1. Artificial fertilization of the eggs of P. inaequivalvis has been carried out and the development followed to the dissoconch stage.

  2. The development is very rapid and metamorphosis occurs in less than 4 days. The veliger spends less than one day in the plankton and this is the first record of a short-lived planktotrophic lamellibranch larva. The speed of development is related to the relatively large (105– 125 μ diameter) size of the egg and its contained food reserves. There is minimal dispersal of the young from the very localized habitat of the adult.

  3. No protonephridia or larval eyes are developed and the gut follows a simple course in comparison with other lamellibranch larvae. The gills do not develop until after metamorphosis and the large palps and foot produce the ciliary currents connected with respiration, feeding, and cleansing between the loss of the velum and the development of the functional gill.

During an investigation into the functional morphology and habits of Pandora inaequivalis (Allen, 1954; Allen & Allen, 1955), successful artificial fertilization of the eggs was carried out and the larvae reared past metamorphosis to the dissoconch stage. As nothing is known of its development, nor that of any other member of the suborder Anomalodesmata, it seems desirable to describe the development and relate it to the habits of the adult. The adults of P. inaequivalvis, on the Brittany coast, are local in their distribution and occur in sheltered sandy bays where they lie at or very close to the surface of the sand at low-water mark. The problem of maintaining each population would seem to depend on some means of limiting the dispersal of the young. Such a limitation was found to occur.

P. inaequivalvis is a hermaphrodite bivalve, the testis and ovary maturing at about the same time. The eggs and sperm are shed directly into the sea. A supply of sperm and eggs was obtained by removing the left, flat, valve of the mature animal and making a cut through the body-wall; first eggs and then sperm were pipetted off and placed in separate finger-bowls in a small quantity of sea-water. About 200 eggs were transferred to another finger-bowl containing 400 ml. of sea-water and a drop of sperm suspension added. The first attempts at fertilization were unsuccessful, probably because the sperm concentration was too high. Many of the eggs showed polyspermy effects. Grave (1927), reporting similar difficulties in rearing Cumingia, relates polyspermy to the lack of a fertilization membrane. This appears to be absent also in Pandora. Successful fertilizations followed on the addition of smaller quantities of sperm suspension. Nine fertilizations were carried out between 22 July and 27 August 1954. Seven were reared to the trochophore stage and five of these further to the settled dissoconch. The water in the finger-bowls was changed daily. It was not filtered and the organisms in it constituted the entire food of the larvae. Periodically the larvae had to be transferred to new bowls owing to the formation of a bacterial film on the glass surface, to which the larvae adhered and then died. The temperature of the water was maintained at 18° C.±1°, which is approximately the temperature experienced by the larvae in the field. In addition to observations on the living embryos, samples of the various stages were fixed for future examination.

Early cleavage

The unfertilized eggs are spherical (Text-fig. 1A). They range in size from 105 μ to 125 μ in diameter, the majority being 125 μ. This is relatively large for eulamellibranch bivalves (Table 1), whose eggs are usually about 70 μ to 80 μ in diameter. There is a large hemispherical germinal vesicle at the animal pole.

TABLE 1.

The relation of egg diameter (μ) to type of larval life in various Eulamellibranchia (data from Thorsen, 1946)

The relation of egg diameter (μ) to type of larval life in various Eulamellibranchia (data from Thorsen, 1946)
The relation of egg diameter (μ) to type of larval life in various Eulamellibranchia (data from Thorsen, 1946)
TEXT-FIG. 1.

Early cleavage, A, unfertilized egg. B, fertilized egg. c, 2-cell (112 hrs. after fertilization). D, 3-cell (2 hrs.), E, 4-cell (212 hrs.), F, 8-cell (312 hrs.). G, posterior view of 10-cell (4 hrs.), H, anterior view of 11-cell (4 hrs.).

TEXT-FIG. 1.

Early cleavage, A, unfertilized egg. B, fertilized egg. c, 2-cell (112 hrs. after fertilization). D, 3-cell (2 hrs.), E, 4-cell (212 hrs.), F, 8-cell (312 hrs.). G, posterior view of 10-cell (4 hrs.), H, anterior view of 11-cell (4 hrs.).

The cytoplasm is light brown, somewhat opaque and granular, presumably owing to the high food content. This opacity proved to be a handicap in that the cell outlines were difficult to see, as was the internal organization of the later embryos in whole-mount preparations. Soon after the eggs are released a gelatinous envelope forms outside the vitelline membrane. The envelope appears to assist in the prevention of polyspermy (see above) and sperm were often seen in its periphery in eggs undergoing normal development. Loosanoff & Davis (1950) record similar observations in the case of Venus mercenaria. The envelope increases the egg diameter to approximately 300 µ Although the eggs may adhere lightly together they do not fuse to form a gelatinous mass as in Loripes lucinalis taken from the same habitat (Pelseneer, 1926). The eggs are heavier than sea-water and remain at the surface of the substratum.

After fertilization the nucleus remains close to the surface of the egg (Textfig. 1B) and there is a slight increase in the diameter of the egg. The maximum recorded diameter of the fertilized egg, not including the gelatinous envelope, was 137 μ. The first cell-division was seen hours after fertilization, and all eggs had divided at least once after hours. The first division is very unequal, with the CD cell more than 4 times as large as the AB cell (Text-fig. 1c). The egg, no longer spherical, measures 155 μ × 125 μ.1 The 3-cell stages were seen between 2 and 3 hours after fertilization, the CD cell dividing before the AB cell (Textfig. 1D). The division of the CD cell is unequal while the following division of the AB cell is equal. Cells A, B, and C are of the same size (Text-fig. 1E). Cell-division is rapid and most eggs were at the 8-cell stage hours after fertilization, when length and breadth measurements were much like those of the 2-cell stage. The division producing the first 4 micromeres is dexiotropic.

The ‘irregular’ pattern of cleavage, apparent at the 3-cell stage, is continued in later divisions. The quartettes are not produced simultaneously; and while the alternation between dexiotropic and laeotropic division, the basic feature of spiral cleavage, can be observed, it is not clear. Lillie (1895), in his beautiful Study of development in Unio, and Meisenheimer (1901), in thecase of Dreissensia, were able to work out a complete cell lineage for these freshwater bivalves. In this they were helped by the relatively large size of the eggs, transparency of the larvae, and differences in size of the cells that make orientation and identification possible even though their cleavage is irregular (see Lillie, 1895, from p. 9). In the case of marine lamellibranchs cell lineage is extremely difficult to determine and there is no clear account of it in the literature. This is due partly to the atypical cleavage pattern and partly to the similar size of the cells, other than the D cell and its later derivatives (see below). The D cell, being so large, causes displacement of the other cells. In the case of Pandora identification is made even more difficult because of the opacity of the eggs.

The effect of the D cell on the arrangement of the other cells can be seen as early as the 8-cell stage. In addition, cells A, B, and C are little different in size to a, b, c, and d (Text-fig. IF). Following the 8-cell stage the D cell is the first to divide, giving rise to 2D and 2d (Text-fig. 1G). The latter cell, like that of Unio and Dreissensia, is much larger than the 2D macromere and eventually gives rise to the ectoderm of the foot and the shell gland. It has been termed the primary somatoblast (A) (Lillie, 1895). In Pandora 2d (X) remains the largest cell until bilateral symmetry is assumed at about the 40-cell stage. The 2D cell is also larger than the remaining cells. The latter are of much the same size and it was found impossible to label the cells after the 12-cell stage. As in Unio the second quartette is produced before the first quartette of micromeres divide, and they are produced in the order 2d, 2c, and then 2b and 2a together, or nearly so (Text-fig. 1H). Four hours after fertilization 12 to 14 cells are present and at 5 hours 16 to 22. Although irregular in the sense that there is no simultaneous division of the cells of a particular quartette, and in fact the quartettes were not recognized as such, the division of the cells follows a specific and regular pattern so that drawings of different eggs from the same sample with the same number of cells are similar.

Most eggs were at the 32-cell stage 6 hours after fertilization and at this stage a small blastocoele is present (Text-fig. 2A). Subsequently the blastocoele never becomes extensive. The egg is almost spherical and 150 μ in diameter. Unlike Unio, the vitelline membrane is closely applied to the surface of the embryo and individual cells rarely project above the general surface level. As in other lamellibranchs (Raven, 1958), the molluscan cross was not apparent. The first sign of bilateral symmetry was noted 7 hours after fertilization (Text-figs. 2 B, C) when the primary somatoblast1 and primary mesoderm (4d) have divided equally to the right and left. This occurs at about the 40-cell stage. Half an hour later gastrulation commences and takes about hours to complete. Gastrulation is by epiboly (Text-fig. 2D). The gastrula is somewhat pear-shaped and a temporary depression marks the position of the blastopore (Text-fig. 2E). The long axis corresponds to the long axis of the future shell. The position of the future stomadeum is situated in the same plane as the blastopore but is closer to the apical region than the last observed position of the latter. It is highly probable that, owing to the rapid development of the shell gland, the cells of the blastopore region are displaced apically and correspond to the position of the stomadeum. Internally the endoderm cells form a compact and solid mass. At the telotrochal end there is a pair of primary mesoderm cells which in later stages, owing to their dense cytoplasm, stand out clearly from the other cells. Little can be seen of the blastocoele. Gastrulae were first recorded 9 hours after fertilization; all larvae had gastrulated by hours.

TEXT-FIG. 2.

Blastula to trochophore. A, 32-cell, the extent of the blastocoele is indicated by the cross-hatching (6 hrs.), B, dorsal view of 40-cell (7 hrs.), c, antero-ventral view of 40-cell, D, section through a gastrulating embryo 1012 hrs.), E, gastrula with the blastopore still visible (11 hrs.), F, shell gland invagination and formation of the archenteron (12 hrs.). G, H, semidiagrammatic lateral and frontal views of a trochophore recently released from the egg membranes (21 hrs.). AR, archenteron; AT, apical tuft; M, mesoderm; MC, mesenchyme; MO, mouth; OP, oral plate; PT, prototroch; S, shell; SG, shell gland; TT, telotroch; X, primary somatoblast cells.

TEXT-FIG. 2.

Blastula to trochophore. A, 32-cell, the extent of the blastocoele is indicated by the cross-hatching (6 hrs.), B, dorsal view of 40-cell (7 hrs.), c, antero-ventral view of 40-cell, D, section through a gastrulating embryo 1012 hrs.), E, gastrula with the blastopore still visible (11 hrs.), F, shell gland invagination and formation of the archenteron (12 hrs.). G, H, semidiagrammatic lateral and frontal views of a trochophore recently released from the egg membranes (21 hrs.). AR, archenteron; AT, apical tuft; M, mesoderm; MC, mesenchyme; MO, mouth; OP, oral plate; PT, prototroch; S, shell; SG, shell gland; TT, telotroch; X, primary somatoblast cells.

Trochophore

Between the end of gastrulation and the release of the first advanced trochophore from the egg membrane, hours after fertilization, development largely concerns the rapid expansion of the apical ectoderm, the differentiation of the cilia, the formation of the shell gland, and the formation of the stomadeum and the lumen of the gut.

Cilia are not formed until gastrulation is complete. The first appearance of the cilia was difficult to determine owing to the close proximity of the vitelline membrane to the surface of the egg. The first movement was noticed hours after fertilization and by 14 hours all the larvae were rotating within the membrane. The first cilia are relatively short and are present in the region of the future prototroch, but within a short time cilia cover most of the larva. Differentiation of the cilia is rapid and those of the apical and prototrochal regions are denser and longer than the remainder. When looked at from its right side the larva revolves in a clockwise direction (Text-fig. 2G).

At about the time of the first movement of the larva, the large cells, derivatives of the primary somatoblast which form the shell gland, invaginate. The shell gland invagination is a crescent-shaped depression with the deepest part in the centre (Text-fig. 2F). It is not as extensive as in many lamellibranchs (Raven, 1958), possibly because of the large mass of endoderm cells that fill the blastocoele. At about 16 hours after fertilization the shell gland evaginates and forms a disk of cells outlined at the periphery by a slight depression. The disk remains small (50 μ) until the embryo is released from its membranes. While the shell gland is invaginated a cavity, the archenteron, appears in the upper half of the endoderm mass opposite the position of the future stomadeum (Text-fig. 2F). The stomadeal invagination occurs at about the time of the evagination of the shell gland and eventually connects with the archenteron. The proctodeum is not formed until much later (p. 260).

During the period up to the trochophore stage the embryo does not remain in the centre of the gelatinous envelope but gradually sinks through it, so that when the larva is due to be released it is at the periphery. The rotation of the larva inside the vitelline membrane stretches the latter until it is about 200 μ in diameter before it finally ruptures. Release of the larvae commences after 17 hours and all are free 22 hours after fertilization. At this stage the maximum length of the larva is 150 μ. Between 17 and 18 hours the shell gland secretes a disk-shaped pellicle. A ridge and groove from the anterior periphery of the gland marks the lower limit of the developing prototroch (2 G, H). A similar groove extends from the postero-lateral edge of the gland towards the position of the future proctodeum; this does not follow a line of cilia. Following the release of the trochophore, further rapid differentiation of the cilia takes place with the formation of the prototroch composed of a double ring of long (25 μ) cilia (Text-fig. 2 G, H). The cilia of the pretrochal area become concentrated into an apical tuft, the short cilia between the tuft and the prototroch are lost. There is a terminal cap of cilia in the postrochal region which later develops into a telotroch surrounding a terminal tuft. The remaining cilia between prototroch and telotroch become concentrated around the mouth and the position of the future foot. After its release, the larva continues to revolve in the same manner as when in the vitelline membrane. Later, as the cilia become further differentiated, it commences a jerky forward movement with the apical tuft leading. The forward movement becomes smoother and revolutions are now spiral, the movement being in a clockwise direction when viewed apically, as in other Eulamellibranchia (Knight-Jones, 1954). Locomotion is relatively weak and after a few hours at the surface of the water the larvae tend to remain at the bottom of the bowl, momentarily resting on the bottom. The periods of rest become more frequent and longer as development proceeds. As the cilia differentiate there are changes in body form that are associated with the rapid development of the shell gland and its surrounding ectoderm, the latter being concerned with the production of the mantle. The prototroch and telotroch are tilted towards the mouth, and stand out from the general body surface (Text-fig. 2 G, H).

Soon after the trochophore is released the bivalve shell is produced. Each valve originates as an oval-shaped, lateral secretion of the shell gland. Although no ligament was seen at this stage, the secreting epithelium appears to be continuous over the mid-line, i.e. looking from above it appears as a dumb-bell shape. The stomadeum and mid-gut become continuous 22 to 24 hours after fertilization. Slight lateral swellings of the mid-gut indicate the future position of the paired digestive diverticula (Text-fig. 3A). The endoderm cells that will form the hind-gut lie close to the telotrochal band below the mouth. On either side of the developing hind-gut two large, dark, mesoderm cells, can be clearly seen. These are not the original paired primary mesoderm cells; division has taken place and dorsally there are two or three paler and smaller cells from which the posterior larval musculature develops. In addition, there is a group of cells that forms a cap over the stomach below the apical tuft and may be referred to as mesenchymal (Text-fig. 2G). Their point of origin is not certain, but they form the velar musculature and possibly the anterior adductor muscle. At 24 hours the larva still measures 150 μ.

TEXT-FIG. 3.

Veliger, A, lateral view showing the bivalve prodissoconch and the early formed anterior adductor muscle (25 hrs.), B, lateral view showing the now flattened apical region and the distended body-walls due to the contraction of the valves (32 hrs.), c, lateral view showing the dorsal and median velar muscles (34 hrs.), D, diagrammatic apical view of B, the cilia are not shown, E, dorsal view of the larval musculature, in particular showing the position of the muscles extending from the body-wall to the shell (34 hrs.), F, lateral view. The velum can be completely retracted within the shell ventral to the mouth (38 hrs.). G, dorso-lateral view (38 hrs.). AA, anterior adductor muscle; AN, anus; CP, cerebral pit; DD, digestive diverticula; DV, dorsal velar muscle; FB, body-wall shell muscles; FG, fore-gut; FT, foot; HG, hind-gut; LP, outpouching of the body-wall; MV, median velar muscle; P, proctodeum; PA, posterior adductor muscle; PR, limit of the pellicle; RP, pedal retractor muscle; ST, stomach; VV, ventral velar muscle. Other lettering as before.

TEXT-FIG. 3.

Veliger, A, lateral view showing the bivalve prodissoconch and the early formed anterior adductor muscle (25 hrs.), B, lateral view showing the now flattened apical region and the distended body-walls due to the contraction of the valves (32 hrs.), c, lateral view showing the dorsal and median velar muscles (34 hrs.), D, diagrammatic apical view of B, the cilia are not shown, E, dorsal view of the larval musculature, in particular showing the position of the muscles extending from the body-wall to the shell (34 hrs.), F, lateral view. The velum can be completely retracted within the shell ventral to the mouth (38 hrs.). G, dorso-lateral view (38 hrs.). AA, anterior adductor muscle; AN, anus; CP, cerebral pit; DD, digestive diverticula; DV, dorsal velar muscle; FB, body-wall shell muscles; FG, fore-gut; FT, foot; HG, hind-gut; LP, outpouching of the body-wall; MV, median velar muscle; P, proctodeum; PA, posterior adductor muscle; PR, limit of the pellicle; RP, pedal retractor muscle; ST, stomach; VV, ventral velar muscle. Other lettering as before.

Veliger

Between 24 and 36 hours after fertilization the ectoderm and its associated structures are still rapidly developing. The prototroch expands and forms a velum which is bilaterally lobed to a slight extent (Text-fig. 3G). The velum, unlike that of gastropods and many lamellibranchs (Lebour, 1938) remains relatively small. The apical region, domed in the trochophore, flattens and later becomes concave. Close to the oral side of the apical tuft an unpaired cerebral pit develops (Text-fig. 3c). Soon afterwards (34– 35 hours) the apical tuft is lost. Two rows of long cilia (30 μ) are present on the edge of the velum. The upper row at the rim of the velum is much denser and somewhat longer than the lower row.

The shell (prodissoconch) enlarges and soon encloses the body. Its rate of growth reaches a maximum at this time (Text-fig. 4). The prodissoconch is equivalve. About 36 hours after fertilization the retracted velum is almost covered by the larval shell. The ligament was first seen at 32 hours. Before the formation of the ligament the valves are joined by a thin membrane that is probably the mid-region of the primary pellicle (Trueman, 1951).

TEXT-FIG. 4.

Graph showing the growth rate of the larval shell during the first week after fertilization.

TEXT-FIG. 4.

Graph showing the growth rate of the larval shell during the first week after fertilization.

Associated with the development of velum and shell is the early formation of a series of larval muscles used to control these structures. Among the first to be formed are two pairs of velar retractors, a dorsal and a median pair. The dorsal pair are attached to the postero-dorsal part of the shell and the median pair are attached postero-laterally (Text-fig. 3 C, E, F). Later, a third pair of velar retractors are developed which are larger than the first two pairs and which serve the ventral part of the velum. These are attached mid-laterally to the shell (Text-fig. 3F). The order of their appearance seems to be of functional significance. Except for the area anterior to the anterior adductor muscle and stomach there is little space in the early veliger into which the velum can retract (Textfig. 3c). Later, when the mantle and shell have more than covered the body, the retracted velum can be accommodated in the space ventral to the mouth (Textfig. 3F). The musculature associated with the movement of the prodissoconch consists of the anterior adductor muscle and the muscles joining the body and the shell. The anterior adductor is by far the most prominent and the largest of the larval muscles and its position at this stage is far removed from the mouth Text-fig. 3F). The posterior adductor muscle cannot be seen and does not play any part in shell movement until the prodissoconch has almost reached its maximum size. There are three lateral muscles on each side of the body. The anterior lie below and in the same plane as the anterior adductor muscle. The other two muscles are attached to the shell close to the origin of the dorsal and median velar muscles and to the body behind the level of the digestive diverticula (Text-fig. 3E). Like the velar muscles these are primary larval muscles and it is doubtful whether they survive metamorphosis. The first contraction of the shell was observed 32 hours after fertilization. The shell does not cover the body at this stage, and it was noticed that the contraction of the valves often caused each side of the body-wall to bulge outwards in a characteristic fashion (Text-fig. 3 B, D). At this time the ventral body-wall is expanding and it invaginates to form the mantle cavity, and it is this tissue that is forced out when the valves contract.

The mesoderm on either side of the hind-gut is still largely undifferentiated, although the cells on the dorsal side will eventually form the kidney rudiments. No larval nephridium was seen. A small proctodeal invagination was observed 30 hours after fertilization although the anus may not be open for another 6 hours. Also at this time the digestive diverticula enlarge to form a simple sac on each side of the stomach.

The foot first becomes apparent as a slight thickening of the ciliated epidermis between mouth and anus. The telotroch is reduced to a ciliated area around and behind the anus.

Veliconcha

From the release of the larva from the egg membrane until 36 to 40 hours after fertilization swimming movements are upwards to the surface of the seawater. Towards the end of this period the movement is less definite and after 40 hours the larvae tend to keep to the bottom. Thus the active swimming phase lasts for less than 18 hours or, more significantly, just over one tidal movement. The larvae that have returned to the bottom of the container do not swim continuously, but every so often rest, or possibly test, for a short while. A ‘testing’ period characterized by momentary settlement on the bottom has been described for many larvae (Thorsen, 1946) and the presence or absence of a suitable substratum may have a considerable influence on the duration of the pelagic phase. Settlement may be postponed until suitable conditions are found. The ‘rest and test’ periods in Pandora become more frequent and longer in duration, and 60 hours after fertilization about 80 per cent, of the larvae are permanently settled on the bottom. This was found to be a critical period in the rearing and many larvae died. Those that survived this period of settlement thrived up to the conclusion of the observations. It was found that a scattering of sand grains on the bottom of the container encouraged settlement and there were far fewer deaths than without this addition.

The rate of shell growth falls rapidly, and between 60 and 100 hours after fertilization there is little or no increase in the size of the prodissoconch (Textfig. 4). At the end of this period the first growth of the dissoconch is seen. As in P. gouldiana (Sullivan, 1948), the prodissoconch is very long (175 μ) in proportion to its height (120 μ).

Forty hours after fertilization the ventral velar muscle is well developed and the velum can be completely retracted within the mantle cavity. The posterior adductor muscle can now be recognized. The most anterior of the larval muscles from shell to body-wall is soon lost; the others remain visible until about 60 hours. In addition, the posterior and anterior retractor muscles become recognizable at this time (Text-fig. 5 A, B). The foot itself is increasing in size.

TEXT-FIG. 5.

Veliconcha and Spat, A, veliconcha (48 hrs.). B, late veliconcha with the apical plate close to the foregut. The style-sac and the first two gill filaments are forming (90 hrs.), inset of the ligament as seen from above, c, drawn from a living spat with the foot extended and indicates the extent of the first three gill filaments (cf. preserved specimen D) (140 hrs.), D, spat (240 hrs.). AP, apical plate; AOL, anterior outer layer; CG, cerebral ganglion; EA, exhalent aperture; GF, gill filaments; GP, byssus gland; GS, gastric shield; HC, hydrocoele; HR, heart rudiment; IA, inhalent aperture; IL, inner layer; KR, kidney rudiment; L, ligament; MG, mid-gut; OD, opening to the digestive diverticulum; PG, pedal ganglion; POL, posterior outer layer; PP, palp; SE, shell edge; SS, style-sac; VE, velum : VG, visceral ganglion. Other lettering as before.

TEXT-FIG. 5.

Veliconcha and Spat, A, veliconcha (48 hrs.). B, late veliconcha with the apical plate close to the foregut. The style-sac and the first two gill filaments are forming (90 hrs.), inset of the ligament as seen from above, c, drawn from a living spat with the foot extended and indicates the extent of the first three gill filaments (cf. preserved specimen D) (140 hrs.), D, spat (240 hrs.). AP, apical plate; AOL, anterior outer layer; CG, cerebral ganglion; EA, exhalent aperture; GF, gill filaments; GP, byssus gland; GS, gastric shield; HC, hydrocoele; HR, heart rudiment; IA, inhalent aperture; IL, inner layer; KR, kidney rudiment; L, ligament; MG, mid-gut; OD, opening to the digestive diverticulum; PG, pedal ganglion; POL, posterior outer layer; PP, palp; SE, shell edge; SS, style-sac; VE, velum : VG, visceral ganglion. Other lettering as before.

The anus is now open and the gut functional. Although the hind-gut tends to overlap the stomach on the left side when the velum is fully retracted, there is no development of a large loop to the left side such as is present in most lamellibranch late-larvae (Thorson, 1946). When the velum is extended the gut takes a direct course from stomach to anus. The digestive diverticula remain as a pair of lateral sacs that obscure much of the stomach. Above and on either side of the hind-gut the remaining mesoderm is still largely undifferentiated and remains so until after metamorphosis. Quayle (1952) reports a similar condition in Venerupis pullastra. The three main ganglia can now be identified but they are not yet joined by the connectives (Text-fig. 5A).

Spat

All the larvae had settled on the bottom 70 hours after fertilization. At this time the foot is used for locomotion even though the velum is still present. Loss of the velum usually occurs soon after final settlement; but in some cases it was retained for as long as 100 hours after fertilization, though it does not appear to be used in locomotion after the foot has become functional. The byssus gland is also functional at this time; later it must degenerate, for it is absent in the adult. For much of their time the larvae remain attached by the byssus to the substratum, with the foot partially retracted so that the ciliated tip lies close to the mouth (Text-fig. 5 B, D). When disturbed the byssus is released and the foot expanded immediately (Text-fig. 5c), and the larva moves to another position. This is contrary to the behaviour of the adult (Allen & Allen, 1955), in which there is a long interval before the foot extends following a disturbance. The habit of resting with the foot close to the mouth was particularly noticeable between 70 and 130 hours after fertilization, i.e. during the time when the larva has settled but is without gills. The ciliation of the foot appears to be important, not so much for the purpose of locomotion, but for feeding and respiration. Apart from the large palps, the foot remains the only strongly ciliated surface within the mantle cavity after the velum is lost. Cilia do not appear to be present on the mantle at this stage. The cilia on the tip of the foot beat towards the proximal part so that when the foot is lying close to the palps a current will be drawn into the mantle cavity that will come close to, if not across, the surface of the palps.

As in the case of other lamellibranch larvae (Meisenheimer, 1901; Cole, 1938; Quayle, 1952), the loss of the velum is intimately connected with the formation of the palps. The process in P. inaequivalvis closely follows that described by Cole (1938) for Ostrea edulis. The apical plate invaginates to lie over the oesophagus, close to the mouth, the dorsal half of the velum sinks below the anterior adductor while the remainder still lies ventral to the mouth (Text-fig. 5B). Disintegration of all but the apical region follows rapidly. Apart from a few larvae with a small amount of debris close to the anterior adductor and mouth, none were seen between the stage described above and that with fully formed upper labial palps. The latter grow rapidly in size to form typical broad triangular flaps, and the mouth comes to lie close to the anterior adductor muscle. The origin of the lower palps was not observed, but, as it is probable that the latero- and post-oral cilia in the veliger form the ciliated surfaces of the oral grooves, the lower palps may be outgrowths of the post-oral epithelium and may not be split off from the upper palps as some authorities have suggested (Quayle, 1952).

The mantle edge remained bilobed throughout the observations (no larva older than 18 days was examined). Inhalent and exhalent apertures without papillae were seen 170 hours after fertilization (Text-fig. 5D), i.e. soon after the formation of the first three gill filaments. The early dissoconch is somewhat inequivalve although the left valve is not as flat as it will later be. There is no tendency for the spat to lie on one side more than on the other. The ligament is composed of two layers and the periostracum. The inner layer is spoon-shaped and, as in the adult Pandora, exceptionally broad (40 μ) in comparison with its length (50 μ) (Text-fig. 5B). Anterior and posterior parts of the outer layer can be distinguished. As early as 90 hours after fertilization the ligament is entirely comparable to the basic primary ligament of an adult bivalve (Owen, Trueman, & Yonge, 1953). Posterior and anterior growth is similar up to 150 hours after fertilization and the ligament can be termed amphidetic at this stage. Later, growth posteriorly greatly exceeds that anteriorly with the consequent formation of the adult opisthodetic ligament.

The gills originate as a slight longitudinal thickening of the mantle epithelium where it joins the body above the foot (Text-fig. 5B). TWO papillae appear when the velum is lost and, shortly afterwards, a third is produced at the posterior end of the thickening. These grow downwards and by 140 hours are ciliated and extend from the anus to the mouth (Text-fig. 5 c, D). NO further filaments were formed during the period of observation.

No larval muscles were seen after the velum is lost, although it is likely that some may go to form the musculature of the palps. The palps reach their maximum size soon after the gill papillae form, but as the latter develop the palps shrink to adult proportions. The large size of the palps at this stage must be related to the problem of feeding before the gills become functional. Further development of the gut largely concerns the stomach, mid-gut, and style-sac. The latter with its broad lumen was first recognized at the end of the veliconcha stage as a downwardly directed pouch of the stomach, posterior to the mid-gut (Text-fig. 5B). In Pandora the style-sac is combined with the mid-gut (Allen, 1954 . In the early spat the style-sac is short but the typhlosoles which divide the lumen of the mid-gut from that of the style-sac can be recognized. The walls of the style-sac are very strongly ciliated. The gastric shield was first seen 140 hours after fertilization lying against the dorsal and posterior sides of the stomach above the style-sac (Text-fig. 5 c, D). The digestive diverticula open laterally close to the junction of the stomach with the mid-gut. The original pair of simple sac-like diverticula slowly enlarge by the formation of pouches in the original wall. By elongation of the neck of the pouches and by further division of the distal end the digestive ducts and tubules are formed. Alternate muscular pulsations of the diverticula similar to that described for the oyster larva (Millar, 1955 were not observed. It is possible that, owing to the rapidity by which the simple sac-like diverticulum develops into the adult form, pulsations do not occur in Pandora.

Cerebro-pedal and cerebro-visceral connectives were first seen 90 hours after fertilization (Text-fig. 5B). The paired kidney rudiments lie in front of the posterior pedal retractor muscle, one on each side of the hind-gut. By 240 hours each has a well-developed central lumen which connects with its fellow posteriorly below the hind-gut. Apart from identifying the rudiments as lying between the kidney and the posterior wall of the stomach, the development of the heart and pericardium was not followed.

The main interest in this development lies in the adaptations to ensure the minimum dispersal from the localized habitats of the adult. The rate of development is remarkably rapid, metamorphosis occurring less than 4 days after fertilization. Less than one day is spent in the plankton. Thus P. inaequivalvis represents a case of a short-lived planktotrophic eulamellibranch larva (Tables 1 and 2). Although there may be others, for relatively little is known of the developments of the Eulamellibranchia, reference to the survey of Thorsen (1946) shows that it is so far unique in this respect.

TABLE 2.

Comparison of development rates of various lamellibranchs (Costello et al., 1957) with that of P. inaequivalvis

Comparison of development rates of various lamellibranchs (Costello et al., 1957) with that of P. inaequivalvis
Comparison of development rates of various lamellibranchs (Costello et al., 1957) with that of P. inaequivalvis

Comparison with other accounts (Costello et al., 1957) shows that the eggs of lamellibranchs with planktonic larvae, including P. inaequivalvis, develop at a similar rate up to the veliger stage. It is in the period between the veliger stage and metamorphosis that the development of Pandora is so rapid, i.e. during the planktonic phase. This is reflected in the absence of protonephridia and larval eyes, in the simple course of the hind-gut and in the absence of development of the gills until after metamorphosis. The speed of development of the veliger is undoubtedly related to the large size of the egg. The lamellibranch larvae with a long planktotrophic life which have developed from small eggs with few food reserves have of necessity to feed on the plankton for their continued development. The additional food reserves in the egg of Pandora make planktonic feeding of secondary importance.

The early metamorphosis of Pandora is not accomplished without problems. In particular, there is the lack of the gill in the early spat. This lack is probably more significant in terms of feeding and of water-flow through the mantle cavity than of the absence of respiratory surface. The general epithelium acts as a respiratory surface. In the period before development of the gill the inhalent feeding, respiratory, and cleansing current is produced by the combined use of the ciliated tip of the foot and the large palps characteristic of this stage. It seems possible that this mechanism will be found to exist in other larvae, for Yonge (1947) points out that the filaments cannot function as food collectors until the food groove is present, and the latter does not appear until the filaments have become reflected. Ciliation of the foot appears to be universal in lamelli-branch larvae.

Le Développement de Pandora inaequivalvis (Linné)

  1. Après la fécondation artificielle des œufs de Pandora inaequivalvis, le développement est suivi jusqu’au stade dissoconque.

  2. Le développement est très rapide et la métamorphose se produit en moins de 4 jours. La larve véligère passe moins d’un jour dans le plancton; c’est la première fois que l’on rapporte le cas d’une si courte phase planctonique pour une larve de lamellibranche. La vitesse du développement est en rapport avec la taille relativement grande (105– 125 μ de diamètre) de l’œuf et à son contenu en réserves nutritives. La dispersion des jeunes est minimum à partir de l’habitat très localisé de l’adulte.

  3. Il ne se forme ni protonéphridies ni yeux larvaires et l’intestin suit une évolution simple, en comparaison avec d’autres larves de lamellibranches. Les branchies ne se développent qu’après la métamorphose; les grands palpes et le pied produisent les courants ciliaires qui assurent la respiration, la nutrition et le nettoyage entre la disparition du velum et le développement de la branchie fonctionnelle.

Allen
,
J. A.
(
1954
).
On the structure and adaptions of Pandora inaequiva vis and P. pinna. Quart
.
J. micr. Sci
.
95
,
473
82
.
Allen
,
M. F.
, &
Allen
,
J. A.
(
1955
).
On the habits of Pandora inaequivalvis (Linné)
.
Proc, malac. Soc. Lond
.
31
,
175
85
.
Cole
,
H. A.
(
1938
).
The fate of the larval organs in the metamorphosis of Ostrea edulis
.
J. Mar. biol. Ass. U.K
.
22
,
469
84
.
Costello
,
P. D.
,
Davidson
,
M. E.
,
Eggers
,
A.
,
Fox
,
M. H.
, &
Henley
,
C.
(
1957
).
In Methods for Obtaining and Handling Marine Eggs and Embryos
, p.
247
.
Marine Biological Laboratory, Wood’s Hole, Mass
.
Grave
,
B. H.
(
1927
).
The natural history of Cumingia tellinoides
.
Biol. Bull., Wood’s Hole
,
53
,
208
19
.
Knight-Jones
,
E. W.
(
1954
).
Notes on the invertebrate larvae observed at Naples during May and June
.
Pubbl. Staz. zool. Napoli
,
25
,
135
44
.
Lebour
,
M. V.
(
1938
).
Notes on the breeding of some lamellibranchs from Plymouth and their larvae
.
J. Mar. biol. Ass. U.K
.
23
,
119
44
.
Lillie
,
F. R.
(
1895
).
The embryology of the Unionidae
.
J. Morph
.
10
,
1
100
.
Loosanoff
,
V. K.
, &
Davis
,
H. C.
(
1950
).
Conditioning V. mercenaria for spawning in winter and breeding its larvae in the laboratory
.
Biol. Bull., Wood’s Hole
,
98
,
60
65
.
Meisenheimer
,
J.
(
1901
).
Entwicklungsgeschichte von Dreissenia polymorpha Pall
.
Z. wiss. Zool
.
69
,
1
137
.
Millar
,
R. H.
(
1955
).
Notes on the mechanism of food movement in the gut of the larval oyster, Ostrea edulis. Quart
.
J. micr. Sci
.
96
,
539
44
.
Owen
,
G.
,
Trueman
,
E. R.
, &
Yonge
,
C. M.
(
1953
).
The ligament in the Lamellibranchia
.
Nature
,
171
,
73
.
Pelseneer
,
P.
(
1926
).
Notes d’embryologie. Ponte et développement de Cypraea europa, Triforis perversa et Lucina lactea
.
Bull. Biol
.
60
,
88
112
.
Quayle
,
D. B.
(
1952
).
Structure and biology of the larva and spat of Venerupispullastra (Montagu)
.
Trans, roy. Soc. Edinb
.
62
,
255
97
.
Raven
,
C. P.
(
1958
).
Morphogenesis: the Analysis of Molluscan Development
, p.
316
.
London
,
Permagon Press
.
Sullivan
,
C. M.
(
1948
).
Bivalve veliger larvae of Malpeque Bay, P.E.I
.
Bull. Fish. Res. Bd. Can
., No.
77
,
1
36
.
Thorsen
,
G.
(
1946
).
Reproduction and larval development of Danish marine bottom invertebrates, with special reference to the planktonic larvae in the Sound (Oresund)
.
Medd. Komm. Havunder-50 g., Kbh., Ser. Plankton
,
4
,
1
523
.
Trueman
,
E. R.
(
1951
).
The structure, development and operation of the hinge ligament of Ostrea edulis
.
Quart. J. micr. Sci
.
92
,
129
48
.
Yonge
,
C. M.
(
1947
).
The palliai organs in the aspidobranch Gastropoda and their evolution throughout the Mollusca
.
Phil. Trans. B
.
232
,
443
518
.
1

Unless otherwise stated all measurements are the maxima recorded.

1

The original primary somatoblast ≡ 2d has probably by this time divided 4 times giving four (x1x4) small somatoblast cells (see Lillie, 1895).