AbstractUsing only amino acids occurring in proteins it should, in principle, be possible to synthetize peptide molecules capable of complexing metal cations. The approach described in this contribution is to build cages of oxygen atoms that could contain A‐cations of specific diameter (in a manner similar to nigericin, nonactin, valinomycin, and other antibiotics) by binding together face‐to‐face and in the correct distance from one another two homodetic cyclic peptide rings. Using a link that may be formed and broken in a reversible manner (for example oxidativeformationon the outside of the cell membrane and reductive cleavage on the inside), one should even be in a position to construct a model for active, energy‐coupled transport of cations across the cell membrane.An example of this type of synthetic products is the bicyclic peptide S, S′‐bis‐cyclo‐glycyl‐L‐hemicystyl‐glycyl‐glycyl‐L‐prolyl. This compound, the first to contain two homodetic peptide rings bound together, was prepared by stepwise synthesis from the C‐terminal of the S‐protected pentapeptide active ester, glycyl‐(S‐diphenylmethyl)‐L‐cysteyl‐glycyl‐glycyl‐L‐proline‐p‐nitro‐phenyl ester, using active esters of protected amino acids, followed by cyclization of this inter‐mediate, cleavage of the diphenylmethyl protecting group, and final oxidation of the thiol group to the disulfide function. EMF.‐Measurements on membranes by Simon and Pioda according to the method of Stefanac & Simon [4] revealed cation specificity in the order: K+ > Na+ > Li+ > Ca++. Optical rotatory dispersion shows that there is a conformational transition during complexation. Circular dichroism points to a 75–90° dihedral angle and right‐handed helicity of the disulfide bond. On this basis the models shown in Fig. 6 were constructed as hypothetical representations of the potassium complexes with cages of eight or of ten oxygen atoms.