A vertebrate disk membrane is composed of 40 % lipids and 60 % proteins. Its fracture faces have been classed into the plasmic (PF) and exoplasmic faces (EF), complementary with each other, like those of most other types of cell membranes. The hypothesis assuming the PF particles as representing membrane-associated proteins has been challenged by serious questions if they in fact emerge from the crystalline formation or decoration effects during freezing and shadowing processes. This problem seems to be yet unanswered, despite the remarkable case of the purple membrane of Halobacterium, partly because most observations have been made on the replicas from a single face of specimen, and partly because, in the case of photoreceptor membranes, the conformation of a rhodopsin and its relatives remains yet uncertain. The former defect seems to be partially fulfilled with complementary replica methods.
