IN spite of a remarkable similarity of the general polypeptide chain structure of immunoglobulins, their molecular weights vary from 150,000 to 950,000 because some occur in monomeric (IgG, IgD and IgE) and others in monomeric or polymeric forms (IgA and IgM) (ref. 1). The largest immunoglobulin, IgM, appears in a stellate configuration of five disulphide-linked subunits (IgM8), each consisting of two pairs of covalently linked heavy (μ) and light (κ or λ) chains2. Recent evidence shows that polymeric immunoglobulins, whether composed of dimers, tetramers or pentamers, contain one molecule of an additional polypeptide, termed J chain3-5. Produced within the same cells as immunoglobulins, J chain becomes attached to polymeric IgM and IgA during the final stages of intracellular assembly6-8. Available results indicate that this polypeptide is essential for polymerisation of monomeric immunoglobulins 9,10, has a molecular weight of about 15,000 (refs 11-13), and is covalently attached by disulphide bonds to the Fc region of IgM (refs 14-16). Although the covalent structure of μ chain has been determined17, the exact location of the disulphide bond(s) connecting J chain to μ chain has not been known. To investigate this question, we used available information concerning fragmentation of IgM and of μ and J chains by cleavage with cyanogen bromide (CNBr) (refs 17-20). Because this procedure leaves disulphide bridges intact we reasoned that the portion of μ chain(s) linked by these bridges to J chain would be obtainable as a CNBr fragment. © 1974 Nature Publishing Group.