chapter three groups of land-dwelling Crustacea--the Land Crabs of the

family Gecarcinidæ, the Land Hermits (Coenobitidæ), and the Land Isopods, or Woodlice (Oniscoidea).

The Gecarcinidæ are abundant in the tropics of the Old and New Worlds. Some of the species at least, probably all, visit the sea at intervals for the purpose of hatching off the eggs carried by the females, and the larval stages are passed in the sea. In the case of _Gecarcinus ruricola_ (Plate XXVI.), a species very common in the West Indies, the migration to the sea takes place annually during the rainy season in May. The Crabs are described as coming down from the hills in vast multitudes, clambering over any obstacles in their way, and even invading houses, in their march towards the sea. Stebbing states that "The noise of their march is compared to the rattling of the armour of a regiment of cuirassiers." The females enter the sea to wash off the eggs which they carry attached to their abdominal appendages, or rather, probably, to allow the young to hatch out. The Crabs then return whence they came, and are followed later by the young, which, having passed through their larval stages in the sea, leave the water, and are found in thousands clinging to the rocks on the shore.

On Christmas Island, in the Indian Ocean, Dr. C. W. Andrews studied the habits of another Land Crab, of which the proper name seems to be _Gecarcoidea lalandii_. He says: "This is the commonest of the Land Crabs inhabiting the island, and is found in great numbers everywhere, even on the higher hills and the more central portion of the plateau. In many places the soil is honeycombed by its burrows, into which it rapidly retreats when alarmed. These Crabs seem to feed mainly on dead leaves, which they carry in one claw held high over the back and drag down into the burrows. From their enormous numbers, they must play a great part in the destruction of decaying vegetable matter and its incorporation into the soil."

"Once a year, during the rainy season, they descend to the sea to deposit [or, rather, to hatch out] their eggs, and during this migration hundreds may be seen on every path down steep slopes, and many descend the cliff-face itself. They remain on the beach for a week or two, and ... afterwards gradually make their way back to their accustomed homes."

In the year of Dr. Andrews' first visit to the island (1898) this migration occurred in January. On a subsequent visit to the island in 1908 he obtained specimens of a large Megalopa larva (see p. 70) which occurred in enormous quantities in the sea shortly after the migration, and also of a small Crab which appeared in similar numbers at a slightly later date. It seems practically certain that these larvæ and young are those of _Gecarcoidea lalandii_. A second species of Land Crab, _Cardisoma hirtipes_, found on Christmas Island, has very different habits from the foregoing. Dr. Andrews says of it: "In this island, at any rate, this species must be regarded as a fresh-water form, and, in fact, when a specimen was seen it might be taken as an indication that fresh water was not far off. It lives in deep holes in the mud at the sides and bottom of the brooks." Dr. Andrews tells me that he never saw this species at or near the sea (in marked contrast to _Gecarcoidea_), and this agrees with the observations of other travellers on species of the genus _Cardisoma_, so that the breeding habits remain unknown. There is every probability, however, that in this case, also, the young stages are passed in the sea.

The student will find, in many textbooks on zoology, the statement that some Land Crabs of the genus _Gecarcinus_ develop without metamorphosis. Although it is impossible, with our present knowledge, to state definitely that this is not the case, there is absolutely no evidence to support it, and it is an interesting example of the way in which erroneous statements sometimes gain currency in science.[3] It is based upon the fact that in 1835 Professor J. O. Westwood described the early stages of "a West Indian Land Crab," in a paper "On the Supposed Existence of Metamorphosis in the Crustacea," published in the Transactions of the Royal Society. Professor Westwood found that the embryos extracted from the egg possessed all the appendages of the adult except the swimmerets, and that young specimens clinging to the abdomen of the parent were perfectly-formed little Crabs. The specimens which he described were sent to him by the Rev. Lansdown Guilding, of St. Vincent, who also deals with the subject in a note published in the _Magazine of Natural History_ in the same year. Neither Westwood nor Guilding refers to the Crab as a _Gecarcinus_, although Guilding calls it the "Mountain Crab," a name which Patrick Browne in 1756 gives to the _Gecarcinus ruricola_ of Jamaica. So far as I am aware, the first writer to refer to Westwood's Crab as a _Gecarcinus_, was Professor T. Bell, who in his "British Stalk-eyed Crustacea," published in 1853, states that some of the original specimens had come into his possession. They consisted of the detached abdomens of female Crabs, with eggs and young adhering to them. It would be by no means easy to identify the species of Crab to which a detached abdomen belonged, and there is nothing in the whole history inconsistent with the supposition that these observations really relate to a River Crab of the family Potamonidæ, of which at least one species, _Pseudothelphusa dentata_, is known to occur on the island of St. Vincent. As we have already seen, some of these River Crabs are quite as much land animals as the Gecarcinidæ, and they are known to have a direct development.

[3] I am indebted to Mr. J. T. Cunningham for calling my attention to some of the facts here recorded.

The Gecarcinidæ possess well-developed gills, but in addition the gill chambers are modified for air-breathing, as in some other amphibious Crabs (_Ocypode_, _Gelasimus_, etc.). Each chamber is capacious and vaulted, and the lining membrane is thick and richly supplied with bloodvessels, and is folded so as to divide off the upper part of the chamber as a sort of pocket.

The Land Hermit Crabs of the family Coenobitidæ are found on the coasts of all tropical seas. Like the Gecarcinidæ, they visit the sea periodically for the purpose of hatching off the eggs, and the larval stages are marine. The species of the genus _Coenobita_ (Plate XXVI.) resemble the marine Hermit Crabs in general shape, and like them use the shells of Gasteropod Molluscs as portable shelters. Where shells are scarce, other hollow objects are occasionally utilized; for example, large individuals will sometimes carry about the shell of a broken coconut, and a specimen has been seen to walk off in a cracked test-tube discarded by a naturalist who was investigating their habits. In one instance Professor Alcock saw an individual "so big that it seemed to have given up hope of finding a house, and was wandering about recklessly, with its tail behind it all unprotected."

The Coenobites often climb into bushes in search of food, and Dr. Alcock "once found one of them busy, like a large bee, among the florets of a coconut, which made me wonder whether they may not sometimes play a part in fertilizing flowers." They are, however, by no means exclusively vegetarians. The author just quoted describes a visit to Pitti Bank in the Laccadive Archipelago, the breeding-ground of two species of terns. The ground was everywhere strewn with the dead bodies and clean-picked skeletons of the young birds. "We soon discovered that one great cause of the wholesale destruction of young birds was the voracity of swarms of large Hermit Crabs (_Coenobita_), for again and again we found recently killed birds, in all the beauty of their first speckled plumage, being torn to pieces by a writhing pack of these ghastly Crustaceans. There were plenty of large Ocypode Crabs, too (_O. ceratophthalmus_), aiding in the carnage."

On Christmas Island Dr. Andrews found a species of _Coenobita_ not unfrequently in the higher parts of the island far from the sea, and he remarks that the occurrence of large marine shells high up on the hills seemed very puzzling until it was noticed that they were brought by the Hermit Crabs.

The species of _Coenobita_ possess a very curious adaptation for aerial respiration. The soft skin of the abdomen is traversed by a network of bloodvessels and acts as a kind of lung, and a pair of contractile vesicles at the base of the abdomen serve as accessory hearts in promoting a specially active circulation in that part of the body. The lining membrane of the gill chambers also appears to aid in respiration as in other terrestrial Decapods.

The "Robber Crab" or "Coconut Crab" (_Birgus latro_--Plate XXVII.) also belongs to the family Coenobitidæ, and has attracted much notice from its relatively gigantic size and its singular habits. Although resembling _Coenobita_ closely in essential structure, _Birgus_ differs from it and from most other Hermit Crabs in not making use of a portable shelter, perhaps owing to the difficulty of obtaining one of suitable size. The necessary protection for the abdomen is obtained by a redevelopment of the shelly plates (terga) on the upper surface of the abdominal somites. The abdomen is carried doubled underneath the body to protect the soft under-surface, and the animal, when threatened, seeks a shelter for its vulnerable hinder part in the nearest hole or cranny. The swimmerets are absent in the male sex, and are present only on one side of the abdomen in the female. This unsymmetrical development of the appendages is interesting as indicating the derivation of the Robber Crab from ancestors adapted to living in the unsymmetrical shells of Gasteropod Molluscs. The last pair of abdominal appendages, which in other Hermit Crabs serve to hold the body in the shell, are here much reduced in size, and quite useless for that purpose. The carapace is very broad posteriorly, owing to the great development of the branchial cavities, which are much too capacious for the very small gills. As in the true Land Crabs, the lining membrane of the gill cavity is thick and spongy, and traversed by numerous bloodvessels; but in this case its efficiency as a lung is added to by numerous tufted papillæ, which increase the surface exposed to the air.

As in other Hermit Crabs, the last two pairs of legs are shorter than the others, and they end in small chelæ. The last pair are very slender, and are usually carried folded up within the gill chambers, which they possibly serve to keep clear from foreign bodies. The penultimate pair of legs are stouter, and the two pairs in front of these are long walking legs. The chelipeds are very strong and are of unequal size. When attacked, the animal defends itself, not, as might have been expected, with its chelipeds, but with the first pair of walking legs, the sharp points of which form very efficient weapons.

The statement that the Robber Crab climbs lofty trees was first made by the Dutch naturalist Rumphius, in the beginning of the eighteenth century. Its accuracy has been often doubted or denied since then, and only finally put beyond dispute by a photograph taken on Christmas Island by Dr. Andrews, which shows one of these Crabs in the act of descending the trunk of a sago-palm. It seems not impossible that the habits of the animal may vary to some extent in different localities, and that where food is abundant on the ground the tree-climbing habit may be in abeyance. If this were so, it would explain the very definite statements made by some observers, that _Birgus_ does _not_ climb trees.

In localities where coconut palms abound, _Birgus_ feeds largely on the nuts, tearing off the fibrous outer husk and breaking open the shell by hammering with its powerful claws at one of the "eye-holes." According to Darwin in his "Naturalist's Voyage," the pincers of the penultimate pair of legs are used for extracting the contents of the nut, but this observation does not seem to have been confirmed. In spite of its name of "Coconut Crab," however, _Birgus_ by no means feeds exclusively on coconuts. On Christmas Island, where until recently there were no coconut palms, the Crabs are exceedingly abundant, and, according to Dr. Andrews, they "eat fruits, the pith of the sago-palm and the screw-pines, dead rats and other carrion, and any of their fellows that may have been injured.... They are excellent scavengers, and have a curious habit of often dragging their food long distances before attempting to eat it. I have seen a Crab laboriously pulling a bird's wing up the first inland cliff, half a mile or more from the camp whence it had stolen it."

Large specimens of the Robber Crab may be at least a foot in length of body when the abdomen is straightened out. Their great strength is illustrated by the fact, related by Darwin, that specimens placed in a strong biscuit-tin, of which the lid was secured by wire, escaped by turning down the edges with their claws, and in doing so actually punched holes quite through the tin.

The breeding habits and mode of development of the Robber Crab have often formed the subject of inquiry by naturalists, but it is only recently that Dr. Willey has been able to prove definitely that the female visits the sea for the purpose of hatching off the eggs, and that the young are hatched in the zoëa stage. The larvæ obtained by Dr. Willey have been described by Mr. Borradaile, who finds that, as was to be expected, they closely resemble those of _Coenobita_. There appears, however, to be no such simultaneous migration of the Crabs towards the sea as has been described in the case of the Gecarcinidæ. The statement, quoted by Darwin, that _Birgus_ visits the sea every night for the purpose of moistening its branchiæ, cannot be universally applicable, since the Crabs are often found, as on Christmas Island, at distances from the sea which put a nightly journey to it out of the question.

Of all Crustacea, the most completely adapted to terrestrial life are the Land Isopods, or Woodlice, which may be found in every garden. It is true that most species are found in damp places, although some that inhabit the sandy deserts of Asia and Africa must be content with a very slight degree of humidity; and in no case is their dependence on moisture greater than, for instance, that of many Insects and Arachnids which are regarded as typically terrestrial animals. Since there is reason to believe that the Woodlice have been derived from marine ancestors--they show no special affinities to the fresh-water Isopoda, like _Asellus_--it is interesting to find that the most primitive forms, which have departed least from the general Isopod type, are commonly found on or near the seashore. The "Sea-slater," _Ligia oceanica_ (Fig. 63), which is abundant in rocky places on our own coast, is one of the most primitive forms. It has a broad, flattened, greenish-brown body, about an inch long, and it runs quickly, creeping into narrow crevices of the rocks, so that it is not easy to catch. The antennules, as in the other land Isopods, are very minute, but the antennæ are long, and have, besides the five segments which form the "peduncle," a "flagellum" of about twelve short segments. The uropods or tail appendages are long, each with two slender, pointed branches. On the under-side of the abdomen can be seen the five pairs of pleopods, each with two plate-like branches attached to a very short peduncle. As in most aquatic Isopods, the plates of the pleopods are soft and thin, and appear adapted to act as gills, although the outer plate of each pair is somewhat stiffer than the inner. The Sea-slater is generally found just above high-water mark, probably always within reach of the salt spray, and it is said sometimes to enter the water of rock-pools.

In almost every garden there may be found, under flower-pots and the like, a Woodlouse, about two-thirds of an inch long, of a brown colour, with yellowish blotches arranged in a row on each side of the back. This is _Oniscus asellus_, a species widely distributed in Europe and North America. It has the antennæ shorter than in _Ligia_, and the flagellum is composed of only three segments. The uropods are quite short. The endopodites of the pleopods are membranous gill-plates, which serve for respiration in the moist air in which these animals generally live. The exopodites are stiff plates which cover and protect the delicate endopodites; it is probable that they also aid in respiration, for they contain a system of minute channels, filled with air, where the cuticle is separated from the underlying cells. As these channels are nowhere open to the outside, the air must find its way in by diffusion through the cuticle.

Even more abundant than _Oniscus asellus_, and often found together with it, is _Porcellio scaber_ (see Fig. 20, p. 51). It is usually of a dark bluish-grey, but occasionally it is irregularly mottled with a lighter colour. The flagellum of the antenna has only two segments. The most interesting difference from _Oniscus_, however, is found in the pleopods. If the under-side of the living animal be examined with a pocket lens, a white spot will be seen on each exopodite of the first two pairs of pleopods. When the structure of the pleopods is investigated by means of microscopic sections (Fig. 64), it is found that the white spots are tufts of fine branching tubes radiating into the interior of the exopodite from a slit-like opening on the outer edge. These tubes arise by an in-pushing of the integument, and they are lined throughout by a delicate continuation of the external cuticle. During life they are filled with air, and they serve to aerate the blood circulating in the interior of the appendage.

Another Woodlouse common in England is _Armadillidium vulgare_ (Fig. 65), a little slaty-grey species with a very convex body, which rolls itself into a ball when touched. Like the last-mentioned species, it has two segments in the flagellum of its short antennæ, and it has tufted air-tubes in the exopodites of the first two pairs of pleopods. It is often mistaken for an animal of widely different structure, which it superficially resembles--the Pill Millipede (_Glomeris marginata_). The latter, however, may easily be recognized by having either seventeen or nineteen pairs of walking legs (instead of seven pairs), set close together in the middle line of the body, and by lacking the plate-like pleopods. The resemblance between the two animals can hardly be regarded as a case of "mimicry," since there is no reason to believe that either benefits by its likeness to the other. As in so many other cases of "convergent resemblance" between animals of different structure, it does not seem possible to get beyond the vague suggestion that a similarity in habits may have led, in some way that we do not understand, to a similarity in appearance.

The presence of air-tubes in the pleopods of many Woodlice raises some questions which are of importance with reference to the classification of the Arthropoda as a whole. The Six-legged Insects, most Spiders and many of their allies, the Centipedes and Millipedes, and the worm-like _Peripatus_, all breathe air by means of fine tubes which penetrate throughout the body, and bring the air into close contact with the tissues. These tubes, which are known as "tracheæ," arise as ingrowths of the outer layer of the embryo, and are lined by a delicate continuation of the external cuticle. It has been held by some zoologists that so peculiar a system of breathing organs must indicate a common descent of the animals that possess them, and accordingly it has been proposed to separate the Insects, Arachnids, Myriopods, and _Peripatus_, as a group, Tracheata, from the Crustacea and some other Arthropods which have no tracheæ. The air-tubes of the Woodlice, however, are precisely like tracheæ in structure and function, and only differ from the tracheæ of the other groups in the fact that they are confined to the appendages, and do not penetrate into the body. Since the Woodlice are a small and highly specialized branch of the Crustacea, we can hardly suppose that they derive their tracheæ from any ancestral type which they had in common with the widely different Arachnids, for example; and if tracheæ have been evolved independently in these two groups, there seems no reason why those of the Insects may not have arisen independently of either. This is only one example out of many which go to show that, in attempting to reconstruct the genealogy, or phylogeny, as it is called, of the animal kingdom, we must constantly admit the possibility of "convergent evolution."

Although Woodlice are very common animals, comparatively little is known of their habits. They seem to live chiefly on vegetable food, and sometimes damage seedlings and tender plants in gardens and greenhouses, but occasionally they are carnivorous, and even cannibalistic, in their habits. A few species live as "guests" in ants' nests, and one of these, the little blind white _Platyarthrus hoffmannseggii_, is common in many localities in this country. Why the ants tolerate their presence we do not know, for they do not seem to render any service to their hosts, as do the plant-lice and some other insects that are kept by the ants for the sake of the secretions which they yield.

The Woodlice, like some other Isopoda, have a peculiar method of moulting. Instead of the whole exoskeleton being cast off at one time, as in other Crustacea, that of the hinder half of the body is moulted first, and it is only after two or three days, when the new cuticle has hardened, that the exoskeleton of the anterior half follows. As a result of this arrangement, it occasionally happens that specimens are found with the fore part of the body differing in colour from the hind part, owing to the one having been moulted more recently than the other.

Woodlice occur in most regions of the globe, and one of the most remarkable features of their geographical distribution is the extremely wide range of certain species. This is probably due, at least in many cases, to their accidental transport by human agency. Thus, _Porcellio scaber_, so common in this country, is also found in great abundance in New Zealand; but Professor Chilton notes that it is usually found near buildings, and only rarely in the native bush, so that there can be no doubt that it has been introduced by artificial means.