There have been two ways of approaching the theory of elementary particles: One is to classify interactions according to their theoretical behaviours and another is to look into various systematics governing the inter· actions realized in experiments. It is aimed in this paper to combine both ways of our approach to suggest many strata which seem to exist in the structure of elementary particles. §2 is devoted to a theoretical classification of interacions. The strong interactions are briefly discussed in §3. Experi· ments on the weak interactions and their theoretical analysis are reviewed in detail in §4. In §5 are discussed some features of our standpoint which seem to play their essential roles in an extremely high energy phenomena. §1. The Non-Local and Local Interactions The quantum field theory has long been faced to problems of the ultraviolet catastrophy since the very beginning of its history. This difficulty is mainly due to the effects of the so-called self-fields, which appear closely bound up with the problem of the structure of elementary particles. The problem of the ultraviolet catastrophy is still left open, although many attemps to settle this difficulty have been proposed. It is common to all the attempts to introduce a constant of dimension of length, e. g. a coupling constant of non-linear interactions, a radius of a non-lccal domain, etc. We have many reasons for guessing a new constant of dimension of length to appear in the coming theory for the elementary particles. Indeed, such a constant is necessary for explaining the mass spectrum, because it is impossible to form any quantity of the dimension of mass by using only h and c. Since the ultraviolet catastrophies are due to effects of the high energy particles in the self-fields, the present theory should be modified in a high energy region. Thus, it appears necessary to introduce a constant l= (ti,cjE0 ) discriminating the two regions of low and high energies in such a way that the above-mentioned high energy region corresponds to energies higher than E0 • The Compton wave lengths of the particles and coupling constants of the second kind interactions are the only constants of dimension of length in the present formulation of the quantum field theory. A question thus arises, if we can go