TY - JOUR
AU - THOROGOOD, KEIRA E.
AB - Abstract The gland of Leiblein of the muricid Nucella lapillus and the nassariid Hinia reticulata has been examined by scanning and transmission electron microscopy. The origin and functional significance of its complex organization and its relationship with the rest of the mid-oesophagus in Nucella are discussed. It is absorptive as well as secretory, and a mechanism is proposed by which solute-rich fluids may enter the gland. Its epithelium is composed of occasional mucous cells and two major cell types: ciliated cells engaged in protein metabolism and unciliated cells responsible for uptake and storage of lipids and carbohydrates, both of which show evidence of pinocytotic uptake of solutes and intracellular digestion in lysosomes. Some enzyme activity persists in the residual bodies they shed by apocrine secretion, but they remain intact in a mucous string until they reach the stomach. Preliminary ultrastructural examination indicates that the gland absorbs cadmium not only from the blood but also directly from its lumen and that it may have the capacity to sequester a wide range of toxins. The same types of cell occur in Hinia in which their cyclical activity has been correlated with feeding. Similar cells have been identified in the oesophageal glands of other prosobranchs. The foregut glands of carnivorous caenogastropods are compared with the gland of Leiblein. There is an inverse correlation between the role of the mid-oesophagus in digestion and absorption and the complexity of the stomach. INTRODUCTION Despite the many variations in the mid-oesophagus of different groups of prosobranchs, two of its most constant features—its twisting as a result of torsion, and its possession of lateral glandular pouches that secrete digestive enzymes—have left their indelible mark even on those most highly modified. In the muricoidean and buccinoidean neogastropods the mid-oesophageal glands of the ventrolateral walls are separated from the main tract and concentrated in a discrete ‘gland of Leiblein’ appended to the posterior end of the mid-oesophagus in the cephalic haemocoel. The gland is described by Taylor & Morris (1988) and Strong (2003) as a synapomorphy of neogastropods. Its development has been correlated with that of a pleurembolic proboscis, requiring that the mid-oesophagus is narrow enough to pass through the nerve ring when the proboscis is extended (Fretter & Graham, 1994). It is highly developed in muricids such as Nucella lapillus (L.), which commonly feed on barnacles or by boring the shells of prey such as Mytilus edulis L. to reach the soft parts (reviewed by Crothers, 1985; Hughes, 1986; Fretter & Graham, 1994). In nassariids such as the carrion-feeding Hinia reticulata (L.) it is relatively small and tubular, with poorer development of its glandular tissue, accompanied by a correspondingly greater development of the digestive gland (Graham, 1941). In Nucella the gland fills the cephalic haemocoel dorsal and posterior to the rest of the mid-oesophagus (Fig. 1), salivary glands and nerve ring. It lies to the left of the retracted proboscis and anterior oesophagus, which is folded into a U-shape when the proboscis is not extended. It cradles and lies mainly to the right of the posterior oesophagus, which is long by comparison with that of a neotaenioglossan caenogastropod such as Littorina littorea (L.). The gland is divisible into a broad thick-walled anterior lobe, a narrower twisted middle lobe, and a thin-walled tapering posterior lobe. It is delicate and highly vascularized, and the walls of all three lobes bear vascular folds covered by a glandular epithelium (Figs 2, 4, 11). Amaudrut (1898) was the first to recognize that the gland of Leiblein is the homologue of the mid-oesophageal glands of lower prosobranchs ‘stripped away’ (in an evolutionary sense) from the main part, and connected to it only by a short duct. An embryological study of Nucella by Ball (1994) has since shown that it arises as a diverticulum of the mid-oesophagus that becomes partly reflected over the mid-, partly over the posterior oesophagus. But it was Graham (1941) who recognized that the thin undifferentiated strip of epithelium lining a groove (a narrow cleft between the dorsal folds, Fig. 1; vg) in the mid-oesophageal wall of Nucella marks the failure of the glands to develop in their typical position. It is recognizable on the outer surface of the oesophagus as a hernia-like ‘scar’ (or mark). This strip is less conspicuous in buccinds than in muricids, and in Hinia and other nassariids it is almost indistinguishable. In muricids the enlargement of the food channel and concomitant ventral extension of the dorsal folds is already evident in what Graham (1941) took to be the anterior oesophagus (see Fretter & Graham, 1994, for discussion). The mid-oesophagus is sub-divided into the valve of Leiblein (another autapomorphic character of neogastropods, Taylor & Morris, 1988), anterior to the nerve ring (Fig. 1; vL), a narrow section encircled by the ganglia, and a section posterior to it rich in mucous glands on the hypertrophied dorsal folds (the pre-torsional left larger than the right). This is the ‘glande framboisée’ of Amaudrut (1898). The valve prevents regurgitation from the mid-oesophagus during extension of the proboscis (Graham, 1941), and is the site of torsion in muricids. The dorsal folds, already in a ventrolateral position at its anterior end twist round the right wall of the valve to a mid-dorsal position at its posterior end, and they come closer together so that the ventral groove lined by the undifferentiated strip separating them becomes even narrower. The valve is smaller or absent in buccinids and nassariids, in which torsion is posterior to the nerve ring. In Nucella the glande framboisée extends on either side of the undifferentiated strip as far as the distal end of the duct to the gland of Leiblein (d). It does not occur in buccinids and nassariids. The factors determining the complexity of the organization of the gland in muricids have not been addressed. There also remains confusion over interpretation of its histology. Graham identified one major cell type throughout the gland, possibly showing different phases of activity but, according to Amaudrut (1898), the anterior end is secretory and the posterior ‘poor in glandular elements’. Conversely, Martoja (1971) described the anterior end as a region of storage and the middle and posterior parts as secretory. These contradictions suggest that there may be cyclical changes in the cells correlated with feeding and digestion such as those described by Martoja (1964) in the carrion feeder Hinia reticulata. The role of the gland of Leiblein in the secretion of digestive enzymes was demonstrated by Mansour-Bek (1934) in Murex anguliferus L., in which she identified four proteolytic enzymes. Hirsch (1915) had earlier found a diastase in that of Murex trunculus L. In addition, reports of neutral and acidic lipids and glycogen in the gland of Nucella (Martoja, 1971), and of phospholipids, sulpholipids and glycolipids in Murex (Bolognani Fantin, Ottaviani & Bolognani, 1977), indicate a storage function. In 1986 Minniti identified glycoproteins in the much smaller gland of Leiblein of the nassariids Amyclina tinei (Maravigna) and Cyclope neritea (L). Recently the experimental work of Leung & Furness (1999) showed that in Nucella the gland has the highest concentration of cadmium (Cd) in the body of snails exposed to this heavy metal. This has revived interest in its functioning and possible use in monitoring heavy metal pollution. In this study we explore the fine structure of the gland of Leiblein in relation to function in Nucella lapillus, which requires some consideration of other parts of the mid-oesophagus, and briefly report preliminary observations and X-ray microanalysis on material exposed to cadmium, provided by Dr K.M.Y. Leung. Dr V.K. Dimitriadis carried out the cytochemical tests as part of a parallel study with E.B.A. on the cytochemistry of the digestive gland of Nucella (Dimitriadis & Andrews, 2000). Nucella did not lend itself well to feeding experiments in the laboratory, so Martoja's experiments on the voracious Hinia were repeated using electron microscopy to examine samples of tissue. MATERIAL AND METHODS Specimens of Nucella were obtained from the University Marine Biological Station, Millport. Some were fixed on arrival (within 25 h of collection) and the remainder were placed in an aquarium with circulating seawater at 10°C. One group was denied access to food and a second batch was kept with Mytilus in the same tank. There was evidence from empty shells with boreholes that at least some of these snails fed, but none was seen feeding. Specimens from both groups were prepared for electron microscopy. Scanning electron microscopy Some specimens were fixed without narcotizing while others were anaesthetized in equal parts of 7.5% magnesium chloride and seawater. In some snails of the latter group the proboscis was extended and pinned in this position before fixation, in the remainder it was left in the retracted state. After removal of the shell, the contents of the cephalic haemocoel were either fixed in situ, or the gland of Leiblein was dissected in fixative and placed in fresh fixative for at least a further hour. The specimens were fixed at room temperature in 3% glutaraldehyde in Sörensen's phosphate buffer at pH 7.2, adjusted to 1100 mOsm with 14% sucrose. The glands were either kept intact or left in fixative overnight then cut into transverse, horizontal or sagittal sections using a single-edged razor blade. They were rinsed in buffer, dehydrated in ethanol and critical-point dried before sputter-coating with gold-palladium and examined in Hitachi H2400, 3100 or 3000 scanning electron microscopes. Transmission electron microscopy Small pieces of tissue were removed from glands immersed in fixative. The primary fixative described above was normally used, followed by post-fixation in 1% aqueous osmium tetroxide. The delicate tissue is difficult to fix well, and variations based on methods described by Glauert & Lewis (1998) were also used. The following modifications proved to be the most valuable for particular features: Tissue was dehydrated in ethanol and embedded either in Spurr's resin, or TAAB resin following immersion in propylene oxide and infiltration in a 50:50 mixture of this with the resin. Areas of interest were identified in sections cut at 0.5 µm and stained in aqueous Toluidine blue. Sections 50–100 nm were stained in aqueous uranyl acetate and Reynolds' lead citrate and examined in a Zeiss E.M.109 or Hitachi H600 transmission electron microscope. Primary fixation in 5% glutaraldehyde in 0.1 M sodium cacodylate buffer, pH 7.2, with the addition of 3% (w/v) sodium chloride followed by rinsing in the same buffer before post-fixation (Owen & McCrae, 1976). The use of an electrolyte to adjust osmolarity, with the resultant decrease in the solubility of proteins increased the density of the tissues and the intensity of staining. Primary fixation in a modification of Karnovsky's fixative (1965): 2% formaldehyde+4% glutaraldehyde in Sorënsen's phosphate buffer, pH 7.2 with the addition of 14% sucrose, and rinsing in the same buffer before post-fixation. Rapid penetration of the formaldehyde ensured excellent fixation and was particularly effective in preserving glycogen. Routine primary fixation and buffer rinse followed by post-fixation in 1% osmium tetroxide to which 1.5% aqueous potassium ferrocyanide (w/v) was added. This enhanced the contrast of cell membranes and glycocalyx. Nuclear staining was poor. Cytochemistry Tests for sulphated and carboxylated carbohydrates: Thin slices of tissue were fixed using the routine primary fixative described above. They were then treated using the methods of Spicer, Hardin & Setser (1978) and Sannes, Spicer & Katsuyama (1979) to demonstrate sulphated carbohydrates. Some pieces were incubated overnight in high iron diamine (HID), whilst others were exposed to low iron diamine (LID) (Takagi et al., 1982). They were then post-fixed in 1% aqueous osmium tetroxide and embedded in Spurr's resin after dehydration in ethanol. Control specimens were incubated in 1 M MgCl2 in place of HID or LID. Post-fixation was omitted for tissue used for the periodate-thiocarbohydrazide-silver proteinate staining (TCH-SP) method of Thiéry (1967) to demonstrate periodate-reactive carbohydrates. Sections were stained in a medium containing 2% thiocarbohydrazine in 20% acetic acid. Control sections were stained without exposure to periodate. Test for acid phosphatase: Specimens were fixed in the standard primary fixative, but were not post-fixed. A modification by Lewis & Knight (1992) of the Barka & Anderson (1962) test was applied. Sections were incubated in a medium containing 0.1 M β-glycerophosphate as substrate in 0.2 M Tris-maleate buffer for 15–30 min. The substrate was omitted, or 0.01 M sodium fluoride was substituted for it in controls. Exposure to cadmium Adult females of Nucella, average size 31.3±1.4 mm in shell length were collected by K.M.Y. Leung at Sandgerdi, Iceland, and acclimated in seawater, salinity 33% at 10°C for 1 week. They were then caged and submerged in seawater in 1000-ml beakers for a total of 53 days, during which time they were starved. The control group was exposed to <0.01 µg Cd.l–1, the experimental to 500 µg Cd.l–1. Samples of six and five snails from the control and experimental groups, respectively, were prepared for TEM. The gland of Leiblein was cut into three parts after initial fixation in the routine primary fixative for 20 min, then transferred to buffer after a further 2 h. The blocks were cut into small pieces before post-fixation in aqueous 1% osmium tetroxide. They were dehydrated in ethanol and embedded in TAAB resin. Sections were stained with lead citrate only. Feeding experiment on Hinia Specimens were obtained from Millport and kept in the same conditions as those described for Nucella. Two control snails, known not to have fed for at least 5 days, were set aside for immediate dissection on their arrival and the rest were numbered with a Tip-Ex pen and placed in an aquarium until they became active. A fresh cod's head was placed in the tank and the snails rapidly began to feed. Each was allowed to feed for 50 min and its number recorded before it was placed in another tank without food. The gland of Leiblein from each of two snails was fixed at intervals of 1, 5 and 23 h, and 10 and 20 days, respectively, after feeding. Small pieces were fixed in 3% glutaraldehyde in Sörensen's phosphate buffer containing 14% sucrose. After post-fixation in 1% osmium tetroxide in the same buffer they were dehydrated in ethanol and propylene oxide before embedding in TAAB resin. Sections were prepared and examined as described above. RESULTS Scanning electron microscopy of the gland of Leiblein of Nucella lapillus In Nucella of shell height 30 mm the maximum dimensions of the gland are 5.5 mm long by 4.5 mm wide at the anterior end tapering to 1.5 mm posteriorly (Figs 2, 3, 4). Its anterior lobe forms a hood over the two pairs of salivary glands that surround the valve of Leiblein (Fig. 8), and it overhangs the nerve ring and narrowest part of the mid-oesophagus in the floor of the haemocoel immediately behind the valve. The gland is intimately associated with the glande framboisée and posterior oesophagus, which are bound to it by muscle fibres and connective tissue. Comparison of glands from snails fixed with or without prior relaxation when the proboscis is retracted shows that the shape changes considerably when the snail is withdrawn into the mantle cavity. In snails narcotized in an extended state before fixation the gland is not compressed, the blood vessels are distended and the lumen is capacious (Fig. 5A, B). In those retracted into the mantle cavity it assumes an S-shape viewed from the right side (Fig. 5C). The middle lobe is bent downwards and compressed vertically between the other two lobes. Its floor is also thrown into a series of pockets, more prominent on the left than the right. The posterior lobe is bent upwards at right angles to it. The whole gland pushes the floor of the mantle cavity upwards and presses against its posterior wall (the anterior wall of the kidney). The tip of the posterior lobe (Fig. 3), the ampulla referred to by Strong (2003) thus invades the afferent renal vein. The lumen is almost obliterated and the blood vessels are collapsed (Fig. 5E). The relationship of the gland with other parts of the foregut does not change significantly when the snail is feeding and the proboscis is extended. In dorsal view two fissures delimit the three regions of the gland, produced by the twisting of the anterior aorta around it (Figs 3, 4; aa). The more anterior (af) indents the left wall vertically between the anterior and middle lobes. The fissure accommodates the glande framboisée, which arises ventrally (Fig. 1C at curved arrows) behind the narrow part of the mid-oesophagus, which is encircled by the nerve ring, and extends dorsally, where it gives off the duct (d) to the gland. This marks the junction with the posterior oesophagus. The origin of the afferent artery (afa) from the anterior aorta is also ventral to the gland (Fig. 6A, B). The posterior fissure (Fig. 6B; pf) is an oblique depression in the dorsal wall of the gland that demarcates the middle from the posterior lobe. It embraces the anterior aorta, which crosses over the gland diagonally from the posterodorsal left to the anteroventral right side (Figs 3, 6, 7; aa) as it does in the mid-oesophagus of lower prosobranchs. The aorta then changes course and turns left, ventral to the gland of Leiblein, where it gives rise to the afferent artery of the gland (Fig. 6A; afa). This major branch of the anterior aorta arches over the duct as it enters the anterior lobe (Figs 3, 4). The left side of the anterior lobe of the gland is hypertrophied, so it curves to the right, and the afferent artery describes an arc to the right reflecting this. Its main branches run along the crests of the densely packed glandular folds and subdivide into a reticulum of subepithelial blood spaces over their sides. It turns to the left and ventrally as it enters the middle lobe and runs along its right wall before forming a plexus on the dorsal wall of the straight posterior lobe (Figs 5–7), which partly overlies the middle lobe. The middle lobe, around which the anterior aorta twists, curves abruptly to the left and downwards like a spiral staircase (Fig. 4). The vascular folds become progressively weaker and less extensively branched towards the posterior end. The efferent route of the blood is through peripheral channels in the troughs between the folds. They open to the cephalic haemocoel and form a reticular pattern on the surface of the gland (Figs 2, 10; o). The middle and posterior lobes of the gland bear a ciliated double ventral fold (Fig. 3C), resembling that which occurs in the oesophagus of some primitive gastropods (Haszprunar, 1988). It is responsible for carrying the mucous string containing the secretions of the gland along the duct and posterior oesophagus as far as the stomach (Figs 4, 5B, D; vf). It is absent in the main part of the mid-oesophagus, nor does it develop in the anterior lobe of the gland, which is almost entirely composed of the left wall. At the junction of the anterior and middle lobes the fold emerges from the depths of the middle lobe and turns left into a strongly ciliated medially facing vestibule at the opening of the duct (Fig. 5C inset, E). The epithelium over the fold is composed of many mucous cells and non-glandular strongly ciliated cells similar to those lining the vestibule around the opening of the duct. The cilia on its marginal folds beat into the median groove (Figs 3C, 5D), in which the direction of the ciliary beat is outward. Hitherto the ventral fold has been the only identifiable route by which the lumen of the gland communicates with the oesophagus, and since its ciliary currents are directed outwards to the posterior oesophagus there seemed to be no possibility that any oesophageal contents could enter the gland. Since the specimens examined in this study have undergone far less shrinkage than histological material during preparation it has been possible to show that the contiguity of the dorsal folds subdivides the duct of the gland into two. The larger, ciliated efferent channel on the posterior face of the duct is separated from a narrow unciliated physiologically closed tube on its anterior face (Figs 3, 8, 9; si). It then becomes apparent that the tube extends the whole length of the mid-oesophagus, from the valve to the gland of Leiblein, underlying the ‘scar’ beneath the dorsal folds (Figs 1, 9). It acts as a siphon through which liquid and solutes, but not particles, may enter the gland. At the opening of the duct to the gland the still-swollen ends of the dorsal folds keep the openings of these two conduits separate. The ciliary currents in the gland ensure that incoming liquid is directed into the anterior lobe away from the outgoing mucous string from the middle lobe (Figs 3C, 9). The string, with the residual bodies it contains, remains intact until it reaches the stomach. One function of the valve of Leiblein appears to be to act as a sump for liquid entering it from the oesophagus in both directions. At the junction of the valve and anterior oesophagus the dorsal folds with the cleft (the vestige of the ventral wall) between them dip down into the floor of the inflated anterior end of the valve. The permanently open gutter between them, uninterrupted by the Ω-shaped fold (Fig. 1; cf), allows any liquid from the anterior oesophagus to drain into it. The food string is contained within the funnel-shaped fold and the fringe of long cilia wraps it in a cocoon of glycoprotein secreted by the mucous pad. The glandular pads in the lateral walls of the valve compress and consolidate the string before it enters the narrow part of the mid-oesophagus, which passes through the nerve ring. Behind this the oesophagus describes a right angle at the base of the glande framboisée, so it lies in a dorso-ventral plane, tied to the left side of the gland of Leiblein at the level of the anterior fissure by muscles and connective tissue (Figs 2, 6A, 9). The walls of both the oesophagus and duct have a relatively modest layer of mainly circular sub-epithelial muscles overlying a stout basal lamina. The muscles are particularly thin beneath the siphon, explaining its appearance as a ‘scar’, but Graham (1941) noted that a large bundle of longitudinal muscles runs along the whole length of each dorsal fold. These muscles, which end where the duct joins the gland, help the juxtaposed dorsal folds to form an effective seal, transforming the morphologically ventral groove in the valve of Leiblein into the siphon separate from the main mid-oesophageal channel. As the dorsal folds come closer together and curve (due to torsion) to a dorsal position at the posterior end of the valve the groove, 0.25 mm wide at its anterior end, narrows into the closed siphon (Figs 1, 8, 9). It is 0.1 mm wide when collapsed. A longitudinal band of extrinsic muscles in the haemocoel overlies the siphon from the valve of Leiblein to the base of the glande framboisée (Fig. 8; em). Another extends from the base of the glande framboisée to the duct (Fig. 2; em). Muscle and connective tissue strands are also prominent on the anterior face of the duct. They are particularly well developed at its junction with the oesophagus. The siphon collapses when the muscles are relaxed (Fig. 9). The capsule of the gland is composed of a thin layer of collagen and a scant reticulum of muscle fibres with a tracery of nerve fibres. The vascular folds (=the septa of other prosobranchs; Figs 10–14) have well developed muscle fibres (Figs 10H, 11B, 14C; mf) radial to the surface of the gland and anchored in the superficial connective tissue. The basal lamina is thin and connective tissue sparse, creating little barrier between the blood and overlying epithelium. Circular muscle fibres around the main afferent arteries along the crests of the folds maintain blood flow, and ciliary activity enhances exchange of solutes across the epithelium by preventing the formation of unstirred layers over the apical cell membranes. The currents over the sides of the folds are directed towards the crests, over which ciliated cells are more prominent, and the direction of their ciliary beat, parallel to the long axes of the folds, is clear in fixed specimens. The curvature of the folds over the roof of the anterior lobe has the effect of creating a circulation away from the opening of the siphon and directing the flow of contents posteriorly. The current travels over the main afferent artery past the vestibule on the left into the middle and posterior lobes. The more regular transverse arrangement of folds in these lobes directs material onto the ventral fold (Fig. 4), and thence to the efferent channel of the duct. The epithelium over the folds is composed of two main interdigitating types of cell, one unciliated, the other ciliated (Fig. 10), among which are hidden the openings of occasional mucous cells. The appearance of the luminal surface of the epithelium varies in different specimens, depending on the phase in the cell cycle (interpretation supported by experiments on Hinia, see below). In snails that have not recently fed, the unciliated cells are club-shaped and largely obscure the ciliated cells. Their balloon-like apices bear few or no microvilli (Figs 10A, 15B) in contrast to the dense array on the ciliated cells. The latter bear several apical blebs among which there are a few long straggling cilia with bulbous bases (Fig. 10E, F). In other specimens the apices of both types of cell, devoid of blebs, are clearly visible (Figs 10, 15A), and the short microvilli on the unciliated cells contrast with the long microvilli on the ciliated cells. Some intermediate stages show swelling of the unciliated cells and a reduction in the number of microvilli. Transmission electron microscopy of the gland of Leiblein of Nucella lapillus In revealing the existence of the same two major cell types throughout the gland of Leiblein the SEM has resolved the confusion in earlier descriptions of the epithelium over the vascular folds, but it is the TEM that has demonstrated their functional significance. Both types are involved in absorption, intracellular digestion and storage (Figs 12–14). While the ciliated cells are the sites of protein catabolism, the often club-shaped unciliated cells are primarily concerned with lipid and carbohydrate metabolism. Superficially the wedge-shaped ciliated cells form a ‘honeycomb’ from the chambers of which the pillar-like club-shaped cells protrude, but the latter provide the greater stability and contact with the basal lamina (Figs 13E, 14C). It is also evident that the apparently contradictory earlier descriptions of their structure reflect different phases in the cell cycle of the same two types throughout the gland rather than variations in the histology of the three lobes of the gland. This confusion was largely created by the limitations of 8–10 µm-thick paraffin sections because in the resting phase the balloon-like apices of the unciliated cells obscure the shorter ciliated cells. Unciliated (club-shaped) cells: These columnar cells have electron-dense cytoplasm usually filled with lysosomes and lipid droplets (Figs 11B, 16). They bear the hallmarks of cells involved in pinocytosis, lysosomal degradation and storage of carbohydrates and absorption of lipids. In well-fed snails they are laden with reserves of lipids and glycogen. In snails known not to have eaten for 5 days to 2 months the cells are club-shaped and their apices, laden with late-stage heterolysosomes and residual bodies, bulge over the ciliated cells. In unfixed material they are seen to contain a colourless vacuole with yellow-green contents. The apical cytoplasm is often partially constricted from the rest of the cell, to which it remains attached by a narrow stalk. The stretched apical membranes bear a relatively sparse covering of short stubby microvilli, at the bases of which there are a few pits (Fig. 16A, B). In glands from some recently collected snails, the cells are more variable in height and may have shed their apical spherules. In others, all the cells have undergone apocrine secretion, reducing the cell height and exposing the adjacent less electron-dense cells and their cilia to the lumen (Figs 11–15). In marked contrast to the typical enzyme-secreting cells of the acinous salivary glands of Nucella (Andrews, 1991) the Golgi bodies lying near the basal nuclei are small, and the granular endoplasmic reticulum (Fig. 16E) is not extensive, indicating protein synthesis on a modest scale. There is never evidence of intense secretory activity or zymogen granules in the cells, so the enzymes they secrete must be in the lysosomes and residual bodies shed in the apical blebs. The lysosomes are positive to tests for sulphated and carboxylated carbohydrates and acid phosphatase, and the cytochemical tests have confirmed that they retain acid phosphatase activity in the free spherules in the lumen. The microvilli bear a glycocalyx neither as dense nor as regular as that of the ciliated cells (cf. Fig. 16A, B inset). It is intensely stained in the periodate test for carbohydrates, as are the glycogen deposits throughout the cytoplasm. Pits in the cell membrane are associated with coated and uncoated vesicles and an apical canal system. The canals appear to open to permanent heterophagosomes (Fig. 16A, B), as described by Owen (1972, 1973) in the digestive cells of bivalves. The contents of the heterolysosomes, which vary in frequency in different specimens (Figs 11–15), are heterogeneous and never show concentric layers. Smooth endoplasmic reticulum is extensive throughout the cell, particularly amongst saturated and unsaturated lipids (Figs 13C, 16F) and reserves of glycogen in the basal cytoplasm (Fig. 14D). Droplets of saturated lipid are closely associated with, and often encircled by mitochondria (Fig. 16D), indicating their use as an energy source by these cells. There is infolding and interdigitation of the basolateral processes associated with mitochondria (Fig. 13), suggesting the existence of a glucose transport mechanism. The processes are in close contact with the basal lamina and ramify among the longer branches of the ciliated cells in a way reminiscent of the enterocytes of the mammalian intestine, rather than the regular array of the proximal convoluted tubule. Axonal endings indicative of regulated control of cellular activity are common among them. Ciliated cells: These cells show features associated with receptor-mediated endocytosis and digestion of low molecular weight proteins and amino acids. Their large apical surface area is greatly increased by a dense covering of long microvilli with a prominent glycocalyx (Fig. 16B inset). These are interspersed with a few long cilia, which presumably assist in the prevention of unstirred layers when the overhanging balloon-like apices of the electron-dense cells do not obstruct them. Unlike the dense cells, they never shed the whole cell apex but occasionally pinch off small blebs containing residual bodies, dark brown or black in unfixed material, and the microvilli are always prominent. Clathrin-coated pits lie at the bases of the microvilli, beneath which are an extensive apical canal system, many mitochondria and lysosomes (Fig. 14B inset). There is a narrower mid-region occupied by the nucleus and Golgi bodies, and from which extend basal branches rich in mitochondria. The branches occupy half the height of the epithelium but have limited contact with the basal lamina. The cytoplasm surrounding the lysosomes and residual bodies is pale and contains little if any glycogen or lipid. In an absorptive phase, when the lumen of the gland is filled with finely granular particles, the heterophagosomes of these cells are larger than those of the unciliated cells, but are electron-lucent except for small numbers of similar particles (Figs 13A, 14B). At a later stage, by contrast, the lysosomes and residual bodies are smaller than those of the unciliated cells. They are negative to tests for carbohydrates. The granular contents (Fig. 13A, arrowed) are composed of concentric layers of material hard enough to cause chatter in thin sections, which preliminary TEM X-ray microanalysis of sections shows to be largely calcium phosphate (M.K. Faulkner, personal communication). They resemble the inorganic granules in the basophil cells of the digestive gland of prosobranchs (Mason & Nott, 1981; Gibbs et al., 1998). The latter authors have shown that the digestive gland granules of Nucella contain magnesium, but note that this may be lost during specimen preparation, which may account for the absence of the ion in these sections. Some layers of the granules that are electron-dense in the TEM are black in 0.5 µm-thick resin sections examined with an optical microscope. They probably contain the melanin identified in histochemical tests by Martoja (1971), which is indicative of protein (tyrosine) catabolism. Epithelium of the ventral fold: The densely ciliated cells on the double ventral fold, like those lining the vestibule around the opening of the duct into the gland, lack the features of absorptive or secretory cells. Their cilia are specialized for creation of strong currents. They do not have bulbous bases and their fibrous rootlets converge on a point close to the nucleus, in contrast to the parallel basal bodies just below the cell surface of those on the vascular folds. Mucous cells: Mucous cells are common on the ventral fold, but they are sparse on the vascular folds. They show β metachromasia in resin sections stained in Toluidine blue, and their ultrastructure is that of typical goblet cells. Cadmium-treated snails: In these specimens the endosomes and lysosomes of the unciliated epithelial cells of the gland of Leiblein contain finer electron-dense deposits than any in the controls (Fig. 15F, G), which suggests that there is some direct absorption of ingested cadmium. In all other respects the cells look normal. X-ray microanalysis has confirmed that Cd occurs in the lysosomes of experimental specimens, together with S, Zn and Fe (M.K. Faulkner, personal communication). Cyclical activity of the gland of Leiblein of Hinia Both cell types in the gland of Leiblein of Nucella occur in Hinia (Figs 17, 18), though the ciliated cells of Hinia differ from those of Nucella in that they are columnar, with a regular array of basal infoldings, and their microvilli are relatively longer. The characteristics of the cells in known phases of the cell cycle of Hinia have been used here to interpret the variability in their fine structure in Nucella. Martoja (1964) could identify only the club-shaped cells in histological preparations of Hinia, but was able to correlate their different stages with feeding and digestion. Her results have been confirmed in this study and apply to both cell types. In specimens that had been starved for 5 days or more the cells were all in a resting phase (Fig. 17A, B), typified by the swollen apices of the electron-dense club-shaped cells bulging into the lumen. They remained in this state in experimental animals starved for up to 20 days, in which the elongation of the neck connecting the apical spherule with the cell body is more marked (Fig. 18D, E). In snails fixed 1 h after feeding, some club-shaped cells had shed their apical blebs (Fig. 17C). The membranes of the microvilli, apical canals and endosomes were denser than in the unfed controls (Fig. 18C) and lysosomes were numerous. In those fixed 5 h later, most cells had shed their apices (Figs 17D, 18A) and the lumen of the gland was filled with the detached spherules. Large masses of unsaturated lipids occurred in the unciliated cells (Fig. 18B). Martoja also observed that ‘regeneration’ of the cells was not synchronized, and while some cells begin to elongate after 2–5 h others remained flattened (10 µm). After 23 h the cycle was almost complete. The apices of the unciliated cells were already beginning to bulge again and they contained smaller lipid droplets. The cytoplasm of both cell types contained many late-stage heterolysosomes and residual bodies. DISCUSSION The main issues arising from this study are: (1) the way in which the complexity of the muricid gland of Leiblein may have been derived from the mid-oesophagus of caenogastropods; (2) how the ultrastructural evidence of its absorptive capacity may be reconciled with an organization of the foregut that seems to preclude the possibility; and (3) how its organization compares with that of the mid-oesophagus of carnivorous neotaenioglossans. There remains the further question as to how its role overlaps with those of the stomach and digestive gland in digestion and absorption. The organization and comparative morphology of the gland of Leiblein Scanning electron microscopy has greatly simplified the interpretation of the three-dimensional organization of such a delicate and complex organ as the gland of Leiblein of Nucella. It reveals the effects of the three factors that determine the architecture of the gland in the adult: its posterior displacement relative to the rest of the mid-oesophagus, the hypertrophy of the left side, which is most obvious at the broad anterior end, and its relationship with the anterior aorta, showing a secondary effect of torsion on the middle lobe. Its relationships with other parts of the foregut in a muricid such as Nucella cannot properly be understood without some reference to the comparative anatomy of the oesophagus, put into a phylogenetic context by Fretter & Graham (1962, 1994), summarized below. Posterior displacement and large size of the gland of Leiblein in muricids: In a more generalized caenogastropod such as Littorina, the mid-oesophagus is a capacious crop in which food may be retained for some time. It lies behind the nerve ring and is that part affected by torsion, marked anteriorly by the supra-oesophageal connective of the visceral loop and posteriorly by the anterior aorta crossing it diagonally. The hypertrophy of the glandular tissue of the mid-oesophagus in muricids such as Nucella exacerbates the problem of accommodating the retracted proboscis and second pair of salivary glands. The separation of the glandular pouches from the narrow food channel not only allows movement of the mid-oesophagus through the nerve ring, but also permits the expansion of the gland of Leiblein into such a large organ lying dorsal and posterior to the rest of the foregut. This involves elongation of the posterior oesophagus, although to a far lesser extent than that of the anterior oesophagus. The virtual separation of the delicate gland from the mid-oesophagus also allows it to change shape and be accommodated without undue compression when the proboscis is retracted and the snail withdraws into the mantle cavity. During development the diverticulum that gives rise to the gland of Leiblein is inverted as it extends backwards over the posterior oesophagus (Ball, 1994), reversing the post-torsional inversion of the mid-oesophagus. This is evident in the adult from the position of the double ventral fold in the two lobes posterior to the opening of the duct. It lies in the true ventral plane in the gland, not the topographically dorsal as might be expected after torsion, and as it does in the posterior oesophagus. Haszprunar (1988) and Strong (2003) have stated that the ventral folds characteristic of the anterior and mid-oesophagus of more primitive prosobranchs are lost in the mid-oesophagus of all caenogastropods. It follows that their presence both in some neogastropods and in the tonnoidean Cassis (Weber, 1927) must be autapomorphic in both groups. In Nucella the fold has an important role in providing the strong ciliary current that carries the secretions of the gland into the posterior oesophagus. A ventral fold also runs along the posterior part of the much smaller tubular gland in buccinids. The difference in the size of the gland in Nucella and Hinia suggests that it reflects in some way the diet, method of feeding and length of time food is retained for digestion in the mid-oesophagus. These determine the availability of readily absorbed nutrients in the foregut that are by some means transported to the gland. In Hinia the rapidity with which pieces of food reach the stomach is not consistent with extracellular digestion in the mid-oesophagus, and would explain the small size of its gland of Leiblein. The long posterior oesophagus partly lined by a transporting epithelium (Payne & Crisp, 1989) may compensate for this by its involvement in absorption of solutes, possibly regurgitated from the stomach (discussed below). Hypertrophy of the left side of the gland: In Nucella the hypertrophy of the left side of the gland is a reflection of the normal growth gradient that affects even the body whorl to some extent in a dextrally coiled snail. It is most marked in the asymmetry of the large anterior lobe, which is developed almost entirely on the left side. The limited development of the right side contributes to the dextral twist between the anterior and middle lobes. The influence of the anterior aorta on torsion: The limits of the middle lobe of the gland, defined by the anterior and posterior fissures, coincide with constrictions where the anterior aorta is entwined around the gland, which combines with the hypertrophy of the left side to produce its definitive spiral. The point at which the artery crosses the posterior end of the middle lobe marks the posterior limit of torsion in Nucella, as it does in the mid-oesophagus of other prosobranchs. The posterior lobe behind that limit is straight and almost bilaterally symmetrical. The result is a situation in which both ends of the mid-oesophagus in Nucella (the valve of Leiblein and the gland of Leiblein) are affected by torsion, but the middle, which contains the glande framboisée, is not. Indeed, a twist in this region, with its voluminous mucous glands, could severely impair movement of the food rod, although Maes & Raeihle (1975) found the relatively much longer comparable region to be the site of torsion in the costellariid Thala. Thus, in muricids the primary effect of torsion, restricted to the anterior end of the mid-oesophagus, is clearly differentiated from the secondary consequence that marks the posterior limit of its influence. Buccinids and nassariids, which lack a glande framboisée, retain the typical prosobranch condition in which torsion produces a gradual twist along the whole length of the mid-oesophagus behind the nerve ring. Cell structure and function: absorption and digestion The glandular nature of the mid-oesophageal pouches of prosobranchs is well documented. Secretion of enzymes for extracellular digestion has been regarded as their sole function because there seemed to be no possibility that nutrients could gain access and enable them to participate in any other activity. This view is not consistent with the ultrastructure of the cells as exemplified by those in the gland of Leiblein. They contrast markedly with the serous (enzyme-secreting) cells of the acinous salivary glands of Nucella (Andrews, 1991), which are specialized solely for regulated secretion, like the pancreatic acinar cells of mammals. Although some enzymes may reside in the glycocalyx of the microvilli, the enzymes in the cells of the gland of Leiblein are lysosomal and primarily concerned with intracellular digestion. Graham (1941) predicted this long before lysosomes had been identified, when he addressed the apparent contradiction between the biochemical evidence of Hirsch (1915) and Mansour-Bek (1934) that it produces proteolytic enzymes, despite its lack of typical enzyme-secreting cells. He suggested that ‘what Mansour-Bek has done is to demonstrate the occurrence of these enzymes in the cells (that exist) as part of the normal equipment of intracellular processes which every cell must possess, and not (specifically) as digestive enzymes secreted into the lumen of the oesophagus’. The mucous string from the gland of Leiblein never enters the mid-oesophagus, and the unit membranes of the residual bodies remain intact in the posterior oesophagus, so the enzymes are not free to contribute to extracellular digestion in the glande framboisée (further discussed below). If they do so in more posterior parts of the gut this may have evolved as an incidental consequence of release of still active enzymes by the breakdown of the limiting membranes of the residual bodies in the stomach. Enzyme activity has also been demonstrated in the residual bodies of the digestive gland (Dimitriadis & Andrews, 2000). Since lysosomes are indicative of intracellular digestion, it follows that the glands must be absorptive. Hirsch (1915) showed that cells in the ‘valdrüse’ of Murex trunculus fed on Carcinus maenas filled with iron saccharate absorbed the iron from the lumen at a stage when they did not contain ‘secretory granules’. When granules were evident they contained iron in a diffuse state, which (later?) occurred as granules in vacuoles, suggesting a cycle of absorption and secretion. The gland did not absorb particulate carmine, and ultrastructural evidence confirms that in Nucella it is not involved in phagocytosis. Franc (1952) cited by Fretter & Graham (1994) also showed that the gland of the muricid Tritonalia (=Ocinebrina) aciculata (L.) is rich in alkaline phosphatase, which is indicative of absorptive activity. Bush (1986, 1989) found similar cells lining the oesophageal pouches of Patella vulgata L., which appear to be a major site of digestion and absorption in this species. Periodic muscular contraction is believed to force fluid into the pouches against the direction of ciliary currents. While the salivary glands of Patella do not produce digestive enzymes, the mid-oesophageal glands release an amylase by apocrine secretion in blebs (containing residual bodies). They are presumably retained in the oesophagus long enough to break down and release enzymes there. The glands also take up radioactive glucose and amino acids by sodium-dependent active transport and Bush gave ultrastructural evidence of cyclical absorptive activity. He suggested that the glands might also absorb dissolved organic matter from seawater, and related their absorptive capacity to the unexpected simplicity of the stomach in Patella, which is uncharacteristic of a microphagous herbivore. The same may be true of the sugar glands of chitons (Brimble, unpublished observations). There is a similar epithelium over the ventro-lateral walls of the mid-oesophagus of Littorina littorea (E.B. Andrews, unpublished observations). It suggests that absorption may be widespread to a greater or lesser extent in the oesophageal glands of marine gastropods and that the gland of Leiblein is a specialization of a primitive condition of the gastropod mid-oesophagus. The existence of cells engaged in absorption and intracellular digestion in the foregut of carnivores challenges conventional thinking, but recent work by Lau & Leung (2004) adds further supporting evidence, albeit circumstantial. Their results have shown that the muricid Thais clavigera (Küster) can supplement its diet by taking in seawater, probably by sucking it up with the proboscis, and absorbing from it (at an unidentified site) suspended and soluble organic matter. Their suggestion that ‘drinking’ could be a common phenomenon in marine invertebrates exposed to eutrophic waters is therefore consistent with the ultrastructural evidence of absorption in the mid-oesophagus. In Thais this capacity can provide >10% of overall energy requirement and helps to sustain the snails in a healthy state during sometimes-long intervals between feeding on prey. The large size and storage capacity of the gland of Leiblein in those neogastropods such as Nucella, for which feeding is in addition energetically costly, may also be important in increasing food reserves. Uptake of cadmium. The presence of fine particles in the endocytotic system and identification of cadmium by X-ray microanalysis in lysosomes of the unciliated cells in the gland of Leiblein of snails exposed to this heavy metal have important implications. This highly toxic heavy metal belongs to the group of so-called ‘soft-acid metals’ that form stable complexes with the cysteine-rich low molecular weight proteins, metallothioneins. These proteins are abundant in cells involved in detoxification, such as the digestive cells of the digestive gland of gastropods, and an increase in their concentration occurs during exposure to heavy metals (Simkiss & Mason, 1983). Leung & Furness (1999) showed that the gland of Leiblein of Nucella accumulated the highest concentration of both the metal and metallothioneins in snails exposed to cadmium for 60 days. Although the concentrations later rose in the kidney and digestive gland as expected, even after a ‘depuration period’ of 110 days, the gland of Leiblein retained the highest levels, indicating that it is a major site of sequestration. The initial increase was greatest in the gill followed by the gland of Leiblein, then the gonad, showing that there was re-distribution of cadmium through the blood to other sites. The ultrastructural evidence of its early appearance in the gland and of uptake through the endocytotic system and apical canals of the unciliated cells in experimental specimens supports the view that the metal can be absorbed from the lumen of the gland as well as from the blood. Typically, the digestive gland is the most important site of sequestration and elimination of heavy metals and other pollutants in gastropods (Simkiss & Mason, 1983; Gibbs et al., 1998). The digestive cells contain metallothioneins and ‘soft bases’ (in the lysosomes) that form stable complexes with heavy metals. In Nucella the gland of Leiblein has also assumed importance both as a site of initial sequestration and long-term storage of cadmium (Leung & Furness, 1999), possibly as a consequence of its proximal position and its rich arterial blood supply. The calcium phosphate concretions of the ciliated cells may also have a wider, perhaps major, role in detoxification of other cations and the relative importance of the two glands in this respect is worthy of further investigation. The hard phosphate-containing granules of the basophil cells of the digestive gland readily form highly insoluble salts with a range of ‘hard-acid metals’. The borderline metals such as zinc and copper can also bind to phosphate concretions. Simkiss & Mason (1983) have pointed out that these two cell types therefore may provide a broad spectrum of regulating, detoxifying and sequestering systems. The cell cycle It may be deduced from the feeding experiments on Hinia, that ingestion triggers apocrine secretion by the cells of the gland of Leiblein preparatory to the next wave of absorption and intracellular digestion. Accumulation of indigestible residues in apical residual bodies is not immediately followed by their expulsion, as it is in the digestive gland of molluscs (e.g. J.E Morton, 1956; Nelson & Morton, 1979; Owen, 1972). The fact that the blebs are retained during the period of recovery and rest until the next bout of feeding means that release of any enzymes remaining active in them coincides with intake of food. At the same time it also transforms the cells into a receptive state for absorption, in which they offer maximum surface area for uptake. Synchrony is lost as the first cells to encounter the incoming stream shed their apical blebs and enter the absorptive phase. Martoja (1964) found that in Hinia all cells had shed their blebs about 5 h after the start of a period of activity, which is confirmed in this study. Fretter & Graham (1994) also noted that Nucella required a period of rest after feeding, which presumably coincides with the phase of intracellular digestion and recovery of the cells. The mechanism of absorption in Nucella The mechanism by which a nutrient-rich solution might reach the cells of the mid-oesophageal glands is even more difficult to understand in Nucella than in those prosobranchs in which the glandular pouches retain a connection with the food channel along their whole length. Nevertheless, as there is clear ultrastructural evidence of endocytosis of carbohydrates and proteins and absorption of lipids from the lumen of the gland of Leiblein, liquid must enter it from the oesophagus. Other neogastropods such as Harpa and Morum are known to ingest their crustacean prey as liquid, probably pre-digested by saliva injected through the proboscis by the pumping action of the mid-oesophagus (Hughes & Emerson, 1987), and there is some evidence of external digestion by naticids too. Reid & Friesen (1980) reported that the most common prey of Polinices, small specimens of the bivalve Tresus nuttallii Conrad were not bored but the part being eaten looked soft and pre-digested. Although Andrews (1991) had no evidence that in Nucella proteases from the acinous salivary glands are introduced into prey during feeding, Crothers (1985) stated that the snail secretes digestive enzymes into the body of its prey, subsequently sucking up a rich soup. The selectivity that it shows in feeding could contribute to this. Piéron (1933) and Graham (1941) observed that when Nucella feeds on Mytilus it takes the soft gonad and digestive gland, rich in glycogen and lipids that are conspicuous in the epithelium of the gland, but leaves the muscles. Its chosen food is easily broken down into small readily digestible particles when seized by the radula, perhaps aided by shell fragments. Undoubtedly the food is mixed with saliva as it enters the dorsal food groove in the buccal cavity. Passage of food through the anterior oesophagus is relatively slow so extracellular digestion may begin, or continue, there. In the valve of Leiblein shell fragments and food particles are coated in a sticky white secretion and compressed into a compact string as they are squeezed and possibly triturated by the thick pads in its lateral walls. This may facilitate extracellular digestion in the glande framboisée, through which passage of the string is even slower. More layers of binding and lubricating mucins are added in this section, in which ciliary currents and peristalsis carry the consolidated string dorsally past the duct of the gland of Leiblein into the posterior oesophagus, along which it is conveyed rapidly to the stomach (Graham, 1941). A thin yellowish-brown stream of secretion emerging from the gland never enters the glande framboisée. It adheres to the food string and remains intact until it reaches the stomach. The bend at the junction of the mid- and posterior oesophagus, the bulging dorsal folds and the position of muscles suggest that there could not be reflux into the glande framboisée from the gland of Leiblein, leaving the acinous salivary glands as the most likely source of extracellular enzymes. Muscular contraction in the glande framboisée, probably assisted by suction when the proboscis is extended, would squeeze solute-laden liquid products of digestion out of the food rod and posteriorly directed ciliary currents must produce a counter-current towards the valve of Leiblein. The bend at the base of the glande framboisée (Figs 2, 8, 9) largely prevents the tendency for the food string to be drawn forward when the proboscis is extended. Any fluid draining into the groove in the floor of the valve may be drawn into the siphon and thence to the gland. Extrinsic muscles and connective tissue control the opening of the siphon where it enters the duct of the gland of Leiblein and this, combined with the fact that when empty the siphon is flattened in cross-section, prevents back-flow from the gland. The mechanism by which liquid is conveyed along such a narrow tube is likely to be two-fold. First, a slow steady stream may be drawn into the gland to compensate for the outflow of secretion along the ciliated channel of the duct. The suction created could be enhanced by contraction of the muscle fibres in the vascular folds of the gland (radial to its surface), so increasing the volume of the lumen. Secondly, muscular activity during the periodic extension and retraction of the proboscis and some peristalsis in the anterior oesophagus may drive pulses of liquid to the gland in larger aliquots. Contraction of the extrinsic muscles and longitudinal muscles of the dorsal folds in the mid-oesophagus may be another important factor. Naticids and those tonnoideans that bore have neither a homologue of the valve nor a completely separate mid-oesophageal gland. However, Reid & Friesen (1980) have proposed a mechanism for the circulation of fluid between the food channel and the partially separated (but not detached) oesophageal gland in Polinices that invokes some principles similar to those applied here to Nucella. In members of the Buccinoidea, such as Hinia, in which there is rapid passage of food to the stomach, there is a trend for reduction or loss of both valve and gland of Leiblein, which indicates that they are functionally linked. In Buccinum the valve is not pear-shaped as it is in Nucella, so is unlikely to compress the food into a compact rod, and the circular fold is narrow with lobed edges and short cilia. It is further reduced in the voracious carrion-feeder Hinia, which tears its food into pieces and passes them to the stomach within 1–2 min of ingestion (Payne & Crisp, 1989). Its oesophageal muscles are well developed, in contrast to their relative paucity in Nucella (E.B. Andrews, personal observations). This explains the vestigial nature of the groove and small size of the valve and gland of Leiblein, offset by the relatively larger digestive gland. Presumably, the partially decayed state of the food provides small quantities of readily available nutrients for absorption by the gland. Other sites of absorption and digestion in Neogastropoda Kantor (2003) has maintained the conventional view that the carnivorous diet of neogastropods is accompanied by replacement of intracellular by extracellular digestion. Undoubtedly extracellular digestion is predominant, but intracellular digestion follows absorption (though not phagocytosis of particles) by epithelial cells in the fore-, mid- and hindgut. The relative contribution of the mid-oesophageal glands to uptake of nutrients in different groups of neogastropods must therefore be put into the context of the functioning of the rest of the gut. There are two major sites of extracellular digestion, absorption and intracellular digestion in muricids—the foregut and the midgut. This arrangement contrasts with that in some neogastropods with different feeding methods, in which the digestive gland is predominant and the mid-oesophageal glands are of lesser importance. In the Conidae there is a change in the function of the mid-oesophageal glands from absorption and intracellular digestion to that of venom production. In Nucella the enzymes in both the foregut and mid-gut derive from two different sources, typical serous cells that secrete zymogens (though Mansour-Bek, 1934, doubted this), and the residual bodies shed by ‘multi-functional’ cells. The two sites differ in that release of the residual bodies at the end of an active phase is delayed in the gland of Leiblein until triggered by the next bout of feeding, but they are shed immediately in the digestive gland, which may have important functional implications. The accessory salivary glands of Nucella, which have a single opening onto the surface of the lower lip, have a role in relaxing prey and possibly anchoring the proboscis during boring but they do not contain digestive enzymes (Andrews, 1991; West et al., 1994). However, the enzymes in the acinous salivary glands initiate extracellular digestion in the oesophagus and their optimum pH is probably higher than that in the stomach. Their effectiveness in making available significant quantities of nutrients for absorption and further digestion in the gland of Leiblein depends on the relatively slow passage of the food through the anterior and mid-oesophagus, particularly the glande framboisée. In Hinia, the rapid passage of food through the oesophagus suggests that there is little time for any significant extracellular digestion there. However, the report by Payne & Crisp (1989) that transporting epithelium lines the part of the posterior oesophagus opening to the stomach in this snail raises the possibility that it absorbs partially digested regurgitated stomach contents. Kantor (2003) has proposed that similar reflux may occur in buccinids. It may therefore be significant that a pouch-like caecum identified by Fretter & Graham (1962) in the posterior oesophagus of Buccinum and Neptunea, which they suggested might help in preventing regurgitation, might also have a role in absorption. Neither of these specializations occurs in Nucella, and there is no ultrastructural evidence of an absorptive epithelium in its posterior oesophagus (E.B. Andrews, unpublished observations). Fretter & Graham also reported the observations of Brock (1936) who found that the ‘flaps’ (ends of the dorsal folds) in the valve of Leiblein of Buccinum respond to stomach contents by closing together. This implies that in this genus regurgitated material from the stomach may travel this far forward in natural conditions. There may be some contribution to digestion in the stomach and intestine from residual enzyme activity in spherules from the gland of Leiblein, but the main source of extracellular gastric enzymes is the digestive gland, for which the acid pH in the stomach provides optimal conditions. Mansour-Bek (1934) reported that the proteinase in the gland of Leiblein of Murex anguliferus was active at pH 8.2. There has been much emphasis on the absorptive capacity and phasic activity of the digestive cells of the molluscan digestive gland by a number of authors (e.g. Morton, 1956; Nelson & Morton, 1979; Merdsoy & Farley, 1973; Boghen & Farley, 1974). Absorptive, digestive and fragmentation phases of these cells have been correlated with a feeding cycle that in some marine species coincides with a tidal rhythm. The ‘crypt cells’ are the equivalent of the typical enzyme-secreting cells of the acinous salivary glands, and the ‘multi-functional’ digestive cells parallel the functions of the epithelium of the vascular folds in the gland of Leiblein. Since the residual bodies shed by the digestive gland of Nucella have been shown to retain at least some enzyme activity after expulsion by the cells (Dimitriadis & Andrews, 2000) breakdown and absorption of any food remaining undigested in the stomach may continue as it travels through the intestine. Kantor (2003) maintained that the digestive gland of buccinids cannot be a site of absorption on the grounds that all ciliary currents lead out of it, and that muscular activity is not likely to counteract them. He also argued that a prominent gastric fold isolates the openings of the ducts of the digestive gland from that part of the stomach containing food. His views are supported by Martoja's experiments on Hinia (1964) and Brown's (1969) on Nassarius obsoletus Say. They are contradicted by the anomalous results of McLean (1971) (cited by Kantor, 2003), who used particulate markers too large for the digestive cells to take up. Although these cells are incapable of phagocytosis in Nucella ultrastructural evidence of their capacity for endocytosis is strong (Dimitriadis & Andrews, 2000), so fluid must enter the gland against the direction of the ciliary currents, leaving muscular action as the most probable mechanism. Kantor found the openings of the ducts to be blocked by mucus in specimens he examined, but this cannot be a permanent feature since there is clearly free flow of secretion into the stomach. Kantor's view that the wall of the stomach and intestine are absorptive has been substantiated in several gastropods (Morton, 1956; Nelson & Morton, 1979; Bush, 1988; Forester, 1977). Comparative anatomy of the mid-oesophagus in carnivorous caenogastropods in relation to diet and feeding mechanisms The comparative anatomy of the foregut and associated glands has long been central to the debate, both as to the ways in which it may be correlated with different diets and feeding strategies where known, and its relevance to the origin and evolution of neogastropods. Interpretation has often been hampered by limited information on embryological development and the fine structure and function of the cells. Amaudrut (1898), in addressing these problems, proposed that the mid-oesophageal glands of ‘taenioglossans’, the toxoglossan venom gland, and the gland of Leiblein were homologous. Fretter & Graham (1994) extended this thesis in considering the possibility that all foregut glands of prosobranchs, with the exception of the accessory salivary glands of neogastropods which are invaginations of the lower lip (Ball, 1994), are part of the same glandular tracts. Three of the five characters of neogastropods defined as autapomorphic by Taylor & Morris (1988) (a view upheld in the following argument) relate to the foregut: possession of two pairs of salivary glands, a valve of Leiblein and a gland of Leiblein. However, the explosive adaptive radiation of the group, largely related to feeding strategies, has resulted in the identification of relatively few other morphological characters in the initial radiation, with a consequent weakness in analysis by the cladistic approach (Kantor, 2002). Strong (2003) has argued that past emphasis on the correlation between structure and function in the alimentary system of caenogastropods, particularly in relation to the mid-gut, has resulted in failure to recognize its value for cladistic analysis. She has further suggested that such correlation (reviewed by Fretter & Graham, 1994) is not always as close as it might seem, because it cuts broadly across patterns of feeding, diet and foregut complexity to a degree previously unappreciated. It may equally be explained by our limited understanding of the links between structure and function. This is so in the case of the mid-oesophageal glands and in neogastropods it also applies to the midgut. The present findings, among others (Bush, 1989; E.B. Andrews, unpublished results), challenge traditional views as to the functions of these glands, not only in neogastropods but also throughout prosobranchs. The link between carnivory and the development of the gland of Leiblein in neogastropods, made on the basis of its assumed primary role in production of extracellular enzymes, must therefore be re-assessed. It follows that the role of the gland of Leiblein as a site of absorption, intracellular digestion and storage requires nutrients to be available as solutes in the mid-oesophagus. If the gland is an autapomorphic character this must have been so in the earliest neogastropods. The carnivorous habit probably arose from opportunistic browsing (Fretter et al., 1998), its efficiency enhanced by the development of a proboscis. Kantor (2002) argued persuasively that the most primitive neogastropods had a short proboscis with a basal buccal mass, such as is found in the muricoidean Pseudolividae and Ptychatractidae, and the conoidean Drillidae and Turridae. If the radula could not be used as a food-gathering organ, then suctorial feeding and the presence in the mid-oesophagus of partially digested ‘soup’ would explain the hypertrophy of the mid-oesophageal glands into the gland of Leiblein. This probably arose in the most primitive muricoideans such as the Ptychatractidae, in which the gland is well developed (Kantor, 2002). In such groups as muricids with a terminal buccal mass, shell and tissue fragments are mixed with partially digested solutes and it increased in size, accompanied by a reduction in the role of the stomach as the major site of digestion. Conversely, in those that adopted a macrophagous habit, such as buccinoideans, there was a trend towards reduction and loss of the gland of Leiblein and the stomach usually remained the site of digestion. If food is largely externally pre-digested liquid, as in the Harpidae, both the mid-oesophagus and stomach are greatly simplified. The type and length of proboscis, radular structure, and in some the use of toxins to paralyse prey (West et al., 1996) are correlated with the sort of prey selected and the way in which it is eaten. The narrow pleurembolic proboscis with a terminal buccal mass and narrow radula typical of buccinoideans and some muricoideans is the most advanced. It allows the radula once again to be used in picking up or fragmenting prey. Conoideans retained a short broad proboscis that led at least in some primitive turrids, to a shift in the site of extracellular digestion to the buccal cavity or external digestion in the proboscis sac. It relieved the mid-oesophagus of involvement in extracellular digestion and absorption of solutes. The ventral part could thus be transformed into a venom apparatus accompanying the exploitation of ever more active prey. The decrease in the importance or even loss of the radula as the main food gathering mechanism in the primitive neogastropods was compensated for by the two major advances in the glandular equipment of the foregut: the development of accessory salivary glands and secretion of proteases by the acinous glands. In most prosobranchs the acinous salivary glands are mainly engaged in secretion of mucus, rarely in production of amylases, or acid in the carnivorous Tonnoidea. The acquisition of tubular accessory glands enabled well-protected or mobile prey to be overcome and more easily manipulated and proteolytic enzymes in the saliva may simultaneously have initiated its digestion. These may have been important steps in overcoming the limitations of a basal radula, a change to suctorial feeding, and development of a gland of Leiblein, in such families as the Ptychatractidae. The paralytic toxin serotonin has been identified in the accessory salivary glands of Nucella (West et al., 1994), but enzymes have not (Andrews, 1991). Accessory glands are lost in the Buccinoidea, which are scavengers or opportunistic feeders and shell-wedgers or chippers. Nevertheless, in a few, such as Neptunea, the acinous salivary glands have secondarily acquired the ability to secrete paralytic agents such as tetramine. Little is known of the fine structure of the glands in different groups. In Hinia they possess three cell types whereas in Nucella there are only two (Andrews, 1991). The importance of a valve of Leiblein in preventing regurgitation during proboscis extension varies. It may be secondarily reduced, lost, or hypertrophied. It is most prominent in species with a large gland of Leiblein, possibly linked with the presence of considerable amounts of partially digested fluid in the mid-oesophagus, as in Nucella.Strong (2003) has correlated the presence of a valve with that of a gland of Leiblein connected directly to the mid-oesophagus, but this is not always so since a similar structure occurs in some marginellids (Ponder, 1970a) and some raphitomine Conoidea (Kantor & Taylor, 2002). The following survey of neogastropod families reveals the correlation between the structure and function of the midgut with those of the foregut and the state of the food when consumed: liquid, ‘soup’, or solid (Table 1). Each requires different treatment in the fore- and midgut. The picture is further complicated because the site at which digestion and absorption are initiated varies in different families. In some the foregut becomes an even more important site of digestion than in the primitive forms and the stomach is simple. In others the stomach is the more important or sole site and is more elaborate, resulting in a complex picture of convergence and parallelism. Suctorial feeders on finely fragmented tissue and/or liquid MURICOIDEA: Costellariidae. The problems of unravelling these links between the structure and function of the fore- and midgut are no more clearly illustrated than within the Muricoidea. The similarity of the mid-oesophageal region in Nucella and that of genera such as Thala (Costellariidae), in different families and with apparently different feeding methods seems on the face of it to negate any functional significance, but there are striking parallels. Although Thala invades the shells of gastropods without boring, like Nucella it has a pair of well-developed accessory salivary glands and paralyses its prey before sucking up the soft tissues as small particles in liquid suspension (Maes & Raeihle, 1975). The ducts of these glands open at the tip of a protrusible epiproboscis similar to that of mitrids. By contrast the odontophore is too large to be protracted through the mouth and a minute radula serves only to pass food particles pumped into the buccal cavity to the oesophagus. The gland of Leiblein is large, though relatively smaller than that of Nucella and it appears not to be affected by torsion. There is a highly developed valve of Leiblein and a very long mid-oesophagus behind the nerve ring in which the glandular dorsal folds are hypertrophied as in the shorter glande framboisée of Nucella. It is the site of torsion and the main site of digestion and is correlated with the possession of a small simple stomach. Columbariidae and Coralliophilidae: The Columbariidae and Coralliophilidae emerged as sister groups of the Muricidae in a cladistic analysis by Kantor (2002). The Columbariidae use a long thin proboscis to reach reclusive sedentary prey in crevices and tunnels (Ponder, 1973). The Coralliophilidae feed on corals. Both groups have lost the accessory salivary glands and do not paralyse their prey. In the former the radula is small and it is lost in the Coralliophilidae, so they too are suctorial feeders on partially digested fine particles, using the buccal cavity as a pump. Their relatively massive gland of Leiblein is even larger than in muricids, presumably reflecting a greater absorptive capacity, and the stomach is small. The fluid ingested is devoid of hard material such as shell fragments and the dorsal folds are not glandular. Olividae: The Olividae burrow in sand and Taylor & Glover (2000) found that Oliva tigridella Duclos has a varied diet of gastropods, bivalves and echinoids. It carries prey on the posterior end of the foot, some of which may be partially eaten, indicating that it is not swallowed whole. Oliva sayana Ravenel feeds on small crustaceans and ‘juice’, Olivancillaria (Lintricula) auricularia (L.) on Donax (Marcus & Marcus, 1959), and fragments rasped by the radula are pumped up by peristalsis. Kantor (2003) has reported the secretion of proteolytic enzymes by the epithelium of the sole of the foot in Oliva, but Marcus & Marcus found the exoskeletons of crustaceans in the stomach, so not all the food is pre-digested. Taylor & Glover (2000) also found that the gut contents of O. tigridella included holothurian spicules, crustacean fragments and rarely radular teeth and polychaete chaetae. The mid-oesophagus is short with highly glandular dorsal folds (their secretion presumably protecting against the hard parts of prey). While the gland of Leiblein is well developed, as in Nucella the stomach is larger and more complex, with a gastric shield and sorting area, which Marcus & Marcus perceived to be an anomaly, but the abrasive material found in the gut contents explains this in functional terms. Volutomitridae: In common with olivids, many species of volutomitrids live in mud and sand (Ponder, 1998) but Bouchet & Kantor (2004) reported 14 deep-sea species around New Caledonia with a marked preference for hard substrates. The food of volutomitrids is unknown, but from his observations on the sand-dwelling Peculator hedleyi (Murdoch) Ponder (1972) deduced that they scrape particles of flesh or suck body fluids. Kantor & Harasewych (1992) also speculated that species of Volutomitra feed on fluids. They have a thin jaw that may be folded into a funnel around a radular incision on the prey. The sharp-edged jaw of Microvoluta from New Caledonia cannot be used in this way but it may be a cutting tool (Bouchet & Kantor, 2004). Typically there is a small single accessory salivary gland and the pleurembolic proboscis is short. The walls of the buccal cavity are highly muscular and the odontophore is small, suggesting that there is a largely fluid intake involving the use of the buccal cavity as a pump. The salivary glands are small, as is the gland of Leiblein that is only partially separated from a narrow dorsal food channel. However, the dorsal folds in the anterior part of the long mid-oesophagus are highly glandular and the stomach, which has a gastric shield and style sac, is adapted for some trituration and digestion. These features are usually associated with a diet including some abrasive material, and this is confirmed for Peculator in which Ponder found fine mineral particles, diatom cases and spicule-like fragments in faeces. The structure of the long mid-oesophagus is peculiar to this family. In Peculator there is a clearly defined valve of Leiblein with a ventral groove (the equivalent of the siphon of Nucella). But it lacks the typical cone-like valve (the circular fold of Nucella), possibly correlated with the highly muscular nature of part of the mid-oesophagus, which is divisible into three sections. Torsion occurs at the posterior end of the valve. The section of the mid-oesophagus between the valve and the nerve ring is exceptionally long and has prominent dorsal folds and a tall glandular epithelium. It must be composed mainly of dorsal food channel since the ventral groove is clearly identifiable. In the next highly muscular section low dorsal folds border a minute food channel. The epithelium over most of the wall is irregular and contains greenish granules like that of the gland of Leiblein, which led Ponder to conclude that it is the anterior part of the gland that remains connected to the dorsal food channel. This interpretation is supported by the absence of the ventral groove in this section—this ventral component is hypertrophied and differentiated. It is the converse of the situation in Nucella in which the ventral part of the mid-oesophagus (the siphon) is vestigial, the food channel is well developed along its whole length, and the glande framboisée behind the nerve ring is an important site of extracellular digestion. The groove reappears in the following muscular region, which comprises mainly a dorsal food channel. It gives off a minute diverticulum (the gland of Leiblein) at the point of origin of a narrow posterior oesophagus, in which muscles are poorly developed and there is a ciliated non-glandular epithelium. Harpidae: The Harpidae show no evidence of ingesting hard material. The loss of the gland of Leiblein and great simplicity of the whole gut reflects external, as distinct from extracellular digestion, and ingestion of liquid in a state ready for absorption (Hughes & Emerson, 1987). Harpids smother their prey and inject saliva to digest the tissues; an oesophageal pump probably sucks up the solute-rich liquid. This is consistent with the large size of the acinous salivary glands; there are no accessory glands, and the buccal apparatus is ‘microscopic’. The valve of Leiblein is vestigial. Dorsal and ventral components cannot be identified in the mid-oesophagus, which has an expanded posterior storage section with longitudinally pleated but undifferentiated walls. The simple stomach merges almost imperceptibly with the posterior oesophagus. Marginellidae: The loss of the gland of Leiblein in marginellids (Ponder, 1970a) and their possession of a venom gland are reminiscent of the conoidean condition. The proboscis is usually short. Graham (1966) described the proboscis of Marginella marginata (L.) as large and muscular, and in Volvarina taeniolata (Mörch) it is about the same length as the shell height when extended (Fretter, 1976). The diet is varied in this diverse family, but the indications are that it is largely fluid when ingested. Volvarina ingests the muscle tissue of its prey (Crangon dentipes) partly by using the radula, partly by suction. Other species feed on bryozoans or ascidians. Ponder & Taylor (1992) discovered that two species of Austroginella bore the shells of bivalves by a chemical mechanism different from that of muricids. Another is known to be an ectoparasite of fish. Typically the buccal mass and radula if present are highly modified, lying in a separate sac below the buccal cavity (also reported to appear transiently in the veliger larva of Hinia; Abro, 1969). This is the arrangement in Marginella marginata, but in M. desjardini Marche Marchad (Graham, 1966) and Volvarina cairoma (Brookes) (Ponder, 1970a) the buccal apparatus is absent, so these snails must depend on suctorial feeding. In these species the duct of the bulb-like venom gland opens mid-ventrally in the posterior part of the buccal cavity (anterior to the site of torsion). This gives the impression that it separated from the mid-oesophagus from the posterior end forwards, as in the Volutomitridae and Conidae (Ponder, 1973). There is no valve of Leiblein, which may be well developed in other marginellids such as Mesoginella pygmaea (Sowerby) (Ponder, 1970a). In Volvarina a mid-oesophageal pouch is the site of digestion, which is reflected in the simplicity of the small stomach (Fretter, 1976). The loss of the pouch in M. desjardini (Graham, 1966) may reflect a more liquid diet. CANCELLARIOIDEA: The Cancellarioidea are set apart from other neogastropods in that they have neither a gland of Leiblein nor a venom gland, but Graham (1966) suggested that the glandular ventral wall of the exceptionally long mid-oesophagus of Cancellaria cancellata (L.) and C. lyrata (Brocchi) may be its homologue. In other respects the foregut shows both primitive and specialized features. Most are suctorial feeders on bivalves and sand-dwelling gastropods; Cancellaria cooperi Gabb sucks blood from electric rays after making small wounds in the skin (O'Sullivan, McConnaughy & Huber, 1987). In Cancellaria (Graham, 1966) the long narrow proboscis accommodates an acinous and an accessory salivary gland, and the odontophore with its highly modified radula is some distance from the mouth, with the valve of Leiblein immediately behind it. The anterior oesophagus is lost as in the Conoidea and the whole of the gut behind the valve is greatly simplified, reflecting a mainly liquid diet. The dorsal folds project deeply into the oesophagus in the valve, and they extend almost to the stomach. Strong (2003) supported the view that cancellariids are a sister group to other neogastropods because the branching of the dorsal afferent renal vein is atypical, the sub-radular organ is not glandular, the subradular extension is not cuticularized and there is no gland of Leiblein. In all other neogastropods the absence of the gland is regarded as a secondary loss, as in the Harpidae. Furthermore, Kantor & Harasewych (1992) found similarities between the foregut of some volutomitrids and cancellariids and, on the basis of 28S rRNA sequencing, Rosenberg et al. (1994) (cited by Harasewych & Petit, 1998) concluded that the group is probably an offshoot of the Muricoidea. The balance of evidence strongly favours this view. E.B.A. disagrees with Strong's interpretation of the anatomy and blood supply of the kidney of prosobranchs in several respects and believes that ultrastructural examination of both the foregut and kidney of cancellarioideans are needed to clarify these aspects. BUCCINOIDEA: Colubrariidae. The only report of feeding in the little-known colubrariids is of Colubraria obscura (Reeve) sucking the blood of parrotfish (Bouchet & Perrine, 1996, cited by Kantor, 2003). Colubrariids are exceptions amongst buccinoideans in that the radula if present is minute, the mid-oesophagus long and glandular and the gland of Leiblein absent (Ponder, 1973). The stomach is a simple sac, which on this limited evidence is correlated with a liquid diet. Macrophagous feeders BUCCINOIDEA: Typically the Buccinoidea are truly macrophagous in that they ingest their food whole or in large chunks (Kantor, 2003) and pass it rapidly to the stomach, and in this respect they differ from most muricoideans. They are mostly inhabitants of soft deposits, reaching live prey or carrion with their very long narrow proboscis, and the terminal radula is normally important in tearing and manipulating food. The accessory salivary glands are lost in all buccinoideans, but in Buccinum and Neptunea access to soft parts is aided by pharmacologically active compounds in the saliva of the acinous glands (Endean, 1972; Shiomi et al., 1994). Less specialized species retain a valve and gland of Leiblein, but the gland is tubular and simpler than that of muricids. In Buccinum there are traces of the mid-oesophageal ventral groove, but this may also disappear. The dorsal food channel is large as in other macrophagous species, with well-developed muscles for conveying food to the stomach. The dorsal folds are not glandular. Since the stomach is the main the site of breakdown and digestion of food it is more complex than in muricids, which is most marked in species that have adopted a herbivorous diet such as some Columbellidae. They feed on brown algae, and the U-shaped stomach has a gastric shield and sorting area (Marcus & Marcus, 1962). A crystalline style may occur (Kantor, 2003). In the nassariid Cyclope neritea, which is a scavenger and deposit feeder, the style sac contains a crystalline style (Morton, 1960). The anatomy of the gut remains a focus in the controversy over the relationship between buccinoideans and other neogastropods. On the basis of his observations on Cyclope, Morton regarded nassariids as the most primitive neogastropods. His view was supported by the fact that the Buccinoidea as a whole show the smallest dietary and structural specializations amongst neogastropods and retain other primitive features (free-swimming larvae, site of torsion). The ‘Buccinidae’ also emerge as primitive in the cladistic analysis of Ponder & Lindberg (1997), and a sister group to all others. They took the lack of accessory salivary glands and the presence of a caecum in the stomach to be plesiomorphic. Kantor (2002) concluded that the Bucconioidea are the most derived neogastropods and pointed out inconsistencies in the pattern of stomach structure, since not all members of the ‘Buccinidae’ have a caecum, but conversely it is well developed in some muricoideans. Both buccinids and muricids emerged as the most derived clades in the cladistic analysis of Strong (2003), a conclusion supported here. In Strong's (2003) analysis the marginellids are basal to both clades. Kantor (2002), by contrast, concluded that the muricids are basal to marginellids, which somewhat diminishes his own argument that buccinoideans and muricids share the same advanced type of proboscis. His case is partly based on the great length of the proboscis in buccinoideans, but it does not necessarily follow that this reflects anything other than differences in feeding methods since several groups of muricoideans and cancellarioideans also have long proboscides. He also suggested that the apomorphic condition of the proboscis could easily be derived from the primitive type, as recapitulated in the ontogeny of Nucella (Ball et al., 1997). To this extent it is possible that the transformation might have occurred independently in muricoideans and buccinoideans, but the arguments given above suggest that it is less probable. Information from the fossil record presents a confusing picture. While on the basis of some interpretations it suggests that buccinoideans are among the most ancient groups of neogastropods (Taylor, Morris & Taylor, 1980), the record is incomplete and some identification is possibly unreliable (Kantor, 2002). The results of a microanalytical survey of the composition of the phosphate granules in the digestive gland of marine prosobranchs Gibbs et al. (1998) are also difficult to interpret. It revealed that, uniquely amongst 17 neogastropod species examined, the granules of buccinoideans have high concentrations of zinc. Those of muricoideans have high values for magnesium–calcium and at least one species of Conus for magnesium but not for zinc. The authors tentatively proposed that the difference might reflect ‘palaeoenvironmental influence’. MURICOIDEA: Olivellidae. The loss of the gland of Leiblein in the Olivellidae reflects a macrophagous habit in some but not all species. This family, like the Nassariidae, also includes deposit feeders (Kantor, 1991) and suspension feeders (Seilacher, 1959), while some feed on foraminiferans (Hickman & Lipps, 1983) (sources cited by Taylor & Glover, 2000). Olivella verreauxii (Duclos), a sand-dweller, ingests crustaceans, scaphopods and Donax, tearing off pieces or eating small ones whole. The ganglia of the nerve ring are clearly separated, allowing for considerable distension of the oesophagus. Olivella lacks accessory salivary glands and has a poorly developed dilatation corresponding to the valve of Leiblein (Marcus & Marcus, 1959). Eight similar folds line the indistinguishable mid- and posterior oesophagus. This simplicity is compensated for by the possession of a complex stomach, the site of extracellular digestion, which has a large gizzard and a well developed sorting area. Mitridae: The mitrids inhabit coral reefs and intertidal platforms and feed almost exclusively on sipunculans which they sometimes swallow almost whole (Taylor, 1989, 1993). The foregut is highly specialized. Both accessory salivary glands and the gland of Leiblein are lost (Ponder, 1972). They have a ring of peristomial papillae and a peculiar muscular epiproboscis used to reach their relatively inaccessible soft-bodied prey. The odontophore is large, the radula broad, and the epithelium of the buccal walls bears 50-µm long projections (reminiscent of stereocilia). The valve of Leiblein is small or absent and there is a short mid-oesophagus. Food is stored in a simple post-oesophageal crop but the stomach, which has a muscular triturating gizzard and a modified style sac but no caecum, remains the site of digestion. Volutidae: The Volutidae is another family in which the ventral part of the mid-oesophagus appears to have separated from the dorsal from posterior to anterior. Alcithoe burrows in sand or mud and feeds on large pieces of ‘dead or living flesh’ of other molluscs (Ponder, 1970b). There is a short proboscis and a powerful buccal mass but the radula is reduced, unlike that of mitrids. The muscular anterior oesophagus shows strong peristalsis and the thinner-walled short mid-oesophagus is the site of copious mucous secretion. The dorsal folds are believed to fuse to form the proximal tubular section of the gland of Leiblein sensu lato. It is long, thick and muscular with two histologically distinct glandular sections, the gland cells of which do not resemble those of typical dorsal folds or the gland of Leblein. The acinous glands of the lobulated anterior section secrete digestive enzymes, the more distal section bears folds. This tube connects the oesophagus to a small terminal bulb guarded by a sphincter, which Ponder took to be the only part homologous with the gland of Leiblein of muricids because of their similar cytology and possession of a ciliated ventral fold. The posterior oesophagus is a capacious crop and the main site of digestion. This is reflected in the small size of the stomach, which lacks a gastric shield but has a very efficient sorting mechanism. Ponder (1970b) proposed that the gland of Leiblein sensu stricto of volutids is homologous with the terminal muscular bulb of the venom apparatus of conoideans and the poison gland itself with the tubular section, which he took to be derived by fusion of the glandular dorsal folds. Volutocorbis, one of the most primitive volutes, seems to foreshadow this. A dilated convoluted section of the oesophagus between the nerve ring and the opening of the duct of the gland of Leiblein corresponds to the tubular part in other volutes (Ponder, 1970b). Our observations on Nucella suggest that indeed the contiguity of the dorsal folds could have resulted in their fusion in both volutids and conoideans. However, the glands of the tubular sections (occupying a morphologically ventral position) are here believed to be homologous with parts of the ventral mid-oesophageal glands, as Amaudrut (1898) originally proposed. They are comparable in position with the ‘undifferentiated’ epithelium on the face of the dorsal folds that lines the siphon in Nucella (Fig. 9D). Different patterns of fusion of the dorsal folds in different groups can be explained by heterochrony, which would overcome the complication of envisaging the separation of the ventral from the dorsal part of the oesophagus in two different directions in the evolution of the Conoidea and other neogastropods. The issue should be resolved by the application of modern biochemical techniques. CONOIDEA: The production of venom delivered by the specialized radular teeth allows members of the Conoidea to catch active prey and the breadth of the short proboscis enables them to swallow smaller specimens whole. But paradoxically they may in some respects be more appropriately regarded as feeders on liquid, since salivary enzymes initiate digestion either externally or in the most proximal parts of the foregut. This is consistent with the loss of the gland of Leiblein and the simplicity of the stomach. The structure of the proboscis is very variable but can largely be defined as intra- or polyembolic. Normally the buccal mass lies at the base of the proboscis and the highly modified radular teeth are derived from a taenioglossate type (reviewed by Fretter & Graham, 1994; Kantor, 2002). Kantor (1988) identified members of two families of the Conoidea, the Turridae and the Drillidae, as possessing the characters of the most primitive neogastropods (except a gland of Leiblein). He described two primitive but specialized turrid genera, Pseudomelatoma and Hormospira (Pseudomelatominae) in which the oesophagus, and also part of the buccal tube in the latter, are elongated into a curved section that is probably the site of digestion (Kantor, 1988). They have large salivary glands and saliva is known to contain proteases in ‘toxoglossans’. The stomach is simple and U-shaped in both. In other turrids with an intraembolic proboscis salivary enzymes may initiate digestion in the buccal cavity; in those with a polyembolic proboscis prey is digested in the proboscis sac before ingestion. By contrast in the Drillidae food reaches the stomach whole. Some members of the Conidae prey on fish as large as the predator. In these cases enzymes rapidly digest the prey externally or at the anterior end of the gut. This is correlated with a simple oesophagus and a small, simple U-shaped stomach not clearly demarcated from the oesophagus or intestine. It has neither a gastric shield nor a sorting area. Homologies within carnivorous Neotaenioglossa The tendency for the dorsal folds to isolate the dorsal and ventral parts of the mid-oesophagus has occurred not only in neogastropods but also in carnivorous caenogastropods, and it has developed in different ways in different groups (Fig. 19). It depends on the degree to which the dorsal folds meet or overlap. The diversity in the structure of the foregut of higher neotaenioglossans (Hypsogastropoda) suggests that the trend must have developed in parallel with the development of different types of proboscis at a very early stage in the evolution of carnivorous prosobranchs. It supports Kantor's (2002) conclusion that the Tonnoidea are unlikely to be the sister group of neogastropods. The acquisition of the acrembolic proboscis in the Ptenoglossa and Naticoidea appears not to have resulted in major displacement of the dorsal folds, as has the advanced pleurembolic proboscis of some Neogastropoda. Nor has separation of the dorsal and ventral parts of the oesophagus occurred in the Tonnoidea to the extent that it has in the Neogastropoda. The pleurembolic proboscis of the Tonnoidea differs from that of neogastropods in the position of its retractor muscles. Comparison of Nucella and Cymatium intermedius suggests that there is not such a significant narrowing of the dorsal food channel in the latter as there is in the former (Fig. 1; Andrews, Page & Taylor, 1999). If this is typical of tonnoideans, hypertrophy of the dorsal food channel might have brought the dorsal folds closer together. The odontophore is typically small and terminal (Kantor, 2002), but Riedel (1995) reported a narrow radula only in the personiid Distorsio, which has an exceptionally long thin proboscis. The radulae of the Ranellinae and Bursinae are comparatively large. It seems unlikely that a narrow buccal cavity and oesophagus could occur in a type such as Tonna, which Morton (1991) observed swallowing holothurians whole. Both the Naticoidea and Tonnoidea include borers and may take in food in a fluid or semi-fluid state (Day, 1969; Taylor et al., 1980; Morton, 1990; Fretter & Graham, 1994), and they show similarities in the organization of the mid-gut. In Polinices lewisii (Gould) the mid-oesophageal gland is subdivided from the food channel to which it remains connected along its whole length (Reid & Friesen, 1980; Page & Pederson, 1998). As in other Naticoidea, Ptenoglossa and Tonnoidea the dorsal folds are asymmetrical; the morphologically left fold is wide enough to overlap its partner and guard the slit-like opening. All ciliary currents in the glands are directed into the food groove. It is strikingly similar to the arrangement in Cerithiopsis (Fretter, 1951) and reminiscent of that in some Tonnoidea (Weber, 1927). These glands bear septate folds, as do those of Patella (Bush, 1989), the neritoidean Phenacolepas omanensis Biggs (Fretter, 1984) and other archaic groups (Salvini-Plawen, 1988). The Ranellidae (=Cymatiidae) seem to show a fairly typical arrangement of the foregut glands and their ducts amongst the Tonnoidea (Houbrick & Fretter, 1969; Andrews, Page & Taylor, 1999; Riedel, 1995). In Cymatium, Houbrick & Fretter (1969) described a red-brown oesophageal gland with a series of transverse folds creating compartments open to the narrow main food channel resembling those of naticids, and its secretions are carried into the food channel by ciliary tracts. The epithelial cells over the folds in Cymatium intermedius are the same types as those in the gland of Leiblein (Andrews, Page & Taylor, unpublished observations). Hughes & Hughes (1981) found a similar organization in the gland of Cassis.Day (1969) identified cells in the gland of the ranellid Argobuccinum that produce unspecified digestive enzymes. Polinices lacks salivary glands, but according to Reid & Friesen (1980) there is a distinctive white anterior part of the oesophageal gland that secretes ‘a masked proteinase precursor’. Hirsch (1915) also found proteolytic enzymes in the oesophageal gland of Natica. Tonna is unusual in that its anterior oesophagus bears the Delle Chiaje (DC) organ (Weber, 1927, as Dolium), which bears a strong resemblance to the mid-oesophageal glands of cymatiids, but with the addition of a posterior caecum ending in a bulb. It has a dark septate glandular (morphologically and topographically) ventral wall. Day (1969) and Hughes & Hughes (1981) have suggested the possible homology of the DC organ with the buccal glands of cymatiids, since it opens to the buccal cavity posteriorly, and Houbrick & Fretter (1969) found patches in a comparable position in the oesophageal wall of Bursa. It raises the question as to whether in Tonna there is any significance in the absence of typical (septate) mid-oesophageal glands and the presence of the DC organ, and the absence in ranellids of a DC organ but the presence of mid-oesophageal glands. If these foregut glands were all part of the same glandular tracts, there would be no difficulty in homologizing the DC organ with mid-oesophageal glands. In discussion with E.B.A. some years ago, Graham suggested that this might be possible on the following grounds: (1) they run (morphologically) mid-ventrally along the anterior half of the oesophagus; (2) they are separated from the main oesophageal lumen by longitudinal folds, i.e. the dorsal folds of the food channel that have migrated ventrally as in neogastropods; (3) the cavity separated ventrally by these folds is septate; (4) the posterior end of the glandular oesophageal region separates from the food-conducting part in several prosobranchs. ACKNOWLEDGEMENTS We thank Dr V.K. Dimitriadis and Dr K.M.Y. Leung for their contributions. E.B.A. owes much to Patricia Goggin of the EM Unit, and to Michael Faulkner and Keira Thorogood, formerly of the EM Unit, Royal Holloway, for their technical assistance and personal support. She is also greatly indebted to Professor J.E. Morton, Dr J.D. Taylor and two referees for valuable comments on the manuscript, and to Professor Sir John Enderby, C.B.E., F.R.S., who has been a source of inspiration during a difficult period. DEDICATION At the time when the biological applications of electron microscopy were gaining momentum Alastair Graham's opportunities to pursue his research interests were severely restricted (Andrews & Allen, 2002), and he was unable to embrace this new field in his studies on the functional anatomy and histology of the gut of prosobranchs. By contrast, Gareth Owen was able to take advantage of the new technology, and made a major contribution to malacology in his ultrastructural studies on bivalves (Montgomery et al., 2002). One of us (E.B.A.), as Graham's student (and wife) owes her choice of subject and her making as a malacologist to him, but was also a beneficiary of Owen's expertise in electron microscopy. This paper is dedicated to the memory of these two men who contributed much to Zoology and to academe at large in their long careers. E.B.A. is entirely responsible for the writing of this paper, but K.E.T. brought great patience and skill to the practical work on the gland of Leiblein of Nucella. As a former student both at Royal Holloway and the University College of Wales, Aberystwyth, she represents yet another generation of zoologists who has reaped benefit from their endeavours. Open in new tabDownload slide Figure 1. Based on drawings by Alastair Graham. A. Dorsal view of the junction between the anterior and mid-oesophagus of Nucella lapillus, cut open along the morphologically median dorsal line, with the dorsal folds twisting round the right wall of the pear-shaped valve of Leiblein from a ventral to a topographically dorsal position. The circular fold (the one-way valve formed by extension of the dorsal folds) lies at its anterior end. The long cilia on the fold are omitted; in life they extend to the posterior limit of the valve. The lines labelled BB indicate the position of the TS shown in B. B. Transverse section through the valve of Leiblein showing the circular fold in the open position at its anterior end. The large arrow indicates the drop in level of the ventral groove, the small arrows the ciliary currents. C. The mid-oesophagus and duct of the gland cut open as above. Part of the glande framboisée has been cut away along the duct and one part of the circular fold in the valve has been upturned to show the mucoid pad. Curved arrows at base of glande framboisée mark the point at which the mid-oesophagus turns through 90° to run dorsally. D. Sagittal section through the valve of Leiblein showing its pear-shape. Curved arrow indicates site of torsion. The lines labelled BB indicate the position of the TS in B. E. Dorsal view of a horizontal section through the valve of Leiblein to show the ventral part in which the thickened glandular part of the wall constricts the lumen. Small arrows indicate direction of ciliary currents. Abbreviations: A (arrowed), anterior; aoe, anterior oesophagus; cf, circular fold; d, duct of the gland of Leiblein; df, dorsal folds; ge, thickened part of the wall of the valve of Leiblein covered with glandular epithelium; glf, glande framboisée; L, left; lc, long languidly beating cilia on edge of circular fold (full length not shown); ldf, left dorsal fold; lf, longitudinal folds; moe, mid-oesophagus; mp, mucoid pad; nr, position of nerve ring; poe, posterior oesophagus; rdf, right dorsal fold; udf, upturned end of dorsal fold; V, morphological ventral; vf, double ventral fold; vg, ventral groove; vL, valve of Leiblein. Open in new tabDownload slide Figure 1. Based on drawings by Alastair Graham. A. Dorsal view of the junction between the anterior and mid-oesophagus of Nucella lapillus, cut open along the morphologically median dorsal line, with the dorsal folds twisting round the right wall of the pear-shaped valve of Leiblein from a ventral to a topographically dorsal position. The circular fold (the one-way valve formed by extension of the dorsal folds) lies at its anterior end. The long cilia on the fold are omitted; in life they extend to the posterior limit of the valve. The lines labelled BB indicate the position of the TS shown in B. B. Transverse section through the valve of Leiblein showing the circular fold in the open position at its anterior end. The large arrow indicates the drop in level of the ventral groove, the small arrows the ciliary currents. C. The mid-oesophagus and duct of the gland cut open as above. Part of the glande framboisée has been cut away along the duct and one part of the circular fold in the valve has been upturned to show the mucoid pad. Curved arrows at base of glande framboisée mark the point at which the mid-oesophagus turns through 90° to run dorsally. D. Sagittal section through the valve of Leiblein showing its pear-shape. Curved arrow indicates site of torsion. The lines labelled BB indicate the position of the TS in B. E. Dorsal view of a horizontal section through the valve of Leiblein to show the ventral part in which the thickened glandular part of the wall constricts the lumen. Small arrows indicate direction of ciliary currents. Abbreviations: A (arrowed), anterior; aoe, anterior oesophagus; cf, circular fold; d, duct of the gland of Leiblein; df, dorsal folds; ge, thickened part of the wall of the valve of Leiblein covered with glandular epithelium; glf, glande framboisée; L, left; lc, long languidly beating cilia on edge of circular fold (full length not shown); ldf, left dorsal fold; lf, longitudinal folds; moe, mid-oesophagus; mp, mucoid pad; nr, position of nerve ring; poe, posterior oesophagus; rdf, right dorsal fold; udf, upturned end of dorsal fold; V, morphological ventral; vf, double ventral fold; vg, ventral groove; vL, valve of Leiblein. Open in new tabDownload slide Figure 2. A. Diagram of the gland of Leiblein of Nucella lapillus slightly straightened and drawn from the left side. Based on the scanning electron micrograph (SEM) in Figure 5A. Curved arrows indicate regions of torsion, right to left at the level of the anterior fissure, left to right at the posterior fissure. Arrows above figure indicate the levels of transverse sections shown in the figures specified. B. SEM of a postero-dorsal view of the duct of the gland; the oesophagus is cut at its junction with the duct and the gland is opened to show the inner opening of the duct. Scale bar=500 µm. Abbreviations: A (arrowed), anterior; aa, anterior aorta; af, anterior fissure; al, anterior lobe; ct, connective tissue; ctf, strands of connective tissue; d, duct of the gland; em, extrinsic muscles; glf, glande framboisée; L, left; moe, mid-oesophagus; ml, middle lobe; o, opening of subepithelial blood space into haemocoel; od, opening of duct; pf, posterior fissure; pl, posterior lobe; poe, posterior oesophagus; spn, supra-oesophageal connective of the visceral loop. Open in new tabDownload slide Figure 2. A. Diagram of the gland of Leiblein of Nucella lapillus slightly straightened and drawn from the left side. Based on the scanning electron micrograph (SEM) in Figure 5A. Curved arrows indicate regions of torsion, right to left at the level of the anterior fissure, left to right at the posterior fissure. Arrows above figure indicate the levels of transverse sections shown in the figures specified. B. SEM of a postero-dorsal view of the duct of the gland; the oesophagus is cut at its junction with the duct and the gland is opened to show the inner opening of the duct. Scale bar=500 µm. Abbreviations: A (arrowed), anterior; aa, anterior aorta; af, anterior fissure; al, anterior lobe; ct, connective tissue; ctf, strands of connective tissue; d, duct of the gland; em, extrinsic muscles; glf, glande framboisée; L, left; moe, mid-oesophagus; ml, middle lobe; o, opening of subepithelial blood space into haemocoel; od, opening of duct; pf, posterior fissure; pl, posterior lobe; poe, posterior oesophagus; spn, supra-oesophageal connective of the visceral loop. Open in new tabDownload slide Figure 3. A. Diagram of the gland of Leiblein of Nucella lapillus from the left side, showing its duct, and the course of the anterior aorta twisting round it (dashed lines where it lies to the right of the gland). The outline is based on the contracted specimen in Figure 5C. The lumen is shown in the anterior lobe only. B. SEM of a transverse section (TS) just anterior to the opening of the siphon into the anterior lobe. Scale bar=750 µm. C. Section through the double ventral fold showing the direction of its ciliary currents (arrowed). The blood space is a branch of the afferent artery. Scale bar=50 µm. Abbreviations: A (arrowed), anterior; aa, anterior aorta; af, anterior fissure; afa, afferent artery; al, anterior lobe; bs, blood space; c, cilia; d, duct; df, dorsal fold; em, extrinsic muscles; g, groove; L, left; l, lumen; ml, middle lobe; pf, posterior fissure; pl, posterior lobe; si, siphon; TS d, transverse section of duct; vf, double ventral fold. Open in new tabDownload slide Figure 3. A. Diagram of the gland of Leiblein of Nucella lapillus from the left side, showing its duct, and the course of the anterior aorta twisting round it (dashed lines where it lies to the right of the gland). The outline is based on the contracted specimen in Figure 5C. The lumen is shown in the anterior lobe only. B. SEM of a transverse section (TS) just anterior to the opening of the siphon into the anterior lobe. Scale bar=750 µm. C. Section through the double ventral fold showing the direction of its ciliary currents (arrowed). The blood space is a branch of the afferent artery. Scale bar=50 µm. Abbreviations: A (arrowed), anterior; aa, anterior aorta; af, anterior fissure; afa, afferent artery; al, anterior lobe; bs, blood space; c, cilia; d, duct; df, dorsal fold; em, extrinsic muscles; g, groove; L, left; l, lumen; ml, middle lobe; pf, posterior fissure; pl, posterior lobe; si, siphon; TS d, transverse section of duct; vf, double ventral fold. Open in new tabDownload slide Figure 4. Dorsal view of the gland of Leiblein of Nucella lapillus cut open to display the folds. The posterior oesophagus and posterior part of the anterior aorta are displaced to the left. The curved black arrow indicates the efferent current from the gland. Small arrows show direction of ciliary currents over the crests of the folds. Open arrow indicates where the anterior aorta turns to lie beneath the anterior lobe. Not drawn to scale. Arrows to the right of the figure indicate the levels of the transverse sections shown in the figures specified. Abbreviations: A (arrowed), anterior; aa, anterior aorta; afa, afferent artery to the gland; al, anterior lobe; d, duct of the gland; f, vascular fold; glf, glande framboisée; l, lumen; ml, middle lobe; pl, posterior lobe; poe, posterior oesophagus; vf, ventral fold. Open in new tabDownload slide Figure 4. Dorsal view of the gland of Leiblein of Nucella lapillus cut open to display the folds. The posterior oesophagus and posterior part of the anterior aorta are displaced to the left. The curved black arrow indicates the efferent current from the gland. Small arrows show direction of ciliary currents over the crests of the folds. Open arrow indicates where the anterior aorta turns to lie beneath the anterior lobe. Not drawn to scale. Arrows to the right of the figure indicate the levels of the transverse sections shown in the figures specified. Abbreviations: A (arrowed), anterior; aa, anterior aorta; afa, afferent artery to the gland; al, anterior lobe; d, duct of the gland; f, vascular fold; glf, glande framboisée; l, lumen; ml, middle lobe; pl, posterior lobe; poe, posterior oesophagus; vf, ventral fold. Open in new tabDownload slide Figure 5. SEMs illustrating the gross structure of the gland of Leiblein of Nucella lapillus. The proboscis was retracted in all specimens. Large solid straight arrows indicate anterior. A. The left side of the gland from a relaxed snail, showing the mid- and posterior oesophagus and anterior aorta. Small arrows indicate anterior and posterior fissures;×marks the bend at the base of the glande framboisée. B. Dorsal view of the posterior lobe of the gland with the dorsal wall removed, in a relaxed snail. Curved arrow indicates opening to middle lobe;×marks the change in level at the junction of middle and posterior lobes where the ventral fold drops ventrally and turns to the left. C. Sagittal section of the gland from a contracted snail, viewed from the right. Small arrows indicate direction of ciliary currents. Inset: detail of the opening of the duct into the gland, showing the double ventral fold. D. The ventral fold at the anterior end of the posterior lobe descending into the middle lobe and turning to the left (curved arrows). Inset: detail of cilia on the ventral fold, with stream of mucus and spherules expelled by the cells. Small arrows indicate direction of ciliary currents. E. Dorsal view of a horizontal section through the gland from a contracted specimen, showing the three lobes and the duct entering the gland. F. Dorso-lateral view of the point at which the anterior aorta crosses the gland from left to right. Arrow indicates connection between posterior and middle lobes. Abbreviations: aa, anterior aorta; af, anterior fissure; afa, afferent artery of the gland; al, anterior lobe; aoe, displaced anterior oesophagus; bw, body wall; d, duct; f, vascular fold; glf, glande framboisée; ml, middle lobe; nr, nerve ring; pf, posterior fissure; pl, posterior lobe; poe, posterior oesophagus; s, spherule in mucous string; se, string of secretion entering duct; v, vestibule; vc, channel on floor of anterior lobe from siphon; vf, double ventral fold. Scale bars: A=1.2 mm; B=0. 5 mm; C=1.15 mm, inset=100µm; D=200 µm, inset=50 µm; E=1.5 mm; F=500 µm. Open in new tabDownload slide Figure 5. SEMs illustrating the gross structure of the gland of Leiblein of Nucella lapillus. The proboscis was retracted in all specimens. Large solid straight arrows indicate anterior. A. The left side of the gland from a relaxed snail, showing the mid- and posterior oesophagus and anterior aorta. Small arrows indicate anterior and posterior fissures;×marks the bend at the base of the glande framboisée. B. Dorsal view of the posterior lobe of the gland with the dorsal wall removed, in a relaxed snail. Curved arrow indicates opening to middle lobe;×marks the change in level at the junction of middle and posterior lobes where the ventral fold drops ventrally and turns to the left. C. Sagittal section of the gland from a contracted snail, viewed from the right. Small arrows indicate direction of ciliary currents. Inset: detail of the opening of the duct into the gland, showing the double ventral fold. D. The ventral fold at the anterior end of the posterior lobe descending into the middle lobe and turning to the left (curved arrows). Inset: detail of cilia on the ventral fold, with stream of mucus and spherules expelled by the cells. Small arrows indicate direction of ciliary currents. E. Dorsal view of a horizontal section through the gland from a contracted specimen, showing the three lobes and the duct entering the gland. F. Dorso-lateral view of the point at which the anterior aorta crosses the gland from left to right. Arrow indicates connection between posterior and middle lobes. Abbreviations: aa, anterior aorta; af, anterior fissure; afa, afferent artery of the gland; al, anterior lobe; aoe, displaced anterior oesophagus; bw, body wall; d, duct; f, vascular fold; glf, glande framboisée; ml, middle lobe; nr, nerve ring; pf, posterior fissure; pl, posterior lobe; poe, posterior oesophagus; s, spherule in mucous string; se, string of secretion entering duct; v, vestibule; vc, channel on floor of anterior lobe from siphon; vf, double ventral fold. Scale bars: A=1.2 mm; B=0. 5 mm; C=1.15 mm, inset=100µm; D=200 µm, inset=50 µm; E=1.5 mm; F=500 µm. Open in new tabDownload slide Figure 6. SEMs of posterior faces of transverse sections of gland of Leiblein of Nucella lapillus at levels indicated in Figures 2 and 4. A. TS of posterior part of the anterior lobe through the anterior fissure and duct. B. TS through the posterior fissure in which the anterior aorta crosses the gland from left to right. Abbreviations: aa, anterior aorta; afa, afferent artery; al, anterior lobe; d, duct; em, extrinsic muscles; glf, glande framboisée; L, left; l, lumen; ml, middle lobe; pf, posterior fissure; poe, posterior oesophagus; vf, double ventral fold. Scale bars=375 µm. Open in new tabDownload slide Figure 6. SEMs of posterior faces of transverse sections of gland of Leiblein of Nucella lapillus at levels indicated in Figures 2 and 4. A. TS of posterior part of the anterior lobe through the anterior fissure and duct. B. TS through the posterior fissure in which the anterior aorta crosses the gland from left to right. Abbreviations: aa, anterior aorta; afa, afferent artery; al, anterior lobe; d, duct; em, extrinsic muscles; glf, glande framboisée; L, left; l, lumen; ml, middle lobe; pf, posterior fissure; poe, posterior oesophagus; vf, double ventral fold. Scale bars=375 µm. Open in new tabDownload slide Figure 7. SEM of a transverse section of the posterior lobe of the gland, posterior oesophagus, and anterior aorta in Nucella lapillus, as indicated in Figures 2 and 4. Abbreviations: anterior aorta; L, left; ml, middle lobe; pl, posterior lobe; poe, posterior oesophagus. Scale bar=375 µm. Open in new tabDownload slide Figure 7. SEM of a transverse section of the posterior lobe of the gland, posterior oesophagus, and anterior aorta in Nucella lapillus, as indicated in Figures 2 and 4. Abbreviations: anterior aorta; L, left; ml, middle lobe; pl, posterior lobe; poe, posterior oesophagus. Scale bar=375 µm. Open in new tabDownload slide Figure 8. A. Diagram of the left side of the mid-oesophagus and part of the gland of Leiblein of Nucella lapillus to illustrate the proposed mechanism by which liquid and solutes are believed to enter the gland. B. The valve of Leiblein in the open position with the circular fold forming a funnel, allowing food to enter it from the anterior oesophagus. C. The valve in the closed position when the proboscis is extended, resulting in liquid from the glande framboisée being drawn into the valve. Abbreviations: af, anterior fissure; al, anterior lobe; aoe, anterior oesophagus, cf, circular fold bearing long cilia; cg, cerebral ganglion; cpc, cerebro-pleural connective; ct, connective tissue strands; d, duct of the gland of Leiblein; dch, dorsal food channel; em, extrinsic muscle (that on anterior face of glf is omitted); fr, food rod; ge, pad of glandular epithelium; gL gland of Leiblein; glf, glande framboisée; hdf, hypertrophied dorsal folds in glande framboisée; j, junction of the mid- and posterior oesophagus and the duct to the gland; li, liquid; ldf, left dorsal fold; lf, longitudinal fold; ml, middle lobe of gland; mp, mucoid pad secreting cocoon around food string; plg, pleural ganglion; poe, posterior oesophagus; rdf, right dorsal fold; sbg, suboesophageal ganglion; si, siphon; spg, supraoesopohageal ganglion; t1, site of torsion (which results in the dorsal folds twisting up the right wall of the vL); t2, anterior limit of secondary effect of torsion (right to left); v, vestibule; vf, double ventral fold; vg, ventral oesophageal groove; vL, valve of Leiblein. White arrows indicate direction of movement of liquid, large black arrows that of the food string, small black arrows, the direction of ciliary currents. Arrowheads show direction of movement of the semi-circular fold that open and close the valve. Open in new tabDownload slide Figure 8. A. Diagram of the left side of the mid-oesophagus and part of the gland of Leiblein of Nucella lapillus to illustrate the proposed mechanism by which liquid and solutes are believed to enter the gland. B. The valve of Leiblein in the open position with the circular fold forming a funnel, allowing food to enter it from the anterior oesophagus. C. The valve in the closed position when the proboscis is extended, resulting in liquid from the glande framboisée being drawn into the valve. Abbreviations: af, anterior fissure; al, anterior lobe; aoe, anterior oesophagus, cf, circular fold bearing long cilia; cg, cerebral ganglion; cpc, cerebro-pleural connective; ct, connective tissue strands; d, duct of the gland of Leiblein; dch, dorsal food channel; em, extrinsic muscle (that on anterior face of glf is omitted); fr, food rod; ge, pad of glandular epithelium; gL gland of Leiblein; glf, glande framboisée; hdf, hypertrophied dorsal folds in glande framboisée; j, junction of the mid- and posterior oesophagus and the duct to the gland; li, liquid; ldf, left dorsal fold; lf, longitudinal fold; ml, middle lobe of gland; mp, mucoid pad secreting cocoon around food string; plg, pleural ganglion; poe, posterior oesophagus; rdf, right dorsal fold; sbg, suboesophageal ganglion; si, siphon; spg, supraoesopohageal ganglion; t1, site of torsion (which results in the dorsal folds twisting up the right wall of the vL); t2, anterior limit of secondary effect of torsion (right to left); v, vestibule; vf, double ventral fold; vg, ventral oesophageal groove; vL, valve of Leiblein. White arrows indicate direction of movement of liquid, large black arrows that of the food string, small black arrows, the direction of ciliary currents. Arrowheads show direction of movement of the semi-circular fold that open and close the valve. Open in new tabDownload slide Figure 9. Detail of the duct of the gland of Leiblein of Nucella lapillus. The small arrows indicate where the right and left dorsal folds make contact. A. The duct cut obliquely. The outflow channel is filled with secretion. B. The anterior ‘physiologically closed siphon’ beneath the scar on the anterior face of the duct. C. Detail of the collapsed siphon, with longitudinal muscle blocks along the margins of the overlying dorsal folds. D. Photomicrograph of a TS through the siphon and dorsal folds. Abbreviations: dch, dorsal food channel; df, dorsal fold; ge, epithelium of contiguous dorsal folds (glande framboisée); m, longitudinal muscle block; pc, posterior (efferent) channel; sc, ‘scar’; se, secretion largely obscuring ventral folds; si, ‘siphon’; ue, ‘undifferentiated’ epithelium lining siphon; vf, part of double ventral fold. Scale bars: A=200 µm; B=100 µm; C=50 µm; D=50 µm. Open in new tabDownload slide Figure 9. Detail of the duct of the gland of Leiblein of Nucella lapillus. The small arrows indicate where the right and left dorsal folds make contact. A. The duct cut obliquely. The outflow channel is filled with secretion. B. The anterior ‘physiologically closed siphon’ beneath the scar on the anterior face of the duct. C. Detail of the collapsed siphon, with longitudinal muscle blocks along the margins of the overlying dorsal folds. D. Photomicrograph of a TS through the siphon and dorsal folds. Abbreviations: dch, dorsal food channel; df, dorsal fold; ge, epithelium of contiguous dorsal folds (glande framboisée); m, longitudinal muscle block; pc, posterior (efferent) channel; sc, ‘scar’; se, secretion largely obscuring ventral folds; si, ‘siphon’; ue, ‘undifferentiated’ epithelium lining siphon; vf, part of double ventral fold. Scale bars: A=200 µm; B=100 µm; C=50 µm; D=50 µm. Open in new tabDownload slide Figure 10. A–F. SEMs showing surface detail of the cells over the vascular folds of the gland in Nucella lapillus.A. Apical regions of both types of cell in a resting phase. B. LS of both cell-types, showing vacuolated cytoplasm and hard granules (arrowed) in ciliated cells. C. Surface view of both cell-types at an absorptive stage in the cell cycle. D. Surface view of cells towards the end of a phase of intracellular digestion, showing development of spherules for secretion. E. Detail of swollen bases of cilia on the ciliated cells. F. Detail of the apical surface of a ciliated cell showing many small blebs, and part of a single large bleb of an unciliated (dense) cell. G. The surface of the gland, showing openings of sub-epithelial blood spaces to the haemocoel. H. Transverse section through a vascular fold with a bundle of sub-epithelial muscle fibres projecting from a blood space. Abbreviations: b, blebs on apical surface of ciliated cell; cc, ciliated cell; f, luminal surface of vascular fold; mf, muscle fibres; mv, microvilli; ne, neck connecting cell body and mature spherule; o, opening of subepthelial blood space; s, mature spherule of unciliated cell; ss, small bleb on ciliated cell; uc, unciliated cell. Scale bars: A, B=20 µm; C=15 µm; D=10 µm; E, F=6 µm; G=150 µm; H=50 µm. Open in new tabDownload slide Figure 10. A–F. SEMs showing surface detail of the cells over the vascular folds of the gland in Nucella lapillus.A. Apical regions of both types of cell in a resting phase. B. LS of both cell-types, showing vacuolated cytoplasm and hard granules (arrowed) in ciliated cells. C. Surface view of both cell-types at an absorptive stage in the cell cycle. D. Surface view of cells towards the end of a phase of intracellular digestion, showing development of spherules for secretion. E. Detail of swollen bases of cilia on the ciliated cells. F. Detail of the apical surface of a ciliated cell showing many small blebs, and part of a single large bleb of an unciliated (dense) cell. G. The surface of the gland, showing openings of sub-epithelial blood spaces to the haemocoel. H. Transverse section through a vascular fold with a bundle of sub-epithelial muscle fibres projecting from a blood space. Abbreviations: b, blebs on apical surface of ciliated cell; cc, ciliated cell; f, luminal surface of vascular fold; mf, muscle fibres; mv, microvilli; ne, neck connecting cell body and mature spherule; o, opening of subepthelial blood space; s, mature spherule of unciliated cell; ss, small bleb on ciliated cell; uc, unciliated cell. Scale bars: A, B=20 µm; C=15 µm; D=10 µm; E, F=6 µm; G=150 µm; H=50 µm. Open in new tabDownload slide Figure 11. A. SEM showing part of the epithelial surface and section through a vascular fold of the gland in Nucella lapillus.B. Transmission electron micrograph (TEM) of a longitudinal section through the epithelium of the posterior lobe during absorption and intracellular digestion. Arrows indicate hard granules. Routine fixation. Abbreviations: bi, basal infolding; bs, blood space; cc, ciliated cell; ce, cut epithelium of vascular fold; ds, developing spherule of an unciliated epithelial cell at an early stage in its formation; l, lumen; li, lipid; ly, lysosome; mf, muscle fibres; n, nucleus; rb, residual body; uc, unciliated cell. Scale bars=10 µm. Open in new tabDownload slide Figure 11. A. SEM showing part of the epithelial surface and section through a vascular fold of the gland in Nucella lapillus.B. Transmission electron micrograph (TEM) of a longitudinal section through the epithelium of the posterior lobe during absorption and intracellular digestion. Arrows indicate hard granules. Routine fixation. Abbreviations: bi, basal infolding; bs, blood space; cc, ciliated cell; ce, cut epithelium of vascular fold; ds, developing spherule of an unciliated epithelial cell at an early stage in its formation; l, lumen; li, lipid; ly, lysosome; mf, muscle fibres; n, nucleus; rb, residual body; uc, unciliated cell. Scale bars=10 µm. Open in new tabDownload slide Figure 12. Diagrams to show the two types of epithelial cells in the gland of Leiblein of Nucella lapillus at different stages of the cell cycle. A. Receptive phase at the start of feeding. B. Absorptive phase, in which heterophagic vacuoles are conspicuous. C. Near the end of a phase of intracellular digestion, when the apices of the unciliated cells become dome-shaped. D. Resting phase, in which the unciliated cells become club-shaped and the apical spherules are attached by a narrow stalk; they are shed at the start of the next feeding cycle. Abbreviations: bi, basal infolding; bl, basal lamina; cc, ciliated cell; cf, collagen fibres; ci, cilium; cr, ciliary rootlet; dcy, dense cytoplasm; doa, domed apex; fa, flat apex; gb, Golgi body; ger, granular endoplasmic reticulum; l, saturated lipid; la, lamellae in hard pigmented phosphate granule; ly, lysosome; mf, muscle fibre; mi, mitochondrion; n, nucleus; ne, neck; nf, nerve fibre; nfb, nucleus of fibroblast; pcy, pale cytoplasm; pgr, layered pigmented granule; rb, residual body; rmv, regular array of long microvilli; s, mature spherule; ser, smooth endoplasmic reticulum; smv, short microvillus; uc, unciliated cell; usl, unsaturated lipid; v, heterophagic vacuole. Not drawn to scale. Open in new tabDownload slide Figure 12. Diagrams to show the two types of epithelial cells in the gland of Leiblein of Nucella lapillus at different stages of the cell cycle. A. Receptive phase at the start of feeding. B. Absorptive phase, in which heterophagic vacuoles are conspicuous. C. Near the end of a phase of intracellular digestion, when the apices of the unciliated cells become dome-shaped. D. Resting phase, in which the unciliated cells become club-shaped and the apical spherules are attached by a narrow stalk; they are shed at the start of the next feeding cycle. Abbreviations: bi, basal infolding; bl, basal lamina; cc, ciliated cell; cf, collagen fibres; ci, cilium; cr, ciliary rootlet; dcy, dense cytoplasm; doa, domed apex; fa, flat apex; gb, Golgi body; ger, granular endoplasmic reticulum; l, saturated lipid; la, lamellae in hard pigmented phosphate granule; ly, lysosome; mf, muscle fibre; mi, mitochondrion; n, nucleus; ne, neck; nf, nerve fibre; nfb, nucleus of fibroblast; pcy, pale cytoplasm; pgr, layered pigmented granule; rb, residual body; rmv, regular array of long microvilli; s, mature spherule; ser, smooth endoplasmic reticulum; smv, short microvillus; uc, unciliated cell; usl, unsaturated lipid; v, heterophagic vacuole. Not drawn to scale. Open in new tabDownload slide Figure 13. TEM of epithelial cells over the vascular folds of the posterior lobe of the gland in Nucella lapillus, potassium ferrocyanide added to secondary fixative, except E, for which Owen & McCrae's method of fixation was used. A. LS of both types of epithelial cell. B. Apical region of a ciliated cell. C. Apical region of an electron-dense unciliated cell. D. Basal region of a ciliated cell, the cytoplasm of which is stained by potassium ferrocyanide. E. Basal region of an unciliated cell. Abbreviations: bi, basal infolding; bl, basal lamina; c, cilium; cc, ciliated cell; j, septate junction; l, lumen; li, lipid; ly, lysosome; me, melanin in residual body (=hard granule arrowed); mf, muscle fibres; mi, mitochondria; mv, microvilli; n, nucleus; nen, nerve ending; ser, smooth endoplasmic reticulum; uc, unciliated cell. Scale bars: A=5 µm; B, C=1.25 µm; D, E=1 µm. Open in new tabDownload slide Figure 13. TEM of epithelial cells over the vascular folds of the posterior lobe of the gland in Nucella lapillus, potassium ferrocyanide added to secondary fixative, except E, for which Owen & McCrae's method of fixation was used. A. LS of both types of epithelial cell. B. Apical region of a ciliated cell. C. Apical region of an electron-dense unciliated cell. D. Basal region of a ciliated cell, the cytoplasm of which is stained by potassium ferrocyanide. E. Basal region of an unciliated cell. Abbreviations: bi, basal infolding; bl, basal lamina; c, cilium; cc, ciliated cell; j, septate junction; l, lumen; li, lipid; ly, lysosome; me, melanin in residual body (=hard granule arrowed); mf, muscle fibres; mi, mitochondria; mv, microvilli; n, nucleus; nen, nerve ending; ser, smooth endoplasmic reticulum; uc, unciliated cell. Scale bars: A=5 µm; B, C=1.25 µm; D, E=1 µm. Open in new tabDownload slide Figure 14. TEM of the epithelium over the vascular folds of the gland in Nucella lapillus.A. Posterior lobe fixed by routine method. Cells at an early stage of intracellular digestion. B. Posterior lobe cells at a later stage of intracellular digestion, with many secondary lysosomes and lipid droplets. Modified Karnovsky fixative. Inset: detail of apical region of a ciliated cell to show glycocalyx and endocytotic canals (arrowed). C. Detail of basal regions of cells and muscle fibres in blood space of posterior lobe. Owen & McCrae's method of fixation. D. Basal cytoplasm of unciliated cells of anterior lobe showing extensive lipid deposits. Modified Karnovsky's fixative. Abbreviations: bi, basal infoldings; bl, basal lamina; c, cilium; cc, ciliated cell; gl, glycogen; l, lumen; li, lipid droplet; ly, lysosome; mi, mitochondria; mf, muscle fibres; mv, microvilli; n, nucleus; ser, smooth endoplasmic reticulum; uc, unciliated cell. Scale bars: A, B=10 µm; Inset=0.5 µm; C, D=1.25 µm. Open in new tabDownload slide Figure 14. TEM of the epithelium over the vascular folds of the gland in Nucella lapillus.A. Posterior lobe fixed by routine method. Cells at an early stage of intracellular digestion. B. Posterior lobe cells at a later stage of intracellular digestion, with many secondary lysosomes and lipid droplets. Modified Karnovsky fixative. Inset: detail of apical region of a ciliated cell to show glycocalyx and endocytotic canals (arrowed). C. Detail of basal regions of cells and muscle fibres in blood space of posterior lobe. Owen & McCrae's method of fixation. D. Basal cytoplasm of unciliated cells of anterior lobe showing extensive lipid deposits. Modified Karnovsky's fixative. Abbreviations: bi, basal infoldings; bl, basal lamina; c, cilium; cc, ciliated cell; gl, glycogen; l, lumen; li, lipid droplet; ly, lysosome; mi, mitochondria; mf, muscle fibres; mv, microvilli; n, nucleus; ser, smooth endoplasmic reticulum; uc, unciliated cell. Scale bars: A, B=10 µm; Inset=0.5 µm; C, D=1.25 µm. Open in new tabDownload slide Figure 15. Detail of the apical regions of the epithelial cells over the vascular folds of the gland in Nucella lapillus.A. SEM of the surface of both cell-types of the anterior lobe during a phase of absorption. B. Surface view of the epithelium of the middle lobe during a resting phase, in which the unciliated cells are club-shaped and bear mature spherules. They obscure the ciliated cells. C. TEM of the apical regions of both cell-types from the anterior lobe in an absorptive/digestive phase. D. TEM of the apical regions of the epithelial cells of the posterior lobe at a late stage of intracellular digestion. E. TEM of the epithelial cells of the posterior lobe during a resting phase, showing the narrow neck of a club-shaped unciliated cell. C–E Owen & McCrae's fixation method. F. TEM of the apical region of an unciliated cell in a ‘control’ specimen of an experiment to demonstrate the uptake of cadmium. G. TEM of the same area in an ‘experimental’ specimen showing fine electron-dense particles in the organelles (arrowed). F, G were stained in lead citrate only. Abbreviations: c, cilia; ec, endocytotic canal; l, lumen; li, lipid; ly, lysosome; mv, microvilli; n, nucleus; ne, neck between secretory spherule and cell body; pm, plasma membrane; s, spherule. Scale bars: A=10 µm; B=25 µm; C–E=10 µm; F, G=1 µm. Open in new tabDownload slide Figure 15. Detail of the apical regions of the epithelial cells over the vascular folds of the gland in Nucella lapillus.A. SEM of the surface of both cell-types of the anterior lobe during a phase of absorption. B. Surface view of the epithelium of the middle lobe during a resting phase, in which the unciliated cells are club-shaped and bear mature spherules. They obscure the ciliated cells. C. TEM of the apical regions of both cell-types from the anterior lobe in an absorptive/digestive phase. D. TEM of the apical regions of the epithelial cells of the posterior lobe at a late stage of intracellular digestion. E. TEM of the epithelial cells of the posterior lobe during a resting phase, showing the narrow neck of a club-shaped unciliated cell. C–E Owen & McCrae's fixation method. F. TEM of the apical region of an unciliated cell in a ‘control’ specimen of an experiment to demonstrate the uptake of cadmium. G. TEM of the same area in an ‘experimental’ specimen showing fine electron-dense particles in the organelles (arrowed). F, G were stained in lead citrate only. Abbreviations: c, cilia; ec, endocytotic canal; l, lumen; li, lipid; ly, lysosome; mv, microvilli; n, nucleus; ne, neck between secretory spherule and cell body; pm, plasma membrane; s, spherule. Scale bars: A=10 µm; B=25 µm; C–E=10 µm; F, G=1 µm. Open in new tabDownload slide Figure 16. TEM of the cytoplasm and organelles in the two types of epithelial cells over the vascular folds of the gland in Nucella lapillus.A. Microvilli, glycocalyx (black arrow), coated pits and apical canal system of an unciliated cell (white arrow). Modified Karnovsky's fixative. B. High power detail of an unciliated cell with an endocytotic pit. Small white arrows, endocytotic canals; larger arrow, opening of canals into pit. Routine fixation. Inset: detail of glycocalyx on the microvilli of a ciliated cell (black arrow). Routine fixation. C. Golgi body of a ciliated cell, and lysosomes in an adjacent unciliated cell. Routine fixation. D. Golgi body of an unciliated cell, and lipid droplets associated with mitochondria. E. Detail of rough endoplasmic reticulum in an unciliated cell. F. Detail of mitochondria and lipid droplet in an unciliated cell. D–F: Potassium ferrocyanide added to the secondary fixative. Abbreviations: ep, endocytotic pit; Gb, Golgi body; ger, granular endoplasmic reticulum; l, lumen; li, lipid; ly, lysosome; mi, mitochondrion; n, nucleus; ser, smooth endoplasmic reticulum. Scale bars: A=0.5 µm, B=0.2 µm; C, D=1 µm; E, F=0.5 µm. Open in new tabDownload slide Figure 16. TEM of the cytoplasm and organelles in the two types of epithelial cells over the vascular folds of the gland in Nucella lapillus.A. Microvilli, glycocalyx (black arrow), coated pits and apical canal system of an unciliated cell (white arrow). Modified Karnovsky's fixative. B. High power detail of an unciliated cell with an endocytotic pit. Small white arrows, endocytotic canals; larger arrow, opening of canals into pit. Routine fixation. Inset: detail of glycocalyx on the microvilli of a ciliated cell (black arrow). Routine fixation. C. Golgi body of a ciliated cell, and lysosomes in an adjacent unciliated cell. Routine fixation. D. Golgi body of an unciliated cell, and lipid droplets associated with mitochondria. E. Detail of rough endoplasmic reticulum in an unciliated cell. F. Detail of mitochondria and lipid droplet in an unciliated cell. D–F: Potassium ferrocyanide added to the secondary fixative. Abbreviations: ep, endocytotic pit; Gb, Golgi body; ger, granular endoplasmic reticulum; l, lumen; li, lipid; ly, lysosome; mi, mitochondrion; n, nucleus; ser, smooth endoplasmic reticulum. Scale bars: A=0.5 µm, B=0.2 µm; C, D=1 µm; E, F=0.5 µm. Open in new tabDownload slide Figure 17. TEM of the two types of cell in the gland of Leiblein of Hinia reticulata.A, B. ‘Control’ unfed for >5 days. C. Cells from a specimen fixed 1 h after feeding. D. Cells from a specimen fixed 5 h after feeding. Abbreviations: bl, basal lamina; cc, ciliated cell; fr, fragment of disintegrated cell; l, lumen; li, lipid; ly, lysosome; mf, muscle fibres; mi, mitochondria; mv, microvilli; n, nucleus; ne, neck between cell body and mature spherule; s, spherule; ser, smooth endoplasmic reticulum; uc, unciliated cell. Scale bars: A, B=4 µm, C=5 µm, D=4 µm. Open in new tabDownload slide Figure 17. TEM of the two types of cell in the gland of Leiblein of Hinia reticulata.A, B. ‘Control’ unfed for >5 days. C. Cells from a specimen fixed 1 h after feeding. D. Cells from a specimen fixed 5 h after feeding. Abbreviations: bl, basal lamina; cc, ciliated cell; fr, fragment of disintegrated cell; l, lumen; li, lipid; ly, lysosome; mf, muscle fibres; mi, mitochondria; mv, microvilli; n, nucleus; ne, neck between cell body and mature spherule; s, spherule; ser, smooth endoplasmic reticulum; uc, unciliated cell. Scale bars: A, B=4 µm, C=5 µm, D=4 µm. Open in new tabDownload slide Figure 18. TEM of the epithelial cells of the gland of Leiblein of Hinia reticulata.A. Oblique section of part of the epithelium 5 h after feeding. B. Cytoplasm of an unciliated cell with large mass of lipid 5 h after feeding. C. The microvilli and apical cytoplasm of a ciliated cell 1 h after feeding; cf density of microvilli with those in A. D. The two types of cell in a specimen not fed for 10 days, showing the apex of a club-shaped unciliated cell overlying ciliated cells. E. The apical regions of the two types of cell in a specimen not fed for 20 days. Abbreviations: bi, basal infoldings; cc, ciliated cell; Gb, Golgi body; l, lumen; li, lipid; mv, microvilli; ne, neck of unciliated cell in resting stage; s, spherule; uc, unciliated cell. Scale bars: A, B=5 µm, C, D=2 µm, E=4 µm. Open in new tabDownload slide Figure 18. TEM of the epithelial cells of the gland of Leiblein of Hinia reticulata.A. Oblique section of part of the epithelium 5 h after feeding. B. Cytoplasm of an unciliated cell with large mass of lipid 5 h after feeding. C. The microvilli and apical cytoplasm of a ciliated cell 1 h after feeding; cf density of microvilli with those in A. D. The two types of cell in a specimen not fed for 10 days, showing the apex of a club-shaped unciliated cell overlying ciliated cells. E. The apical regions of the two types of cell in a specimen not fed for 20 days. Abbreviations: bi, basal infoldings; cc, ciliated cell; Gb, Golgi body; l, lumen; li, lipid; mv, microvilli; ne, neck of unciliated cell in resting stage; s, spherule; uc, unciliated cell. Scale bars: A, B=5 µm, C, D=2 µm, E=4 µm. Open in new tabDownload slide Figure 19. Diagrammatic representation in TS of some of the variations in the mid-oesophagus of carnivorous caenogastropods, indicating possible homologies. It does not represent a phylogenetic scheme. The figures are based on those in Fretter & Graham (1994) and on relevant papers specified in the text. Modifications may involve: changes in relative proportions of dorsal and ventral parts; hypertrophy or loss of mucous glands on dorsal folds; hypertrophy or loss of ventral mid-oesophageal glands; separation of ventral glands from the dorsal food channel; presence or loss of ventral fold. Torsion is ignored; all examples are drawn with the morphological ventral shown in that position. Abbreviations: aoe, anterior oesophagus; dch, dorsal food channel; df, dorsal fold; gL, gland of Leiblein; gl, glandular crest of dorsal fold; gl, glandular strip, possibly the homologue of the right dorsal fold; L, left; ldf, left dorsal fold; lf, longitudinal folds; lp, lateral pouch; mb, muscular bulb; moe, mid-oesophagus; mogl, mid-oesophageal gland; poe, posterior oesophagus; R, right; rdf, right dorsal fold; si, siphon; vf, ventral ciliated fold; vgl, venom gland. Not drawn to scale. Open in new tabDownload slide Figure 19. Diagrammatic representation in TS of some of the variations in the mid-oesophagus of carnivorous caenogastropods, indicating possible homologies. It does not represent a phylogenetic scheme. The figures are based on those in Fretter & Graham (1994) and on relevant papers specified in the text. Modifications may involve: changes in relative proportions of dorsal and ventral parts; hypertrophy or loss of mucous glands on dorsal folds; hypertrophy or loss of ventral mid-oesophageal glands; separation of ventral glands from the dorsal food channel; presence or loss of ventral fold. Torsion is ignored; all examples are drawn with the morphological ventral shown in that position. Abbreviations: aoe, anterior oesophagus; dch, dorsal food channel; df, dorsal fold; gL, gland of Leiblein; gl, glandular crest of dorsal fold; gl, glandular strip, possibly the homologue of the right dorsal fold; L, left; ldf, left dorsal fold; lf, longitudinal folds; lp, lateral pouch; mb, muscular bulb; moe, mid-oesophagus; mogl, mid-oesophageal gland; poe, posterior oesophagus; R, right; rdf, right dorsal fold; si, siphon; vf, ventral ciliated fold; vgl, venom gland. Not drawn to scale. Table 1. The state of the food on ingestion in families referred to in the text. Diet . Gland of Leiblein . Stomach . Feeders on ‘soup’ or liquid  MURICOIDEA  Muricidae Big Simple & small  Columbariidae Massive Simple & small  Coralliophilidae Massive Simple & small  Costellariidae* Fairly big Simple & small  Olividae Fairly big More complex than above  Volutomitridae‡ Small Complex  Harpidae† Absent Hardly discernible  Marginellidae (some*, one†) Venom gland Simple & small  BUCCINOIDEA  Colubrariidae† Absent Simple & small  CANCELLARIOIDEA† Long strip‡ Hardly discernible Detritivores  BUCCINOIDEA  Nassariidae (some) Reduced Complex Macrophagous feeders  MURICOIDEA  Olivellidae Absent Complex  Mitridae Absent Complex  Volutidae* Small Simple  BUCCINOIDEA  Buccinidae (most) Reduced Complex  Nassariidae Further reduced Complex  Columbellidae Small Complex  CONOIDEA  Turridae, Conidae† Venom gland Simple Diet . Gland of Leiblein . Stomach . Feeders on ‘soup’ or liquid  MURICOIDEA  Muricidae Big Simple & small  Columbariidae Massive Simple & small  Coralliophilidae Massive Simple & small  Costellariidae* Fairly big Simple & small  Olividae Fairly big More complex than above  Volutomitridae‡ Small Complex  Harpidae† Absent Hardly discernible  Marginellidae (some*, one†) Venom gland Simple & small  BUCCINOIDEA  Colubrariidae† Absent Simple & small  CANCELLARIOIDEA† Long strip‡ Hardly discernible Detritivores  BUCCINOIDEA  Nassariidae (some) Reduced Complex Macrophagous feeders  MURICOIDEA  Olivellidae Absent Complex  Mitridae Absent Complex  Volutidae* Small Simple  BUCCINOIDEA  Buccinidae (most) Reduced Complex  Nassariidae Further reduced Complex  Columbellidae Small Complex  CONOIDEA  Turridae, Conidae† Venom gland Simple *Digestion in oesophageal crop. †Digestion external, in anterior end of foregut, or blood-sucking habit. ‡Diet and feeding habits or homology unknown but deduced. Open in new tab Table 1. The state of the food on ingestion in families referred to in the text. Diet . Gland of Leiblein . Stomach . Feeders on ‘soup’ or liquid  MURICOIDEA  Muricidae Big Simple & small  Columbariidae Massive Simple & small  Coralliophilidae Massive Simple & small  Costellariidae* Fairly big Simple & small  Olividae Fairly big More complex than above  Volutomitridae‡ Small Complex  Harpidae† Absent Hardly discernible  Marginellidae (some*, one†) Venom gland Simple & small  BUCCINOIDEA  Colubrariidae† Absent Simple & small  CANCELLARIOIDEA† Long strip‡ Hardly discernible Detritivores  BUCCINOIDEA  Nassariidae (some) Reduced Complex Macrophagous feeders  MURICOIDEA  Olivellidae Absent Complex  Mitridae Absent Complex  Volutidae* Small Simple  BUCCINOIDEA  Buccinidae (most) Reduced Complex  Nassariidae Further reduced Complex  Columbellidae Small Complex  CONOIDEA  Turridae, Conidae† Venom gland Simple Diet . Gland of Leiblein . Stomach . Feeders on ‘soup’ or liquid  MURICOIDEA  Muricidae Big Simple & small  Columbariidae Massive Simple & small  Coralliophilidae Massive Simple & small  Costellariidae* Fairly big Simple & small  Olividae Fairly big More complex than above  Volutomitridae‡ Small Complex  Harpidae† Absent Hardly discernible  Marginellidae (some*, one†) Venom gland Simple & small  BUCCINOIDEA  Colubrariidae† Absent Simple & small  CANCELLARIOIDEA† Long strip‡ Hardly discernible Detritivores  BUCCINOIDEA  Nassariidae (some) Reduced Complex Macrophagous feeders  MURICOIDEA  Olivellidae Absent Complex  Mitridae Absent Complex  Volutidae* Small Simple  BUCCINOIDEA  Buccinidae (most) Reduced Complex  Nassariidae Further reduced Complex  Columbellidae Small Complex  CONOIDEA  Turridae, Conidae† Venom gland Simple *Digestion in oesophageal crop. †Digestion external, in anterior end of foregut, or blood-sucking habit. ‡Diet and feeding habits or homology unknown but deduced. Open in new tab References ABRO, A.M. 1969 . Studies in the functional anatomy of some British monotocardian veligers, with observations on the musculature of the buccal mass of Littorina littorea (L.). PhD thesis, University of Reading. AMAUDRUT, A. 1898 . La partie antérieure du tube digestif et la torsion chez les mollusques gastéropodes. Annales des Sciences Naturelles Zoologie , 8 : 1 –291. ANDREWS, E.B. 1991 . The fine structure and function of the salivary glands of Nucella lapillus (Gastropoda: Muricidae). Journal of Molluscan Studies , 57 : 111 –126. ANDREWS, E.B. & ALLEN, J.A. 2002 . Alastair Graham. Biographical Memoirs of the Royal Society , 48 : 129 –150. ANDREWS, E.B., PAGE, A.M. & TAYLOR, J.D. 1999 . The fine structure and function of the anterior foregut glands of Cymatium intermedius (Cassoidea: Ranellidae). Journal of Molluscan Studies , 65 : 1 –19. BALL, A.D. 1994 . The ontogeny of the foregut of Nucella lapillus. PhD thesis, Royal Holloway (University of London). BALL, A.D., ANDREWS, E.B. & TAYLOR, J.D. 1997 . The ontogeny of the pleurembolic proboscis in Nucella lapillus (Gastropoda: Muricidae). Journal of Molluscan Studies , 63 : 87 –99. BARKA, T. & ANDERSON, P.J. 1962 . Histochemical methods for acid phosphatase using hexazonium pararosanilin as coupler. Journal of Histochemistry and Cytochemistry , 10 : 741 –753. BOGHEN, A. & FARLEY, J. 1974 . Phasic activity in the digestive gland cells of the intertidal prosobranch, Littorina saxatilis (Olivi) and its relation to the tidal cycle. Proceedings of the Malacological Society of London , 41 : 41 –56. BOLOGNANI FANTIN, A.M., OTTAVIANI, E. & BOLOGNANI, L. 1977 . Primi dati istochimici e biochimici sulla ghiandola di Leiblein di Murex trunculus. Rivista di Istochimica normale e patalogica , 21 : 227 . BOUCHET, P. & KANTOR, Y.I. 2004 . New Caledonia: the major centre of biodiversity for volutomitrid molluscs (Mollusca: Neogastropoda: Volutomitridae). Systematics and Biodiversity , 1 : 467 –502. BOUCHET, P. & PERRINE, D. 1996 . More gastropods feeding at night on parrotfishes. Bulletin of Marine Science , 59 : 224 –228. BROCK, F. 1936 . Suche, Aufnahme und enzymatische Spaltung der Nahrung durch die Wellhornschnecke Buccinum undatum L. Zoologica, Stuttgart , 34 : 1 –136. BROWN, F.C. 1969 . The structure and function of the digestive system of the mud snail Nassarius obsoletus (Say). Malacologia, 9 : 447 –500. BUSH, M.S. 1986 . The ultrastructure and function of the gut of Patella vulgata. PhD thesis, Bedford College, University of London. BUSH, M.S. 1988 . The ultrastructure and function of the intestine of Patella vulgata. Journal of Zoology, London , 215 : 685 –702. BUSH, M.S. 1989 . The ultrastructure and function of the oesophagus of Patella vulgata Linnaeus. Journal of Molluscan Studies , 55 : 111 –124. CROTHERS, J.H. 1985 . Dog-whelks: an introduction to the biology of Nucella lapillus (L.). Field Studies , 6 : 291 –360. DAY, J.A. 1969 . Feeding of the cymatiid gastropod Argobuccinum argus in relation to the structure of the proboscis and secretions of the proboscis gland. American Zoologist , 9 : 909 –916. DIMITRIADIS, V.K. & ANDREWS, E.B. 2000 . Ultrastructural and cytochemical study of the digestive gland cells of the marine prosobranch mollusc Nucella lapillus (L.) in relation to function. Malacologia , 42 : 103 –112. ENDEAN, R. 1972 . Aspects of molluscan pharmacology. In: Chemical Zoology , 7 . Mollusca (M. Florkin & B.T. Scheer, eds), 421–466. Academic Press, New York. FORESTER, A.J. 1977 . The function of the intestine in the pulmonate mollusc, Helix pomatia L. Experientia , 33 : 465 –467. FRANC, A. 1952 . Notes éthologiques et anatomiques sur Tritonalia (Ocinebrina) aciculata (Lk) (Mollusque prosobranche). Bulletin de Laboratoire Maritime Dinard , 36 : 31 –34. FRETTER, V. 1951 . Observations on the life history and functional morphology of Cerithiopsis tubercularis (Montagu) and Triphora perversa (L.). Journal of the Marine Biological Association of the United Kingdom , 29 : 567 –586. FRETTER, V. 1976 . The anatomy and feeding of the volutacean prosobranch Volvarina taeniolata Mörch. Journal of Molluscan Studies , 42 : 327 –336. FRETTER, V. 1984 . The functional anatomy of the neritacean limpet Phenacolepas omanensis Biggs and some comparison with Septaria. Journal of Molluscan Studies , 50 : 8 –18. FRETTER, V. & GRAHAM, A. 1962 . British prosobranch molluscs. Ray Society, London. FRETTER, V. & GRAHAM, A. 1994 . British prosobranch molluscs. Revised and updated edition. Ray Society, London. FRETTER, V., GRAHAM, A., PONDER, W.F. & LINDBERG, D. 1998 . Prosobranchia Introduction. In: Mollusca: The Southern Synthesis. Fauna of Australia , 5 (P.L. Beesley, G.J.B. Ross & A. Wells, eds), 605–638. CSIRO Publishing, Melbourne. GIBBS, P.E., NOTT, J.A., NICOLAIDOU, A. & BEBIANNO, M.J. 1998 . The composition of phosphate granules in the digestive glands of marine prosobranch gastropods: variation in relation to taxonomy. Journal of Molluscan Studies , 64 : 423 –433. GLAUERT, A.M. & LEWIS, P.R. 1998 . Biological specimen preparation for transmission electron microscopy. Practical methods in Electron Microscopy , 17 (A.M. Glauert, ed.). Princeton, New Jersey. GRAHAM, A. 1941 . The oesophagus of the stenoglossan prosobranchs. Proceedings of the Royal Society of Edinburgh, Series B , 61 : 1 –23. GRAHAM, A. 1949 . The molluscan stomach. Transactions of the Royal Society of Edinburgh , 61 : 737 –778. GRAHAM, A. 1966 . The fore-gut of some marginellid and cancellariid prosobranchs. Studies in Tropical Oceanography , 4 : 134 –151. HARASEWYCH, M.G. & PETIT, R.E. 1998 . Cancellarioidea. In: Mollusca: The Southern Synthesis. Fauna of Australia , 5 (P.L. Beesley, G.J.B. Ross & A. Wells, eds), 845–846. CSIRO Publishing, Melbourne. HASZPRUNAR, G. 1988 . On the origin and evolution of major gastropod groups, with special reference to the Streptoneura. Journal of Molluscan Studies , 54 : 367 –441. HICKMAN, C.S. & LIPPS, J.H. 1983 . Foraminifery: selective ingestion of Foraminifera and test alterations produced by the neogastropod Olivella. Journal of Foraminiferal Research , 13 : 108 –124. HIRSCH, G.C. 1915 . Die Ernährungsbiologie fleischfressender Gastropoden (Murex, Natica, Pterotrachea, Pleurobranchaea, Tritonium). Zoologischer Jahrbücher Abteilung Allgemeine Zoologie und Physiologie , 35 : 357 –504. HOUBRICK, R.S. & FRETTER, V. 1969 . Some aspects of the functional anatomy and biology of Cymatium and Bursa. Proceedings of the Malacological Society of London , 38 : 415 –429. HUGHES, R.N. 1986 . A functional biology of marine gastropods. Croom Helm, London. HUGHES, R.N. & EMERSON, W.K. 1987 . Anatomical and taxonomic characteristics of Harpa and Morum (Neogastropoda: Harpidae). Veliger , 29 : 349 –358. HUGHES, R.N. & HUGHES, H.P.I. 1981 . Morphological and behavioural aspects of feeding in the Cassidae (Tonnacea, Mesogastropoda). Malacologia , 20 : 385 –402. KANTOR, Y.I. 1988 . On the anatomy of Pseudomelatominae (Gastropoda, Toxoglossa, Turridae) with notes on functional morphology and phylogeny of the subfamily. Apex , 3 : 1 –19. KANTOR, Y.I. 1991 . On the morphology and relationships of some oliviform gastropods. Ruthenica , 1 : 17 –52. KANTOR, Y.I. 2002 . Morphological prerequisites for understanding neogastropod phylogeny. Bolletino Malacologico , Suppl. 4: 161 –174. KANTOR, Y.I. 2003 . Comparative anatomy of the stomach of Buccinoidea (Neogastropoda). Journal of Molluscan Studies , 69 : 203 –220. KANTOR, Y.I. & HARASEWYCH, M.G. 1992 . Morphology of the digestive system of Volutomitra alaskana Dall, 1902 (Gastropoda, Pectinibranchia, Volutomitridae), with notes on the possible mechanism of feeding. Ruthenica , 2 : 45 –53. KANTOR, Y.I. & TAYLOR, J.D. 2002 . Foregut anatomy and relationships of raphitomine gastropods (Gastropoda: Conoidea: Raphitominae). Bolletino Malacologico , Suppl. 4: 83 –110. KARNOVSKY, M.J. 1965 . A formaldehyde-glutaraldehyde fixative of high osmolarity for use in electron microsocopy. Journal of Cell Biology , 27 : 137A. LAU, D.C.P. & LEUNG, K.M.Y. 2004 . Feeding physiology of the carnivorous gastropod Thais clavigera (Kuster): do they eat ‘soup’? Journal of Experimental Marine Biology and Ecology , 312 : 43 –66. LEUNG, K.M.Y & FURNESS, R.W. 1999 . Induction of metallothionein in dogwhelk Nucella lapillus during and after exposure to cadmium. Ecotoxicology and Environmental Safety , 43 : 156 –164. LEWIS, P.R. & KNIGHT, D.P. 1992 . Cytochemical staining methods for electron microscopy. In: Practical methods in electron microscopy , 14 (A.M. Glauert, ed.), 174–178. Elsevier, Amsterdam. MAES, V.O. & RAEIHLE, D. 1975 . Systematics and biology of Thala floridana (Gastropoda: Vexillidae). Malacologia , 15 : 43 –67. MANSOUR-BEK, J.J. 1934 . Über die proteolytischen Enzyme von Murex anguliferus Lamk. Zeitschrift für vergleichende Physiologie , 20 : 343 –369. MARCUS, E. & MARCUS, E. 1959 . Studies on ‘Olividae’. Boletins de Faculdade de Filosofia, Ciências e Letras, Universidade de Sâo Paulo, Zoologia , 22 : 99 –188. MARCUS, E. & MARCUS, E. 1962 . Studies on Columbellidae. Boletins de Faculdade de Filosofia, Ciências e Letras, Universidae de Sâo Paulo, Zoologia , 24 : 335 –384. MARTOJA, M. 1964 . Contribution á l'étude de l'appareil digestif et la digestion chez les gastéropodes carnivores de la famille Nassaridés. Cellule , 64 : 237 –334. MARTOJA, M. 1971 . Données histologiques sur les glandes salivaires et oesophagiennes de Thais lapillus (L.) (=Nucella lapillus. Prosobranche néogasteropode). Archives de Zoologie Experimentale et Génerale , 112 : 249 –291. MASON, A.Z. & NOTT, J.A. 1981 . The role of intracellular biomineralized granules in the regulation and detoxification of metals in gastropods with special reference to the marine prosobranch L. littorea (L.). Aquatic Toxicology , 1 : 239 –256. Mc LEAN, N. 1971 . On the function of the digestive gland in Nassarius. Veliger , 13 : 273 –274. MERDSOY, B. & FARLEY, J. 1973 . Phasic activity in the digestive cells of the marine prosobranch gastropod, Littorina littorea (L.). Proceedings of the Malacological Society of London , 40 : 473 –480. MINNITI, F. 1986 . Morphological and histochemical study of pharynx of Leiblein, salivary glands and gland of Leiblein in the carnivorous Gastropoda Amyclina tinei Maravigna and Cyclope neritea Lamarck (Nassariidae: Prosobranchia Stenoglossa). Zoologischer Anzeiger , 217 : 14 –22. MONTGOMERY, I., GOTTO, V., ROBERTS, D. & HALTON, D. 2002 . Gareth Owen CBE, DSc, MRIA, FIBiol. 1922–2002. Bulletin of the Malacological Society of London , 39 : 3 . MORTON, B. 1990 . Prey capture, preference and consumption by Linatella caudata (Gastropoda: Tonnoidea: Ranellidae) in Hong Kong. Journal of Molluscan Studies , 56 : 477 –486. MORTON, B. 1991 . Aspects of predation by Tonna zonatum (Prosobranchia: Tonnoidea) feeding on holothurians in Hong Kong. Journal of Molluscan Studies , 57 : 11 –20. MORTON, J.E. 1956 . The tidal rhythm and action of the digestive system of the lamellibranch Lasea rubra. Journal of the Marine Biological Association of the United Kingdom , 35 : 563 –586. MORTON, J.E. 1960 . The habits of Cyclope neritea, a style-bearing stenoglossan gastropod. Proceedings of the Malacological Society of London , 34 : 96 –105. NELSON, L. & MORTON, J.E. 1979 . Cyclic activity and epithelial renewal in the digestive gland tubules of the marine prosobranch Maoricrypta monoxyla (Lesson). Journal of Molluscan Studies , 45 : 262 –283. O'SULLIVAN, J.B., McCONNAUGHEY, R.R. & HUBER, M.E. 1997 . A blood-sucking snail: the Cooper's nutmeg, Cancellaria cooperi Gabb, parasitises the California electric ray, Torpedo californica Ayres. Biological Bulletin , 172 : 362 –366. OWEN, G. 1972 . Lysosomes, peroxisomes and bivalves. Science Progress, Oxford , 60 : 299 –318. OWEN, G. 1973 . The fine structure and histochemistry of the digestive diverticula of the protobanchiate bivalve Nucula sulcata. Proceedings of the Royal Society of London, Series B , 183 : 249 –264. OWEN, G. & McCRAE, J.M. 1976 . Further studies on the latero-frontal tracts of bivalves. Proceedings of the Royal Society of London, Series B , 194 : 527 –544. PAGE, L.R & PEDERSON, R.V.K. 1998 . Transformation of phytoplanktiferous larvae into predatory carnivores during the development of Polinices lewisii (Mollusca, Caenogastropoda). Invertebrate Biology , 117 : 208 –220. PAYNE, C.M. & CRISP, M. 1989 . Ultrastructure and histochemistry of the posterior oesophagus of Nassarius obsoletus (Linnaeus). Journal of Molluscan Studies , 55 : 313 –321. PIÉRON, H. 1933 . Notes éthologiques sur les gastéropodes perceurs et leur comportement avec l'utilisation de méthodes statistiques. Archives de Zoologie Experimentale et Génerale , 75 : 1 –20. PONDER, W.F. 1970 . Some aspects of the morphology of four species of the neogastropod family Marginellidae with a discussion on the evolution of the toxoglossan poison gland. Journal of the Malacological Society of Australia , 2 : 55 –81. PONDER, W.F. 1970 . The morphology of Alcithoe arabica (Gastropoda: Volutidae). Malacological Review , 3 : 127 –165. PONDER, W.F. 1972 . The morphology of some mitriform gastropods with special reference to their alimentary and reproductive systems (Neogastropoda). Malacologia , 11 : 295 –342. PONDER, W.F. 1973 . The origin and evolution of the Neogastropoda. Malacologia , 12 : 295 –338. PONDER, W.F. 1998 . Volutomitridae. In: Mollusca: The Southern Synthesis. Fauna of Australia , 5 (P.L. Beesley, G.J.B. Ross & A. Wells, eds), 842–843. CSIRO Publishing, Melbourne. PONDER, W.F. & LINDBERG, D.R. 1997 . Towards a phylogeny of gastropod molluscs: an analysis using morphological characters. Zoological Journal of the Linnean Society , 119 : 83 –265. PONDER, W.F. & TAYLOR, J.D. 1992 . Predatory shell drilling by two species of Austroginella (Gastropoda: Marginellidae). Journal of Zoology, London , 228 : 317 –328. REID, R.G.B. & FRIESEN, J.A. 1980 . The digestive system of the moon snail Polinices lewisii (Gould, 1847) with emphasis on the role of the oesophageal gland. Veliger , 23 : 25 –34. RIEDEL, F. 1995 . An outline of cassoidean phylogeny (Mollusca, Gastropoda). Contributions to Tertiary and Quaternary Geology , 32 : 97 –132. RIEDEL, F. 2000 . Ursprung und Evolution der ‘höheren’ Caenogastropoda. Berliner Geowissenschaftliche Abhandlungen , E 32 : 1 –240. ROSENBERG, G., KUNCIO, R.S, DAVIS, G.M. & HARASEWYCH, M.G. 1994 . Ribosomal RNA phylogeny of selected gastropod and unionacean bivalve mollusks. Nautilus , 2 (Suppl. 2): 111 –121. SALVINI-PLAWEN, L. v. 1988 . The structure and function of molluscan digestive systems. In: The Mollusca , 11 : Form and function (E.R. Trueman & M.R. Clarke, eds), 301 –379. Academic Press, London. SANNES, P.L., SPICER, S.S. & KATSUYAMA, T. 1979 . Ultrastructural localization of sulfated and complex carbohydrates with a modified iron diamine procedure. Journal of Histochemistry and Cytochemistry , 27 : 1108 –1111. SEILACHER, A. 1959 . Schnecken im Brandungssand. Natur und Volk , 89 : 539 –566. SHIOMI, K., MIZUKAMI, M., SHIMAKURA, K. & NAGASHIMA, Y. 1994 . Toxins in the salivary glands of some marine carnivorous gastropods. Comparative Biochemistry and Physiology , 107B : 427 –432. SIMKISS, K. & MASON, A.Z. 1983 . Metal ions: metabolic and toxic effects. In: The Mollusca , 2 : Environmental biochemistry and physiology (P.W. Hochachka, ed.), 101 –164. Academic Press, New York. SPICER, S.S., HARDIN, J.H. & SETSER, M.E. 1978 . Ultrastructural visualization of sulfated complex carbohydrates in blood and epithelial cells with the high iron diamine procedure. Histochemical Journal , 10 : 435 –452. STRONG, E.E. 2003 . Refining molluscan characters: morphology, character coding and a phylogeny of the Caenogastropoda. Zoological Journal of the Linnean Society , 137 : 447 –554. TAKAGI, M., PARMLEY, T.R., SPICER, S.S., DENYS, F.R. & SETSER, M.E. 1982 . Ultrastructural localization of acidic glycoconjugates with the low iron diamine method. Journal of Histochemistry and Cytochemistry , 30 : 471 –476. TAYLOR, J.D. 1989 . The diet of coral-reef Mitridae (Gastropoda) from Guam, with a review of other species of the family. Journal of Natural History , 23 : 261 –278. TAYLOR, J.D. 1993 . Dietary and anatomical specialization of mitrid gastropods (Mitridae) at Rottnest island, Western Australia. In: Proceedings of the fifth international marine biological workshop: the marine flora and fauna of Rottnest Island, Western Australia (F.E. Wells, D.I. Walker, H. Kirkman & R. Lethbridge, eds), 583 –599. Western Australian Museum, Perth. TAYLOR, J.D. & GLOVER, E.A. 2000 . Diet of olives: Oliva tigridella Duclos, 1835 in Queensland. Molluscan Research , 20 : 19 –24. TAYLOR, J.D. & MORRIS, N.J. 1988 . Relationships of neogastropods. Malacological Review , Suppl. 4 : 167 –179. TAYLOR, J.D., MORRIS, N.J. & TAYLOR, C.N. 1980 . Food specialization and the evolution of predatory prosobranch gastropods. Palaeontology , 23 : 375 –409. THIÉRY, J.P. 1967 . Mise en évidence des polysaccharides sur coupes fines en microscopie électronique. Journal de Microscopie , 6 : 987 –1018. TRACEY, S., TODD, J.A. & ERWIN, D.H. 1993 . Mollusca: Gastropoda. In: The Fossil Record , 2 (M.J. Benton, ed.), 131 –167. Chapman & Hall, London. WEBER, H. 1927 . Der Darm von Dolium galea (L.), ein vergleichend anatomische Untersuchung unter besonderer Berücksichtigung der Tritonium-Arten. Zeitschrift für Morphologie und Ökologie der Tiere , 8 : 663 –804. WEST, D.J., ANDREWS, E.B., McVEAN, A.R., OSBORNE, D.J. & THORNDYKE, M.C. 1994 . Isolation of serotonin from the accessory salivary glands of the marine snail Nucella lapillus. Toxicon , 32 : 1261 –1264. WEST, D.J., ANDREWS, E.B., BOWMAN, D., McVEAN, A.R & THORNDYKE, M.C. 1996 . Toxins from some poisonous and venomous marine snails. Comparative Biochemistry and Physiology , 113C : 1 –10. © The Author 2005. Published by Oxford University Studies on behalf of The Malacological Society of London, all rights reserved © The Author 2005. Published by Oxford University Studies on behalf of The Malacological Society of London, all rights reserved
TI - AN ULTRASTRUCTURAL STUDY OF THE GLAND OF LEIBLEIN OF MURICID AND NASSARIID NEOGASTROPODS IN RELATION TO FUNCTION, WITH A DISCUSSION ON ITS HOMOLOGIES IN OTHER CAENOGASTROPODS
JF - Journal of Molluscan Studies
DO - 10.1093/mollus/eyi036
DA - 2005-08-01
UR - https://www.deepdyve.com/lp/oxford-university-press/an-ultrastructural-study-of-the-gland-of-leiblein-of-muricid-and-mQba0TuJDc
SP - 269
EP - 300
VL - 71
IS - 3
DP - DeepDyve
ER -