AbstractThe widespread recognition of the corpuscular nature of membrane ultrastructure demands re‐evaluation of established concepts of their molecular organization. Many aspects of membrane physiology, composition, and metabolism provide support for the proposal that most membranes consist of two‐dimensional polymers of lipoprotein subunits. Such a model allows the activity, specificity, and adaptability attributed to biological membranes. Evidence which supports this corpuscular model for membranes and some inadequacies of the bimolecular lipid leaflet model are pointed out.The lamellae of plant chloroplasts are membranes which clearly consist of subunits (quantasomes). Their four surfactant lipids and pigments comprise 50% of the lipoprotein subunits. In each of these surfactant lipids there is found a limited and specific group of fatty ester components. This phenomenon suggests that the hydrocarbon chain of the fatty esters may specifically complement certain hydrophobic amino acid sequences in the membrane protein. The protein, then, would determine the sites where the lipid will be most firmly bound. It is proposed that the lipids of membrane subunits are bound by hydrophobic association of the hydrocarbon chains regions within the interior of the protein. The resulting two‐dimensional lipoprotein aggregate would possess the strongly anionic charged groups of the phospholipids on its surface. Metabolically‐driven alterations in conformation of such a flexible lipoprotein ion exchange membrane allows a consistent interpretation of biological membrane transport phenomena.