CHAPTER VIII

CRUSTACEA OF FRESH WATERS

The Crustacean fauna of fresh water is much less rich and varied than that of the sea. Although the number of individuals in a pond or lake may be enormous, they will be found to belong to a comparatively small number of species. All the subclasses of Crustacea with the exception of the Cirripedia have representatives in fresh water, but in most of them only a very few of the families and genera comprise truly fresh-water species. In spite of the comparative poverty of the fauna, however, it is of very great interest, more especially with regard to the problems of geographical distribution; and the ease with which specimens may be collected everywhere, and kept in small aquaria, renders it a particularly attractive subject of study for the amateur naturalist.

The general uniformity of the fresh-water fauna throughout the world has often been remarked. Darwin says: "When first collecting in the fresh waters of Brazil, I well remember feeling much surprise at the similarity of the fresh-water insects, shells, etc., and at the dissimilarity of the surrounding terrestrial beings, compared with those of Britain." This uniformity is well illustrated by many of the smaller Crustacea. In a gathering of Cladocera, Copepoda, and Ostracoda, from Central Africa or from Australia, we find that most of the genera, and even some of the species, are identical with those found in similar situations in this country. It is by no means the case that all the species and genera are thus universally distributed, for there are many, especially among the larger forms, which have a very restricted range; but this does not render less striking the general uniformity of the fauna over very wide areas.

When we consider the physical environment of fresh-water animals, it seems at first sight as if this wide distribution were the reverse of what might have been expected, for the area occupied by them is far more discontinuous than in the case of terrestrial or marine animals. The inhabitants of a pond or lake are to a great extent isolated; and although they may spread to other ponds and lakes by way of communicating streams or rivers, where these are not too swiftly flowing and are not interrupted by falls, yet direct passage from one river system to another is rarely possible. Further, since practically the whole of the fresh water on the surface of the globe is constantly flowing, more or less rapidly, towards the sea, the smaller feebly swimming forms tend to be swept down with the current, and ultimately carried to perish in the sea. It follows that only those forms which possess special adaptations for dispersal are able to flourish in fresh water. In many cases, as will be described below, the eggs of the smaller Crustacea can survive being dried up, and in this state they may be blown about by wind or carried to great distances in mud, adhering to the feet of migratory wading birds. Darwin says: "The wide-ranging power of fresh-water productions can, I think, in most cases be explained by their having become fitted, in a manner highly useful to them, for short and frequent migrations from pond to pond, or from stream to stream, within their own countries; and liability to wide dispersal would follow from this capacity as an almost necessary consequence" ("Origin of Species," sixth edition, chapter xiii.). In accordance with this, we find that it is just those groups of Crustacea which show these adaptations for dispersal that are most universally distributed in fresh water. On the other hand, the larger Crustacea, like the Crayfishes and River Crabs, which cannot so easily be transported from one locality to another, have as a rule a more restricted range. These larger forms, from their size and powers of swimming or creeping, can make their way upstream and spread throughout a river system, and in some cases they can leave the water and journey for short distances overland. On the other hand, since free-swimming larvæ would be liable to be swept out to sea, most of them have a direct development, the young only leaving the protection of the mother when they have attained the form and habits of the adult. When all these factors have been taken into account, however, there still remain many cases where the distribution of individual species or of groups is hard to explain, and shows indications of dating from a time when the outlines of continents and the connections of river systems were different from what they are now.

Before proceeding to mention some of the more characteristic forms of fresh-water Crustacea, it should be mentioned that in large lakes, as in the sea, we can distinguish a littoral fauna in the shallow waters close to the shore, a plankton fauna of the surface waters, and a deep-water fauna. The littoral fauna does not differ in general characters from that found in smaller ponds and gently-flowing rivers; the plankton comprises many peculiar species showing adaptations for flotation, as in the case of the marine plankton; and the deep-water fauna is very poor in species and in individuals, and shows some relations with the subterranean fauna to be mentioned later.

Of all the subclasses of Crustacea, the Branchiopoda are the most characteristically fresh-water animals, only a few Cladocera being found in the sea, and some Anostraca in salt lakes and brine pools.

The larger Branchiopoda (Anostraca, Notostraca, and Conchostraca) are generally found in small, shallow ponds which are liable to be dried up in summer. The "Fairy Shrimp" (_Chirocephalus diaphanus_; see Fig. 10, p. 35) has been found in swarms in the water standing in deep cart-ruts in a country lane in England, and _Apus_ sometimes appears suddenly in rain-water puddles of a few square yards in area, which dry up after a few weeks of hot weather. The eggs of these animals, when dried in the mud, may remain dormant for long periods, and many species have been hatched out from samples of dried mud brought by travellers from distant countries. In such a sample from the Pool of Gihon at Jerusalem, it is recorded that the eggs of _Estheria_ (see Fig. 11, p. 36) were found to be capable of hatching after being kept dry for nine years. In some species it is said that the eggs will not develop unless they have been first dried, but this is not the case with _Chirocephalus_. In favourable conditions development takes place very rapidly. Messrs. Spencer and Hall, in describing the Branchiopoda of Central Australia, say: "Certainly not more than two weeks after a fall of rain, and probably only a few days, numberless specimens of _Apus_, measuring in all about 2-1/2 to 3 inches in length, were swimming about; and, as not a single one was to be found in the water-pools prior to the rain, these must have been developed from the egg."

From what has been said, it is apparent that the larger Branchiopoda are particularly well fitted to be distributed by the agency of birds, and this is no doubt the explanation of the way in which many of the species suddenly appear in localities where they were previously unknown, and, after swarming for a longer or shorter time, sometimes for several successive seasons, as suddenly vanish. A striking example of this is afforded by _Apus cancriformis_ (see Plate II.), which formerly occurred in several localities in the South of England, and appears more or less irregularly in many parts of the Continent of Europe. No British specimens had been recorded for over forty years, and the species was believed to be extinct in this country, when it was found in 1907 by Mr. F. Balfour Browne in a brackish marsh near Southwick, in Kirkcudbrightshire. It can hardly be supposed that so large an animal as _Apus_, and one so easily recognized, would have escaped notice altogether had it occurred regularly in any part of the British Islands. It is much more probable that the Scottish specimens found in 1907 had developed from eggs accidentally transported by some bird from the Continent. In 1908 a careful search in the same locality failed to reveal a solitary specimen.

The Anostraca and Notostraca usually swim with the back downwards. Particles of mud and of animal and vegetable matter are drawn by the currents produced in swimming, into the ventral groove between the pairs of feet, and are passed forwards to the mouth to serve as food. Some species of Conchostraca are said to swim in the same inverted position; but Messrs. Spencer and Hall, in the memoir already quoted, state that the Australian Conchostraca swim back uppermost. They attribute the difference in habit between the Conchostraca and Notostraca to the fact that in the former group the valves of the shell can be rapidly closed to protect the soft and vulnerable appendages, while no such protection is possible in the Notostraca. They found on one occasion a specimen of _Apus_ (Notostraca) attacked by three Water-beetles, which were tearing its soft appendages, and they suppose that _Apus_ generally escapes such attacks by swimming upside down.

The breeding habits of the Branchiopoda are also of interest, from the prevalence in many species of reproduction by unfertilized eggs, or "parthenogenesis." This may go on for many generations, and in _Apus_, for instance, it is possible to examine thousands of specimens before finding a single male, although, for some unexplained reason, males are sometimes comparatively common. It is probable that males must appear sooner or later, otherwise the series of parthenogenetic generations will come to an end; but it is not certain that this is the case, and there are some species of Conchostraca of which the males have never been seen.

The genus _Artemia_ (Fig. 55), among the Anostraca, is peculiar in its habitat; for, while most of the Branchiopoda inhabit fresh or brackish water, it flourishes in concentrated brine. In the South of Europe it is found, as it was formerly in England, in the shallow ponds in which sea-water is exposed to evaporation for the manufacture of salt, and in these it occurs in such numbers as to give the water a reddish colour. It is also found in salt lakes, like the Great Salt Lake of Utah, in the United States, and in many other parts of the world. The specimens from different localities often differ considerably, especially in the form of the tail-lobes; but it has been shown that these differences are more or less directly correlated with the degree of salinity of the water in which the animals live, and it is probable that the forms which have been described are all variations of a single cosmopolitan species ranging from Greenland to Australia, and from the West Indies to Central Asia. _Artemia_ is the only one of the Anostraca that is known to be parthenogenetic, some colonies consisting entirely of females, while in others males are abundant. The reddish colour above alluded to is found also in _Branchipus_, _Apus_, and other Branchiopoda, and is due, as Sir Ray Lankester first showed, to the presence in the body-fluids of hæmoglobin, the red colouring matter of the blood of Vertebrates, which is important in the process of respiration.

The smaller Branchiopoda known as "Water-fleas," forming the order Cladocera, are abundant everywhere in fresh water. _Daphnia pulex_ and other species of the genus, and the little Lynceidæ, of which _Chydorus sphæricus_ (Fig. 56) is the commonest species, are to be found in ponds and ditches, and often swarm in farmyard ponds where the water is foul with decaying matter. In most gatherings from such localities only female specimens will be found, and nearly all of these will be seen to carry a cluster of eggs or of developing embryos in the "brood-chamber" between the back part of the body and the shell. In _Daphnia pulex_ (see Fig. 12, p. 37) a single brood may consist of thirty young, and occasionally of more than twice that number. As the broods may succeed each other at intervals of two or three days, it will be seen that the multiplication of the species in favourable circumstances may be exceedingly rapid. It has been calculated that in sixty days the progeny of a single female might amount to about 13,000,000,000. In addition to these parthenogenetic eggs, which hatch at once while still within the brood-chamber, the Cladocera produce, at certain seasons, another kind of egg which requires to be fertilized by the male before it will develop. These eggs are dark in colour and are enclosed in a thick shell, and they do not hatch at once, but are cast off when the shell of the female is moulted. Very commonly these "resting eggs," as they are called, are produced in the autumn and lie dormant until the following spring, and they can survive drying or freezing without injury, while the thin-shelled parthenogenetic eggs within the brood-chamber of the mother are easily killed. In addition to having thick shells, the resting eggs are further protected in most, but not in all, cases by the moulted carapace of the parent, which is specially thickened for the purpose. This modification of the carapace is most highly developed in the family Daphniidæ (Fig. 57), where a saddle-shaped area on the dorsal side, known as the "ephippium," becomes thickened, and on moulting separates from the rest of the carapace to form a compact case enclosing the two resting eggs. The outer wall of the ephippium is divided up into small hexagonal cells, which become filled with air, causing the ephippium to float at the surface of the water. In this position the ephippia readily become entangled in the feathers of birds, and in some cases the shell is provided with spines or hooks, which facilitate transport to other localities by such means.

The appearance of males and the production of ephippial eggs--in other words, the "sexual period"--is generally more or less restricted to one season of the year. In most species, particularly in those which live in lakes, the sexual period occurs in the late autumn, and the ephippial eggs lie dormant during the winter, and hatch in the spring. In species living in small ponds exposed to the risk of overheating or of drying up during summer, there is often a distinct sexual period in the spring, when ephippial eggs are produced to tide over the unfavourable conditions of the warmer months of the year. Although no species is known to be exclusively parthenogenetic, yet it appears that purely parthenogenetic colonies of certain species may be found in favourable localities, where they may reproduce from year to year without males ever being found.

Certain species of Cladocera belong to the plankton of lakes and large ponds, and show modifications which adapt them for a floating life. Some of these belong to the genus _Daphnia_, and differ from the species found in other situations by their glassy transparency. As in the case of many marine plankton Crustacea, this transparency is probably due to the thinness of the shell and to the general watery condition of the body, giving the necessary buoyancy to enable the animal to remain constantly afloat. The same effect is no doubt produced by the long terminal spine of the carapace and by the great helmet-shaped crest into which the upper part of the head is often produced. A form very characteristic of the plankton of large lakes is _Bythotrephes_ (Fig. 58), which is found in the lakes of Scotland, Ireland, Wales, and the Lake District of England. In _Bythotrephes_ the carapace does not enclose the body, but is reduced to a small brood-sac; the abdomen, however, is drawn out into a long spine, which may be two or three times as long as the body. A further point of interest is the division of the eye into a dorsal and a ventral portion, differing in structure in much the same way as do the two divisions of the eyes in certain marine plankton Crustacea (see p. 152). Another very remarkable lacustrine form is _Leptodora_, the largest of all the Cladocera, being sometimes more than half an inch in length. In this case also the carapace is very small, and does not enclose the body. The swimming antennæ are very large, and the abdomen is long and divided into several segments.

_Leptodora_ is further remarkable on account of its mode of development. The parthenogenetic eggs, as in other Cladocera, develop directly, but the resting eggs give rise to larvæ of the nauplius type.

_Holopedium_, which is found in similar situations, surrounds itself with a mass of a jelly-like substance which it secretes. A similar envelope of jelly is found in some marine plankton animals, though not, so far as is known, in any Crustacea, and it no doubt serves to give buoyancy to the animal.

The Copepoda of fresh water are as abundant and universally distributed as the Cladocera. Species of the genus _Cyclops_ (see Fig. 14, p. 39), easily recognized by the pear-shaped body and the two egg-packets carried by the female, are to be found in almost every pond and ditch. The genus _Canthocamptus_ comprises species of smaller size, with slender, flexible body, and carrying only a single egg-packet. The plankton of lakes and ponds includes species of _Diaptomus_ (Fig. 59), which have a narrow body and very long antennules. The latter are held out stiffly while the animal swims by rapid movements of the antennæ and mouth parts, making occasional sudden leaps by means of its oar-like feet. In this genus also the egg-packet is single. The development can easily be studied by keeping egg-carrying females of _Cyclops_ in a jar of water, when the nauplius larvæ will soon hatch out.

Although the Copepoda, unlike the Cladocera, are not parthenogenetic, it has been found that certain species of _Diaptomus_ produce resting eggs capable of surviving freezing or drying. In the early part of the breeding season the eggs have thin shells, and they hatch after a short time. In the autumn, however, thick-shelled eggs are produced, which lie dormant in the mud until the following spring. It has recently been discovered that species of _Cyclops_ and _Canthocamptus_ pass through a resting stage, in which the animal surrounds itself with a cocoon-like capsule of mud held together by a glutinous secretion produced by glands on the surface of the body and limbs. The encapsuled animals, in the cases observed, lie dormant in the mud during the summer, to resume active life in the colder months of the year. It is very probable that they can also be dried without injury, and that the "cocoons" serve the same purpose as the resting eggs of other species.

Numerous species of Ostracods, belonging to the genus _Cypris_ (see Fig. 13, B, p. 38), and other closely related genera, occur in fresh water. Like the Cladocera, they reproduce largely by parthenogenesis, and the males of many species are rarely found, while in some species they have not yet been discovered. In Professor Weismann's laboratory at Freiburg a colony of _Cypris_ was kept in an aquarium for eight years, and during the whole of that time no males made their appearance, the colony reproducing exclusively by parthenogenesis. Probably in all species the eggs survive drying.

The common "Freshwater Shrimp" (_Gammarus pulex_), which has already been described, may be taken as a type of a large number of Amphipoda, for the most part closely allied, which are widely distributed in most regions of the world, with the exception of the tropics. _G. pulex_ itself ranges from the British Islands to Mongolia. As the eggs are carried, till they hatch, in the brood-pouch of the parent, and are not known to survive drying, it is difficult to understand in what way _Gammarus_ and its allies contrive to spread from one locality to another.

The little fresh-water Isopod _Asellus aquaticus_ (Fig. 60) is common in ponds and canals in this country. It may be recognized by its general resemblance to a Woodlouse, with very long antennæ, and with a pair of long, slender, forked uropods projecting behind. The species is widely distributed in Europe, and other species of the same and closely related genera are found in North America.

In Australia and New Zealand the Isopoda are represented in fresh waters by a very peculiar group of species, forming the suborder Phreatoicidea, which have more the aspect of Amphipods than of Isopods, since the body is more or less flattened from side to side, instead of from above downwards.

With regard to the mode of distribution of the fresh-water Isopoda, there is the same difficulty as in the case of the Amphipoda, for the eggs are carried in a brood-pouch, and do not seem to be in any way protected against drought. It is no doubt in consequence of this that the fresh-water species and genera of both Amphipoda and Isopoda, though widely distributed, do not have the world-wide range of many of the more minute Crustacea described above.

The common Crayfish, _Astacus_ (or _Potamobius_) _pallipes_, is the only truly fresh-water Decapod found in England, although a small Prawn, _Palæmonetes varians_, which usually inhabits brackish water, may occasionally be found in places where the water is practically fresh. The structure of the Crayfish is very similar to that of the Lobster, but, as already mentioned, it differs in its mode of development, having no free-swimming larval stage. From its size, and from the fact that the eggs are carried by the female, the Crayfish cannot be transported from one locality to another by the agencies which distribute the smaller fresh-water Crustacea. On the other hand, the adult animals can live out of the water for days, or even weeks, if they are kept moist, and the English species is stated to leave the water occasionally, and to make short excursions on land. Many species found in foreign countries are still more truly amphibious in their habits. It is clear, however, that the means of dispersal of the Crayfishes are very limited, and on this account the problems connected with their geographical distribution are of great interest. An admirable discussion of the subject will be found in Professor Huxley's book on the Crayfish, and the conclusions reached by him have hardly been modified by thirty years of subsequent research. Only a very brief outline can be attempted here.

Crayfishes are found in the fresh waters of the Northern and Southern Hemispheres (Fig. 61), but in each case they are practically confined to the temperate regions, and are absent from a broad intervening tropical zone. The Northern Crayfishes, forming the family Astacidæ (or Potamobiidæ) are distinguished, among other characters, by having a pair of appendages on the first abdominal somite, at least in the male sex; the Southern Crayfishes have no appendages on that somite, and for this and other reasons are regarded as constituting a distinct family--Parastacidæ. There is thus a general correspondence between the geographical distribution of the Crayfishes and the more important structural differences expressed in their classification. There can be no doubt that the two families have been derived from a common stock of marine lobster-like animals, and it is reasonable to suppose that two branches of this stock became independently adapted to a fresh-water habitat in the North and in the South, giving rise to the Astacidæ and the Parastacidæ respectively.

The distribution of the individual genera is, however, not so easy to understand. The species found in Europe all belong to the genus _Astacus_, which also penetrates into Asia as far as Turkestan and the basin of the River Obi.

Throughout the greater part of Asia no Crayfishes are found until we come to the Far East, where we find an isolated colony in the river-system of the Amur, in Korea, and in the north of Japan. These far eastern Crayfishes, however, differ so much from the typical species of _Astacus_ that they are now placed in a subgenus (sometimes regarded as a distinct genus), _Cambaroides_. Curiously enough, the typical genus _Astacus_ reappears again on the other side of the Pacific, where several species occur in that part of North America which lies west of the Rocky Mountains. East of the Rockies, again, numerous species are found belonging to a distinct genus, _Cambarus_, which ranges from Canada to Central America and Cuba, and this genus is allied in certain respects to the _Cambaroides_ of Eastern Asia. If the systematic relations of these genera have been properly interpreted, it is by no means easy to understand in what way their present distribution has been brought about.

The Southern Crayfishes have an even more scattered and discontinuous range. In New Zealand the genus _Paranephrops_ occurs, in Australia and Tasmania the genera _Astacopsis_ (Plate XX.), _Cheraps_ and _Engæus_ (Plate XX.). A single species of _Cheraps_ has been recorded from New Guinea, but no Crayfishes are found in any part of the Malay Archipelago, in Southern Asia, or on the continent of Africa, although, curiously enough, a single species of a peculiar genus (_Astacoides_) is found in Madagascar. In South America species of _Parastacus_ are found in Southern Brazil, Argentina, and Chili. It is evident that these various genera of Parastacidæ, which are now so widely isolated from each other, must have reached their present habitats when the relative distribution of land and sea in the Southern Hemisphere was very different from what it is now. What exactly the nature of the land connection between the various islands and continents was, whether by way of an Antarctic continent or otherwise, is a question that can only be suggested here. To attempt to answer it would involve the consideration of the distribution of many other groups of animals besides Crayfishes.

Before leaving the Crayfishes, it may be mentioned that certain species have become adapted to almost terrestrial habits. A number of species of _Cambarus_ in North America are often found at considerable distances from open water, burrowing in damp earth, their burrows reaching down to the ground-water. In many cases they throw up chimney-like piles of mud at the mouths of their burrows, and in places their chimneys are so numerous as to "hamper farming operations by interfering with the harvesting machines, clogging and ruining them." The species of _Engæus_ (Plate XX.), found in Tasmania, are there known as "Land Crabs," and burrow in marshy places and in the forests up to an elevation of 4,000 feet.

The broad equatorial belt which separates the regions inhabited by the Northern and the Southern Crayfishes is characterized by the presence of several other groups of fresh-water Decapoda. The large River Prawns, which are found nearly everywhere within the tropics, belong to the genus _Palæmon_ (Plate XXI.), which is very closely related to the common marine Prawns (_Leander_) of our own coasts. Some of these Prawns grow to a foot or more in length of body, and the large claws may measure as much again. From the Crayfishes, for which they are sometimes mistaken, they may be easily distinguished by the fact that the large pincer-claws are not the first, but the second pair of legs. Another widely-spread group of River Prawns, for the most part of small size, is the family Atyidæ (Plate XXII.), in which the two pairs of pincer-claws are feeble, and have the fingers tipped with brushes of long hairs, used in sweeping up minute particles of food from the mud. The distribution of these Prawns presents many difficult problems, as an example of which we may mention the presence of identical or closely related species in the fresh waters of West Africa and of the West Indies.

The Brachyura (or Crabs) include many species that live in fresh water. Some of these, like the species of _Sesarma_ (see Plate XXIII.) and some other genera of the family Grapsidæ, are common throughout the tropics, passing up the rivers from the brackish water of estuaries, and being often found long distances inland in quite fresh water. The true River Crabs, however, belong to the family Potamonidæ, and are very common throughout the warmer regions of the globe. One species, _Potamon edule_ (Plate XXIII.), formerly called _Telphusa fluviatilis_, is found in the South of Europe (Italy, Greece, etc.). Very numerous species, as yet only imperfectly known, occur throughout the whole of Africa, in Southern Asia, and in the Malay Islands, extending to Australia in the south and Japan on the north. In the New World the River Crabs are found in South America, and extend north to Mexico and the West Indian Islands. Many of the River Crabs are amphibious in habits, and may be found burrowing in marshy ground or in damp forests. The young are hatched from the egg with all the appendages developed, and they remain clinging to the abdomen of the mother until after the first moult, when they are perfectly-formed little Crabs (see Fig. 31, p. 78).

The groups which have been mentioned are all characteristic inhabitants of the fresh waters over considerable areas of the surface of the globe. There are, however, in addition to these, certain Crustacea which occur in isolated localities, and have no close allies in fresh waters elsewhere. In the streams of Southern Brazil and Chili there is found a small Crustacean (_Æglea lævis_--Plate XXIV.), not unlike the Galatheas of our own coasts, which is interesting as being the only species of the Anomura found in fresh water. Still more remarkable are the Syncarida, which are represented by two species of "Mountain Shrimps" (see Fig. 84, p. 264) in Tasmania, and by a third species found near Melbourne. These forms have no near allies among living Crustacea, but appear to be related, as will be shown in a later chapter, to certain fossil Crustacea found in Palæozoic rocks.

Belonging to a different category from any of those mentioned are certain Crustacea closely allied to, or identical with, species living in the sea, which inhabit inland lakes where no direct passage from the sea is now possible. Attention was first called to these in the case of some of the large lakes of Sweden, in which Professor Lovén found some Crustacea--_Mysis relicta_ (see Fig. 16, p. 47), _Mesidotea entomon, Pontoporeia affinis_--almost or quite identical with species inhabiting the Baltic, the Arctic Ocean, and the North Atlantic. There is geological evidence to show that these lakes were once fjords, or arms of the sea, and have become cut off from communication with the Baltic by gradual elevation of the land. The marine animals which they contained would thus be imprisoned, and as the water became less and less salt, by the inflow of rivers, certain species which were able to accommodate themselves to the altered conditions would survive. Some of the species living in the Swedish lakes have since been found to have a wider distribution. Thus, _Mysis relicta_, which should perhaps be reckoned as only a variety of the _Mysis oculata_ of Arctic seas, has been found in lakes in Russia, North Germany, and North America (Lake Superior and others), and has lately been discovered in Lough Neagh and some other lakes in Ireland.

The brackish waters of the Caspian Sea contain a very remarkable assemblage of animals, including many Crustacea, which, although now quite isolated from the oceans, are certainly of marine, and in part of Arctic, origin. Among these are some species closely allied to or identical with those of the Swedish lakes already mentioned, together with a great variety of species of Mysidacea, Cumacea, and Amphipoda, which appear to have been evolved from marine forms since the Caspian was cut off from communication with the Arctic Ocean.

To such assemblages of animals derived from marine species and isolated in inland lakes the name of "relict" faunas has been given. It is necessary to use caution, however, in extending this explanation of their origin to every case of peculiar lake faunas. For example, there are difficulties in the way of supposing that Lake Baikal was ever in open and direct communication with the sea, although it contains many animals, such as seals, which are certainly of marine origin. The chief Crustacea of the lake are numerous species of Amphipods belonging to the genus _Gammarus_, and other genera closely related thereto, and for these, at all events, there is no need to assume a "relict" origin.

One of the most remarkable lakes in the world from a zoological point of view is Lake Tanganyika in Africa. When it was found that this lake contained a fauna very different from that of the other great lakes of Africa, it was rashly assumed that it must be of relict origin, and some remarkable speculations were indulged in as to the former connection between the lake and the sea. Further research, while it has greatly emphasized the peculiar nature of the fauna, has entirely disposed of the view that it originated in this way. The Crabs and Prawns, for example, are not nearly related to marine forms, but belong to groups that are characteristic of fresh waters in the tropics. While Nyassa and the Victoria Nyanza have as yet only yielded a single species of Prawn, and that one of enormously wide distribution (from the Nile to Queensland), Tanganyika contains no fewer than twelve species, all of which are peculiar to the lake, while all except one belong to genera unrepresented elsewhere. Similarly, the Crabs found in the other great lakes of Africa belong to commonplace types of River Crabs of the genus _Potamon_; in Tanganyika, in addition to some of these, there are three species of a remarkable genus, _Platytelphusa_, not known from any other locality. The Copepoda and Ostracoda of Tanganyika comprise a remarkably large number of species, many of them peculiar to the lake. A most unusual feature is the entire absence of Cladocera. It is not easy to explain the occurrence of this remarkable fauna in Tanganyika, but the evidence from other groups of animals, such as Mollusca and fishes, tends to suggest that the lake must have been, until recently, completely isolated from the other lakes and river-systems of Africa, that it had no outlet, and that the water was consequently more or less brackish. Under these conditions the fauna of the lake, originally similar to that of the other African lakes, has evolved along lines of its own.

A very interesting division of the fresh-water fauna is constituted by those animals which inhabit underground waters. In the South of England there is found not unfrequently in the water of wells a small colourless transparent Amphipod known as the "Well Shrimp" (_Niphargus aquilex_--Fig. 62), distinguished from the common fresh-water _Gammarus_ by the slenderness of its body, by the elongation of the last pair of tail appendages (uropods), and by the absence of eyes. The proper habitat of _Niphargus_ is not actually in the wells, but in the subterranean reservoirs and streams by which the wells are fed. These subterranean channels intercommunicate over wide areas, and are now known in many parts of the world to contain a peculiar assemblage of animals which become accessible to the naturalist in wells and in the streams and lakes of large caves. Further, the scanty "abyssal" fauna of deep lakes is partly made up of species which enter the lakes by subterranean channels, and find a suitable habitat in the deep water. Species of _Niphargus_, for example, have been dredged in Lough Mask in Ireland and in some of the Swiss lakes.

Several species of blind Crayfishes have been found in caves in North America, the best known being one (_Cambarus pellucidus_--Plate XXV.) found in the Mammoth Cave in Kentucky; and blind Prawns belonging to various genera have been discovered in caves in America and Europe.

A very remarkable feature of the subterranean fauna is that a number of the animals appear to be more closely allied to marine species than to any known from fresh waters above-ground. This is especially the case with some of the Isopoda belonging to typically marine families like the Cirolanidæ and Anthuridæ, and it has been suggested that these have been derived from marine species which have entered the underground waters directly from the sea by way of submarine fissures in the crust of the earth.

The environment in which these subterranean animals live resembles that of the deep-sea animals in the absence of light, and the consequent absence of plant-life. They must ultimately depend for food on animal and vegetable débris washed down from the surface, but the food-supply must be scanty, for the water in which they live is usually very clear and free from organic matter. It is not surprising to find that nearly all of them are blind, and the few species provided with visual organs which have been described, from caves, are probably only temporary or accidental immigrants. Whether the degeneration of the eyes is the direct effect of disuse, or is due to natural selection ceasing to keep the eyes up to the standard of usefulness, is a question which has been much debated, and its answer, were we sure of it, would settle some of the most fundamental problems of the evolution theory.

At all events, we do not find in any truly subterranean species large and peculiarly modified eyes like those of many deep-sea animals, and this may be associated with the complete darkness of their habitat, not lighted by phosphorescent organisms as the deep sea is. In another respect these animals differ from those of the deep sea, for they are all colourless or nearly so; while many of the inhabitants of the deep sea, as we have already seen, are brilliantly coloured.