ABSTRACT. The laminar dorsal valve of living Neocrania consists of: a primary layer of rhombohedral tablets, composed of granular calcite and commonly forming slats orthogonal to the margin, associated with polysaccharides and a ®brous 60 kDa protein; and a secondary layer of spirally growing (10.4) rhombohedra, doped with the 60 kDa protein and interleaved with membranes of a ®brous 44 kDa protein. The ventral valve consists exclusively of a primary layer with the same composition and basic structure as that of the dorsal valve. Investigation of selected antecedents shows that the chemico-structure of the Neocrania shell has been virtually unchanged since the ®rst appearance of the stock in the Early Ordovician (Arenig). The greatest phylogenetic change affected the ventral valve that varied, even in Ordovician genera, from a ®lm of calcitic blades to a complete succession of primary and secondary laminae. The most profound diagenetic changes occurred before the Late Cretaceous with proteins degrading into peptides that were dispersed, with the loss of less stable amino acids, during laminar recrystallization. Palaeozoic shells suffered further recrystallization but, even after pressure solution, the original laminar fabric was replicated long after it had lost its constraining organic membranes.KEY WORDS: organoclastic shell structures, shell degradation, protein diagenesis.T H E ultrastructure of the organocalcitic shells of cranioid brachiopods was ®rst described by Williams and Wright (1970) and Schumann (1970). The accounts dealt almost exclusively with the mineral components of the shell in relation to an outer organic cover, the periostracum, and an underlying secreting epithelium, the mantle, and its papillose outgrowths, caeca, that permeate the skeletal succession. Since then the organic components of skeletal systems have come under increasing scrutiny, including those described as`intracrystalline' which are considered more likely to withstand the degrading effects of fossilization (Towe 1980). Initially this kind of inclusion seemed unimportant in the preservation of organic residues in cranioid fossils on the assumption that the calcitic laminae of the secondary shell have always grown spirally with their membranous substrates entrapped, and therefore immured, by screw dislocations on a nanometric scale (Williams 1970). However, the discoveries by Weedon and Taylor (1995) and Brown (1998) that rhombohedra, growing by screw dislocation on organocalcitic laminae of bryozoans and Neocrania, could be sectorally pitted by bleach, prompted Williams et al. (1999) to investigate the process of polymer adsorption.Con®rmation that a protein, with a molecular weight of 60 kDa, is adsorbed into rhombohedra along determinable crystallographic planes and sectors, provided a means for interpreting patterns of ultrastructural kinked crevasses, hillocks and cavities (Markov 1995), that could arise through the degradation of intralaminar organic residues. With this prospect in mind, it was decided to induce degrad...