The question of self-recognition by the host and problems of autoantibodies and their specificity.

Since Ehrlieh and Morgenroth (5) first published their famous horror autotoxicus concept hundreds of observations have been made supporting the idea that the formation of deleterious antibodies against an organism's own antigen is an exceptional event, if possible at all. Investigations on human ABO blood groups have furnished the best. known example of obedience to the biological principle of horror autoloxicus. According to Landsteiner's law (13), every human being produces antibodies against A and/or B antigens whenever they are absent from the red cells--but autoantibodies directed against the antigens present on the individual's own erythrocytcs never occur. This and other obscrv,qtions led many investigators to believe tha t the inabilit 5 of autoantibody formation is a genetically established charaetemstic of every living organism. According to Furuhata (8), the presence or absence of natural blood group antibodies is genetically determined in a fashion similar to the blood group antigens. The observation of Owen (18) on cattle chimeras opened up a new era of experimental approach to the problem of horror autotoxicus. His basic observations were confirmed in certain instances of human chimeras. The biological importance of Owen's observations was recognized by Burnet and Fenner (3), who came to the eonelusion that the horror autotoxicus principle does not depend upon a genetieal mechanism but may be acquired during ontogenesis. This new eoneept was proved to be eorreet by the basic experiments on immune tolerance performed by Medawar, Billingham, and Brent (see recent review paper by Medawar [16]). The faet tha t the fetus elaborates the mysterious ability to distinguish between an organism's own and foreign antigens is no longer questioned. According to the elonal selection theory of Burner (1), elones which would otherwise be capable of producing antibodies are destroyed by eontaet with the eor-