discontinuous signals. This is because biochemical 

 processes and interactions are essentially continuous ones, unlike the binary 

 nature of a nerve impulse which either occurs or does not occur. When such 

 an extension has been made, and it will certainly not be long in coming for 

 there is intense activity in this field, then the analysis of biochemical processes 

 in terms of such constructs as majority organs and the notions of redundancy 



