It is the generally accepted opinion among men of science that all life originated in the sea. Not that all parts of the sea are equally favourable to the development of forms of life. The ocean surface, with its entire absence of shelter or resting-place, and the deep sea, whose abysses are always dark and cold and changeless, offer little encouragement to plant or animal life, as an original starting-point. True, both the surface and the depths of the sea have become colonised by myriads of forms, Mollusca amongst them, but these quarters are in the truest sense colonised, for the ancestors of those who inhabit them in all probability migrated from elsewhere.

It was no doubt the littoral region and the shallow waters immediately below it, a region of changeable currents, of light and shade, of variation, within definite limits, of temperature and tide effects, which became the scene of the original development of plant life, in other words, of the food-supply which rendered possible its colonisation by higher animals. But the littoral region, besides the advantages of tenancy which it offers to animal life, has also its drawbacks. The violence of the surf may beat its inhabitants in pieces, the retreat of the tide exposes them, not merely to innumerable enemies in the shape of predatory birds and beasts, but also to a change in the atmospheric medium by which they are surrounded. Hence, in all probability, have arisen the various forms of adaptation which are calculated to bring about the ‘survival of the fittest’; hence, to narrow our point of view to the Mollusca, the development of hard shells, or exoskeletons, hence the sand-burrowing, rock-boring, rock-clinging instincts of various genera and species.[1]

What was the primitive form of molluscan life is little likely to be ever positively known, although, on grounds of comparative anatomy, something approaching to the archi-mollusc is often constructed, with more or less probability, by careful observers. From one of the oldest known geological strata, the Cambrian, nearly four hundred species of Mollusca are known, which include representatives of nearly all the great Orders as they exist at the present day, and without the slightest sign of approximation to one another. With regard to the origin of the land and fresh-water Mollusca some definite conclusions can be arrived at, which will be given in their proper place.

Scarcely any portion of the coast-line of the world is destitute of molluscan life, except in regions where extreme cold forbids its existence. Thus along the shores of Northern Asia there is no proper littoral fauna, the constant influence of travelling ice sweeping it all away; animal life begins at about three fathoms. But in every coast region not positively hostile to existence Mollusca make their home. Each description of habitat has its own peculiar species, which there flourish best, and exist precariously, if at all, elsewhere. Thus the sandy waste of estuaries, the loose and shingly beaches, the slimy mud-flats beset with mangroves, the low stretches of jagged rock, and even the precipitous cliffs, from whose base the sea never recedes, have all their own special inhabitants. The same is true of the deep sea, and of the ocean surface. And when we come to examine the land and fresh-water Mollusca, it is found not merely that some Mollusca are terrestrial and others fluviatile, but that certain species haunt the hills and others the valleys, some the recesses of woods and others the open meadow sides, some prefer the limestone rocks, others the sandy or clayey districts, some live only in still or gently moving waters, while others are never found except where the current is rapid and powerful.

It is within the tropics that the Mollusca become most numerous, and assume their finest and quaintest forms. A tropical beach, especially where there is a good tide-fall and considerable variety of station, abounds in molluscan life to an extent which must literally be seen to be believed. The beach at Panama, to select an instance familiar to the present writer, is astonishingly rich in species, which probably amount in all to several hundreds. This is due to the immense variety of habitat. On the rocks at high-water mark, and even above them, occur Truncatella, Melampus, Littorina, and Siphonaria; where a mangrove-swamp replaces the rock, on the branches overhead are huge Littorina, while three species of Cerithidea crawl on the mud, and Cyrena and Arca burrow into it. Lower down, in the rock pools at half-tide mark are Cerithium, Purpura, Omphalius, Anachis (2 sp.), Nassa, and several Crepidula. At low-water mark of ordinary tides, under stones half buried in clean sand, are Coecum and Vitrinella; under the blocks which rest on solid rock are Cypraea (4 or 5 sp.), Cantharus, more Anachis, Columbella (3 sp. including the graceful C. harpiformis), and Nitidella. Where the blocks of rock are rather muddy, Conus lurks, and with it Turritella and Latirus. Where the rocks form a flat-topped platform 2 or 3 feet high, with here and there a deep crack, huge Chitons 3 inches long conceal themselves, with two species of Turbo, Purpura, and Clavella. At extreme low-water mark of spring tides, on the isolated rocks are Monoceros, Leucozonia, and Vermetus, in them are Pholas and a burrowing Mytilus, under them are more Conus, Dolium, and huge frilled Murices. Patches of clean gravelly sand here produce Strombus; on the operculum of the great Str. galea is sure to be a Crepidula, exactly fitting its breadth. On the liquid mud-flats to the north glide about Marginella, Nassa, and Truncaria, in the clean sandstretch to the west Olivella ploughs about by hundreds with several species of Natica, and Tellina and Donax bury themselves deep, while farther down are Artemis, Chione, and, where mud begins to mix with the sand, Mytilus and more Arca. Each of these species has its own habitat, often circumscribed to a few square feet at the most, and it would be utterly useless to seek for it anywhere except in its own special domain.

Equally abundant are the land Mollusca of the tropics. Prof. C. B. Adams relates that within the limits of a single parish in Jamaica, named Manchester, which measures no more than four miles long and one mile broad, he obtained no fewer than one hundred species. Mr. J. S. Gibbons, in a description of the Mollusca he obtained near St. Ann’s, Curaçao, gives a lively picture of their abundance in an exceptionally favoured locality:—[2]

“Near the outskirts of the town a waste piece of ground supplied me with occupation for all the time I had to spare. Neither grass nor water was to be seen, the only vegetation consisting of a few stunted cacti and still fewer acacia bushes. This, however, was so rich in shells that of several species enough specimens could have been collected in a few yards to supply, I should suppose, all the shell cabinets in the world.... The stones, plants, and ground were covered with Strophia uva L., Tudora megacheila, P. and M., was in equal abundance, suspended by its silk-like thread from acacia boughs, or strewed thickly on the ground underneath. A Bulimulus (B. multilineatus var. sisalensis) abounded on the smaller boughs, while under masses of coral Macroceramus inermis Gundl., Pupa parraiana d’Orb, and P. pellucida Pfr., were abundant. In the loose soil Cylindrella Raveni Bland, Cistula Raveni Bland, and a curious Cionella were so numerous that a spade would have been the best instrument with which to collect them. I wasted a good deal of valuable time in separating them from the soil, when by simply taking away a few handfuls of mould, I might have obtained a larger number of specimens. A species of Stenogyra and a Succinea complete a list, all of which might have been gathered from almost any square yard of ground on the hillside.”

Position of Mollusca in the Animal Kingdom.—Up to very recent times it was usual to regard the Mollusca as one of the four subdivisions of a great family known as Malacozoa, the subdivisions being (1) Mollusca, (2) Tunicata, (3) Brachiopoda, (4) Polyzoa or Bryozoa. This classification is still retained in the leading modern manual on the subject.[3] The progress, however, of investigation leads to the belief that the Mollusca are not so closely related to these other groups as such a classification would seem to imply. The Tunicata, for instance, appear, from the whole course of their development, to occupy a position near to the Vertebrata. The relations of the Brachiopoda and Polyzoa will be more particularly referred to in that part of this History which deals especially with those groups. The position of the Mollusca is, in many respects, one of considerable isolation. Any attempt, therefore, definitely to relate them to one group or another, is, in all probability, to go further than the present state of our knowledge warrants. Especially to be deprecated are systems of classification which confidently derive the Mollusca in general from this or that group. The first undisputed traces of animal life, which appear in the Cambrian epoch, exhibit the same phyletic distinctions as now exist. Sponges, Echinoderms, Mollusca, and Worms, formed already, in those immeasurably remote ages, groups apparently as generally distinct from one another as they are at the present time. It would seem that any theory of development, which confidently teaches the derivation of any one of these groups from any other, is, in the present state of the evidence before us, hazardous in the extreme.

Some indications of relationship, which must not be pushed too far, may be drawn from a consideration of embryonic resemblance. An especial characteristic of the Mollusca is the possession of a particular form of larva, which occurs in one of the stages of development, known as the trochosphere (see p. 130). This form of larva is shared with two orders of Annelida, the Chaetopoda and the Gephyrea armata, and, in all probability, with the Polyzoa as well. It may also be significant that the adult form in Rotifera bears a close resemblance to the trochosphere larva in those groups.

Basis of Classification.—The Mollusca are divided into four great Orders—Cephalopoda, Gasteropoda, Scaphopoda, and Pelecypoda.[4] Each name, it will be noticed, bears reference to the ‘foot,’ i.e. to the organ of motion which corresponds in function to the foot in the Vertebrata.

In the Cephalopoda the feet, or, as they are more frequently termed, the ‘arms,’ are arranged symmetrically round the head or mouth. The common forms of ‘cuttle-fish’ (Octopus, Loligo) are familiar examples of Cephalopods.

The Gasteropoda crawl on the flat under-surface or ‘sole’ of the foot. Snails, slugs, sea-hares, whelks, periwinkles, and coats-of-mail or chitons are examples of this Order.

The Scaphopoda possess a long tubular shell open at both ends; with their small and elongated foot they are supposed to dig into the mud in which they live. The common Dentalium or tusk-shell of our coasts is a representative of this Order.

The Pelecypoda[5] are enclosed in a bivalve shell fastened by a muscular hinge, the adjacent part of the valves being generally more or less toothed; the foot is as a rule roughly comparable to the shape of an axe-head.

To these four Orders is frequently added a fifth, the Pteropoda, whose exact position is at present not absolutely settled. The Pteropoda[6] are ‘pelagic,’ i.e. they live in the open waters of the ocean, rising to the surface at night, and sinking into cooler water by day. They are provided with a pair of wing-like appendages or ‘feet,’ on each side of the head, by means of which they are enabled to swim. Some authorities regard the Pteropoda as a subdivision of Gasteropoda, others as forming a separate Order, of equivalent value to the other four. The question will be further discussed below (see chap. xv.), but for the present it will be sufficient to state that the weight of evidence appears to show that the Pteropoda are modified Gasteropoda, with special adaptations to pelagic life, and are therefore not entitled to rank as a separate Order.

Some writers conveniently group together the first three of these Orders, the Cephalopoda, Gasteropoda, and Scaphopoda, under the title Glossophora,[7] or Mollusca furnished with a radula or ribbon-shaped ‘tongue,’ set with rows of teeth and situated in something of the nature of a head, as distinguished from the Aglossa (or Lipocephala),[8] i.e. those Mollusca which have no radula and no head. To the latter belong only the fourth Order, the Pelecypoda. This view postulates, for the primitive ancestral Mollusc, a body with a more or less developed head, and possibly the rudiments of an apparatus for grinding or triturating food. This form, it is held, either developed or degenerated. In the former case, in consequence of the more active mode of life upon which it may be supposed to have entered, it gave rise to all the more highly organised forms which are grouped under the three great Orders. When, on the other hand, the ancestral form associated itself with an inactive or sedentary life, it was, we may believe, modified accordingly, and either lost by atrophy or failed to acquire those special points of organisation which characterise the highly-developed form. Hence the Pelecypoda, or bivalves, whose characteristic is the absence of any definite cephalic region or masticatory apparatus. It is a remarkable fact in support of this theory of the origin of the Aglossa that certain of their larvae are known to possess traces of higher organisation, e.g. an external mouth and eyes, the former of which becomes covered by the mantle lobes, while the latter disappear long before the adult stage is reached.

Thus we have

Classification of Gasteropoda.—The Gasteropoda are numerically very largely in excess of the two other Orders of the Glossophora, far more complicated as regards classification, and contain a large proportion of those examples of the Mollusca which are most familiar to the ordinary observer. It will therefore be convenient to postpone for the present a fuller discussion of the subdivisions of the Cephalopoda and Scaphopoda, as well as of the Aglossa, returning to them again in special chapters (chaps. xiii. and xvi.), and to devote a few introductory words to the classification and relations of the Gasteropoda.

The Gasteropoda are divided into four Classes, Amphineura, Prosobranchiata, Opisthobranchiata, and Pulmonata.

(1) The Amphineura[9] are bilaterally symmetrical Mollusca, i.e. with organs either single and central, or paired and disposed on either side of the longer axis of the animal. The shell, when present, is never spiral, but consists of eight overlapping plates, kept together by an elliptical girdle. The Amphineura are divided into (a) Polyplacophora,[10] or Chitons, and (b) Aplacophora (Chaetoderma and Neomenia).

(2) The Prosobranchiata[11] are so named from the fact that the breathing organ (branchia or ctenidium[12]) is as a rule situated in front of the heart, the auricle at the same time being in front of the ventricle. They are asymmetrical, almost always furnished with a shell, which is at some time spiral, and with an operculum. The sexes are separate. They are either marine animals, or can be shown to be more or less directly derived from genera which are marine. They are divided into (a) Diotocardia[13] (Haliotis, Fissurella, Trochus, Nerita, Patella), which have, or whose immediate ancestors are believed to have had, two auricles to the heart, two sets of breathing organs, two kidneys, but no proboscis, penis, or siphon, and (b) Monotocardia,[14] in which the heart has only one auricle, the true breathing organ is single, and there is a single kidney. To this division belong the great majority of marine univalve Mollusca, e.g. Cypraea, Buccinum, Murex, Littorina, Ianthina, all the land and fresh-water operculates (Cyclostoma, Melania, Paludina, etc.), as well as the Heteropoda, which are a group of Prosobranchiata which have betaken themselves to a pelagic life.

(3) In the Opisthobranchiata[5] the breathing organs (when present) are behind the heart, and the auricle of the heart is consequently behind the ventricle. They are asymmetrical marine animals; usually, but by no means always, without a shell, scarcely ever with an operculum in the adult state. The sexes are united in the same individual. The Opisthobranchiata fall into two divisions: (a) Tectibranchiata, in which the breathing organ is more or less covered by the mantle, and a shell is usually present, which is sometimes rudimentary, e.g. Bulla, Aplysia, Umbrella, and the whole group of Pteropoda; (b) Nudibranchiata, or sea slugs, which have no shell and no true ctenidia, but breathe either by the skin, or by ‘cerata’ or papilliform organs prominently developed on the back: e.g. Doris, Aeolis, Dendronotus.

(4) The Pulmonata[15] are asymmetrical air-breathing non-marine Mollusca, generally, but not always, furnished with a shell. The sexes are always united in the same individual, and the operculum is always wanting, except in Amphibola. They are conveniently divided into Stylommatophora,[16] in which the eyes are at the tip of the upper tentacles, which are retractile (Helix, Limax, Bulimus, and all true land slugs and snails), and Basommatophora, in which the eyes are at the base of the tentacles, which are not retractile (Limnaea, Planorbis, Physa, and all the Auriculidae).

Thus we have

The relation of the four great Orders to one another will be better discussed when we come to deal with each Order separately. The problem of the origin and mutual relationship of the various forms of molluscan life is of extreme subtlety, and its solution can only be approached after a comprehensive survey of many complicated anatomical details. But there is one branch of the Mollusca—the land and fresh-water genera—whose origin is, comparatively speaking, of recent date, and whose relationships are therefore less likely to have suffered complete obliteration.

Origin of the Land and Fresh-water Mollusca.—The ultimate derivation of the whole of the land and fresh-water molluscan fauna must, as has already been remarked, be looked for in the sea. In certain cases the process of conversion, if it may be so termed, from a marine to a non-marine genus, is still in progress, and can be definitely observed; in others the conversion is complete, but the modification of form has been so slight, or the date of its occurrence so recent, that the connexion is unmistakable, or at least highly probable; in others again, the modification has been so great, or the date of its occurrence so remote, that the actual line of derivation is obscured or at best only conjectural.

This passage from a marine to a non-marine life—in other words, this direct derivation of non-marine from marine genera—is illustrated by the faunal phenomena of an inland brackish-water sea like the Caspian, which is known to have been originally in connexion with the Mediterranean, and therefore originally supported a marine fauna. The Mollusca of the Caspian, although without exception brackish- or fresh-water species, are in their general facies distinctly marine. Of the 26 univalve species which inhabit it 19 belong to 4 peculiar genera (Micromelania, Caspia, Clessinia, Nematurella), all of which are modified forms of the marine Rissoidae. The characteristic bivalves belong to the genera Adacna, Didacna, and Monodacna, all of which can be shown to be derived from the common Cardium edule. We have here a case where complete isolation from the sea, combined no doubt with a gradual freshening of the water, has resulted in the development of a number of new genera. The singularly marine facies of several of the fresh-water genera now inhabiting Lake Tanganyika, has given rise to the belief, among some authorities, that that lake was at one time an inlet of the Indian Ocean. In the upper waters of the Baltic, marine and fresh-water Mollusca flourish side by side. So complete is the intermixture, that an observer who had lived on no other shores would probably be unable to separate the one set of species from the other.[18] Thus between Dagö and Papen-Wiek[19] Mytilus edulis, Cardium edule, Tellina balthica, Mya arenaria, Littorina rudis, and Hydrobia balthica are the only true marine species; with these live Unio, Cyclas, Neritina, Limnaea, and Bithynia. The marine species and Neritina live down to 15–20 fath., the rest only down to 3 fath. Under stones close to the shore of the Skärgård at Stockholm[20] are found young Cardium and Tellina, and at 3 to 6 fath. Limnaea peregra, and Physa fontinalis. Near Gothland Limnaea is found in the open sea at 8–12 fath., and with it occur Cardium and Tellina. At the Frisches Haff[21] Mya arenaria is the only marine species, and lives in company with 6 sp. Limnaea, 1 Physa, 9 Planorbis, 1 Ancylus, 4 Valvata, 2 Sphaerium. Were the Sound to become closed, and the waters of the Baltic perfectly fresh, it would be inevitable that Mya arenaria, and such other marine species as continued to live under their changed conditions, should in course of time submit to modifications similar in kind to those experienced by the quondam marine species of the Caspian.

It seems probable, however, that the origin, at least in a great part, of the land and fresh-water Mollusca need not be accounted for by such involuntary changes of environment as the enclosure of arms of the sea, or the possible drying up of inland lakes. These cases may be taken as illustrations of the much more gradual processes of nature by which the land and fresh-water fauna must have been developed. The ancestry of that fauna must be looked for, as far as the Gasteropoda are concerned, in the littoral and estuarine species; for the Pelecypoda, in the estuarine alone. The effect of the recess of the tide, in the one case, and the effect of the reduced percentage of salt, in the other, has tended to produce a gradual adaptation to new surroundings, an adaptation which becomes more and more perfect. It may be safely asserted that no marine species could pass into a land or fresh-water species except after a period, more or less prolonged, of littoral or estuarine existence. Thus we find no land or fresh-water species exhibiting relationships with such deep-sea genera as the Volutidae, Cancellariidae, Terebridae, or even with genera trenching on the lowest part of the littoral zone, such as the Haliotidae, Conidae, Olividae, Capulidae. The signs of connexion are rather with the Neritidae, Cerithiidae, and above all the Littorinidae, which are accustomed to live for hours, and in the case of Littorina for days or even weeks, without being moistened by the tide. Similarly the fresh-water Pelecypoda exhibit relationships, not with genera exclusively marine, but with genera known to inhabit estuaries, such as the Mytilidae, Corbulidae, Cardiidae.

It would be natural to expect that we should find this process of conversion still going on, and that we should be able to detect particular species or groups of species in process of emigration from sea to land, or from sea to fresh water. Such species will be intermediate between a marine and a land or fresh-water species, and difficult to classify distinctly as one or the other. Cases of Mollusca occupying this intermediate position occur all over the world. They inhabit brackish swamps, damp places at high-water mark, and rocks only at intervals visited by the tide. Such are Potamides, Assiminea, Siphonaria, Melampus, Hydrobia, Truncatella, among the univalves, and many species of Cyrena and Arca among the bivalves.

Origin of the Fresh-water Fauna

(a) Pelecypoda.—Estuarine species, which have become accustomed to a certain admixture of fresh water, have gradually ascended the streams or been cut off from the sea, and have at last become habituated to water which is perfectly fresh.

Thus Dreissensia (rivers and canals throughout N. Europe and N. America) and Mytilopsis (rivers of America) are scarcely modified Mytili (Fig. 8); Scaphula is a modified Arca, and lives in the Ganges, the Jumna, and the Tenasserim at a distance of 1600 miles from the sea (Fig. 9). Pholas rivicola is found imbedded in floating wood on the R. Pantai many miles from its mouth. Cyrena, Corbicula, and probably Sphaerium and Pisidium are derived, in different degrees of removal, from the exclusively marine Veneridae; Potamomya (rivers of S. America), and Himella (R. Amazon) are forms of Corbula. The Caspian genera derived from Cardium (Adacna, Didacna, Monodacna), have already been referred to. Nausitora is a form of Teredo, which lives in fresh water in Bengal. Rangia, Fischeria, and Galatea probably share the derivation of the Cyrenidae, while in Iphigenia we have one of the Donacidae which has not yet mounted rivers, but is confined to a strictly estuarine life. The familiar Scrobicularia piperata of our own estuaries is a Tellina, which lives by preference in brackish water.

The great family of the Unionidae is regarded by Neumayr[22] as derived from Trigonia, the points of similarity being the development of a nacreous shell, the presence of a strong epidermis, and the arrangement of the muscular scars. It is remarkable, too, that on many Uniones of Pliocene times there is found shell ornamentation of such a type as occurs elsewhere among the Pelecypoda only on Trigonia.

The genera of fresh-water Pelecypoda are comparatively few in number, and their origin is far more clearly discernible than that of any other group. This is perhaps due to the fact that the essential changes of structure required to convert a marine into a fresh-water bivalve are but slight. Both animals “breathe water,” and both obtain their nutriment from matter contained in water. Similar remarks apply to fresh-water operculate Gasteropoda. But the passage from a marine to an aerial life involves much profounder changes of environment, which have to be met by correspondingly important changes in the organism. This may be in part the reason why the ancestry of all Pulmonata, whether land or fresh-water, is so difficult to trace.

(b) Gasteropoda.—(1) Operculate. Canidia and Clea are closely allied, with but little modification, to the marine Cominella[23] (Fig. 11), as is also Nassodonta to Nassa. They occur (in fresh water) in the rivers of India, Indo-China, Java, and Borneo, associated with essentially fresh-water species. Potamides, with its various sub-genera (Telescopium, Pyrazus, Pirenella, Cerithidea, etc.), all of which inhabit swamps and mud-flats just above high-water mark in all warm countries, are derived from Cerithium (Fig. 12); Assiminea, Hydrobia, and perhaps Truncatella, from Rissoa. It is a remarkable fact that in Geomelania (with its sub-genera Chittya and Blandiella) we have a form of Truncatella which has entirely deserted the neighbourhood of the sea, and lives in woody mountainous localities in certain of the West Indies. Cremnoconchus, a remarkable shell occurring only on wet cliffs in the ghâts of southern India, is a modified Littorina. Neritina and Nerita form a very interesting case in illustration of the whole process. Nerita is a purely marine genus, occurring on rocks in the littoral zone; one species, however, (N. lineata, Chem.) ascends rivers as far as 25 miles from their mouth, and others haunt marshes of brackish water. Neritina is the fresh-water form, some species of which are found in brackish swamps or even creeping on wet mud between tide marks, while the great majority are fluviatile, one group (Neritodryas) actually occurring in the Philippines on trees of some height, at a distance of a quarter of a mile from any water. Navicella is a still further modified form of Neritina, occurring only on wet rocks, branches, etc., in non-tidal streams (Fig. 13).

The great family of the Melaniidae, which occurs in the rivers of warm countries all over the world, and that of the Pleuroceridae, which is confined to North America, are, in all probability, derived from some form or forms of Cerithium. The origin of the Paludinidae, Valvatidae, and Ampullariidae is more doubtful. Their migration from the sea was probably of an early date, since the first traces of all three appear in the lower Cretaceous, while Melaniidae are not known until Tertiary times. Ampullaria, however, shows distinct signs of relationship to Natica, while the affinities of Paludina and Valvata cannot as yet be approximately affirmed.

(2) Pulmonata.—Intermediate between the essentially fresh-water and the essentially marine species come the group sometimes known as Gehydrophila, consisting of the two families Auriculidae and Otinidae. These may be regarded as Mollusca which, though definitely removed from all marine species by the development of a true lung or lung cavity in the place of a gill, have yet never become, in respect of habitat, genuine fresh-water species. Like Potamides, they haunt salt marshes, mangrove swamps, and the region about high-water mark. In some cases (Otina, Melampus, Pedipes) they live on rocks which are moistened, or even bathed by the spray, in others (Cassidula, Auricula) they are immersed in some depth of brackish water at high tide, in others again (Scarabus) they are more definitely terrestrial, and live under dead leaves in woods at some little distance from water. Indeed one genus of diminutive size (Carychium) has completely abandoned the neighbourhood of the sea, and inhabits swampy ground almost all over the world.

To this same section Gehydrophila have been assigned two remarkable forms of air-breathing “limpet,” Siphonaria and Gadinia (see page 151), and the aberrant Amphibola, a unique instance of a true operculated pulmonate. Siphonaria possesses a pulmonary cavity as well as a gill, while Gadinia and Amphibola are exclusively air-breathing. Siphonaria lives on rocks at or above high-water mark, Gadinia between tide marks, Amphibola (Fig. 15) in brackish water at the estuaries of rivers, half buried in the sand. There can be little doubt that all these are marine forms which are gradually becoming accustomed to a terrestrial existence. In Gadinia and Amphibola the process is so far complete that they have exchanged gills for a pulmonary cavity, while in Siphonaria we have an intermediate stage in which both organs exist together. A curious parallel to this is found in the case of Ampullaria, which is furnished with two gills and a pulmonary chamber, and breathes indifferently air and water. It is a little remarkable that Siphonaria, which lives at a higher tide level than Gadinia, should retain the gill, while Gadinia has lost it.

The ultimate affinities of the essentially fresh-water groups, Limnaea, Physa, Chilina, cannot be precisely affirmed. The form of shell in Latia, Gundlachia, and perhaps Ancylus, may suggest to some a connexion with the Otinidae, and in Chilina, a similar connexion with the Auriculidae. But, in a question of derivation, similarities of shell alone are of little value. It is not a little remarkable, for instance, that we should find a simple patelliform shell in genera so completely distinct from one another in all anatomical essentials as Ancylus, Patella, Siphonaria, Propilidium, Hipponyx, Cocculina, and Umbrella.

Some recent authors, on grounds of general organisation, regard the Limnaeidae and their allies as Opisthobranchs adapted to an aerial life. It is held[24] that the Nudibranchiate Opisthobranchs have given birth to the Pulmonata Stylommatophora or land snails, and the Tectibranchiate Opisthobranchs to the Pulmonata Basommatophora or fresh-water snails. Such a view seems at first sight open to some objection from other views than those which deal simply with anatomy. The Opisthobranchiata are not, to any marked extent, littoral genera, nor do they specially haunt the mouths of rivers. On the contrary, they inhabit, as a rule, only the very lowest part of the littoral zone, and are seldom found, except where the water is purely salt. In other cases, when the derivation of land or fresh-water genera is fairly well established, intermediate forms persist, which indicate, with more or less clearness, the lines along which modification has proceeded. It has, however, recently been shown that Siphonaria[25] and Gadinia,[26] which have, as has been already mentioned, hitherto been classified as Pulmonata, are in reality modified forms of Opisthobranchiata, which are in process of adaptation to a life partly marine, partly on land. They may therefore be regarded as supplying the link, hitherto missing, between the land Pulmonata and the marine groups from one or other of which the latter must have been derived. The general consensus of recent opinion inclines towards accepting these views, some writers[27] being content to regard the Pulmonata, as a whole, as derived from the Tectibranchiate Opisthobranchs, while others[28] go further and regard the Stylommatophora as derived directly from the Basommatophora.

Origin of the Land Fauna

Gasteropoda.—(1) Operculate. On a priori grounds, one might predict a double origin for land operculates. Marine species might be imagined to accustom themselves to a terrestrial existence, after a period, more or less prolonged, of littoral probation. Or again, fresh-water species, themselves ultimately derived from the sea, might submit to a similar transformation, after a preliminary or intermediate stage of life on mudbanks, wet swamps, branches overhanging the water, etc. Two great families in this group, and two only, seem to have undergone these transformations, the Littorinidae and the Neritidae. The derivation of almost all existing land operculates may be referred to one or other of these groups.