Idiotypy shares with allotypy the characteristic that antigenic specificities proper to it in one category of proteins, as one given class of immunoglobulins, are not carried by the protein molecules of this category in all the individuals of the same animal species. But, in contrast to allotypy which has been observed in several other kinds of proteins after its discovery among immunoglobulins, idiotypy is by definition restricted to antibodies.One of the most two important features of idiotypy--also the most striking difference between idiotypy and immunoglobulin allotypy--is that an idiotypic specificity found in antibodies against one given antigen has been found neither in the serum of the nonimmunized animal, nor in antibodies against another, noncrossreacting antigen. It was this striking difference that first made the authors conscious that what they observed in rabbits was definitely different from conventional allotypy (1).A second important feature of idiotypy, which contrasts also with conventional allotypy, is that each idiotypic spedfieity, carried by antibodies of one given rabbit against Salmonella typki, has not been found yet (at least not exactly under the same form) in antisera of other rabbits against the same antigenic material, x * Aided by a grant (67-00-605) of the Deq~gation G~n&ale ~ la Recherche Scientifique e t Technique (Comit~ de Biologic Mol&-ulaire).x The reluctance of the authors to propose a new term for the new phenomenon before being quite sure that this was necessary made them present their first results as a peculiar kind of allotypy in which each allotypic specificity would be restricted (a) to a single antibody, and (b) to a single individual. Actually, the newly observed phenomenon was fundamentally different from aUotypy, and the use of the same term for both would have been extremely confusing in the necessarily elaborate discussions which involve the three different kinds of antigenic specificities carried by the same immunoglobulin molecule (i.e. isotypic, allotypie, and idiotypic specificities). Consequently, it was found necessary to coin a new word of the same family as those proposed in 1956 (2) and in 1960 (3, 4, 5). In the same way as in "allotypic" or "isotypic," where "typic" (from rwro~, print or type) stands for antigenic specificity, "allo" (from tXko~, other) stands for different, and "iso" (from ~ros, the same) stands for similar, "idiotypic" was proposed (6, 7). The first part of the word (idio, from ~8~o,, peculiar) is justified by the extreme peculiarity of the antigenic specificities in question. An idiotype is a peculiar kind of protein antigen defined by its idiotypic specificity. In this paper, we will 595 on