The same kind of organs occur in Amblystoma.^[18] They appear, previous to the breaking open of the gill-clefts, as protrusions of epiblast, long before any of the external gills on the branchial arches. When the clefts have broken open, the quadrate sends out laterally a tiny crescent-shaped process a little above the jaw-joint, and this process extends to the base of the balancer, but not into it, and a bundle of muscle-cells grows into the balancer. It is easy to recognise the same organ in the extremely long thread-like structures of the larva of Xenopus. In the Apoda they are likewise present, but are retained permanently as highly specialised, probably tentacular organs (cf. p. 86, Apoda).

One of the most unexpected features is the suppression of the lungs in various kinds of Salamandridae. The lungs are either reduced to useless vestiges or they are quite absent. This occurs in aquatic and terrestrial, American and European forms, and it is noteworthy that the reduction of the lungs does not apply to all the species of the various genera, nor is it restricted to one sub-family.

The following list is due to the researches of H. H. Wilder,^[19] L. Camerano,^[20] E. Lönnberg,^[21] and G. S. Hopkins^[22]:–All the Desmognathinae and Plethodontinae; Amblystomatinae, Amblystoma opacum; Salamandrinae, Salamandrina perspicillata. In Triton and other Salamandrinae the length of the lungs varies; in some they extend more, in others less, than half way down the distance between head and pelvis. Hopkins remarks: "Two questions are naturally suggested by this apparently aberrant condition of the respiratory organs. First, what structures or organs have taken on the function of the lungs and branchiae; and secondly, is there any modification in the form or structure of the heart which in any way may be correlated with the above-mentioned peculiarities of the lungless forms?" Wilder concluded that respiration was probably carried on by the skin, and perhaps, to some extent, by the mucosa of the intestine. Camerano thinks that, at least in the European forms, respiration is effected by the bucco-pharyngeal cavity, and that the skin affords no efficient aid. The left auricle in the lungless forms is much smaller in comparison than the right, and there is no pulmonary vein. The auricular septum has a large aperture, the communication between the auricles being larger than even in Necturus (which breathes essentially by gills). The sinus venosus, instead of opening into the right auricle only, opens more freely into the left than into the right, and the latter communicates more directly with the ventricle than the left, instead of about equally. In short, the heart of these creatures appears almost bilocular, instead of being trilocular, at least functionally.

The lungs of the Urodela are always simple, extremely thin-walled bags. They are highly developed in the Anura, the walls being modified into numerous air-cells, whereby the respiratory surface is considerably increased. The lungs are filled with air by the pumping motion of the throat while the mouth is closed, the nostrils being provided with muscular valves. A muscular apparatus assists the filling of the lungs in the Anura.^[23]

Most, if not all, Anura and some Urodela have a voice produced by the larynx, which, especially in the Anura, is provided with a complicated cartilaginous and muscular apparatus and with vocal cords. The voice of the Urodela is at the best a feeble squeak. The females of the Anura are either mute or they produce a mere grunt, but that of many males is very loud, and, moreover, in many species it is intensified by vocal sacs which act as resonators. These sacs are diverticula of the lining of the mouth-cavity, and bulge out the outer skin and the muscles, chiefly the mylo-hyoid, of the throat. The nostrils and the mouth are firmly closed during the croaking. "The sacs are called internal when they are covered by the unmodified gular integument, however much this may be distended; external when their membrane projects through slits at the sides of the throat, as in Rana esculenta (Fig. 52, p. 269), or when the skin is thinned and converted into a bladder-like pouch, as in Hyla arborea."^[24] These sacs exhibit many modifications. They may be unpaired and median, and open by two slits into the mouth, on either side below the tongue; in Bufo one of the slits or openings, either the right or the left, is obliterated. They may be paired and symmetrical, and open one on each side of the head, below and near the posterior angle of the jaws. These modifications differ in closely allied species. They reach their greatest complication in Rhinoderma and in some of the Cystignathidae by extending far back beneath the skin into the wide lymphatic spaces. In Rhinoderma they are put to the unique use of nurseries for the young (see p. 228). Leptodactylus typhonius has a very distinct pair of outer vocal sacs and a well-marked unpaired sac which extends into the belly and communicates with each outer sac. Several species of Paludicola, e.g. P. fuscomaculata and P. signifera, have a similar arrangement, in addition to an unpaired gular sac which can be inflated independently of the rest (see Fig. 45, p. 220).

Urino-Genital Organs

The kidneys and the male generative glands are still intimately connected with each other. The general plan is as follows:–

The kidneys consist of a large number of glomeruli, produced by the coiled segmental tubes, each of which is composed of a nephrostome or funnel opening into the body-cavity, a Malpighian body and an efferent canal. The latter combine to form the segmental duct which opens into the cloaca. The testes, composed of a large number of sperm-producing glands, are drained by transverse canals which combine into a longitudinal canal, and this again sends off numerous efferent canals which open into the efferent canals of the kidney, so that the segmental duct (Leydig's duct of many authors) conveys both sperma and urine.

In the female the network of transverse and longitudinal canals, which originally connect the generative glands with the kidney's efferent canals, is deduced in so far as the connection is interrupted and the vestiges of the transverse canals are no longer functional. The eggs fall into the body-cavity and are caught up by the ostium or inner abdominal opening of a special duct, the oviduct (Müllerian duct of many authors). Vestiges, more or less complete, of these oviducts persist in the males of most Amphibia.

This general scheme presents some modifications in the various groups of Amphibia.

The Apoda retain the most primitive conditions. The kidneys are still long and narrow, and the glomeruli are, at least in the anterior part of the organ, still strictly segmental, agreeing in number and position, each with a vertebral segment; later, the number of the glomeruli is greatly increased, and the former agreement becomes quite disturbed. The generative glands still retain their segmental arrangement, but they are restricted to a much shorter region than the kidneys. In the male Apoda a considerable portion of the cloaca can be everted by special muscles, and acts as an intromittent organ. Both sexes possess a ventral urinary bladder.

In the Urodela both kidneys and testes are much concentrated, the testes especially have lost all outward appearance of segmentation, and their efferent canals, connecting them with the longitudinal collecting canal, are much reduced in numbers. The greater portion of the kidneys, at least their anterior half, has all the appearance of a degenerating organ and is on the way to losing its urinary function, although it still possesses Malpighian bodies and complete ducts; the main function of the latter is now the conveyance of the sperma. In the Perennibranchiata, and in some others, e.g. Spelerpes variegatus, the longitudinal collecting canal, between testis and kidney, is sometimes suppressed, a very simple, but pseudo-primitive arrangement. A urinary bladder is present. The cloaca is not eversible.

In most Anura, e.g. Rana and Bufo (Fig. 7; 4, 5), the same scheme is adhered to. The efferent canals of the testis form a network, with a longitudinal canal, and open into the efferent canals of the kidney, in the substance of which they are more or less deeply imbedded. The ducts which lead out of the kidney to compose Leydig's duct, are frequently dilated, or the latter duct is much elongated, convoluted or varicated, and this whole portion is enclosed in a sheath of connective tissue, giving an appearance as if the single duct itself were dilated in the greater part of its length; hence the occasional name of vesicula seminalis. Such means of storing the sperma enable the latter to be ejected suddenly in great quantities.

In Bombinator (6) some of the most anterior seminal canals do not perforate the kidney, but run over it superficially and open directly into a branch of Leydig's duct. This branch, no doubt equivalent to a number of segmental canals which have lost their uriniferous function, is curved round the upper end of the permanent kidney, while its forward continuation, ending blindly, is the remnant of its former headward extension. This arrangement of Bombinator is carried further in Discoglossus (7). The testis conveys its sperma through a wide duct directly into Leydig's canal, without interfering with the kidney, and all the testicular efferent network is lost. The anterior end of Leydig's duct still extends headwards; its middle portion acts solely as a vas deferens, while the lower portion still behaves like a typical segmental duct, conveying both sperma and urine. Lastly, in Alytes (8) the functional division of the old segmental duct has been carried to an extreme. The kidney is drained by one canal only, now a true ureter, and this is of course produced by a consolidation of the multiple exclusively uriniferous canals of the lower half of the kidney. The whole of the segmental duct is now in the service of the testis, and near its junction with the ureter it forms a large diverticulum or true vesicula seminalis.

Remnants of oviducts, or Müllerian ducts, are common in the male Anura; they are best developed in Bufo, much reduced, and individually absent, in Rana. In Bombinator each duct is restricted to its upper or abdominal portion, and is attached to the vestigial headward extension of Leydig's duct. Lastly in Discoglossus and in Alytes all traces of oviducts seem to have vanished, at least in the adult males.

It is interesting to note that in the arrangement of the urino-genital ducts the Discoglossidae are the most advanced of all Amphibia, instead of showing the most primitive conditions. This is rather unexpected, but is paralleled by the epichordal type of the vertebral column.

The oviducts of the Apoda and Urodela remain more or less straight; in the viviparous species they form uterus-like dilatations. In the Anura they become greatly elongated during the breeding season and form many convolutions. As a rule each oviduct opens separately into the cloaca, but in Hyla they have one unpaired opening, while in Bufo and Alytes the lower parts of both oviducts are themselves confluent.

All Amphibia possess Fat-bodies. They consist of richly vascularised lymphatic tissue, the meshes of which are filled with lymph-cells, globules of fat and oil. In the Apoda these bodies lie laterally to the generative glands, and along the posterior half of the kidneys. In the Urodela they accompany the anterior half of the kidney. In the Anura they are lobate, and are placed upon the anterior end of the testes or ovaries. Their exact function is still doubtful, but it is intimately connected with that of the generative glands. The old notion, that they are simply stores of fat for the nourishment of the animal during hibernation, is quite untenable. The fat-bodies do not decrease during this period, on the contrary they attain their fullest size in the spring at the time of the rapidly awaking activity of the reproductive organs, and they enable considerable quantities of sperma and of eggs to be produced and ripened without detriment to, or utter exhaustion of, the animals, which often spawn before they have had time or opportunity to feed. After the spawning season the fat-bodies have dwindled down to inconspicuous dimensions.

Lastly, there is in some Anura, hitherto observed in Bufo only, a mysterious organ, intercalated between the fat-body and the testis or ovary. This is "Bidder's organ" and it seems to be a rudimentary ovary, or rather that upper, anterior portion of the whole organ which undergoes retrogressive metamorphosis. It disappears in old female toads, but in the males it sometimes assumes a size equal to, or surpassing that of the testes. The males are in this respect hermaphrodite, and cases are known in which parts of the generative glands have developed testes and egg-bearing ovaries.

The spermatozoa of the Apoda and Urodela have an undulating membrane along the tail, while the head-end is either pointed or truncated. Those of Spelerpes fuscus and of Ichthyophis glutinosa measure about 0.7 mm. in total length, those of the other Urodela being much smaller. A peculiarity of the Urodela is that their spermatozoa are massed together in or upon spermatophores, an arrangement which undoubtedly facilitates the internal fecundation of the female without actual copulation. The female takes up such a deposited spermatophore with the cloacal lips, squeezes the sperma out of the capsule which remains behind, and either conveys the former into a special receptaculum seminis, e.g. in Salamandra atra and in Triton, or the spermatozoa wriggle their way, thanks to the undulating tail, directly up the oviducts to the ova.

The spermatophores are composed of a colourless, soft, gelatinous mass, which is probably produced by the cloacal gland. The shell of jelly is in fact a cast of the cloacal cavity, reproducing all its ridges, furrows and folds, while a toad-stool-shaped papilla of the cloaca makes the inside lumen of the cast, e.g. in Triton. Those of Salamandra maculosa are much simpler, consisting, in conformity with the absence of a cloacal papilla, merely of a cone with a globular mass of sperma on the top. Those of Amblystoma are similar.

The spermatozoa of the Anura show considerable differences in the various genera, of which, however, only the European forms have been properly examined. The "head" is wound like a corkscrew in Discoglossus, Pelobates, and Pelodytes; spindle-shaped, more or less curved, in Rana temporaria and R. agilis, Hyla, Bufo and Bombinator, in the latter with an irregular membrane on one side; cylindrical in Rana esculenta and R. arvalis. The tail is mostly long and filiform, but in Bufo vulgaris and Discoglossus it is provided with an undulating membrane. Their size is generally very small, only about 0.1 mm., excepting those of Discoglossus which reach the astonishing length of 3 mm. These differences in shape, especially that of the head, explain why species of the same genus, e.g. Rana temporaria and R. arvalis, cannot fertilise each other.

The eggs differ much in size, colour, and numbers. They are holoblastic, with unequal cleavage, but those species which possess an unusual amount of food-yolk, for instance Rhacophorus schlegeli and the Apoda, approach the meroblastic type of segmentation. As a rule, the greater the amount of yolk, the smaller is the number of eggs produced. But the number which is laid during one season is not only difficult to calculate, but it varies individually, old females laying more than young specimens. Moreover, some kinds, e.g. the Discoglossidae, spawn several times in one year. Alytes, Rhinoderma, Hylodes, Rhacophorus, Pipa, in fact those kinds which are remarkable for special nursing habits, lay only a few dozen eggs at a time. Hyla arborea produces up to 1000, Rana temporaria about 3000, Bufo vulgaris averages 5000, Bufo viridis and Rana esculenta up to 10,000 and more. T. H. Morgan^[26] has observed a Bufo lentiginosus which laid 28,000 eggs within ten hours! The number of eggs produced by the Apoda and Urodela is comparatively moderate, in the average a few dozen, Amblystoma alone laying about 1000.

The eggs possess a gelatinous mantle of variable thickness and consistency. In Amphiuma they are strung together like the beads of a rosary, and the envelope hardens into a kind of shell. Many Newts and some Anura fasten their eggs singly on to plants and other objects in the water, with or without threads of stiffening mucus. In many Anura, e.g. Bufonidae, they pass out as closely-set strings of beads, one string out of each oviduct; in others, e.g. Ranidae, they are disconnected, and form large, lumpy masses, especially when the gelatinous mantle swells up in the water. The use of this mantle seems to be chiefly the protection of the growing embryo, which in many species, when hatched out of the egg proper, drops into and remains for some time in the softened jelly. Possibly the latter affords some nutriment to the early larva.

Concerning the mode of fecundation it is to be remarked that copulation proper takes place only in the Apoda. For the Urodela Boulenger^[27] has given the following summary. In no case does actual copulation take place. The male deposits the spermatophores which it is the office of the female to secure:–

The act of fecundation of most of the other kinds of Urodela, notably Cryptobranchus, Amphiuma, Proteus, has not yet been observed.

Embracing of the two sexes is the universal rule with the Anura, the male creeping on to the back of the female and clasping her firmly with the arms and hands either in the inguinal region, higher up, or under the armpits. See the numerous statements in the systematic part. This often extremely forcible, pressing embrace seems to be necessary, although the females can deposit the eggs without the help of the male, but in such cases the expulsion takes place at irregular intervals instead of at one time. When the eggs appear at last, and this happens in many species many hours, or even some days, after the beginning of the embrace, the male voids the contents of its seminal vesicles over them. Fertilisation is consequently external, with the possible exception of Pipa, q.v. p. 152.

Deposition of the eggs and nursing habits.–The majority of the Amphibia are oviparous, but some Apoda and Urodela are viviparous. It is unnecessary to call the latter condition ovo-viviparous, since this is really a distinction without a difference.

Viviparous forms:–amongst Urodela; Salamandra maculosa, the young burst the egg-membrane in the act of being born, and are provided with long gills; S. atra, the young undergo their whole development and metamorphosis within the uterus (see p. 119); Spelerpes fuscus, the young are likewise born in the perfect condition: amongst Apoda; Typhlonectes compressicauda and Dermophis thomensis.

The oviparous Apoda, at least Ichthyophis and Hypogeophis, and a few of the Urodela, as Desmognathus and Amphiuma, take care of their eggs by coiling themselves around them in a hole underground.

Nursing habits are very common amongst the Anura. Boulenger^[28] has summarised the various conditions concerning the deposition and care that is taken of the eggs, in the following list, in which more recent discoveries have been interpolated.

The development and metamorphosis of many species have been described in the systematic part. The following is a short general account of some of the more important features. Metamorphosis in the Apoda and Urodela is restricted chiefly to the reduction of the gills, the closing of the clefts, and the loss of the gill-chamber and the finny margins of the tail; but the change from the tadpole to the final Anurous animal implies an almost entire reorganisation.

In the earliest condition the embryo consists of a large head and body, while the tail is still absent. Behind the beginnings of the future mouth appears a transverse crescentic fold, with the convexity looking backwards, which develops into the paired or unpaired adhesive apparatus. This consists of large complex glands, developed in the Malpighian layer, originally covered by the cuticula, which soon disappears, whereupon the sticky secretion enables the larva to attach itself to the gelatinous mantle of the egg, later on to weeds or other objects in the water. The name of suckers, often applied to this apparatus, conveys a wrong idea, there being neither muscles nor any suctorial function. The shape of this organ undergoes many changes during the early life of the individual, and differs much in the various genera, affording thereby diagnostic characters.^[29] At first a crescent, it divides into a right and a left oval or disc, which either remain asunder and behind the mouth (Rana, Bufo), or they move forwards to the corners of the mouth (Hyla) or further back, and unite again more or less completely, as in Discoglossus and Bombinator. It is mostly of short duration, and disappears by the time that the larva, by the proper development of the gills and the tail and the functional mouth, changes into the tadpole. But in a few species these discs transform themselves into an elaborate ventral disc. Such an organ persists throughout the greater part of the tadpole-stage in certain Oriental species of Rana, all of which, when adult, possess fully webbed toes and strongly dilated discs on the fingers and toes, e.g. Rana whiteheadi, R. natatrix, and R. cavitympanum of Borneo, R. jerboa of Java (this larva was originally described and figured as that of Rhacophorus reinwardti), and R. afghana of the Himalayan system. These tadpoles, at least those of R. jerboa, are further remarkable for having the "spiracular" opening very far back on the left side, nearer to the base of the tail than to the snout, so as to be well out of the way when the creature has attached itself by the adhesive disc.

The mouth of the tadpoles of Anura is furnished with horny armaments, substitutes for teeth. Their development and that of the mouth in general has been well described by Gutzeit.^[30] In the young larvae of Rana temporaria, one or two days after hatching, a shallow groove appears above the conspicuous pair of adhesive organs. The groove becomes rhombic in outline, and when the mouth has been formed in its centre, the jaws appear in the median corners of the rhombus. The epidermis then rises like a circular wall around the jaws, and divides into an upper and lower lip; furrows appear on them, and between these various papillae and comb-like transverse plates of teeth. The papillae are possibly tactile organs, but although nerves enter them, nerve-endings of a sensory nature have not yet been discovered. On the fourth day the jaws become black, by the tenth day horny teeth have appeared upon all the plates of the mouth-armature, and on the seventeenth day the mouth-apparatus has reached the configuration typical of the tadpole, which is now about 14 mm. long. The number of horny teeth in R. temporaria amounts to about 640. These teeth are not cuticular products, but cornified cells; they are very small, and consist each of one horny cell, which is shaped like a nightcap, the apex of which is curved back and serrated. The little teeth are shed continuously, the renewal taking place by successive cells growing into the bases of the older series. The shape and size differ much in the various genera and species. The comb-like plates, composed of those teeth which surround the lips, seem to be used chiefly for the fixing or hooking of the food, while those which compose the horny beak proper, the armature of the jaws, are used like the radulae of snails. These beaks are likewise composed of a great number of individual teeth, closely packed together in several rows, but the teeth themselves are simple and not serrated.

In Hyla arborea there are in all about 560 teeth. The development of the mouth does not begin before the eleventh day; the horny teeth break through, and the jaws get black edges, on the eighteenth. In Pelobates fuscus the number of horny teeth is increased to about 1100. In Borborocoetes taeniatus the horny teeth form series of five bells, which fit into each other like the joints of a rattlesnake's tail.

One of the most extraordinary kinds of tadpoles is that of Megalophrys montana.^[31] Mr. Annandale (Skeat Expedition) found it at Bukit Besar, Malay Peninsula, from 2000 to 3000 feet above the level of the sea. The tadpoles (Fig. 11) were found in the beginning of the month of May 1899 in sandy streams and in pools of rain-water; they floated in a vertical position, the peculiar membranous funnel-shaped expansion of the lips acting as surface-floats. The inside of the funnel is beset with radiating series of little horny teeth, and the whole apparatus is possibly used for scraping the under-surface of the leaves of water-plants in search of food. Total length of the tadpoles 1 inch.^[32]

The gills, the formation of the operculum, and the modifications of the branchial arterial arches have been described fully on p. 43; those of the hyo-branchial skeleton on p. 31. Fusion of the opercular fold with the skin of the neck, across the branchial region, causes the head to become confluent with the trunk (cf. Fig. 9, 3, p. 57). The body becomes oval, more or less globular, and the alimentary canal is greatly elongated and stowed away in the shape of a neat, very regular spiral, shining through the ventral wall of the body; the anus opens at the end of a somewhat protruding tube, either in the median line, just in front of the ventral fin (Discoglossidae, Pelobates, Bufo), or it assumes an asymmetrical position by turning to the right side (Hyla, Rana).

Although both pairs of limbs begin to bud simultaneously, or the fore-limbs even earlier, the hind-limbs are hurried on, and appear first, long before the fore-limbs. The latter lie ready beneath the skin of the gill-chamber, and the right always breaks through the skin, while the left does the same in the Mediogyrinidae, while in the Laevogyrinidae it is generally pushed through the left-sided spiracular opening, immediately behind the outer gills. According to Barfurth the right limb appears, in about 80 per cent. of Rana esculenta, from two to eight hours before the left.

Meanwhile the lungs are being developed, and the tadpole occasionally rises to the surface to breathe air. The gills, which, as has been explained elsewhere, are less ancestral than they are larval organs, degenerate, and all the organs are modified for the coming terrestrial life. The fins of the tail are absorbed, the horny armature of the mouth and lips is shed in pieces and makes room for the true teeth, the eyes receive lids, and the whole cranium, especially the apparatus of the jaws, undergoes the final modifications–widening and lengthening of the mouth, arresting of the mento-Meckelian cartilages, elongation of the Meckelian cartilages or lower jaw proper, shifting backwards of the suspensorium, and lengthening of its orbital process to form the pterygo-palatine bridge.

The tadpole ceases to feed, the whole intestinal canal is voided of its contents, and by "histolysis" is thoroughly rebuilt, becoming wider and shrinking to about one-sixth of its original length,–undoing thereby the spiral–preparatory for the coarser food, which consists of insects, worms, and other strictly animal, living matter. Hitherto the tadpoles have lived on "mud," confervae, Diatoms, rotting vegetable and animal matter. The anal tube collapses, becomes ultimately absorbed, and a new vent is formed at and below the root of the tail.

Barfurth^[33] has made interesting observations and experiments with regard to the absorption of the tail and other organs which disappear during the metamorphosis. This is retarded by low temperature; it is accelerated by rest and freedom from mechanical disturbances, as, for instance, concussion of the water. Hunger shortens or hurries on the last stages of metamorphosis, the absorption of the tail taking place in four instead of five days. Amputation of the tail has no retarding influence; it is followed at once by regeneration, although the tadpole may be on the verge of reducing the tail. Whilst hungering the whole organism draws upon its available store of material, naturally first upon those parts which sooner or later are to become superfluous. This applies eminently to the tail, which represents a considerable amount of "edible" matter, and also to that portion of the skin which still covers the fore-limbs. The elements of the cutis are resorbed, thereby thinning the skin; and consequently the limbs break through earlier in fasting than in well-fed specimens. Nature herself seems to apply hunger as an accelerator. Mlle. von Chauvin found that the larvae of Urodela normally fast during the transformation, and according to Barfurth the larvae of Rana temporaria eat less after their hind-limbs are fully developed. This is, however, also preparatory for the reorganisation of the gut, which has to be more or less empty during the shortening process.

The loss of the tail is not due to a sudden dropping off of this organ–a crude but by no means uncommon belief–but is brought about by a very gradual process of resorbtion. When the fore-limbs begin to break through the skin, the tip of the tail shrinks and becomes black, owing to an increase, or rather concentration, of the pigment cells. The reduction proceeds from the tip forwards until on about the fifth day there remains only a short, conical, black stump. From the beginning of this process of reduction the tail is scarcely used for locomotion, the tadpole rowing with its legs, or it crawls and hops about, although the tail may still be 20 mm. long. The cells of the epidermis atrophy, shrink, and peel off, while those of the cutis, blood-vessels, nerves, muscles, and chorda dorsalis become disintegrated, often undergoing fatty degeneration. The leucocytes eat up the débris and other dissolved tissue, and carry it away through the lymphatic vessels, to be used as new building material in the rest of the animal.

Barfurth asks very properly, Why do these tissues degenerate and die? Because the vasomotor nerve-fibres cease to regulate the circulation. And why does this trophic influence of the central nervous system stop? Because the function of the tail becomes superfluous through the appearance of the fore-limbs. The tail is doomed, and degenerates like any other organ without a function. The whole process is, of course, a recapitulation of ancestral, phylogenetic evolution.

CHAPTER III

NEOTENY–REGENERATION–TEMPERATURE–GEOGRAPHICAL DISTRIBUTION

Neoteny.–It has long been known that the larvae of the Spotted Salamander occasionally attain the size of 80 mm. or about 3 inches, whilst the majority undergo metamorphosis when they are only 40 mm. long. Again, larvae of Triton have been found, in the months of April and May, 80 to 90 mm. long, still with functional gills, but with the sexual organs fully developed. De Filippi^[34] found in one locality in Lombardy, besides a few normal fully metamorphosed specimens of only 30 mm. in length, more than forty specimens, which, although they had attained full size, about 55 mm., and were sexually mature, still retained their gills. According to him such gill-breathing, otherwise mature specimens, occur constantly in a small lake in the Val Formazzo, on the Italian slope of the Alps, in the province of Ossola. Later Duméril^[35] astonished the world by his account of the metamorphosis of the Mexican gill-breathing Axolotl into an entirely lung-breathing and terrestrial creature, hitherto called Amblystoma, and supposed to be not only a different species, but to belong to a different family from the Axolotl, which was known as Siredon axolotl s. pisciforme, and naturally classed with the Perennibranchiata.

This discovery led to a series of observations and experiments, chiefly conducted by Marie von Chauvin, instigated thereto by Koelliker and by Camerano.^[36] It was then found that many, if not most of the European Amphibia, both Urodela and Anura, occasionally postpone their metamorphosis, and also that such Urodela sometimes become adult for all practical purposes, but retain their gills.

This retardation, the retention of larval characters beyond the normal period, was called Neotenie by Kollmann^[37] (νέος, young; τείνω, extend, stretch). He distinguished further between:–I. Partial Neoteny, namely, simple retardation of the metamorphosis beyond the normal period, for instance, the wintering of tadpoles of Pelobates fuscus, Bombinator pachypus, Pelodytes punctatus, Alytes obstetricans, Hyla arborea, Rana esculenta, R. temporaria, Bufo vulgaris, and B. viridis: II. Total Neoteny, where the animal retains its gills, but becomes sexually mature; hitherto observed in Urodela only, e.g. Triton vulgaris, T. alpestris, T. cristatus, T. boscai, T. waltli and Amblystoma. Intermediate stages between these two categories are not uncommon.

A satisfactory explanation of the meaning of neoteny is beset with difficulties. Some authorities look upon the phenomenon simply as the result of adaptation to the surroundings, which make it advantageous for the creature to retain its larval features. Others think that the surroundings somehow or other retard or prevent the assumption of the adult characters. Undoubtedly there are many cases in which larvae have been reared in water-holes with steep walls, so that they could not change from aquatic to terrestrial life, and it stands to reason that abnormally forced and prolonged use of the gills and of the tail may stimulate these organs into further growth at the expense of the limbs and other organs which are intended for terrestrial life. But not unfrequently typical neotenic and overgrown specimens occur side by side with others which have completed their metamorphosis, and the same is true of larvae of newts which were reared, for experimental purposes, under exactly the same conditions–for instance, in a high-walled glass vessel.

Weismann tried to explain neoteny as cases of reversion to atavistic ancestral conditions, but this idea is based upon an assumption which is probably wrong. His idea necessitates the supposition that all the Amphibia were originally gill-breathing, aquatic, and limbless animals, and that every feature seen in a larva must necessarily indicate an ancestral phylogenetic stage. It is, on the contrary, much more probable that the external gills of the Urodela have been developed in adaptation to their embryonic and larval, essentially aquatic, life. Consequently the possession of such gills would be a secondary, and not, strictly speaking, an atavistic feature. Normal loss of these gills, exclusively pulmonary respiration, and preponderating terrestrial life characterise the final adult Amphibian. These cases of neoteny are therefore instances of more or less complete retardation, or of the retention, of partially larval conditions.

The whole problem is, however, by no means simple. Salamandra atra has become viviparous, and the whole metamorphosis takes place within the uterus; in fact, the young have an embryonic, but no larval period, if by the latter we understand the free swimming and still imperfect stage. Similarly, various Anura–for instance, Hylodes martinicensis–pass rapidly through their metamorphosis, and have suppressed the stage of free swimming tadpoles. On the other hand, in many newts, the duration of the larval period is much prolonged, and moreover is very subject to individual variation. In the Axolotl this larval period is continued until and after sexual maturity is reached. The extreme condition would then be represented by the Perennibranchiate genera. It may seem reasonable to look upon these as the youngest members of the Urodela, and the loss of the maxillae in the Sirenidae and Proteidae supports this idea. But it so happens that the majority of the most neotenic genera are more primitive in the composition of the skull and the vertebral column than the typically terrestrial and rapidly metamorphosing genera. Witness the amphicoelous vertebrae, the completeness of the pterygoids, the separate nature of the palatine bones, and the separate splenials, as mentioned in detail in the description of their skull.

We have therefore to conclude, first, that the various Perennibranchiate genera do not form a natural group, but are a heterogeneous assembly; secondly, that they have become Perennibranchiate at a phylogenetically old stage–in fact, that they are the oldest, and not the newest, members of the present Urodela. At the same time, it would be erroneous to suppose that the first Urodela were aquatic creatures, provided with a finny tail, with small, ill-developed lungs, and with epidermal sense organs. All these features are, on the contrary, directly correlated with aquatic life, and are larval acquisitions, not ancestral reminiscences. It would be equally wrong to allude to the absence of lungs in many newts as a piscine and therefore ancestral feature. The development of the typical pentadactyloid limb, the connexion of the pelvic girdle with the vertebral column, the development of the lungs, and absolute suppression of internal gills point without doubt to terrestrial creatures. What then, may we ask, were the first Amphibia like? and how about the external gills? They were undoubtedly akin to the less specialised Lepospondylous Stegocephali, in particular the gill-less Microsauri, and the various stages may perhaps be reconstructed as follows:–

(1) Terrestrial, with two pairs of pentadactyloid limbs; breathing by lungs only; with a fully developed apparatus of five pairs of gill-arches, which during the embryonic life perhaps still carried internal gills; with or without several pairs of gill-clefts. Reduction of the dermal armour and of the cutaneous scutes had taken place.

(2) Additional respiratory organs were developed by the embryo, in the shape of external gills; these were at first restricted to embryonic life (as in the existing Apoda), but were gradually used also during the aquatic life of the larva. These external gills, together with the lungs, have superseded the internal gills, of which there are now no traces either in Urodela or in Anura.

(3a) Some Urodeles, retaking to aquatic life, retained and further enlarged the external gills into more or less permanent organs (cf. also Siren, p. 136).

(3b) The majority of Urodela hurried through the larval, aquatic stage, and some–e.g. Salamandra atra–became absolutely terrestrial. The possession of unusually long external gills by this species and by the Apoda indicate that these organs are essentially embryonic, not larval, features.

Regeneration.–Most Amphibia possess the faculty of regenerating mutilated or lost limbs. This takes place the more certainly and quickly the younger the animal. The amputation necessary to study these phenomena need not be experimental. Axolotls and other Urodelous larvae frequently maim each other fearfully, by biting off the gills or one or more limbs. The gills do not even require amputation. If the larvae are kept in stagnant water the gills often shrivel up or slough off and grow again. The same applies to the larvae of viviparous species, e.g. Salamandra atra, which, when cut out of the uterus and put into water, soon cast off their long, tender gills and produce a stronger set. In an Axolotl,^[38] two years old, a hand was cut off. After four weeks there was a conical stump; after the sixth week this stump had two points; in the eleventh week three or four fingers were discernible, and a week later the complete hand. Frequently these creatures reproduce five instead of the normal four fingers. But the more proximal the cut, the more liable is the new limb to reproduce supernumerary fingers, or even extra hands and feet. Complete regeneration of the limb, cut off in the middle of the humerus, took place within five months.

Triton taeniatus, adult, reproduces cut fingers within five or six weeks, and if the hand be cut above the carpus, new finger-stumps appear in about one month. Götte has observed that an adult Proteus did not completely reproduce its whole leg until after eighteen months; and, according to Spallanzani, more than one year elapses before the limb, bones, and cartilages of Triton regain their normal strength.

The Anura are likewise capable of regenerating their limbs, the more readily the younger the specimens. For instance, in a tadpole of Rana temporaria, in which the fore-limbs were still hidden, the hind-limb, cut at the middle of the thigh, reproduced nineteen days later a knee, followed by a short two-toed stump. Ultimately the whole limb became completed. The tail of tadpoles regenerates very quickly and completely, even if it be cut off shortly before the final metamorphosis, when the tail would in any case be reduced. Metamorphosed Anura have almost entirely lost this faculty, but not absolutely. I myself have kept two specimens of Rana temporaria, which, when already adult, had each lost a hand at the wrist. First there was only the clean-cut stump with a scar, but within a year this changed into a four-cornered stump, and two of the protuberances developed a little further, reaching a length of about 4 mm. These specimens lived for four years without further changes.

Temperature.–Amphibia, like Fishes and Reptiles, are, as a rule, classed as cold-blooded animals, in opposition to the warm-blooded Birds and Mammals. This distinction is one of degree only. The terms poikilothermous and homothermous (ποίκιλος, variable; ὅμος, equable) are based upon a sounder principle, but are likewise liable to exceptions. Those creatures which, like Birds and Mammals, possess a specific temperature of their own under normal conditions, that of hibernation being excepted, are homothermous. Cold-blooded creatures have no specific temperature; they more or less assume that of their surroundings. Frogs and newts, for instance, when living in the water, naturally assume its temperature, which is, of course, many degrees lower in a cold spring than in a shallow pond warmed by the sun on a hot summer's day. The same applies to the changes from day to night. Dark-coloured tortoises basking in the sun are sometimes so hot that they are disagreeable to touch, since they possess but little mechanism for regulating their heat. The same individual cools down during a chilly night by perhaps 40° C. Anura are, however, very susceptible to heat; most of them die when their temperature rises to about 40° C. Under such conditions they die quickly when in the water, but in the air their moist skin counteracts the heat, lowering it by evaporation; otherwise it would be impossible for a tree-frog to sit in the glaring sun in a temperature of 120° F. Toads and others with drier skins seek the shade, hide under stones, or bury themselves in the coolest spots available, and many Amphibia and Reptiles aestivate in a torpid condition during the dry and hot season. Many of them can endure a surprising amount of cold, and during hibernation their temperature may sink to freezing-point. This power of endurance does not apply to all alike; tropical species can stand less than those which live in temperate and cold regions. In spite of many assertions to the contrary, it may safely be stated that none of our European frogs, toads, and newts survive being frozen hard. They may be cooled down to nearly -1° C., and they may be partially frozen into the ice. Circulation of the blood is suspended in such cooled-down frogs; their limbs may become so hard that they break like a piece of wood, but the citadel of life, the heart, must not sink much below freezing-point, and must itself not be frozen, if the animal is to have a chance of recovering. The protoplasm resists a long time, and so long as some of it is left unfrozen the rest will recover. Hibernating frogs are lost if they are reached by prolonged frost during exceptionally severe winters. Every frog will be killed in an artificial pond with a clean concrete bottom, but if there is sufficient mud, with decaying vegetable matter, the creatures survive, simply because they are not absolutely frozen. A severe winter not infrequently kills off all the younger creatures, while the older and more experienced hide themselves more carefully and live to propagate the race.

Geographical Distribution.

There is a very ably written chapter on the geographical distribution of the Amphibia by Boulenger in the Catalogue of Batrachia Gradientia, pp. 104-118. He came to the important conclusion that the geographical distribution of the Amphibia agrees in general with that of the freshwater fishes. Günther's division into a Northern, Equatorial, and Southern zone is modified only in so far as the last two are combined into one, "Tasmania and Patagonia not differing in any point regarding their Frog Fauna from Australia and South America respectively."

Boulenger recognises–