Form, habit and evolution in the chamidae (bivalvia) with reference to conditions in the rudists (hippuritacea)

The Superfamily Chamacea contains the one Family Chamidae with free and attached species. Species of Echinochama are briefly attached by the right valve. Permanently attached individuals are cemented by one or other valve. Since some species may attach indifferently by either valve, separate genera cannot be erected on the basis of the valve of attachment, but it remains convenient to refer to ‘Chama' and ‘Pseudochama' denoting individuals attached by left and right valves respectively. The tangential component in shell growth causes anterior separation of the umbones and splitting of the ligament, the valves becoming spirally coiled as in the unattached Glossus (Isocardia). Major bilateral asymmetry occurs in ‘Chama' and ‘Pseudochama', the under valves being the larger and deeper component. Since the work of Munier-Chalmas (1882) it has been known that dentition in the Chamidae is either normal (in ‘Chama') or inverse (in ‘Pseudochama'), i.e. that dentition on all attached and on all free valves is similar, conditions in 'Chama' and ‘Pseudochama' being mirror images of one another. Right or left valves of the one cannot be compared with right or left valves of the other. This research is based on study of two intertidal Californian species, Chama pellucida and Pseudochama exogyra, all individuals reported on having, respectively, normal and inverse dentition. The shell is particularly dense—in striking contrast to that of the extinct rudists. The primary ligament is greatly modified by the tangential component which anteriorly causes splitting and posteriorly overgrowth of the calcareous layers of the valves by the posterior outer, and the inner, ligament layers. Unlike an opisthodetic ligament, growth is exclusively in a posterior direction. It follows that growth represents a process of slow rotation on the substrate, in an anticlockwise direction in ‘Chama' clockwise in ‘Pseudochama’. Posteriorly directed growth along the hinge line causes longitudinal extension of the two cardinal teeth present in each valve. Difficult to homologize, these teeth are designated a1 and a2 on the attached, and f1 and f2 on the free, valve. The first, and more ventrally situated, teeth which originate more anteriorly, are much the larger. Small posterior lateral teeth occur in each valve. The occurrence of inversion is undoubted and is fully illustrated. The mantle on the under valve is the more extensive with umbonal regions curved anteriorly under the hinge plate. The mantle margins are united, by fusion of the inner folds exclusively, (Type A of Yonge 1957a) except for the pedal gape, through which the small foot occasionally protrudes, and inhalant and exhalant apertures. The latter are separately extended on short siphonal tubes. The adductors are large, in correlation with the habitat (exposure to the air and to violent water movements), and extended dorso-ventrally, the mouth being carried dorsally and the anus ventrally and the visceral mass extended dorso-ventrally. The valves separate only slightly; pseudofaeces with much mucus are readily extruded by contraction of the ‘quick’ region of the adductors. The foot must assist in cleansing. The inner demibranchs of the dorso-ventrally extended ctenidia are much the larger. Elaborately plicate, the ctenidia furnish the maximum of current-producing and straining surface; consequent danger of blockage is countered by combined ciliary and muscular means. The pattern of ciliation is Type C (1) of Atkins (1937). The marginal groove along the inner demibranch is exceptionally deep and narrow and guarded by tufts of long cilia, only the finest particles enter. The small palps are asymmetrical with the longer proximal oral groove on the under side. Highly selective, they reject material from the anterior, instead of the usual posterior, end. Pseudofaeces collect at the base of the inhalant siphon. The gut is very short, and, as in all Bivalvia, the stomach is highly asymmetrical; this is not influenced by inversion (e.g. it is the same on the right side in both ‘Chama’ and ‘Pseudochama' Inversion is thus revealed as affecting only pallial structures, i.e. dentition and the form of the mantle and valves. There are no unusual features of an internal anatomy already adequately described by earlier workers. Statocysts are retained. Sexes are separate and gonads penetrate the mantle. Nothing is known about development, although attachment is probably at the end of the dissoconch stage. Bilateral asymmetry affects primarily the mantle/shell including the pallial organs (ctenidia and palps); a great part of the food collecting surface may lie in the larger mantle cavity on the under side. There are secondary effects on the visceral mass and on the pedal retractors. Cementation in the Bivalvia is shown to occur at four stages in the life history in different groups. In the Anisomyaria the valve of attachment is predetermined by previous byssal fixation but not in the Chamidae. It was presumably always preceded by change from a vertical to a horizontal disposition, another probable preconditioning factor being absence of united siphons (and usually of teeth). The shell became inequivalve with the body bilaterally asymmetrical and rounded. The problem of growth when attached by one valve, usually solved by ventralward growth involving continual increase to the under side of the ligament, is met by rotation in the Chamidae. Mechanisms are developed for efficient cleansing of the horizontally fixed mantle cavity. Form in the Chamidae is influenced by (1) the tangential component acting in the longitudinal vertical plane with the consequent separation and coiling of the umbonal regions modifying the transverse component dorsally, and (2) cementation causing bilateral asymmetry and so further modification of the transverse component. The effects of these on the median and antero-posterior axes of the body and the hinge and demarcation lines of the mantle/shell are discussed in relation to the longitudinal and transverse vertical planes. The pattern of tooth and socket formation on either side of the crest of pallial epithelium ventral to the mantle isthmus,' normally constant in relation to right and left valves, is related in the Chamidae to attached and free valves, i.e. is associated with cementation. This problem does not arise in edentulous cemented bivalves. Articulation between an upper opercular and a lower deeply concave valve involves a rolling back of the hinge plate on the former and corresponding over-arching of the latter. The hinge plate is always bounded ventrally by a tooth on the attached and by a socket on the free valve and this appears to be functionally necessary. Inversion would seem to be a consequence of cementation and to be the concern only of the mantle/shell. Form in the extinct rudists (Hippuritacea) was basically similar to that in the Chamidae. Although attached by either valve (only the Diceratidae indifferently by either) inversion here appears to have been a phylogenetic not an ontological process. Comparisons are made between conditions in free Glossus, normal and inverse Chamidae and representatives of the rudist families Diceratidae, Requieniidae, Capulinidae, Hippuritidae and Radiolitidae. These exhibit full effects of the tangential component with often striking bilateral asymmetry, the umbones separated to maximum extent and the ligament consequently moved at right angles from the longitudinal to the transverse axis. The direction of the teeth is similarly changed, e.g. in the Hippuritidae and Radiolitidae those on the opercular upper valve become laterally (i.e. vertically) extended to fit into deep sockets on the cone-shaped under valve. Enormous shells of a unique cavernous character and consequently very light were rapidly secreted to raise these bivalves high above the rocky substrate; the rudists were the supremely adapted epifaunistic bivalves. Functional loss of the ligament must have preceded its final disappearance; the foot may then have raised the upper valve. It is suggested that the mantle lobes were united, as in the Chamidae, by the inner mantle margins with siphonal tubes at least as short. Wide separation of the umbones would carry with them the anterior ends of the ctenidia with development of long proximal oral grooves; in cases of great asymmetry much the greater current producing and straining surface would be in the under mantle cavity. The general disposition of the body is deduced from the known association of the mouth with the anterior, and the anus with the posterior, adductor. Pillars and oscules in the Hippuritidae could not have been ‘siphona'. The inhalant and exhalant openings were raised high by shell secretion. The pillars (and so the oscules) are here regarded as a consequence of the formation of pallial folds bounding a waste canal which, as in modern bivalves, would discharge through the inhalant opening. This would also explain the great antero-posterior asymmetry in the Hippuritidae. The greater symmetry of the Radiolitidae is suggested as due to the ventral displacement of the inhalant aperture with the waste canal bounded by pseudopillars. Cleansing of the deep under mantle cavity must have been of supreme importance. The majority of rudists must have occupied very similar shallow and intertidal rocky areas as do the cemented Chamidae although often raised higher and becoming gregarious to form reefs. Especially in the Radiolitidae large species lost attachment to lie recumbent in deeper water with the posterior (siphonal) surface uppermost. The increase in numbers, size and elaboration of form throughout the Mesozoic indicates long continued stable conditions. Natural selection must then have favoured variations in these laterally extended rudists which raised them higher above the rocky substrate. The final rudist forms, Caprinidae, Hippuritidae and