The principle of respiration is the same in the Mollusca as in all other animals. The blood is purified by being brought, in successive instalments, into contact with pure air or pure water, the effect of which is to expel the carbonic acid produced by animal combustion, and to take up fresh supplies of oxygen. Whether the medium in which a mollusc lives be water or air, the effect of the respiratory action is practically the same.

Broadly speaking, Mollusca whose usual habitat is the water ‘breathe’ water, while those whose usual habitat is the land ‘breathe’ air. But this rule has its exceptions on both sides. The great majority of the fresh-water Mollusca which are not provided with an operculum (e.g. Limnaea, Physa, Planorbis), breathe air, in spite of living in the water. They make periodic visits to the surface, and take down a bubble of air, returning again for another when it is exhausted. On the other hand many marine Mollusca which live between tide-marks (e.g. Patella, Littorina, Purpura, many species of Cerithium, Planaxis, and Nerita) are left out of the water, through the bi-diurnal recess of the tide, for many hours together. Such species invariably retain several drops of water in their branchiae, and, aided by the moisture of the air, contrive to support life until the water returns to them. Some species of Littorina (e.g. our own L. rudis and many tropical species) live so near high-water mark that at neap-tides it must frequently happen that they are untouched by the sea for several weeks together, while they are frequently exposed to a burning sun, which beats upon the rocks to which they cling. In this case it appears that the respiratory organs will perform their functions if they can manage to retain an extremely small amount of moisture.[265]

The important part which the respiratory organs play in the economy of the Mollusca may be judged from the fact that the primary subdivision of the Cephalopoda into Dibranchiata and Tetrabranchiata is based upon the number of branchiae they possess. Further, the three great divisions of the Gasteropoda have been named from the position or character of the breathing apparatus, viz. Prosobranchiata, Opisthobranchiata and Pulmonata, while the name Pelecypoda has hardly yet dispossessed Lamellibranchiata, the more familiar name of the bivalves.

Respiration may be conducted by means of—(a) Branchiae or Gills, (b) a Lung or Lung-cavity, (c) the outer skin.

In the Pelecypoda, Cephalopoda, Scaphopoda, and the great majority of the Gasteropoda, respiration is by means of branchiae, also known as ctenidia[266], when they represent the primitive Molluscan gill and are not ‘secondary’ branchiae (pp. 156, 159).

In all non-operculate land and fresh-water Mollusca, in the Auriculidae, and in one aberrant operculate (Amphibola), respiration is conducted by means of a lung-cavity, or rarely by a true lung, whence the name Pulmonata. The land operculates (Cyclophoridae, Cyclostomatidae, Aciculidae, and Helicinidae) also breathe air, but are not classified as Pulmonata, since other points in their organisation relate them more closely to the marine Prosobranchiata. Both methods of respiration are united in Ampullaria, which breathes indifferently air through a long siphon which it can elevate above the surface of the water, and water through a branchia (see p. 158). Siphonaria (Fig. 57) is also furnished with a lung-cavity as well as a branchia. Both these genera may be regarded as in process of change from an aqueous to a terrestrial life, and in Siphonaria the branchia is to a great extent atrophied, since the animal is out of the water, on the average, twenty-two hours out of the twenty-four. In the allied genus Gadinia, where there is no trace of a branchia, but only a lung-cavity, and in Cerithidea obtusa, which has a pulmonary organisation exactly analogous to that of Cyclophorus,[267] this process may be regarded as practically completed.

Respiration by means of the skin, without the development of any special organ, is the simplest method of breathing which occurs in the Mollusca. In certain cases, e.g. Elysia, Limapontia, and Cenia among the Nudibranchs, and the parasitic Entoconcha and Entocolax, none of which possess breathing organs of any kind, the whole outer surface of the body appears to perform respiratory functions. In others, the dorsal surface is covered with papillae of varied size and number, which communicate with the heart by an elaborate system of veins. This is the case with the greater number of the Aeolididae (Fig. 58, compare Fig. 5, C), but it is curious that when the animal is entirely deprived of these papillae, respiration appears to be carried on without interruption through the skin.

In the development of a distinct breathing organ, it would seem as if progress had been made along two definite lines, each resulting in the exposure of a larger length of veins, i.e. of a larger amount of blood, to the simultaneous operation of fresh air or fresh water. Either (a) the skin itself may have developed, at more or less regular intervals, elevations, or folds, which gradually took the form of papillae, or else (b) an inward folding, or ‘invagination,’ of the skin, or such a modification of the mantle-fold as is described below (p. 172) may have taken place, resulting in the formation of a cavity more or less surrounded by walls, within which the breathing organs were ultimately developed. Sometimes a combination of both processes seems to have occurred, and after a papilliform organ has been produced, an extension or prolongation of the skin has taken place, in order to afford a protection to it. Respiration by means of a lung-cavity is certainly subsequent, in point of time, to respiration by means of branchiae.

The branchiae seem to have been originally paired, and arranged symmetrically on opposite sides of the body. It is not easy to decide whether the multiple form of branchia which occurs in Chiton (Fig. 59), or the simple form as in Fissurella (Fig. 60), is the more primitive. Some authorities hold that the multiple branchia has gradually coalesced into the simple, others that the simple form has grown, by serial repetition, into the multiple. There appears to be no trace of any intermediate forms, and, as a matter of fact, the multiple branchia is found only in the Amphineura, while one or rarely two (never more) pairs of branchiae, occur, with various important modifications, in the vast majority of the Mollusca.

Amphineura.—In Chiton the branchiae are external, forming a long row of short plumes, placed symmetrically along each side of the foot. The number of plumes, at the base of each of which lies an osphradial patch, varies from about 70 to as few as 6 or 7. When the plumes are few, they are confined to the posterior end, and thus approximate to the form and position of the branchiae in the other Amphineura. In Chaetoderma, the branchiae consist of two small feather-shaped bodies, placed symmetrically on either side of the anus, which opens into a sort of cloaca within which the branchiae are situated. In Neomenia the branchiae are still further degraded, consisting of a single bunch of filaments lying within the cloaca, while in Proneomenia there is no more than a few irregular folds on the cloaca-wall (Fig. 61).

In the Prosobranchiata, symmetrically paired branchiae occur only in the Fissurellidae, Haliotidae, and Pleurotomariidae, in the former of which two perfectly equal branchiae are situated on either side of the back of the neck. These three families taken together form the group known as Zygobranchiata.[268] In all other families the asymmetry of the body has probably caused one of the branchiae, the right (originally left), to become aborted, and consequently there is only one branchia, the left, in the vast majority of marine Prosobranchiata, which have been accordingly grouped as Azygobranchiata. Even in Haliotis the right branchia is rather smaller than the left, while the great size of the attachment muscle causes the whole branchial cavity to become pushed over towards the left side. In those forms which in other respects most nearly approach the Zygobranchiata, namely, the Trochidae, Neritidae, and Turbinidae, the branchia has two rows of filaments, one on each side of the long axis, while in all other Prosobranchiata there is but one row (see Fig. 79, p. 169).

In the great majority of marine Prosobranchiata the branchia is securely concealed within a chamber or pouch (the respiratory cavity), which is placed on the left dorsal side of the animal, generally near the back of the neck. For breathing purposes, water has to be conveyed into this chamber, and again expelled after it has passed over the branchia. In the majority of the vegetable-feeding molluscs (e.g. Littorina, Cerithium, Trochus) water is carried into the chamber by a simple prolongation of one of the lobes or lappets of the mantle, and makes its exit by the same way, the incoming and outgoing currents being separated by a valve-like fringe depending from the lobe. In the carnivorous molluscs, on the other hand, a regular tube, the branchial siphon, which is more or less closed, has been developed from a fold of the mantle surface, for the special purpose of conducting water to the branchia. After performing its purpose there, the spent water does not return through the siphon, but is conducted towards the anus by vibratile cilia situated on the branchiae themselves. In a large number of cases, this siphon is protected throughout its entire length by a special prolongation of the shell called the canal. Sometimes, as in Buccinum and Purpura, this canal is little more than a mere notch in the ‘mouth’ of the shell, but in many of the Muricidae (e.g. M. haustellum, tenuispina, tribulus) the canal becomes several inches long, and is set with formidable spines (see Fig. 164, p. 256). In Dolium and Cassis the canal is very short, but the siphon is very long, and is reflected back over the shell.

The presence or absence of this siphonal notch or canal forms a fairly accurate indication of the carnivorous or vegetarian tendencies of most marine Prosobranchiata, which have been, on this basis, subdivided into Siphonostomata and Holostomata. But this classification is of no particular value, and is seriously weakened by the fact that Natica, which is markedly ‘holostomatous,’ is very carnivorous, while Cerithium, which has a distinct siphonal notch, is of vegetarian tendencies.

In the Zygobranchiata the water, after having aerated the blood in the branchiae, usually escapes by a special hole or holes in the shell, situated either at the apex (Fissurella) or along the side of the last whorl (Haliotis). In Pleurotomaria the slit answers a similar purpose, serving as a sluice for the ejection of the spent water, and thus preventing the inward current from becoming polluted before it reaches the branchiae (see Fig. 179, p. 266).

In Patella the breathing arrangements are very remarkable. In spite of their apparent external similarity, this genus possesses no such symmetrically paired plume-shaped branchiae as Fissurella, but we notice a circlet of gill-lamellae, which extends completely round the edge of the mantle. It has been shown by various authorities that these lamellae are in no sense morphologically related to the paired branchiae in other Mollusca, but only correspond to them functionally. The typical paired branchiae, as has been shown by Spengel, exist in Patella in a most rudimentary form, being reduced to a pair of minute yellow bodies on the right and left sides of the back of the ‘neck.’ A precisely similar abortion of the true branchiae, and special development of a new organ to perform their work, is shown in Phyllidia and Pleurophyllidia (see below under Opisthobranchiata). This circlet of functional gills in Patella has therefore little systematic value, being only developed in an unusual position, like the eyes on the mantle in certain Pelecypoda, to supply the place of the true organs which have fallen into disuse. Accordingly Cuvier’s class of Cyclobranchiata, which included Patella and Chiton, has no value, and has indeed long been discarded. In Chiton the gills never extend completely round the animal, but are always more or less interrupted at the head and anus. They are the true gills, the plumes being serially repeated in the same way as the shell plates.

In the land Prosobranchiata (Cyclostomatidae, Cyclophoridae, Aciculidae, Helicinidae) which, having exchanged a marine for an aerial life, breathe air instead of water, the branchia has completely disappeared, and breathing is conducted, as in the Pulmonata, by a lung-cavity. In certain genera of land operculates, e.g. Pupina, Cataulus, Pterocyclus, a slight fissure or tube in the last whorl (see Fig. 180, p. 266) serves to introduce air into the shell, which is perhaps otherwise closed to air by the operculum. In Aulopoma, which has no tube, the operculum admits free circulation of air. In certain other Cyclostomatidae the apex is truncated, and air can enter there. De Folin closed with wax the aperture of Cycl. elegans, and found that on placing it in a pneumatic machine, the shell gave off air through its whole surface. On the other hand, Cylindrella and Stenogyra decollata, on being submitted to the same test, showed that the truncated part alone was permeable by air.

Fischer and Bouvier have made some interesting observations on the breathing of a species of Ampullaria (insularum Orb.). The species has, in common with all Ampullaria, two siphons, but while the right siphon is but slightly developed, the left is very long, almost twice as long as the shell (see Fig. 65). The animal, when under the water, lengthens its siphon, brings the orifice to the surface, and by alternately raising and depressing its head produces in the pulmonary sac movements of ex- and inspiration; these are repeated about ten or fifteen times at regular intervals of from six to eight seconds, a method of respiration strongly resembling that of the Cetacea. At the same time, branchial respiration takes place. If powdered carmine is added to water, the particles are seen to enter the branchial cavity by the siphon and pass out by the short right siphon. Sometimes the animal remains under water for hours without rising to the surface to inspire air. In Valvata (Fig. 66) the branchia is very large, and projects like a leaf or fan above the shell on the left side; on the corresponding position on the right side is a long filiform appendage, which some have regarded as representing the other branchia.

Opisthobranchiata.—A true branchia occurs only in the Tectibranchiata and the Ascoglossa. It lies on the right side, and is usually more or less external, being partly covered sometimes by the shell (as in Umbrella, Fig. 5), sometimes by a fold of the mantle. In the Pteropoda (which are probably derived from the Tectibranchiata), all the Thecosomata, with the exception of Cavolinia, have no specialised branchia, but probably respire through portions or the whole of the integument. In the Gymnosomata an accessory branchia has in many cases been developed at the posterior end of the body. Pneumodermon alone has both lateral and posterior branchiae well developed, Clione and Halopsyche are destitute of either, while the four remaining families have one branchia, sometimes lateral, sometimes posterior.[269]

Certain of the Nudibranchiata possess no special breathing organs, and probably respire through the skin (Elysia, Limapontia, Cenia, Phyllirrhoë). The majority, however, have developed secondary branchiae, in the form of prominent lobes or leaf-like processes (the cerata), which are carried upon the back, without any means of protection. These cerata are, as a rule, of extreme beauty and variety of form, consisting sometimes of long whip-like tentaculae, in other cases of arborescent plumes of fern-like leafage, in others of curious bead-like appendages of every imaginable shape and colour. In Doris they lie at the posterior end of the body, in a sort of rosette, which is generally capable of retraction into a chamber. In Phyllidia and Pleurophyllidia these secondary branchiae lie, as in Patella, on the lateral portions of the mantle.

The Scaphopoda in all probability possess neither true nor secondary branchiae.

Pulmonata.—When we use the term ‘lung,’ it must be remembered that this organ in the Mollusca does not correspond, morphologically, with the spongy, cellular lung of vertebrates; it simply performs the same functions. The ‘lung,’ in the Mollusca, is a pouch or cavity, lined with blood-vessels which are disposed over its vaulted surface in various patterns of network. The pulmonary sac or cavity is therefore a better name by which to denote this organ.

It seems probable, as has been already shown (pp. 18–22), that all Pulmonata are ultimately derived from marine forms which breathed water by means of branchiae. Thus we find intermediate forms, such as Siphonaria, possessed of both a branchia and a pulmonary sac, the former being evanescent, while in Gadinia and Amphibola it has quite disappeared. In the vast majority of Pulmonata no trace of a branchia remains; its function is performed by a chamber, always situated at the right side of the animal, and generally more or less anterior, admitting air by a narrow aperture which is rhythmically opened and closed. In Arion and Geomalacus (Fig. 69) this aperture is in the front of the right side of the ‘shield,’ in Limax (Fig. 71) in the hinder part, in Testacella (Fig. 20) it is near the extremity of the tail, under the spire of the shell; in Janella it is on the middle of the right edge of the shield (Fig. 70). If a specimen of Helix aspersa, or better, of H. pomatia, is held up to the light, the beautiful arborescent vessels, with which the upper part of the pulmonary chamber is furnished, can be clearly seen by looking through the aperture as it dilates. It is only in the Auriculidae that an actual spongy mass of lung material appears to exist. When in motion, a Helix inspires air much more frequently than when at rest. Temperature, too, seems to affect the number of inspirations; it appears doubtful whether, during hibernation, a snail breathes at all. In any case, the amount of air required to sustain life must be small.

With regard to the respiration of fresh-water Pulmonata there appears to be some difference of opinion. It is held, on the one hand, that the Limnaeidae only respire air, making periodic visits to the surface to procure it, and that they perish, if prevented from doing so, by asphyxiation. If, we are told,[270] as a Limnaea is floating on the surface of the water in a glass jar, a morsel of common salt be dropped upon its outstretched foot, it will sink heavily to the bottom, emitting a stream of air from its pulmonary orifice. On recovering from the shock, it will anxiously endeavour to regain the surface, but will have some difficulty in doing so, owing to its now much greater specific gravity. When it succeeds, it creeps almost out of the water, and exposes its respiratory orifice freely to the air. If the experiment is repeated several times on the same individual, it becomes so much weakened that it has to be taken out of the water to save its life. Moquin-Tandon, on the other hand, is strongly of opinion[271] that there is no absolute necessity for Limnaea to obtain air by rising to the surface, and that, if prevented from emerging, it can obtain air from the water. When covered in by a roof of ice, Limnaea has not been observed to suffer any inconvenience. Moquin-Tandon kept L. glabra and Planorbis rotundatus in good health under 20 mm. of water for eighteen and nineteen days, and relates a case in which Physa was kept alive under water for four days, and Planorbis for twelve. Young specimens, both of Limnaea and Planorbis, do not rise to the surface for a supply of air; they are hatched with the pulmonary cavity full of water.

It is probable, therefore, that Limnaeidae are capable, on occasion, of respiration through the skin. Some authorities are of opinion that certain long and narrow lamellae, situated within the pulmonary sac, are employed for the purpose of aqueous respiration. Ancylus, which never makes periodic excursions to the surface, perhaps respires by receiving into its pulmonary chamber the minute quantities of oxygen given off by the vegetation on which it feeds.

Limnaeidae taken from a great depth of water, e.g. from 130 fathoms in the lake of Geneva, have been examined by Forel.[272] The pulmonary sac is full of water, but there is no transformation of organs, no appearance of a branchia, to meet the changed circumstances of their environment. Doubtless a good deal of respiration is done by the skin; being soft and vascular, it respires the air dissolved in the water. Forel cites cases of Limnaea living at much shallower depths, which come to the surface once, and then remain below for months. The oxygen of this supply must soon have become exhausted, and the animals, discontinuing for a time the use of the pulmonary chamber, must have respired through the skin. Shallow-water Limnaea, according to the same authority, remain beneath the surface during cold weather; when warm weather returns they rise to the surface to take in a supply of air. Since the water at great depths is always very cold, there is no need for the Limnaea living there to rise to the surface at all.

It is a curious fact that Limnaea, which have been respiring by the skin for the whole winter, should suddenly, on the first warm days of summer, take to rising to the surface and breathing air. But exactly the same phenomenon is shown in the case of Limnaea from great depths. Placed in an aquarium, they immediately begin rising to the surface and inspiring air; in other words, they experience instantaneously a complete transformation of their respiratory system.

In Onchidium, a land pulmonate which has retrogressed to an amphibious or quasi-marine mode of life, there is no organ which represents the pulmonary or branchial cavity, the so-called lung being only a cavity of the kidney. Respiration is, however, conducted by the skin as well, and by the dorsal papillae.[273]

Land Mollusca can sustain, for a considerable time, complete deprivation of atmospheric air. Helices placed in an exhausted receiver show no signs of being inconvenienced for about 20 hours, and are able to survive for about two or three days. If detained under water, they are very active for about 6 hours, then become motionless, the body swells, owing to the water absorbed, and death ensues in about 36 hours. Immersion for only 24 hours is generally followed by recovery. In the latter case, the cause of death is not so much deprivation of air as compulsory absorption of water by the skin. The amount of water thus taken up is surprising. Spallanzani found that a Helix which weighed 18 grammes increased in weight by 13½ grammes after a prolonged immersion. Even slugs enclosed in moist paper gained more than 2 grammes in the course of half an hour. Experiment has shown that the amount of carbonic acid gas produced by respiration stands in direct relation to the amount of food consumed. Four pairs of snails were taken which had recently awakened from their winter sleep and had eaten heartily, and an equal number, under the same circumstances, which had been prevented from eating. It was found that the first four pairs produced, in consuming a given amount of oxygen, 11, 9, 10, and 13 parts respectively of carbonic acid, while the second set produced, in consuming the same amount of oxygen, only 4, 8, 7, and 9 parts of carbonic acid.[274] Hibernating Helices, if weighed in December and again in April, will be found to have lost weight, due to the expiration of carbonic acid. Owing to the difficulty of experiment, opinions vary as to the absolute temperature of snails. It appears to be established that several snails, if placed together in a tube, raise the temperature one or two degrees C., but as a rule, the temperature of a solitary Helix differs very slightly from that of the surrounding air. Increased activity, whether in respiration or feeding, is found to raise the temperature.

W. H. Dall, writing of the branchia in Pelecypoda, remarks[275] that there can be no doubt that its original form was a simple pinched-up lamella or fold of the skin or mantle. This, elongated, becomes a filament. Filaments united by suitable tissue, trussed, propped, and stayed by a chitinous skeleton, result in the forms, wonderful in number and complexity, which puzzle the student to describe, much more to classify.

In Pelecypoda the branchiae are placed on each side of the body, between the mantle and the visceral mass. They lie in a chamber known as the branchial cavity. Leading into this cavity, and behind it, are, as a rule, two tubes or siphons, one of which conducts water to the branchiae, while the other carries it away after it has passed over them. The lower is known as the branchial or afferent siphon, the upper as the anal or efferent siphon (see Figs. 72 and 73). The action of these siphons can readily be observed by placing a little carmine in water, near to the siphonal apertures of an Anodonta or Unio. In many cases (e.g. Psammobia, Tellina, Mya, genera which burrow deeply in sand) both the siphons are exceedingly long, sometimes considerably longer than the whole shell. In some cases the two tubes are free throughout their entire length, in others they become fused together before their entrance within the shell (Fig. 74). In other genera, which do not burrow (e.g. Ostrea, Pecten, Arca, Mytilus), the siphons are rudimentary or altogether absent (Fig. 75).

The number and arrangement of the branchiae varies considerably. It appears probable that the different degrees of complication of the gill indicate degrees of specialisation in the different groups of Pelecypoda, in other words, assuming that a simpler form of gill precedes, in point of development, a more complicated form, the nature of the gill may be taken as indicating different degrees of removal from the primitive form of bivalve.

1. The simplest form of gill (Nucula, Leda, Solenomya, etc.) is that which consists (Fig. 76, A, compare Fig. 100, p. 201) of two rows of very short, broad, not reflected filaments, the rows being placed in such a way that they incline at right angles to one another from a common longitudinal axis. The filaments are not connected with one another, nor are the two leaves of each gill united at any point. (Protobranchiata.)

2. In the Anomiidae, Arcadae, Trigoniidae, and Mytilidae each gill consists of two plates or rows of much longer filaments, which consequently occupy a much larger space in the mantle cavity (Fig. 76, B). Unable to extend beyond the limits of the mantle, filaments are reflected or doubled back upon one another, those of the external plate being reflected towards the outside, those of the internal plate towards the inside. Each separate filament is not connected with the filament next adjacent, except by surface cilia situated on small projections on the sides of the filaments, and interlocking with the cilia of the adjacent filament. The two superposed plates or leaves of the gill may or may not be united by cords running between the two parts of a filament. (Filibranchiata.)

3. In the Pectinidae, Aviculidae, and Ostreidae a further development takes place. The filaments of each gill are reflected in the same way as in the Filibranchiata, but the part thus reflected may become completely united or ‘concresce’ with the mantle on the exterior and with the base of the foot on the interior side. The leaves of each gill plate, which have thus become doubled (the gills being apparently two instead of one on each side), are folded or crumpled, and the filaments are modified at the re-entrant angles of the fold. (Pseudolamellibranchiata.)

4. In all the remaining Pelecypoda, except class 5, in other words, in the very large majority of families, the filaments are either reflected, as in (3), or simple; but the process of concrescence is so far advanced that the adjacent filaments are always intimately connected with one another in such a way as to admit the passage of the blood; and the leaves of each gill-plate (Fig. 76, C) are united by cross channels in a similar way. (Eulamellibranchiata.)

5. In certain of the Anatinacea alone (Cuspidaria, Lyonsiella, Poromya, Silenia) the gills are transformed into a more or less muscular partition, extending from one adductor muscle to the other (Fig. 76, D), and separating off the pallial chamber into two distinct divisions, which communicate by means of narrow slits in the partition. (Septibranchiata.)

Thus the process of gill development in the Pelecypoda appears to lead up from a simple to a very complex type. In its original form, at all events in the most primitive form known to us, the gill is a series of short filaments, quite independent of one another, strung in two rows; then the filaments become longer and double back, while at the same time they begin to show signs of adhesion, as yet only superficial, to one another. In a further stage, the reflected portions become fused to the adjacent surfaces of the foot and mantle, while the interlamellar junctions serve to lock the two gill-plates together; finally, the mere ciliary junction of adjacent filaments is exchanged for intimate vascular connection, while the gill-plates as a whole become closely fused together in a similar manner.

This theory of origin is strengthened by closer observation of the phenomena of a single group. Taking the Septibranchiata as an instance, we find that in Lyonsiella the branchiae unite with the mantle in such a way as to form two large pallial chambers, the structure of the branchiae being preserved, and their lamellae covering the partition. A further stage is observed in Poromya. There, a similar partition exists, but it has become muscular, preserving, however, on each side two groups of branchial lamellae, separated one from the other by a series of slits, which form a communication between the two pallial chambers. A further stage still is seen in Silenia. There the same muscular partition exists, but the branchial lamellae on either side have disappeared, the slits between the two chambers, which occur in Poromya, still persisting, but separated into three groups. Cuspidaria represents the last stage in the development. In the ventral chamber there appears nothing at all corresponding to a branchia; the surface of the partition appears perfectly uniform, but on careful examination three little separate orifices, remains of the three groups of orifices in Silenia, are observed.[276]

Relation between Branchiae and Heart.—The object of the branchiae being, as has been already stated, to aerate the blood on its way to the heart, we find that the heart and the branchiae stand in very important structural relations to one another. When the branchiae are in pairs, we find that the auricles of the heart are also paired, the auricle on the right and left sides being supplied by the right and left branchiae respectively. This is the case with the Dibranchiate Cephalopods (Argonauta, Octopus, Loligo, etc.), the Zygobranchiate Prosobranchs (Fissurella, Haliotis), and all Pelecypoda. In the Amphineura (Chiton, etc.) there are two auricles corresponding to the two sets of multiple branchiae. In the case of the Tetrabranchiate Cephalopods (Nautilus) there are four auricles corresponding to each of the four branchiae. Compare Fig. 79, A, B, C, D, E.

On the other hand, when the branchia is single, or when both branchiae are on the same side, and one is aborted and functionless, the auricle is single too, and on the same side as the branchia. This is the case with the Tectibranchiate Opisthobranchs (Philine, Scaphander, etc.), all the Pectinibranchiate Prosobranchs (Rachiglossa, Taenioglossa, and Ptenoglossa), and the other Azygobranchiate Prosobranchs (Trochidae, Neritidae, etc.). In the last case the right auricle exists, as well as the left, but is simply a closed sac, the coalescing of the two gills on the left side having thrown all the work upon the left auricle. Compare Fig. 79, F, G, H.