Local Approximations in Renormalizable and Nonrenormalizable Theories. II

Abstract

The investigations of a previous paper are generalized to two‐point matrix elements. A principle is formulated, which yields unique finite Feynman rules in the renormalizable case, i.e., permits a unique separation of counterterms. For nonrenormalizable theories this principle yields uniqueness up to a ``scaling'' parameter. The results are generalized to a large class of Feynman graphs. For this subset of graphs, field‐theoretical principles do not determine this scaling parameter.