On the Mantle Cavity and Its Contained Organs in the Loricata (Placophora)

On the Mantle Cavity and its Contained Organs in the Loricata (Placophora). By C. M. Yonge, DJSe., University of Bristol.

With 6 Text-figures.

CONTENTS. PAGE 1. INTRODUCTION 367 2. THE MANTLE CAVITY 368 (a) Lepidochitona oinereus, p. 370; (6) Tonieella marmorea and AcanthocMtona crinitus, p. 373; (c) Lepidopleurus asellus, p. 374. 3. THE GILLS 376 4. Mucous GLANDS 382 5. OSPHRADIA AND OTHER SENSE OBGANS 383 6. DISCUSSION 386 7. SUMMARY 388 8. BEFERENCES 389

1. INTRODUCTION. The investigations described in this paper represent a con- tinuation of previous work on the mantle cavity and its contained organs in the Gastropoda (Yonge, 1937 c, 1938) and in the Scaphopoda (Yonge, 19376). They are intended to form part of a comprehensive survey, from the functional aspect, of these organs throughout the Mollusca. The structure and the anatomical relations of the organs contained in the mantle cavity of the Loricata are well known, as a result, in the main, of the morphological investigations of Haller (1882, 1884), Pelseneer (1898, 1899), and Plate (1898-1901). But only Arey and Crozier (1919) have made observations on the water currents in the mantle cavity, and they failed to point out how these were brought about, giving no account either of the ciliation of the gills or of the manner in which the mantle cavity is divided into inhalant and exhalant chambers. 368 0. M. YONGB The greater part of this work has been carried out on two common British species, Lepidochitona cinereus (Lepi- dochitonidae) and Lepidopleurus asellus, which is one of the few representatives of the primitive order Lepidopleurida. Comparative data only has been obtained from Tonicella marmorea (Lepidochitonidae) and Acanthochitona crinitus (Oryptoplacidae).1 All of these species live between tide-marks with the exception of Lepidopleurus asellus, which, in the. English Channel, occurs at depths ranging from 15 to 43 fathoms. Comparative observations on the gills of the Gastropod, Haliotis tuberculata, were carried out at Naples during the course of other work. Acknowledgements are due to Dr. S. Kemp, F.E.S., and members of the staff of the Plymouth Laboratory, for assistance during a period spent there in connexion with this work, to Professor E. Dohm, Director of the Stazione Zoologica, Naples, to the Colston Eesearch Society of the University of Bristol for financial assistance, also to Mr. H. P. Steedman, Laboratory Steward in the Department of Zoology, University of Bristol, for cutting sections.

2. THE MANTLE CAVITY. The mantle cavity in the Loricata consists of lateral pallial grooves in communication posteriorly, and bounded internally by the sides of the foot and externally by the inner margin of the girdle. The mouth opens anteriorly and the anus posteriorly (Text-figs. 1 and 3, A.), both in the middle line. The sole of the foot frequently extends under the anus in life. Eeproductive and excretory pores open into the pallial groove on either side, near the posterior end but always in the region occupied by gills. The former openings are always the more anterior. The gills vary widely in number throughout the Loricata and are actually not constant for any one species. Thus, of the species here examined, Tonicella marmorea has between 19 and 26, Lepidochitona cinereus between 16 and 19, Acanthochitona crinitus about 15, and Lepido- pleurus asellus between 11 and 13. The gill series may be 1 The names used are those adopted by Wlnckworth (1932). MANTLE CAVITY IN LORICATA 869 holobranch or merobranch according as to whether or not they extend the full length of the pallial grooves. None of the species examined possesses gill series of the former type, although Tonicella marmorea and Lepidochitona cinereus (Text-fig. 1) approach this condition. The first-formed gill occurs in the region between plates seven and eight. According to Pelseneer (1898, 1899) the excretory pore invariably opens immediately anterior to this gill, which is morphologically, and also functionally as the present work reveals, of great significance. It will be termed throughout the post-renal gill. Pelseneer also states that it is invariably the largest gill, but, as Plate (1901) had pointed out, this is not always the case, although, if not actually the largest, it is always one of a group of especially large gills. This post-renal gill may be the last of the series, as in Lepidochitona cinereus (Text-fig. 1, GP.), in which case the condition is known as abanal. But, during development, gills may be added posteriorly as well as anteriorly to this gill. The former are termed adanal gills. There may or may not be a space between the last adanal gill and the anus. Of the species examined only Lepidopleurus asellus (Text-fig. 3) possesses adanal gills, and these extend up to the anus. Finally, the thin inwardly projecting ridge which bounds the inner surface of the girdle is dilated on each side in the region opposite to the posterior margin of the foot. The pair of inwardly projecting mantle folds (Text-figs. 1-3, GP.) SO formed have been described by both Pelseneer and Plate.1 This investigation has revealed their function. Observations on the nature of the currents in the mantle cavity were carried out by placing animals on glass slides. When the animals had attached themselves the slides were inverted on. two pillars of plasticine in a shallow glass dish, the water in which just covered the slides. The dish was then placed on the

1 Plate (1901) states that the folds are absent in some species, but he includes amongst these Lepidopleurus asellus and Tonieella marmorea where it certainly occurs. It is probably of universal occur- rence, Plate being misled by contraction in the fixed material which he exclusively studied. NO. 323 B b 870 0. M. YONGE stage of a binocular microscope, carmine added to the water, and the nature of the water currents observed with the animal ventral side uppermost. The general observations of Arey and Crozier (1919) on Chiton tuberculatuj (a Bermudan species attaining a length of 9 cm.) were confirmed. The gills create a current of water which runs backward along the pallial grooves and out in the mid-line posteriorly (Text-figs. 1 and 8). The regions of intake vary, being created by local liftings of the girdle. When the animals are completely submerged these are usually anterior (Text-figs. 1 and 8, i.) but may be lateral, and there may be more than one opening on either side. When the anterior ©nd of an animal is out of water inhalant openings are created by a lifting of the girdle in the region still submerged. In this way a shorter, but still efficient, respiratory current is produced. The exhalant opening (Text-figs. 1-8, E.) is also created by a local raising of the girdle, in this case always at the posterior end. The full discussion of the mechanisms concerned in the maintenance of the respiratory currents demands separate consideration of the various species. (a) L©pidochitona cinereus.—Although this species belongs to the order Chitonida, and so is less primitive than Lepidopleurus asellus (order Lepidopleurida), the de- scription of conditions in both species will be easier if Lepi- dochitona cinereus is considered first. In the animal shown in Text-fig. 1 there are seventeen gills on each side extending along some four-fifths of the pallial grooves. They are attached to the roof of the grooves. All but the last and largest of these, the post-renal gill (GP.), bend inwards towards the sides of the foot, but the post-renal gill extends backwards with its (morphologically) posterior surface applied to the side of the girdle fold (GF.) and its anterior surface to the side of the foot. It thus blocks the pallial groove in this region. The five gills immediately anterior to it also bend backwards to a greater or less extent, but they lie against the sides of the foot only. Throughout the whole series the sides of adjacent gills are alwayi closely applied. Lateral cilia on the gills (described later) create a current of water which passes from their outer to their inner surfaces. IC

o GF

TBXT-HG. 1. Lepidochitona cinereus, ventral aspect, drawn from life: gills and boundaries of shell plates shown in left pallial groove, osphradium and division between inhalant and exhalant chambers (denoted by broken line) shown in right pallial groove, x 10. A., anus; E., exhalant current; EC, exhalant chamber; p., foot; o., girdle; GI., most anterior gill; GF., girdle fold; GP., post-renal gill; I., inhalant currents; M., mouth; o., osphradium. Arrows indicate direction of currents, broken arrows those in exhalant ohamber. 372 C. M. YONGB Owing to the arrangement of the post-renal gill across the pallial groove there is a complete separation of the mantle cavity into inhalant and exhalant chambers (Text-fig. 1, ic, BO.). The former consists of the pallial groove anterior to the first gill and the region outside the gills behind this. The exhalant chamber consists anteriorly of the region between the gills and the sides of the foot (and so bounded ventrally by the lateral extensions of the sole) and posteriorly of the entire pallial groove behind the post-renal gills. The boundary between these chambers is indicated by the broken line in the right pallial groove in Text-fig. 1. A powerful backwardly directed current runs along the inhalant chamber, water being drawn through the gills laterally into the exhalant chamber and also posteriorly between the filaments of the post-renal gills. This inhalant current is so powerful that suspended particles do not tend to settle on to the roof of the pallial groove when the animal is inverted for inspection. Particles too large to pass between the gill filaments are carried to the tips of the gills, and so into the exhalant chamber. The gills are muscular and may assist the respiratory current by moving outwards and then drawing inwards. The post-renal gills are particularly active, repeatedly extending and then retracting, while their tips move from side to side. In the exhalant chamber the water flows backwards, as indicated by the broken arrows in Text-fig. 1, and the streams from both sides finally pass out by way of the exhalant opening. This may be formed either in the mid-line posteriorly or a little to either side of this, but it is always confined to the space between the tips of the post-renal gills. Particles carried in this current become concentrated in a narrow rejection current which runs along either side of the anus (see broken arrow to the right of this in Text-fig. 1). The eggs or sperms liberated from the reproductive pores situated between the second and third gills (counting from the post-renal gill) on either side, the outflow from the excretory pores immediately anterior to the post-renal gills (GP.) and the faeces from the anus, all pass out with the exhalant current (E.) which is responsible for their transport. Both the genital and the excretory pores open on the inner side MANTLE CAVITY IN tOEICATA 373 of the attachment of the gills, i.e. into the exhalant chamber, the apertures being so disposed that material is directed inwards towards the sides of the foot. The roof of the pallial groove is throughout richly ciliated, and cilia also occur on the outer but not on the inner sides. These cilia assist to some extent in the transport of particles posteriorly. The walls of the exhalant chamber are more generally ciliated notably in association with the paffial mucous tracts (see below).

TEXT-MG. 2. Acanthochitona crinitus, ventral aspect of posterior region, drawn from life. X 8. Lettering as before. (b) Tonicella marmorea and Acanthochitona crinitus.—Both of these species belong to the order Chitonida, the former being included in the family Lepidochitonidae and the latter in the Cryptoplacidae. As already noted they resemble Lepidoehitona cinereus in the absence of adanal gills, and, although the number of gills differs in both species from that in Lepidoehitona cinereus, they function in essentially the same manner with the large post-renal gill extending backwards and filling the space between the girdle fold and the edge of the foot. In Acanthochitona crinitus, where the gills only extend for some two-fifths the length of the pallial grooves, there is a tendency for inhalant openings to be formed laterally, near the anterior end of the gills, rather 374 0. M. YONGE than near the anterior end of the pallial groove as in the other two species with their greater and more extended series of gills. Moreover, in this species, as shown in Text-fig. 2, the post-renal gills extend farther back than in Lepidochitona cinereus, leaving a more restricted posterior exhalant cavity. The exhalant opening is usually confined to a narrow region in the mid-line, but may move a little to one side or the other when the animal makes turning movements. (c) Lepidopleurus asellus.—Conditions here, with adanal gills extending posteriorly up to the anus (see Text-fig. 3), are in significant respects different from those in the three representatives of the Ohitonida. The usual number of gills, as shown in Text-fig. 3, is thirteen, although it may be less. The gills are attached nearer to the base of the foot than are those of Lepidochitona cinereus, and they extend forward only as far as the junction between plates six and seven (GI.). The last six are posterior to the post-renal (GP.), and so are adanal gills. The anterior six gills bend inwards towards the foot. The first four of these, as indicated by the dotted lines in Text-fig. 3, are actually obscured ventrally by the sole of the foot. The last seven gills, i.e. the post-renal gill and all the adanal gills, bend outwards, their tips making contact with the edges of the girdle folds. The girdle folds have the same form as those in the Chitonida, but extend farther posteriorly (see Text-fig. 3). Thus the pallial grooves are divided into inhalant and exhalant chambers, but in a somewhat different manner than in Lepidochitona cinereus and the other members of the Chitonida. The inhalant chamber is relatively larger, com- prising the whole of the pallial grooves anterior to the seventh shell plate, and the outer portion of the pallial cavity beneath the seventh plate, and the outer dorsal region of the grooves in the most posterior region. The bending outwards of the gills in the posterior region is responsible for this dorsal extension of the inhalant chamber. Water passes laterally through the gill filaments in the anterior region and downwards (upwards as the animal lies in Text-fig. 3) posteriorly. The exhalant chamber is thus on the inner side of the pallial groove in the MANTLE CAVITY JUS LOBICATA §75

TEST-BIG. 3. Lepidopleurus asellus, ventral aspect, drawn from life, arrangement as in Test-fig. 1. x 12. Lettering as before. region beneath plate seven extending over the complete ventral area posterior to that. Inhalant openings are normally formed anteriorly. The exhalant opening is restricted to the region immediately posterior 376 C. M. YONGE to the anus, which is here raised on a papilla carrying it almost to the inner edge of the girdle. Owing to the smaller number of gills the water current is slower than in Lepidochitona cinereus. In correlation with this particles drawn in with the inhalant current are carried in the main by ciliated tracts which run along the edges of the pallial groove, on the sides of the foot and of the girdle. Particles carried in the first of these are conveyed to the posterior margin of the foot where they congregate in mucus-laden masses, and are from time to time transferred to the surface over which the animal is moving. Particles are carried in the outer current to the region of the girdle fold where they are conveyed outwards to the exterior. The course of both of these currents is indicated by arrows in Text-fig. 3. They have no counterpart in any species of the Chitonida examined where all particles pass into the exhalant chamber. Particles which remain in suspension in the respiratory current are carried into the postero-dorsal extension of the inhalant chamber above the outwardly extended adanal gills. They finally emerge into the exhalant chamber between the second and third gills counting from the posterior end (see broken arrows on either side of B. in Text-fig. 3). Transport of this material is assisted by cilia on the roof of the pallial groove and dorsally along the sides of the foot, i.e. in the regions of the neural and pedal mucous tracts (see below). The reproductive pores open between the bases of the eighth and ninth gills counting from the posterior end, and the excretory pores in front of the post-renal gills. The openings of both pairs of pores are directed downwards, and not also inwards as in Lepidochitona cinereus, but their products will be carried outwards in the exhalant current. Owing to the posterior extension of the anus (A.) faeces are deposited outside the body with the minimum of assistance from the relatively weak exhalant current. 3. THE GILLS. The structure of the gills of the Loricata has been described by a number of workers, notably Burne (1896), Pelseneer (1898, 1899), and Plate (1898-1901). Pour views have been advanced as to their nature. (1) Spengel (1881) considered them all MANTLE CAVITY IN.tOBICATA 377 homologous with the etenidia of other Mollusca and suggested the name Polybranehiata for the group. (2) Clans (quoted by Plate, 1901) thought that each gill row represents a ctenidium with a greatly extended longitudinal axis, each gill being homologous with a single filament of the etenidium. (3) Pelseneer (1898, 1899) considered that the post-renal gills were true etenidia, but that the remaining gills were secondary structures. (4) Ihering (1877), Thiele (1890), and Plate (1901) al regarded the gills as secondary structures, outgrowths of the mantle like the gills of Patella or of the Nudibranchia. But all of these theories are based on exclusively morphological examinations of the gills; apart from determining the circulation of blood through them no attempt has been made to study the gills functionally. In particular the arrangement and nature of their ciliation has never been determined and compared with that of the etenidia of the prosobraneh Gastropoda. The nature of the ciliation on the gills of Lepidoehitona cinereus has been examined in detail. That of the gills of the other species, including Lepidopleurus asellus, agrees in all essential details. Within the inner margin of the axis of each gill (Text-fig. 4) passes an afferent blood-vessel (Text- figs. 4 and 5, AV.), and within the outer margin an efferent vessel (BV.). On the inner side of each of these extend strands of longitudinal muscle (LM.), and on the outer side the internal and external branchial nerves respectively (N.). The contraction of the longitudinal muscle on the outer (efferent) side of the gill axis will pull the gill away from the side of the foot, the contraction of the other muscles will draw it back again. This provides the mechanism for the regular inward and outward movements executed by the gills. Contraction of both muscles will draw the entire gill downwards—or anteriorly in the case of the post-renal gills in Lepidochitona cinereus and the other Chitonida, or inwards in the ease of the adanal gills of Lepidopleurus asellus. Extension will follow owing to pressure of blood when the muscles relax. From either side of the axis extend a series of filaments which diminish in size from the base to the apex of the gill (Text-fig. 4). The most striking feature about their ciliation is the presence on the sides 378 0. M. YONGE of the filaments, extending somewhat nearer to the inner than to the outer surface, of a broad band of cilia (AC.) which attain a length of some 35/L*. The function of these cilia is that of

TEXT-ITO. 4. Lepidoohitona cinereus, posterior aspect of a gill from the left pallial groove, drawn from life. x90. AC, area of long attaching cilia on filaments; A.V., afferent blood-vessel; EV., efferent blood-vessel; IM., longitudinal muscle. Arrows indicate direction of ciliary currents (horizontal arrows respiratory current due to lateral cilia), dotted arrows flowi n blood-vessels. attachment to the corresponding region on the next gill. It has already been noted how closely adjacent gills are applied in life. On either side of this region, and on both sides of the axis, cilia are present which beat to the tip of the gill to which all larger particles are carried (see arrows in Text-fig. 4). MANTLE CAVITY IN LOBICATA 879 The opposed surfaces of the filaments may be divided into three zones (see Text-fig. 5). The region near the outer surface is practically devoid of cilia. The middle region, bounded at the margin by the zone of attaching cilia (AC), possesses long lateral cilia (LC.) which cause a powerful current to pass between the filaments from the outer to the inner side of the gill (see horizontal arrows in Text-fig. 4). The innermost region possesses

AV- LM

EV- •LM

TEXT-ITG. 5. Lepidochitona cinereus, lateral view ofpair of gill filaments, drawn from life. X125. LC., lateral cilia; IT., branchial nerves. Other lettering as before. Arrows indicate direction of beat of cilia, crosses beat to apex of gill. scattered cilia which beat in the same direction. The current created by the lateral cilia is responsible for the respiratory stream, and is so powerful that all smaller particles are conveyed by it between the filaments, only the largest ones passing to the tips of the gills. The structure of the gills is essentially similar to that of ctenidia, while the ciliation clearly represents no more than a modification of that present on the ctenidia of the most primitive living Gastropoda, i.e. species of zygobranchiate genera such as Diodora (Fissurella) orHaliotis. 380 C. M. YONGE In the latter (Text-fig. 6) a narrow band of lateral cilia (LC.) on the elongated filaments creates a water current which passes from the efferent to the afferent side of the axis, here in the primitive positions, ventral and dorsal respectively. The broad band of interlocking cilia (Text-figs. 4 and 5, AC.) on the sides of the filaments in Lepidochitona cinereus corresponds to the cluster of long terminal cilia (TC.) at the apex of each filament in Haliotis. The cilia on the margins of the filaments

AFC

LC R

TEXT-ITG. 6. Haliotis tnberculata, lateral view of pair of gill filaments drawn from life. xl5. AFC., abfrontal cilia; CK., chitinous supporting rod (shown right side only); re, frontal cilia; LC, lateral cilia (shown left side only); TC, long cilia at tips of filaments. Other lettering and arrows as before. on the efferent and afferent sides correspond to the frontal (FC.) and abfrontal (AFC.) cilia respectively in Haliotis. • The differences between the filaments in the two cases are not difficult to explain. The shortening of the filaments in the Loricata, and the presence of the broad band of interlocking cilia, are both due to the increased number of gills and the closeness with which these are applied to each other. In order to offset the effect of the consequent decrease in the lateral surface, the breadth of the zone occupied by lateral cilia has been greatly increased. But there is no longer any need for the chitinous supporting rods (Text-fig. 6, CR.), characteristic of the elongated filaments of the ctenidia of both Gastropoda MANTLE CAVITY IN I.OKICATA 381 and Lamellibranchia, in the shortened filament of the Loricata. Sections reveal the absence of this although the filaments are everywhere strengthened by strands of connective tissue which run across the internal cavity. Finally, the difference between the direction of beat of the frontal and abfrontal cilia in Haliotis and corresponding cilia in Lepidochitona einereus is due to the different manner in which sediment brought in with the respiratory current is disposed of in the two animals. Cleansing is the primitive function of all sach cilia, and the direction of their beat actually varies considerably in different Gastropoda (Yonge, unpublished work). There is thus good reason for believing that the gills of the Loricata are homologous with those of the prosobranch Gastropoda, i.e. are true ctenidia, and that, of the four views as to their nature, the original suggestion of Spengel, which has hitherto found no support, is true. The ciliation of the secondary gills of Patella has been examined and does not correspond in the slightest with that of ctenidia, whereas that of the gills of the Loricata unquestionably does while their structure is fundamentally similar. The one point of real difference is the multiplication of the ctenidia in the Lorieata, and that is to be associated with the elongation of the body and the consequent anterior extension of the originally posterior mantle cavity into a pair of narrow elongated pallial grooves. The different orientation of the ctenidia, the axes of the majority of which extend downwards (horizontally outwards in the case of the post-renal and adanal gills of the Lepidopleuridae), is due to the same cause. But, like the ctenidia of the prosobranch Gastro- poda and of the Lamellibranchia, they effectively divide the mantle cavity into exhalant and inhalant chambers which the secondary gills of Patella do not. The identical structure of the gills lends no support to Pelseneer's theory that only the post-renal gill is a ctenidium, while Plate (1901) has con- clusively disproved the inherently improbable theory of Glaus that the gill series represents a single ctenidium. Plate (1901) advanced as one objection to the view that the gills are ctenidia the fact that the auricles and kidneys are not multiplied correspondingly as they are in Nautilus. But conditions are 382 C. M. YONGB essentially different in the two groups. In the Tetrabranchiata the duplication of the gills may reasonably be ascribed to the greater respiratory needs of the animals which have not been met, as they have in the dibranchiate Cephalopoda, by a more efficient circulatory system including the provision of branchial hearts, and by a more powerful respiratory current created by the muscular action of the entire mantle (confined in Nautilus to the siphon). In the Loricata the respiratory needs are not greater than in the primitive mollusc, with the single pair of ctenidia in a posterior mantle cavity, from which they certainly evolved. Multiplication is due, as already stated, to a reduction in size of the individual ctenidia owing to the change in shape and mode of functioning of the mantle cavity. The effective respiratory surface is not increased. There is, therefore, no need for similar multiplication of either auricles or kidneys.

4. Mucous GLANDS. Plate (1898-1901) noted the presence in the Loricata of tracts of unicellular mucous glands in the mantle cavity. He dis- tinguished between neural, pedal, pallial, and branchial tracts on the roof of the pallial groove, sides of the foot, inner wall of the girdle, and inner axis of the gills respectively. In none of the large number of species which he examined did he find more than three of these tracts, while some species had only two, others one, and some none at all. Various com- binations of the possible series of tracts were found in different species. These tracts are easily seen in sections, the gland-cells being usually greatly elongated and the contained, somewhat granular, secretion staining readily with eosin and allied stains. They are always interspersed with ciliated cells. In Lepi- dochitona cinereus and Acanthochitona crinitus (Tonicella marmorea has not been sectioned) they are confined to a pallial tract on the dorsal haK of the outer walls of the pallial groove in the exhalant chamber posterior to the gills. In Lepidopleurus asellus Plate states that only branchial and pedal tracts are present, although he found neural tracts associated with these in two other species of the same genus, Lepidopleurus cajetanus and Lepidopleurus MANTLE CAVITY IN 10BICATA 383 medinae. But in Lepidopleurus aseilus neural tracts, anterior to the gills, do exist although the cells are not so elongated as in the pedal tracts. The latter extend the full length of the pallial grooves, i.e. in both the anterior, inhalant, and in the posterior, exhalant, chambers. The arrangement and histology of the tracts has been wel figured by Plate (1898- 1901) and Knorre (1925). The function of their secretion is clearly to consolidate the particles of sediment brought in with the respiratory current. They are thus analogous with the hypobranehial glands of the prosobranch Gastropoda (Yonge, 1937 a, 1938). The pallial mucous tracts on the outer sides of the roof of the exhalant chamber in the two species of the Chitonida examined may possibly be homologous with the hypobranehial glands because they occur in the same relative position. Plate recognized the essential function of the tracts, although bis ignorance as to the nature of the water circulation in the mantle cavity was such that he suggested that the secretion of the pallial tracts in the exhalant chamber was to protect the gills from contamination with the faeces. Plate noted that mucous tracts are most wide- spread in the more primitive genera with merobraneh gill series. This is to be correlated with the greater water current produced by the gills of the more highly evolved species, e.g. Lepi- dochitona cinereus with its extensive series of gills has only one pair of mucous tracts, while Lepidopleurus aseilus with fewer and relatively smaller gills has three, of which one, the neural, is entirely in the inhalant chamber and the greater part of the pedal tract also. In many of the holo- braneh species the water current is presumably great enough to carry particles through the pallial grooves in suspension so that mucous glands become superfluous.

5. OSPHBADIA AND OTHER SBNSB OEGANS. Pour sets of sense organs have been identified in the mantle cavity of the Loricata. These are (1) the osphradia, (2) the anterior 'olfactory organs', (3) the branchial'olfactory organs', and (4) the lateral sense organs. The osphradia occur in the majority of the Loricata being 384 C. M. YONGB apparently absent only in the Lepidopleurida and a few species of the Chitonida (Plate, 1901). They are present in the three species of Chitonida here examined. They are elongated structures lying along the roof of the pallial groove and extending from the anus as far as the post-renal gills (Text-figs. 1 and 2, o.). Each consists of an area of long epithelial cells containing ciliated and sense cells, and bounded by a well-defined cuticle through which penetrate the sense hairs. Prom the cells numerous nerve-fibres pass into the pallial cord which runs immediately below. They have been figured by both Plate (1898-1901) and Pelseneer (1899). Both of these workers regarded them as homologous with the osphradia of the prosobranch Gastropoda because they occur in the same relative position. While, as indicated below, this may be true, it should be noted' that, whereas in the Gastropoda they are in the inhalant region of the mantle cavity, in the Loricata they lie in the exhalant chamber. This is also true of the 'osphradia' in the Lamellibranchia, as Stork (1934) has pointed out. Plate assumed that the osphradia have an olfactory function, for' testing' water, as originally suggested by Bernard (1890) for the osphradia of the Gastropoda. Arey and Crozier (1919) come to the same conclusion. But an organ so situated that the excretory products of the animal flow over it would not appear to be ideally situated for such a function, moreover it cannot ' test' the water until this is on the point of leaving the mantle cavity. It has been pointed out elsewhere (Hulbert and Yonge, 1937) that the osphradia of the Gastropoda may with good reason be regarded as tactile organs concerned with estimating the amount of sediment brought in with the respiratory current. These organs are always present in ctenidiate Prosobranchia which are all exposed to the danger of having the mantle cavity blocked with sediment, whereas only some shore-living or estuarine species are ever exposed to danger from impure water. In the merobranch Chitonida examined sediment is all passed into the exhalant chamber, and, as shown by the broken lines lateral to the anus in Text-figs. 1 and 2, actually passes along the surface of the osphradia before being ejected. There thus appears as good reason for considering these structures to be concerned MANTLE CAVITY IN 10RICATA 385 with estimating the amount of sediment brought into the mantle cavity as for regarding them as olfactory organs. It is also easier on this basis to account for their retention in the Loricata in the same position as in the Gastropoda. In the latter group sediment tends to settle out on to the osphradium in the inhalant region (Yonge, 1938) which is posterior and ventral, in the former it is accumulated in the same relative position, but this is here in the exhalant chamber owing to the formation of inhalant openings anteriorly and the different orientation of the ctenidia. Plate denied the presence of anterior'olfactory organs' which had previously been described by Blumrich (1891). More recently, however, Knorre (1925) has confirmed Blumrieh's conclusions by describing these organs in Ischnochiton herdmani and Ischnochiton aequigranulatus. They certainly occur in Lepidochitona cinereus but not in Lepidopleurus asellus. They consist of longitudinal strips of sensory epithelia anterior to the gills and lying along the roof of the palh'al groove. They are innervated from the pedal cord and resemble the osphradia histologically. Knorre regards them as olfactory in function, but they may with equal reason be regarded as tactile organs concerned with estimating sediment. They would certainly be in a position to detect any abnormally great amount of sediment carried in with the inhalant currents and enable the animal to respond at once by closing the inhalant openings. The branchial and lateral sense organs occur only in the order Lepidopleurida where, as a result probably of the posterior extension of the gills, osphradia are absent. The former consist of sensory patches of epithelia running along the efferent axes of the gills, i.e. on the topographically outer side in the case of the anterior gills, and on the upper side in the case of the post- renal and all adanal gills. Burne (1896), who originally described them, gives a detailed account of their position and innervation. He regarded them as true osphradia, but Plate, with good reason, considered them secondary structures. They lie in the path of the water current which passes through the gills but also of the sediment which it carries. The lateral sense organs NO. 323 C c 886 C. M. YONGE were first described by Tbiele (1895). They consist of a series of small patches of elongated sensory epithelium about the middle of the pallial wall and so project slightly into the cavity of the pallial groove. They extend along the entire pallial groove, and their number varies according to the age and size of the animal. They have not been counted in Lepidopleurus asellus but in Lepidopleurus cajetanus they vary from 26 to 35 pairs. The epithelium is of the same type as in the other sense organs, and, like them, they are innervated from the palh'al cord. Their function also must remain problematical, Plate regarded them as olfactory, but they lie in the path of sediment carried along the outer walls of the pallial groove and may therefore with equal reason be regarded as being concerned with the estimation of this. In either case the branchial and lateral sense organs of the Lepidopleurida very probably meet the needs served by the osphradia and anterior sense organs in the majority of the Chitonida.

6. DISCUSSION. The Loricata are a group of Mollusca in which the body has become elongated and the original single shell valve converted into a series of eight overlapping plates. This change has fitted them admirably for creeping over the uneven surfaces of rocks, particularly within the tidal zone where the encrusting algae on which they feed grow in greatest abundance. Pretter (1937) has shown how highly adapted for this diet are the organs of feeding and digestion in the Chitonida. The extensive ventral surface provided by the foot and girdle enables them to cling closely to the uneven surface when the tide is out and to resist dislodgement during stormy weather. The articulating plates permit protection by curling up should the animal become detached. The capacity for creating inhalant apertures over a wide area on either side of the body, and for producing a respiratory current even when the anterior half of the body is out of water, are both of considerable biological significance. The Loricata are amongst the most highly specialized of shore- living animals. The anterior elongation of the mantle cavity into a pair of MANTLE CAVITY IK LOBICATA 387 narrow pallial grooves and the consequent formation of inhalant- openings anterior to the gills instead of a single ventral posterior opening have been responsible, as already noted, for the modification of the organs normally present in the mantle cavity of Mollusca. The ctenidia have been altered in position and reduced in size but correspondingly increased in number, while their filaments have been modified in shape and citation. As a result the necessary functional division of the mantle cavity into inhalant and exhalant chambers has been main- tained. The mucous glands have been in some cases increased in number, in others remain paired structures on the roof of the mantle cavity near the anus, and in others are entirely suppressed, all in correlation with the number and position of the ctenidia. The osphradia—assuming these structures to be homologous with those so designated in the Gastropoda—per- sist in the majority of cases, but may share their problematic function, tactile or olfactory, with an anterior sense organ of similar structure. Where absent, as in the Lepidopleuridae owing to the posterior extension of the ctenidia, they are replaced by branchial and lateral sense organs. The nature of the modified mantle cavity has further involved the appearance of new structures, the girdle folds, which enable the all-important post-renal gills (aided by adanal gills where these occur) to complete the functional division of the pallial grooves into inhalant and exhalant chambers. It is along such lines that we may explain, on a functional basis, the arrangement in the Loricata of these organs, the structure of which has been so admirably described by morpho- logists such as Haller, Plate, and Pelseneer. There remains for discussion the significant differences which exist between the members of the orders Lepidopleurida and Chitonida. The former are characterized by the absence, or presence in a smooth condition only, of insertion plates. They are phylogenetically older than the Chitonida, the extinct family Gryphoehitonidae being Paleozoic and the surviving family Lepidopleuridae appearing in the early Tertiary. At the present time the Chitonida are, with a few isolated exceptions, shore- living animals occurring universally between tide-marks, where- 388 C. M. YONGE as the Lepidopleurida seldom occur on the shore. They extend, unlike the Chitonida, into deep water, having been taken from depths exceeding 3,000 fathoms. There appears, as outlined above, good reason lov believing that the characteristic form of the Loricata evolved under shore-living conditions. The evolution of the more highly specialized Chitonida with their more efficient respiratory system and more complicated shell plates, possibly led to the migration into deeper water of the species of the four surviving genera of the Lepidopleurida. Their descendants retain not only the primitive shell structure, but also a smaller number of gills producing a relatively weak respiratory current. The latter will be no disadvantage when they are continually under water, and so do not need to interrupt the current when the tide is out. Eejection of waste material from the inhalant cavity and the associated presence of neural and pedal mucous tracts is probably also a primitive loricate feature, although acquired after these animals evolved from the original molluscan stock. But the presence of adanal gills and the substitution of branchial and lateral sense organs for the osphradium must be considered secondary developments, although certain of the Chitonida have acquired adanal gills independently.

7. SUMMARY. 1. The course of the water currents in the mantle cavity of three species of the Chitonida, and one species of the Lepido- pleurida, has been determined. 2. Inhalant openings are created anteriorly or laterally by local raising of the girdle. The single exhalant opening is always posterior and confined to the region between the last pair of gills. 3. The exhalant current carries with it the genital and excretory products, and, in the Chitonida, the faeces. 4. The bridging of the pallial grooves in the region of the girdle folds by the post-renal gills (and adanal gills in the Lepidopleurida) completes the functional division of the pallial grooves into inhalant and exhalant chambers. 5. The gills possess the typical structure of ctenidia, and their MANTLE CAVITY IN LOBICATA 889 ciliation is a modification only of that of ctenidia. They are to be regarded as multiplied ctenidia and not as secondary structures. 6. The individual filaments are shortened, attached to those of adjacent gills by long interlocking cilia, and have a broad band of lateral cilia which create the respiratory current. 7. Four possible tracts of mucous glands in the pallial grooves are concerned with the consolidation of sediment. The pallial tracts may be homologous with the hypobranchial glands of the Prosobranchia; all are analogous with these. In the Chitonida sediment is rejected only from the exhalant chamber, in the Lepidopleurida mainly from the inhalant chamber. 8. Osphradia,possiblyhomologouswiththoseinthe Gastropoda, occur in the majority of the Chitonida. With them may be associated anterior sense organs. In the Lepidopleurida they are replaced by branchial and lateral sense organs. All are similar in structure and innervation. They have been considered olfactory in function, but with equal reason may be regarded as tactile organs concerned with the estimation of sediment. 9. The Loricata probably evolved between tide-marks, their characteristic structure being admirably adapted for life on the shore. 10. The reasons for the differences between the structure and habits of the Lepidopleurida and the Chitonida are discussed.

8. BEFEKENCES. Arey, L. B., and Crozier, W. J., 1919.—"Sensory Responses of Chiton", ' Journ. Bxp. Zo6l.', 29. Bernard, F., 1890.—"Rech. s. 1. organes palleaux des Gasteropodes Proso- branches", 'Ann. Sci, nat. Zool.', (7), 9. Blumrich, J., 1891.—"Integument der CMtonen", 'Zeit. wiss. Zool.', 52. Bume, R. H-, 1896.—"Anatomy of Hanleyi abyssorum, M. Sars", 'Proo. Malac. Soc. Lond.', 2. Fretter, V., 1937.—"Structure and Function of the Alimentary Canal of some Species of Polyplacophora (Mollusca)", 'Trans. Roy. Soc. Edin.', 59. Haller, B., 1882.—"Organisation der CMtonen der Adria, 1", 'Arb. zool. Inst. Univ. Wien', 4. • 1884.—"Organisation der CMtonen der Adria, 2", ibid., 5. 390 C. M. YONGE

Hulbert, G. C. E. B., and Yonge, C. M., 1937.—"A Possible Function of the Osphradia in the Gastropoda", 'Nature', 139. Ihering, H. von, 1877.—'Vergl. Anat. des Nervensystems und Phylogenie der Mollusken.' Leipzig. Knorre, H. von, 1925.—"Schale und Riickensinnesorgane von Trachyder- mon (Chiton) einereus u. d. ceylonischen Chitonen der Sammlung Plate", 'Jena. Zeit. Naturw.', 61. Pelseneer, P., 1898.—"Morphologie des branchies et des orifices renaux et genitaux des Chitons", 'Bull. Sci. France Belg.', 31. 1899.—"Rech. morphologiques et phylogenetiques sur les Mollusques archaiques", 'Mem. cour. Acad. Belg.', 57. Plate, L. H., 1898.—"Anatomie u. Phylogenie der Chitonen (a)", 'Zool. Jahrb.', Suppl. 4, Fauna Chilensis (1). 1899.—"Anat. u. Phylog. der Chitonen (6)", ibid., 5, Fauna Chilen- sis (2). 1901.—"Anat. n. Phylog. der Chitonen (c)", ibid., 5, Fauna Chilen- sis (2). Spengel, J. W., 1881.—"Geruchsorgane und Nervensystem der Mollusken. Beitr. z. Erkennt. der Einheit des Molluskentypus", 'Zeit. wiss. Zool.', 35. Stork, H. A., 1934.—"Beitr. z. Histol. u. Morphol. des Osphradiums", 'Arch, neerl. Zool.', 1. Thiele, J., 1890.—"tJ. Sinnesorgane der Seitlinie und das Nervensystem von Mollusken", 'Zeit. wiss. Zool.', 49. 1895.—"tJ. d. Verwandtschaftsbeziehungen der Amphineura", 'Biol. Centralb.', 15. Winckworth, R.? 1932.—"BritishMarine Mollusea", 'Journ. Conchol.', 19. Yonge, C. M., 1937a.—"Biology of Aporrhais pes-pelecani and A. ser- resiana", 'Journ. Mar. Biol. Assoc.', 21. 19376.—"Circulation of Water in the Mantle Cavity of Dentalium entalis", 'Proc. Malac. Soc. Lond.', 22. 1938.—"Evolution of Ciliary Feeding in the Prosobranchia, Capulus ungaricus", 'Journ. Mar. Biol. Assoc.', 22.