ABSTRACT
The renormalization group (RG) does not produce exact solutions of the Onsager type, and its application involves quite drastic approximations. But unlike the closed-form approximations of the classical theory discussed in Chapters 2 and 3, which do not reflect physical behaviour near the critical temperature, Tc, except in a small number of specific cases, the RG endeavours to retain all the essential physical characteristics of the problem near Tc, whilst rejecting those features which are not significant. This requires continuous thought and attention - in Wilson's graphic language (1975a): 'One cannot write a renormalization cook book.' In his Nobel prize lecture, Wilson (1983) describes how hard it was in the initial stages of his work to find approximations which would be computable in practice.
