[Criticism of the current theories concerning the formation of the immunological repertory].

Application of the probability calculus to the variational structure of the immunoglobulins leads to demonstrate the impossibility of the random production of the repertory's antibodies. Physical and Cybernetic considerations lead, instead, to demonstrate the necessary existence of aspecific recognition of the antigen, i.e. of the antigen's activity. Application of the theory of stochastic convergence, at last, leads to demonstrate that, if there would be a connection between this aspecific recognition and the antibody's production, the latter would be possible. This connection, as a consequence, must exist.