or the Pure Town, a town of India, in the Punjab, stands on a mound 40 feet high, in the midst of a plain, 5 miles W. of the Ravree, and 98 S.S.W. of Lahore. It derives its name from having been long the residence of a famous Mohammedan saint, whose tomb, a plain edifice in a depression below the general level of the town, is much frequented by pilgrims, both Hindu and Mohammedan. Pak Pattan is believed to be the site of the colossal altars erected by Alexander the Great to mark the limits of his conquests.

a town of Hungary, county of Tolna, stands on the right bank of the Danube, 62 miles S. of Buda. The inhabitants are chiefly employed in the culture of the vine, and in sturgeon-fishing. There are two churches and a synagogue, a cattle market, and it has some trade in corn. Pop. (1846) 7,910.

PALÉOGRAPHY (παλαιός, ancient, and γράφειν, I write), is that branch of knowledge which has to do with the interpretation of ancient inscriptions and documents. (See Archæology, Diplomatics, Hieroglyphics, and Egypt. For an account of the catalogues of manuscripts in the British Museum and elsewhere, see Libraries.)

PALÉOLOGUS, the name of an illustrious Byzantine family, first mentioned in history in the eleventh century, from which period it played an important part in the affairs of the empire till its downfall. The family of Palæologus occupied the throne of Constantinople without interruption from 1260 to 1453, when that city was taken by the Turks. (See Constantinopolitan History. A full account of this powerful house will be found in the Familia Byzantina of Ducange, pp. 230-348; and a stemma of the family is given under "Palaiologen," in Ersch and Gruber's Encyclopædie; also in Smith's Dictionary of Greek and Roman Biography and Mythology.) PALÆONTOLOGY

Is the science which treats of the evidences of organic beings in the earth's strata; evidences mainly consisting of petrified or fossil remains of plants and animals belonging to species that are mostly extinct.

The endeavour to interpret such evidences has led to comparisons of the forms and structures of existing plants and animals, which have greatly and rapidly advanced the science of comparative anatomy, especially as applied to the animal kingdom, and herein more especially to the hard and enduring parts of the animal frame, such as corals, shells, crusts, scales, bones, and teeth.

In applying the results of these comparisons to the restoration of extinct species, physiology has benefited by the study of the relations of structure to function requisite to obtain an idea of the food and habits of such species. It has thus been enriched by the well-defined law of "correlation of structures."

Zoology has gained an immense accession of subjects through the determination of the nature and affinities of extinct animals, and its best aims have been proportionally advanced. Much further and truer insight has been gained into the natural arrangement and subdivision of the classes of animals since palæontology expanded our survey of them. Thus a few hard-scaled fishes,—Polypterus, Lepidosteus, &c.—which represent a subordinate group of the herring family (Clupeidae), in the second edition of Cuvier's Règne Animal, have been found to be the remnants of an almost extinct order, equivalent to the whole Malacoptygii of that naturalist; and the Ruminantia, which Cuvier deemed to be a very natural and well-defined order, has since become known to be a peculiarly modified subdivision of a wider and more natural group of hoofed quadrupeds, the Artiodactyla.

The knowledge of the type or fundamental pattern of certain systems of organs, e.g., the framework of the Vertebrata and the teeth of the Mammalia, has been much advanced by the more frequent and closer adherence to such type discovered in extinct animals, and thus the highest aim of the zoologist has been greatly promoted by palæontology.

But no collateral science has profited so much by palæontology as that which teaches the structure and mode of formation of the earth's crust, with the relative position, time, and order of formation of its constituent stratified and unstratified parts. Geology has left her old hand-maiden mineralogy, to rest almost wholly upon the broad shoulders of her young and vigorous offspring, the science of organic remains.

By this science the law of the geographical distribution of animals, as deduced from existing species, is shown to have been in force during periods of time long antecedent to human history, or to any evidence of human existence; and yet, in relation to the whole known period of life-phenomena upon this planet, to have been a comparatively recent result of geological forces determining the present configuration and position of continents. In this relation, palæontology throws light upon a most interesting branch of geographical science, that, viz., which relates to former configurations of the earth's surface, and to other dispositions of land and sea than prevail at the present day.

Finally, palæontology has yielded the most important facts to the highest range of knowledge to which the human intellect aspires. It teaches that the globe allotted to man has revolved in its orbit through a period of time so vast that the mind, in the endeavour to realize it, is strained by an effort like that by which it strives to conceive the space dividing the solar system from the most distant nebula.

Palæontology has shown that, from the inconceivably remote period of the deposition of the Cambrian rocks, the earth has been vivified by the sun's light and heat, has been fertilized by refreshing showers, and washed by tidal waves; that the ocean not only moved in orderly oscillations regulated, as now, by sun and moon, but was rippled and agitated by winds and storms; that the atmosphere, besides these movements, was healthily influenced by clouds and vapours, rising, condensing, and falling in ceaseless circulation. With these conditions of life, palæontology demonstrates that life has been enjoyed during the same countless thousands of years; and that with life, from the beginning, there has been death. The earliest testimony of the living thing, whether coral, crust, or shell, in the oldest fissiliferous rock, is at the same time proof that it died. At no period does it appear that the gift of life has been monopolized by contemporary individuals through a stagnant sameness of untold time, but it has been handed down from generation to generation, and successively enjoyed by the countless thousands that constitute the species. Palæontology further teaches, that not only the individual, but the species perishes; that as death is balanced by generation, so extinction has been concomitant with the creative power which has continued to provide a succession of species; and furthermore, that, as regards the various forms of life which this planet has supported, there has been "an advance and progress in the main." Thus we learn, that the creative force has not deserted the earth during any of the epochs of geological time that have succeeded to the first manifestation of such force; and that, in respect to no one class of animals, has the operation of creative force been limited to one geological epoch; and perhaps the most important and significant result of palæontological research has been the establishment of the axiom of the continuous operation of the ordained becoming of living things.

In entering upon the present survey of the evidences of organic beings in the earth's crust, it is proposed to commence with the lowest or most simple forms, and, as the subject of fossil plants has been ably dealt with under the head PALEONTLOGICAL BOTANY, in vol. v., p. 232, to treat chiefly of the remains of the animal kingdom.

A reference to the subjoined "Table of Strata" (fig. 1) will indicate the relative position of the geological formations cited. The numerals opposite the right hand give the approximate depth or vertical thickness of the strata.

Organisms, or living things, are those which possess such an internal cellular or cellulo-vascular structure as can receive fluid matter from without, alter its nature, and add it to the alternative structure. Such fluid matter is called "nutritive," and the actions which make it so are called "assimilation" and "intus-susception." These actions are called "vital," because, as long as they are continued, the organism is said "to live." When the organism can also move, receive the nutritive matter by a mouth into a stomach, inhale oxygen and exhale carbonic acid, develop tissues the proximate principles of which are quaternary compounds of carbon, hydrogen, oxygen, and nitrogen, it is called an "animal." When the organism is rooted, has no mouth or stomach, exhales oxygen, has tissues composed of "cellulose" or of binary or ternary compounds, it is called a "plant." But the two divisions Protozoa, of organisms called "plants" and "animals," are specialized members of the great natural group of living things; and there are numerous organisms, mostly of minute size and retaining the form of nucleated cells, which manifest the common organic characters, but without the distinctive superadditions of true plants or animals. Such organisms are called "Protozoa," and include the sponges or Amorphozoa, the Foraminifera or Rhizopods, the Polycestinea, the Diatomaceae, Desmidiae, and most of the so-called Polygastria of Ehrenberg, or infusorial animalcules of older authors.

PROTOZOA.

CLASS I.—AMORPHOZOA.

Fossil sponges take an important place among the organic remains of the former world, not only on account of their great variety of form and structure, but still more because of the extraordinary abundance of individuals in certain strata. In England they specially characterize the chalk formation,—extensive beds of silicified sponges occur in the upper greensand, and in some beds of the oolite and carboniferous limestone. In Germany a member of the Oxford colite is called the "spongitenkalk," from its numerous fossils of the present class.

Existing sponges are divided into horny, flinty, and limy, or "ceratose," "silicious," and "calcareous," according to the substance of their hard sustaining parts, which parts are commonly in the shape of fine needles, or spicula, of very varied forms, but in many species of sufficient constancy to characterize such species. The soft organic substance is for the most part structureless and difficult; it is uncontractile and impassive. The larger orifices on the surface of a sponge are termed "oscula," and are those out of which the currents of water flow; these enter by more numerous and minute "pores."

The calcareous sponges abound in the oolitic and cretaceous strata, attaining their maximum of development in the chalk; they are now almost extinct, or are represented by other families with calcareous spicula. The horny sponges appear to be more abundant now than in the ancient seas, but their remains are only recognizable in those instances where they were charged with silicious spicula.

M. D'Orbigny enumerates 36 genera and 427 species of fossil sponges; and this is probably only a small proportion of the actual number in museums, as the difficulty of determining the limits of the species is very great, and many remain undescribed.

Paleospongia and Acanthospongia occur in the lower Silurian; and Stromatopora, with its concentrically laminated masses, attains a large size in the Wenlock limestone. Steganoecyium, Sperispongia, and species of Sephyla, are found... Protozoa in the Devonian; and Bothroconis, Mamillopora, and Tragos, in the Permian or magnesian limestone. Several genera are common to the trias and oolites; and several more are peculiar to the latter strata. The Oxfordian sponges belong chiefly to the genera Eudela, Hippolimus, Cribripsongia, Stellipsongia, and Cupulipsongia. Their fibrous skeleton appears to have been entirely calcareous, and often very solid; their form is cup-shaped, or mammillated, or incrusting; and many have a sieve-like appearance, from the regular distribution of the excurrent orifices (oscula) over their surface.

The greensand of Farigdon in Berkshire is a stratum prolific in sponges, chiefly cup-shaped and calcareous, of the genera Siphonia and Cheniendopora; or mammillated, like Cheniendion and Verticillopora. The Kentish rag is full of sponges, which are most apparent on the water-worn sides of fissures. Some beds are so full of silicious spicula as to irritate the hands of the quarrymen working those beds. The greensand of Blackdown is famous for the number and perfect preservation of its pear-shaped Siphonia (fig. 2, 1); whilst those of Warminster are ornamented with three or more lobes. The latter locality is the richest in England for large cup-shaped and branching sponges (Polypothecia), which are all silicified. The sponges, chiefly Siphonia, of the upper greensand of Farnham are infiltrated with phosphate of lime, and have been used in agriculture.

The sponges of the chalk belong to several distinct families. Choanites resembles the Siphonia, but is sessile, and exhibits in section, or in weathered specimens, a spiral tube winding round the central cavity. It is the commonest sponge in the Brighton brooch-pebbles. Others are irregularly cup-shaped and calcareous; and many of the Wiltshire flints have a nucleus of branching sponge (S. elavellata). The chalk flints, arranged in regular layers, or built up in columns of "Paramoebula," all contain traces of sponge structure, and their origin is in some measure connected with the periodic growth of large crops of sponges. Frequently the crust or outer surface only of the sponge has been silicified, while the centre has decayed, leaving a botryoidal or stalactitic cavity. The cup-shaped sponges are almost always more or less enveloped with flint, which invests the stem and lines the interior, leaving the rim exposed. The sponges of the Yorkshire chalk are of a different character: some are elongated and radiciform, others horizontally expanded, but they contain comparatively little silica; while those belonging to the genus Manon (fig. 2, 4) having prominent "oscula," are superficially silicified, and will bear immersion and cleaning with hydrochloric acid. The largest group of chalk sponges, typified by Ventricutes (fig. 2, 3), have the form of a cup or funnel, slender or expanded, or folded into star-like shape (Guetterdia, fig. 2, 2), with processes from the angles to give them firmer attachment. Some have a tortuous or labyrinthine outline, and others are branched or compound, like Brachiolites. Curious sections of these may be obtained from specimens enveloped with flint or pyrites. The burrowing-sponge, Cliona, is commonly found in shells of the tertiaries and chalk. The great cretaceous Exogona of the United States are frequently mined by them; and flint casts of Belemnites and Inocerasi are often covered by their ramifying cells and fibres. Thin sections of chalk flints, when polished and examined with the microscope, sometimes exhibit minute spherical bodies (Spiniferites) covered with radiating and multicuspid spines. From their close resemblance to the little fresh-water organism Xanthidium, they long bore that name; but they are certainly marine bodies, and probably the spores of sponges.

Class II.—FORAMINIFERA.

The organisms of this class are small, and for the most part of microscopic minuteness,—of a simple gelatinous structure, protected by a shell. They grow by successive gemmation from a primordial segment, sometimes in a straight line, more commonly in a spiral curve; and each segment so developed has its own shelly envelope. As, however, they are organically connected, the whole seems to form a "chambered" or "polythalamous" shell. The last-formed segment is usually distinguished by very long, slender, pellucid, colourless, contractile filaments, like rootlets; whence the name "Rhizopods," sometimes given to the class. But both the outer wall and the septa of the compound shell are perforated by minute apertures, through which either connecting or projecting filaments of the soft organic tissue pass; whence the name Foraminifera. The several segments or jelly-filled chambers are essentially repetitions of each other; and there is no proof that the inner and earlier segments derive their nourishment from the outer and last-formed one. A Foraminifer may therefore be regarded either as a series of individuals, organically united, or as a single aggregate being, compounded according to the law of vegetative repetition.

The minute, chambered shells of Rhizopods enter largely into the composition of all the sedimentary strata, and are so abundant in many common and familiar materials, like the chalk, as to justify the expression of Buffon, that the very dust had been alive. The deep-sea soundings of the Atlantic Telegraph Company have shown that the bed of that great ocean, at a depth approaching, or even exceeding, two miles, is composed of little else than the calcareous shells of a Globigerina and a few other Rhizopods, with the silicious shields of the allied Polycystinea. The composition of the chalk is extremely similar: when the finer portion, amounting to half or even less, has been washed away, the remaining sediment consists almost entirely of foraminated shells, some perfect, others in various stages of disintegration. They have also been found in other marine formations which are soft enough to be washed, down to the Lower Silurian; and in the hard limestones and marbles they can be detected in polished sections, and thin slices laid on glass. The greater part of these shells are microscopic, but some of the large extinct Foraminifera, called, from resembling a piece of money, Nummulites, are two inches in diameter. The generic divisions in use for these shells were mostly invented by M. D'Orbigny, but on artificial grounds,—viz., exclusively upon the plan of growth, or mode of increase in the number of chambers. The structure and anamorphoses of these complex atoms have been recently investigated by Messrs Williamson and Carpenter, and especially by Mr W. K. Parker. Six hundred and fifty-seven fossil species, belonging to seventy-three genera, are described,—commencing in the palaeozoic age, increasing in number and variety with each successive stratum, and attaining their maximum in the present seas. Most of these genera, and even some of the species, pass through many formations; indeed, if correctly observed, they are the oldest known living organisms. *Dentalina communis*, *Orbitolites complanatus*, *Rosalia italica*, and *Rotolina globulosa*, all living species, are said to be found in the chalk; *Rotolina umbilicata* ranges to the gault; and *Webbina rugosa* is common to the upper lias, the chalk, and present sea. It has, however, been observed that fossil Rhizopods, set free by the disintegration of rocks, are mingled with the recent shells on every beach; and Mr M'Andrew has obtained them in this condition from great depths of the mid-channel.

The earliest important form is the *Fusulina* (fig. 2, 5), which forms layers many inches, or even feet, in thickness in the carboniferous limestone of Russia. The recent genera *Dentalina* and *Textularia* are found in the magnesian limestone; *Nodosaria*, *Cristellaria*, and *Rotalia*, in the lias. *Flabellina* (fig. 2, 6) is peculiar to the chalk; *Orbitoides* (fig. 2, 9) to the chalk and tertiary series; *Operculina*, *Orbitolites*, and *Alveolina* appear first in the tertiary, and are still living. The *Lituola* (fig. 2, 7) occurs in the chalk and in chalk flints, and has been described as a species of "Spriolite." Many of the cretaceous Foraminifera contain a brown colouring matter, which remains after the shell has been dissolved with weak acid, and has been regarded as the remains of the animal substance which once filled all the cells.

The "calcaire grossier," which is employed at Paris as a building-stone, contains *Foraminifera* in such abundance that one may say the capital of France is almost constructed of those minute and complex shells.

But it is in the middle eocene, or "nummulitic period," that the Rhizopods attained their greatest size, and played their most important part. Wherever limestones or calcareous sands of this period are met with, these Foraminifera abound, and literally form strata which in the aggregate become mountain masses. These "nummulitic limestones" are found in Southern Europe, in Northern Africa, and in India; they also occur in Jamaica. The commonest form is the *Nummulite* (fig. 2, 8), which occurs in the building-stone of the Great Pyramid. These *Nummulites* were evidently sedentary; and in the large thin species, one side is moulded to the inequalities of the sea-bed on which it grew.

**Polycystinea.**—The tertiary marls of Barbadoes afforded to Ehrenberg an extensive series of novel and extraordinary microscopic organisms, composed of silica, but foraminated like the shells of the Rhizopods. The same forms, and others similar to them, have been met with in the deep-sea mud of the Gulf of the Erebos and Terror, and more recently in the mud of the Atlantic soundings. They are quite distinct in form and character from most of the silicious-shelled *Diatomacea*, but some of them resemble the *Coccolithus* and *Actinoecyclus*. No less than 282 forms, grouped in 44 provisional genera, have been described.

**Class III.—Infusoria.**

*(Polygnatha, Ehrenberg.)*

Numerous genera and multitudes of so-called species of free and locomotive microscopic organisms, which, because they do not present the distinctive characters of plants or animals, have been by turns referred to one or other kingdom, possess shells of flint, and consequently enter Protozoa largely into the domain of fossil evidences of former life. The silicious shells of these infusory organisms present under the microscope the most definite as well as beautiful characters of form and sculpture, which are as recognisable and distinctive as those of the calcareous shells of Mollusca. The plates of the incomparable works and memoirs of Ehrenberg abound with exact figures of the delicate sheaths, shells, and shields of the loricated Infusoria of past and present areas of life, the deposits of which, by reason of their pure, flinty, atomic constitution, were known in the arts long before science had detected their nature and vital origin. In 1836 portions of the stone called "tripoli" or "polierschiefer" (polishing-slate of lapidaries) were microscopically examined by Ehrenberg, who discovered it to be wholly composed of the silicious shells of Infusoria, and chiefly of an extinct species called *Gaiella distans*. At Bilin in Bohemia there is a single stratum of polierschiefer, not less than 14 feet thick, forming the upper layer of a hill, in every cubic inch of which there are forty-one thousand millions of the above-named organic unit. This mineral likewise contains shells of *Nucula*, *Bacillaria*, *Actinoecyclus*, and other silicious organisms. The lower part of the stratum consists of the shells compacted together without any visible cement; in the upper masses the shells are cemented together, and filled by amorphous silicious matter formed out of dissolved shells. At Egea in Bohemia there is a stratum of 2 miles in length, and averaging 28 feet in thickness, of which the uppermost 10 feet are composed wholly of the silicious shells of Infusoria, including the beautiful *Campylodiscus*; the remaining 18 feet consist of the shells mixed with a pulverulent substance. Corresponding deposits of the silicious cases of Infusoria have since been discovered in many other parts of the world, some including fresh-water species, others marine species of Infusoria.

The conditions of such deposits will be readily understood by examining the sedimentary deposits of bogs and stagnant or slow-flowing sheets of water. In warm latitudes and seasons, such water swarms with infusorial life, and the indestructible cases of the loricated kinds are found in great quantities in the sedimentary deposits. Beneath peat bogs they have been found to form strata of many feet in thickness, and co-extensive with the turbaries, forming a silicious mart of pure whiteness. A quantity of pulverulent matter is deposited upon the shores of the lake near Uranien in Sweden, which, from its extreme fineness, resembles flour: this has long been known to the poorer inhabitants under the name of "berg-melk," or mountain-mead, and is used by them, mixed up with flour, as an article of food. It consists in great part of silicious shells of Infusoria, with a little organic matter. With regard to the source of fossil infusorial remains in sea-water, the following evidence is given in the *United States Coast Survey*, 1856:

Soundings of the gulf-stream near Key Siscayne, Florida, varying in depths from 147 fathoms to 205 fathoms, give a light greenish-grey mud composed chiefly of Foraminifiers, Diatoms, Polycystins, and Geolites, in a profusion only surpassed by the fossil polycystinous strata of Barbadoes. The Foraminifers compose the largest part of these muds, including *Textularia Americana*, *Marginalia Bachei*, and other forms, particularly many species of the *Plicatilla* of Ehrenberg, which had been supposed to live only in shallower haunts. The silicious shells of Diatoms abound in the residue, after the calcareous Foraminifers have been dissolved by acid. The inorganic portion of the soundings is chiefly quartz sand, and its proportion is quite small. Such manifestations of life, with its mineral results, have been detected from the earliest sedimentary deposits to the present time; but as regards the Infusoria, they are given on the grandest scale in formations of the tertiary age. The town of Richmond in Virginia, United States, is built on barren silicious strata of marine origin and tertiary age. The strata are 20 feet in thickness, composed chiefly of infusorial flint-shells, including the well-known and beautiful microscopic objects, *Actinoecetus* and *Coccolithus*.

Most of the infusorial formations, as the polishing-slates at Cassel, Planitz, and Bilin, are astounding monuments of the operation of microscopic organisms at former periods of the history of this planet. The minute size, elementary structure, tenacity of life, and marvellous reproductive power of the Infusoria have enabled them to survive as species those destroying causes which have exterminated contemporaneous higher forms of organism. Species of *Bacillaria* still exist which were in being at the period of the deposition of the chalk. Existing species of *Diatomaceae* have been detected as low down as the oolite. The discovery by Ehrenberg of more than twenty species of silicious-shelled Infusoria, fossil, in the chalk and chalk-marls, which are identical in species with some now living in the bed of the Baltic, is an instructive addition to the obscure history of the introduction of species of living things in this planet, and must add greatly to the interest of the infusorial class in the eyes of the geologist and philosopher.

"For these organisms," writes Ehrenberg, "constitute a chain which, though in the individual link it be microscopic, yet in the mass is a mighty one, connecting the life-phenomena of distant ages of the earth, and proving that the dawn of the organic nature co-existent with us reaches further back in the history of the earth than had hitherto been suspected." "The microscopic organisms are very inferior in individual energy to lions and elephants, but in their united influences they are far more important than all these animals." If it be ever permitted to man to penetrate the mystery which enshrouds the origin of organic force in the wide-spread mud-beds of fresh and salt waters, it will be by experiment and observation on the atoms which manifest the simplest conditions of life.

**ANIMALIA.**

**INVERTEBRATA.**

**FOSSIL INVERTEBRATA.**

Remains of invertebrate animals occur in strata of every age, from the partially metamorphic and crystalline rocks of the Cambrian system to the deposits formed by the floods of the last winter, and the tides of yesterday. They are found in every country, from the highest latitude attained by Arctic voyagers to the extremities of the southern continents, and at the greatest elevation hitherto climbed in the Andes or Himalaya. If some classes,—e.g., *Tunicata*, *Acaulepha*,—seem not to be represented in stratified deposits, they are such as, either from the soluble and perishable tissues composing the entire frame, could not be expected to be fossilized under any conceivable circumstances; or from the same cause, are only not so recognisable at one of their metagenetic phases. Evidence of compound Hydrozoa,—i.e., of the polypes which Ellis called "Corallines,"—and especially of the genus *Campanularia*, would show that the acalypheal type and grade of organization had been manifested at the period of the formation of the strata containing such fossil Polypi. With the above seeming exceptions, every class of invertebrate animal is represented by fossil remains.

They consist of corals and shells, of the petrified skeletons of star-fishes and sea-urchins, of the hard coverings of crabs and insects, of the tracks and shelly habitations of worms, and of impressions of surfaces, and casts of cavities of organisms, retained, after these have perished, by the matrix.

The condition in which invertebrate fossils occur depends on the nature of the matrix and other accidental circumstances; for while some are scarcely altered in composition, or even in colour, others are silicified or infiltrated with carbonate of lime. Some may be cleared by the action of acid or exposure to the weather, and some require the chisel of the mason or the mill of the lapidary for the proper exhibition of their structure.

Multitudes of recent species are fossilized in the newer tertiarys whose history can be made out perfectly from living specimens; but the number of these diminishes gradually in each older stratum, while the proportion of extinct forms is ever on the increase. No living species more highly organized than a Rhizopod is found in the secondary rocks. Recent genera extend further back in time; indeed a few may be recognised in strata of palaeozoic age, shedding a light on the probable affinities and conditions of their associates. Many of the smaller groups of genera, called families, disappear in the secondary, and still more in the palaeozoic period, and are to a limited extent replaced by groups which no longer exist. But, as to the larger groups of Protozoa and true invertebrate animals, it may be affirmed that every known fossil belongs to some one or other of the existing classes, and that the organic remains of the most ancient fossiliferous strata do not indicate or suggest that any earlier and different group of beings remains to be discovered, or has been irretrievably lost, in the universal metamorphosis of the oldest rocks.

**Province I.—RADIATA.**

**Sub-Province POLYP.**

A polype is a small soft-bodied aquatic animal which generally presents a soft cylindrical oval or oblong body, with an aperture at one of its extremities, which is surrounded by a crown of radiating filaments or "tentacles." This aperture leads to the digestive cavity, which, in most Polypes, is without intestine or vent. A very large proportion of these animals has organs of support, called "polyparies" or corals, of various forms and substance, but for the most part consisting of carbonate of lime; and, as a general rule, locomotion is lost with the development of the polypary, which usually attaches the polype to some foreign body. The organization of the soft tissues is in general simple; the faculties of the Polypes are very limited; and the vital phenomena, save those of irritability and contractility, are inconspicuous. Nevertheless, the influence of the combined powers of some of the species, in adding to and modifying the crust of the earth, is neither slight nor of limited extent.

**Class I.—HYDROZOA.**

Char.—Polypary, when present, flexible, external; for the most part developing cells for the polypes according to regular patterns.

**Family I.—GRAPTOLITIDAE.**

To this class may probably belong the organic remains called "Graptolites," which are exclusively and characteristically Silurian fossils. A certain knowledge of their affinities would require examination of the soft parts; and the family has long been extinct. Indications of the flexible consistency of the polypary, and M. Barrande's statement of the existence of a cylindrical canal in its axis, which he conjectures to have contained the common connecting tissue of the polypes, have weighed with the writer in placing the Graptolites provisionally in the present class Polyp. The axis of the polypary is sometimes straight (fig. 3, 3), sometimes spiral (fig. 3, 6). The ordinary form, as given by the Graptolites priodon (fig. 3, 3), is serrated on one side only, and is found abundantly in the Cambrian or older Silurian beds of Scotland and Wales; it occurs also in the Ludlow rocks. The double Graptolites (Diplogranus, fig. 3, 5, and Didymograptus, fig. 3, 4) are Cambrian forms. Rastrites (fig. 3, 6) had the polypes only in one side, and they are less crowded: it characterizes Barrande's division E of the lower Silurian beds of Bohemia, and has not yet been found in Britain. The Graptolites occur in argillaceous strata, especially in the mud-stones of Wales and Cumberland, and in the alum-slates of Sweden. These beds remind one of the mud bottoms in which the Virgularia and other long and slender graptolite forms of "Pennatulidae" flourish in forest-like crowds. The primeval Graptolite may have presented a more generalized polype structure than is now met with in the specially differentiated Sertularians and sea-pens.


**Fig. 3.**

Hydrozoa: Anthozoa: Bryozoa.

1. Protovirgularia dichotoma, M'C.; Silurian, Denmark. 2. Oldhamia antiqua, Forsk.; Carboniferous, Wicklow. 3. Gorgonia priscula, Brongniart; Silurian, Britain. 4. Didymograptus Murchisoni, Heck; L. Silurian, Wales. 5. Diplogranus falzani, His; L. Silurian, Britain. 6. Rastrites peregrinus, Barr; Silurian, Bohemia. 7. Comites juniperinae, Eich; U. Silurian, Wales. 8. Pseudograptus antiquus, Lonsdale; L. Silurian, Torrington. 9. Archimedes, Archimedes, Lamarck; Carboniferous, Kentucky. 10. Pillpora plumosa, M'C.; Carboniferous, Ireland. 11. Fenestrella membranacea, Phil.; Carboniferous, Britain.

**Class II.—Anthozoa.**

In this class of Polypes the tentacles are hollow, and, in most, with pectinated margins. The polypary is usually internal, and forms the bodies more properly called "corals" and "madreporites."

*Asteroida.*—Great doubt attaches to some of the fossils referred to this class of Polyphy. The terms "Gorgonia" and "Alcyonium" have been applied to objects not well understood, and usually proving to be Bryozoa and sponges. The lower Silurian fossil called *Pyritonema* consists of a fasciculus of silicious fibres, and has been supposed to be related to the glass zoophyte (*Hyalonema*). The miocene deposits of Piedmont contain a species of the Mediterranean genus *Corallium*, an *Antipathes*, and an *Isis* (or *Isisina*, D'Orb.), which is also found in Malta. The London clay contains one coral (*Graphuloria*), referred to the *Pennatulidae*, and two *Gorgonidae* (*Mopsa* and *Websteria*).

**Anthozoa or Actinoidea.**—The lamelliferous or stony corals are (next to the Testacea) the largest and most important class of invertebrate fossils. They attained a great development in the earliest seas, and were perhaps more widely diffused and individually abundant in the Silurian age than at any subsequent period. "Reef-building" corals are now confined to warm seas, and are wanting even on great tracts of tropical coast. The *Oculina* is the only large coral now found in the north. But in palaeozoic times the representatives of the modern Astraeas and Caryophyllias Radista, extended as far northward as Arctic voyagers have penetrated; and at a much later period they formed reefs of considerable thickness and extent in the area of the coralline oolite. The Silurian limestone of Wenlock Edge is itself a coral reef thirty miles in length; and the Plymouth limestone and carboniferous limestone have frequently the aspect of coral-banks skirting the older regions of Cambrian slate and Devonian "killas." The structure of coral-banks may be studied in the lofty limestone cliffs of Cheddar, and in the wave-worn shores of Lough Erne, as well as in the upheaved coral islands of the southern seas. In the fields about Steeple Ashton, every stone turned up by the plough is a coral; and our inland quarries and chalk-pits afford to the palaeontologist materials for the study of a class almost wholly wanting on the present sea-shores of Europe. The history of the British fossil corals, as given by Milne Edwards and Haime in the *Transactions of the Paleontographical Society*, exhibits, equally with that of the fossil shells by other authors, a transition from a state very different from that which now subsists in our part of the world, and a gradual approximation to the present order of things.

In the palaeozoic strata the corals belong chiefly to two extinct orders; those of the secondary period more resemble living corals of warmer climates than ours; and the few tertiary genera and species resemble those of Southern Europe and our own coast.

The distinction between one large group of the palaeozoic Anthozoa (*Cyathophyllidea*) and more modern corals consists in the quadripartite character of their plaited cups or stars, whereas the lamellae (or septa) of the other families are developed in multiples of 6. A remarkable exception exists in the *Holocyatis* (fig. 5, 8), an Astrea-like coral with quadripartite stars, which is found in the lower greensand. The old-rock corals are also remarkable for the manner in which they are partitioned off by horizontal "tabulae" (fig. 4, 3), like the septa of the *Nautilus* and *Spondylus*. This character obtains not only in the *Cyathophyllidea*, but also in the *Milleporidae*, *Farositidae*, and other palaeozoic families. Of the 129 Silurian corals, 121 belong to the tabulated divisions.


**Fig. 4.**

Palaeozoic Corals (Anthozoa).

1. Amplexus Sewerbyi, Phil.; Carboniferous, Ireland. 2. Cyathophyllum turbinatum, Lin.; L. Silurian, Wenlock. 3. Cyathophyllum suborbiculatum (section); L. Silurian, Wenlock. 4. Cyathophyllum Silurianum, Lonsd.; L. Silurian, Wenlock. 5. Zaphrentis Phillipps, M. Edw.; Carboniferous, England. 6. Lithostrotion striatum, Flem.; Carboniferous, Europe. 7. Aeserraria luxuriana, Eich.; U. Silurian, Europe. 8. Heliolites intertexta, Wahl.; L. Silurian, Europe. 9. Syringopora ramulosa, Goldf.; Carboniferous, Europe. 10. Halysites semilunatus, L. Silurian, North American Regions. 11. Favosites Gothlandicus, Lons.; Silurian, North.

The Devonian system contains about 150 described corals, the carboniferous limestone 76, and the magnesian limestone only 5 or 6. The commonest forms of simple, turbinated corals are *Cyathophyllum* (fig. 4, 2 and 3), which exhibits four slight fossulae in its cup, and is often supported by root-like processes. In Zaphrentis (fig. 4, 5) there is but one deep fossula. Amplexus (fig. 4, 1) is a characteristic carboniferous fossil, nearly cylindrical, and often so straight and regular in its growth as to have been originally described as a chambered shell. The radiating septa are very slight, and the horizontal partitions simple, flat, and almost as regular as the septa of the Orthoceras.

In the Silurian Cystiphyllum (fig. 4, 4) the lamellae are also evanescent; but the tabulae are represented by numerous vesicular plates. The corals of these genera are not always solitary, or merely in groups; some species of Cyathophyllum constantly form compound masses, with cups rendered polygonal by contact, like C. regium of the Bristol limestone.

The allied genus Accervularia (fig. 4, 8) resembles an Astrea, and exhibits in a remarkable manner the multiplication of its corallites by calicular gemmation. The genus Lithostrotion (fig. 4, 7) of the carboniferous limestone is also compact and astriform, but the new corallites are produced by lateral gemmation. Corals with the same structure, but not compact, are known by the name Lithadendron (fig. 4, 6). The "chain-coral" (Halyssites, fig. 4, 11) and Syringopora (fig. 4, 10) resemble at first sight the recent asteroid Tubiporidea; in Halyssites the radiating septa are quite rudimentary; and in Syringopora the tabulae are funnel-shaped, forming a central axis to each tube. The Forositidae (fig. 4, 12) are mostly very regular both as to their polygonal shape and transverse tabulae; the cells of adjacent corallites are connected by pores, either in the sides or angles of the walls; the septa are rudimentary. In the genus Chatetes the tubes are always slender, and much elongated, and their walls imperforate. Michelinia resembles the fruit of the Nelumbium; it has vesicular tabulae and root-like processes to its basal plate. Heliolites (fig. 4, 9), of which many species are found in the Silurian and Devonian limestones, is related to the recent Millepora. The radiating septa are distinct, and the tabulae regular; the interspaces between the stars are filled up with fine and regular tubes. One genus of Fungidae (Paleocyclus) occurs in the upper Silurian.


Secondary and Tertiary Corals (Anthozoa).

1. Turbinaria sublata, Lam.; M. Eocene, Europe. 2. Diploctenium lunatum, Brug.; Chalk, France. 3. Mirabulus coronula, Goldf.; L. Greensand, Europe. 4. Aspidium cristatum, Lam.; Cretaceous (O). Algeria. 5. Cyclolites ellipticus, Lam.; L. Chalk, France. 6. Pachygyra labyrinthica, Mich.; L. Chalk, England. 7. Pachygyra labyrinthica, Mich.; L. Chalk, France. 8. Holocystis elegans, Locard.; L. Greensand, Isle of Wight. 9. Monticulina encyphylitica, Lam.; Great Oolite, France. 10. Styela De la Bechei, M. Edw.; Corallium, Wales. 11. Thecosomilla annulata, Flem.; Corallium, Wales.

The British secondary corals are not very numerous; for although specimens abound in the coral-rag districts, only 14 species are found in that formation. Altogether, 65 species are found in the English oolites, and 22 in the chalk and greensands. These are mostly Astraeidae, vol. xvii.

Palaeontology.

Three common forms in the Radiata, oolites are Monticulita (fig. 5, 9), Styelina (fig. 5, 10), and Thecosomilla (fig. 5, 11). The English cretaceous strata afford the Holocystis (fig. 5, 8), which is the most recent coral with quadripartite septa; Trochogyra and Parasomilla (fig. 5, 6), resembling the recent Cyathina; and the little "Fungia" coronula (fig. 5, 3), described in two genera of distinct orders (Micrabacia and Stephano-phyllia) in the Monograph of the Palaeontographical Society. The lower chalk of France and Germany contains many other corals, especially Cyclolites (fig. 5, 5), Pachygyra (fig. 5, 7), and Diplocentrum (fig. 5, 2). The Aspidiscus (fig. 5, 4) was sent by Dr Shaw from Algeria.

The English eocene strata contain 25 corals, all extinct, and belonging to 15 genera. These include an Astraea (Litharea Websteri), which grows on the water-worn flint pebbles; a Balanophyllia, similar to the existing coral; a Dendrophyllia, which is the oldest member of the genus; an Oculina; and 8 species of the genus Turbinolina (fig. 5, 1). The corals of the English pliocene are mostly Bryozoa; only four true corals have been found in the coralline crag, belonging to the genera Sphenatrochus, Flabellum, Cryptangia, and Balanophyllia, all reputed extinct, although the first is very closely related to the living Sphenatrochus Macandrewi.

The total number of fossil corals enumerated by M. D'Orbigny in the Prodrome de Paleontologie amounts to 1135, grouped under 216 genera. But notwithstanding all the labour which has been bestowed on this branch of palaeontology by Goldiuss, Michelini, Lonsdale, and Milne Edwards, species are continually discovered or brought home from abroad which are altogether new, and cannot be placed in any of the constituted genera.

Class III.—Bryozoa.

Char.—Tentacles of the polype hollow, with ciliated margins; alimentary canal with stomach, intestine, and anus; polypary, when present, external, horny, and calcareous.

The metamorphoses which the Bryozoa undergo are like those of the lower Polypa; the embryo developed from the ovum is an oval, discoid, or subdepressed body, with a general or partial ciliated surface, by which it enjoys a brief locomotive life after its liberation from the parent. The Bryozoa are allied to the compound Ascidia; but not one of the ascidian Molluscoids quits the ovum as a gemmule swimming by means of cilia; and no Bryozoan quits the ovum in the guise of a Cercarian or tadpole, to swim abroad by the alternate inflexions of a caudal appendage. In a progressive and continuous series of teachings, by pen or word of mouth, the place of an osculant or transitional group is governed by convenience, by considerations of how best to teach by comparison and easy gradation. The real merits of the man who would make scientific capital by changing the position of such group, and by imputing error or ignorance to the author from whom he may differ in this respect, are easily weighed and soon understood.

The Bryozoa, whether regarded as the highest organized Polypes, or as the lowest organized Mollusca, or as an intermediate type, are treated of in systematic palaeontology in the position here assigned to them. The practical palaeontologist finds himself compelled to arrange and study the fossil Bryozoa along with the corals, if only on account of the difficulty he in many cases experiences of determining to which class of Polypa his specimens belong. M. D'Orbigny, who has devoted much attention to this class, enumerates 544 fossil species, distributed in 73 genera. This number must be very far below the real one, since the Bryozoa of the chalk, which alone have been carefully examined, amount to 213 species; while only two species are known from the trias, none at all from the lias, and only five from the upper oolites, so rich in corals and sponges. In the Cours Élémentaire, the fossil Bryozoa are stated to amount to 1678.

Of the 19 or 20 paleozoic genera, none extend into the secondary strata; but of the 18 oolitic genera, Entalophora and Defranceia range onwards to the ter- taries; and Alecto, Idmoncea, and Eschara still survive. The oldest known fossil, Oldhamia (fig. 3, 2), is supposed to be a Bryozoon. The most common paleozoic form is Fenestrella (fig. 3, 11), resembling the recent "lace-coral;" there are 35 species, ranging from the lower Silurian to the Permian. One of its modifications resembles a feather (Ptilopora, fig. 3, 10), and is found in the carboniferous limestone. Another, more remarkable, has a spiral axis (Archimediopora, fig. 3, 9), and occurs in the same formation in Kentucky. One of the oldest genera is Ptilodictya (fig. 3, 8), of which 7 species are found in the lower Silurian formations. The slabs of Silurian limestone obtained at Dudley are covered with myriads of small and delicate fossils, including many Bryozoa. Some of these are spread like a film over other fossils, and have been doubtfully referred to the modern genera Discopora and Berenicea; others, with slender branches, and erect or creeping, are called Milleporas, Heteroporans, and Escharinas. The genus Cenites (fig. 3, 7) perhaps belongs here. The magnesian limestone contains several large "lace-corals" of the genera Fenestrella, Synocladia, and Phyllopora; and two branching species of Thamniscus and Acanthocladia. The oolites afford many small in- crusting species related to Diastopora, and branching forms like Terebellaria and Chrysosora. In the chalk, the Escharas are most numerous, and Lunulites and Cul- pularia first appear. Some thin beds of the lower chalk are almost composed of Bryozoa, mingled with Forami- nifera. The coralline crag of Suffolk takes its name from the great abundance of Bryozoa it contains, among which Eschara, Cellepora, Fascicularia, Theonoo, Hornera, Idmoncea, Flustra, and Tubulipora are the most important.

CLASS IV.—ECHINODERMATA.

(Sea-Fishes, Sea-Urchins.)

Char.—Marine; commonly free, repent animals, with the integument in most perforated by erectile tubular ten- tacles, hardened by a reticulate deposit of calcareous salts, and in many armed with spines.

The fossil Radiata present a view of comparatively un- exhausted richness to the palaeontologist. More difficult of study than shells, and less uniformly present in all strata, the enduring remains of Echinodermata and corals are unsurpassed in beauty of form and structure, and in the value of the evidence they afford.

The present summary of the extinct forms of Echino- dermata will commence with

Order I.—Crinoidea.

Char.—Body with ramified rays, supported temporarily or permanently on a jointed calcareous stem; alimentary canal, with mouth and vent, both, as in Bryozoa, approximated.

The "stone-lilies," or crinoid star-fishes, formed a nu- merous and important group in the paleozoic seas, where they obtained their maximum number and variety. M. D'Orbigny describes 31 paleozoic genera, 2 triassic, 10 oolitic, and 4 cretaceous—of which latter 3 (Pentacrinus, Bourguetierinus, and Comatula) are found in the tertiarys and modern seas. The Crinoidea differ from the other Echinoderms in having the generative organs combined with the arms, and opening into special orifices near their base. Nearly all the genera, except Comatula and Mur- susites (fig. 6, 9), appear to have been attached either by the expanded base of the column, as in Apiocrinus, or by jointed processes, as in Bourguetierinus. In many instances the lower part of the column throws out innumerable root-like side-arms, which strengthen and support it. The column is comparatively short in Apiocrinus Parkinsoni, and extremely elongated in Pentacrinus Hiemeri. It is round in nearly all the paleozoic Crinoids; and when five-sided, the articular surfaces of the joints are simply radiated, as in the rest. These joints are perforated in the centre, and when detached, are the "St. Cuthbert's beads" of story (fig. 6, 5). In Platycrinus the stem is compressed, and the articular surfaces are elliptical. In the genus Pentacrinus, which commences in the lias, the sculpturing of the articulations is more complex (fig. 6, 8), but it is quite simple in the other modern genera. The body of the Crinoid is composed of polygonal plates forming a cup, which is covered by a canopy of smaller plates. The mouth is often proboscisiform; the anal orifice is near it. The five arms which crown the cup are sometimes nearly simple, but feathered with slender, jointed fingers; in other genera they divide again and again, dichotomously; and in two remarkable Silurian forms, Anthocrinus and Crotalocrinus (fig. 6, 4), these subdivisions are extremely numerous, and the successive ossicles are articulated to each other laterally, forming web-like expansions, similar in appearance to the coral Fenestrella (fig. 3, 11.) Other remarkable Silu- rian Crinoids belong to the genera Glyptocrinus, Euclyp- tocrinus, Geocrinus (the "Dudley Encrinite"), and Caryo- crinus. Several are common to the Silurian and Devonian, as Melocrinus, Cyathocrinus, and Rhodocrinus; the two last, and Poterocrinus, extend into the carboniferous forma- tions. Cupressocrinus and some others are peculiarly Devonian; Platycrinus, common to Devonian and coal formations; and many genera (including the "nave Encrinite,"—Actinocrinus, Gilbertocrinus, and Woodocrinus), are proper to the carboniferous limestone. The famous "lily Encrinite" (Encrinites entrocha, fig. 6, 9) is charac- teristic of the middle trias, or "muschel-kalk;" the "clove Encrinite" (Eugeniacrinus, fig. 7, 9) abounds in the Ox- Radiata. forlorn oolites of Germany; *Apicerinus*, *Millericrinus*, and several forms related to *Comatula*—e.g., *Pterocoma* and *Saccosoma*—are also peculiarly oolitic. The "tortoise Encrinite" (*Marsupites*, fig. 6, 9) is found only in the chalk, along with *Bourguetcrinhus* (fig. 7, 10); and the bodies of *Comatula*, which, when they have lost their arms and claspers, are called "Glenotremites." (Fig. 7, 7.—upper surface with sockets of the five arms; 8 under surface, showing articulations of claspers, and the scar of the larval stem.)

**Order 2.—Cystoidea.**

This order was established by Von Buch for a small group of palaeozoic Echinoderms formerly included with the *Crinoidea*. They have a globular body covered with close-fitting polygonal plates attached by a simple, jointed stem. The mouth is minute, and opposite to the stalk; close to it is the small anal opening; and a little more distant the generative orifice, covered by a pyramid of five or six little valves. Some of the genera, like *Pseudocrinus* (fig. 6, 2), have two or four tentaculiferous arms, bent down over the body and lodged in grooves, to which they are anchored. Others, like the *Spheronites* (fig. 6, 1), have only obscure indications of tentacles situated close to the mouth. In *Pseudocrinus* and some other genera two or three pairs of lamellated organs, called *pectinated rhombs*, are placed on the contiguous margins of certain body-plates. They are supposed not to penetrate the interior, and no office has been conjecturally assigned to them; but Professor Forbes suggested that they might represent the "epaulettes" of the larval *Echinidae*, to which group he supposed the Cystidean bore the same relation as the Crinoids hold to the star-fishes. There are 9 genera, of which 8 are found in the British strata—4 in the upper and 4 in the lower Silurian.

**Order 3.—Elastoidea.**

A separate order has been proposed for another small group of palaeozoic fossils typified by *Pentremites* (fig. 6, 3). The body is globular or elliptical, and supported on a small, jointed stalk, with radiated articular surfaces and irregular side-arms. It is composed of solid polygonal plates, with a minute oral orifice at the summit surrounded by five other openings, four of which are double and ovarian, the fifth rather larger and anal. There are five petaloid ambulacra of variable length, converging to the mouth, furrowed down the centre, and striated across. According to the observations of Dr Ferdinand Romer, these supported numerous slender, jointed tentacula, indicated by the rows of marginal pores. One species is found in the upper Silurian, 6 in the Devonian, and 24 in the carboniferous, which has received the name of "pentremitite limestone" in the United States, on account of the abundance of these fossils it contains.

**Order 4.—Asteroidea.**

(*Sea-Stars, Brittle Stars.*)

Char.—Body radiate; integument hardened by calcareous pieces, and more or less armed with spines; no dental apparatus.

*Asteriidae* and *Ophiuridae*—Fossil star-fishes, though less common, have a wider range than their allies the fossil urchins, being found amongst the earliest organic forms. *Paleaster*, *Protaster* (fig. 7, 6), and *Lepidaster* (fig. 7, 5), are Silurian star-fishes, presenting many anomalies, and scarcely referable to any existing families. *Tropidaster*, *Pleuraster*, *Aspidura*, *Ophiurella*, and *Amphiura* are oolitic genera; *Ophioderma*, *Ludia*, and *Astropecten* range from the lias to the present seas; *Stellaster* and *Arthaster* are peculiar to the cretaceous; and *Ophiura*, Radiata.

---

**Fig. 7.**

*Galeritidae*; *Asteriidae*; *Crinoidea*.

1. *Pygaster semialatus*, Phil.; *Jaf*. Oolite, Cheltenham. 2. *Anambytes oratus*, Lam.; *U*. Chalk, Europe. 3. *Galerites allochrous*, Lam.; *U*. Chalk, Kent. 4. *Scutella subrotunda*, Miocene, Malta. 5. *Lepidaster* Grayi, Forbes; *U*. Purbeck, Dudley. 6. *Protaster* Milleri, Miller; Llandovery rock, Salop. 7. *Comatula* (Glenotremites), upper surface of body. 8. *Comatula* (lower surface); Chalk, Sussex. 9. *Eugenaerina quinquedactyla*, Schil.; Oxfordian, Wurtemberg. 10. *Bourguetcrinhus ellipticus*, Mill.; Chalk, Kent.

*Astrogonium*, *Oreaster*, and *Gonioides* are both cretaceous and living.

**Order 5.—Echinidea.**

(*Sea-urchins.*)

Char.—Body spheroid or discoid, incased in a crust of inflexibly-jointed calcareous plates, and armed with spines; dental system complex, arranged so as to resemble a "lantern."

The *Echinidea* appear first in the carboniferous limestone, and attain their maximum in the cretaceous strata. In all secondary and more modern *Echinidae*, the shell is composed of five double rows of ambulacral plates, and five inter-ambulacral; but in the *Paleichnus* (fig. 8, 1), of the carboniferous limestone there are six rows of inter-ambulacral plates, and in *Perischoodon* five. Only detached plates of *Archocidarids* (fig. 8, 2) have been seen, and these, by their six-sided form, seem also to have been arranged in more than double series. Normal *Echinidae*; of the existing genus *Cidaris*, abound in the upper trias.

---

**Fig. 8.**

*Echinidae*; *Spatangidae*.

1. *Paleichnus spharicus*, Scouler; Carboniferous, Ireland. 2. *Archocidarids* Urli, Flem.; Carboniferous, Ireland. 3. *Cidaris glaberrima*, Gmelin; Carboniferous, England. 4. *Hemicyclaria incrassilla*, Flem.; Carboniferous, Calais. 5. *Salenia petillana*, Desm.; *U*. Greensand, Wilts. 6. *Diasaster ringens*, Ag.; *Inferior Oolite*, Dorset. 7. *Hemicyclaria Grencovii*, Forbes; *U*. Greensand, Blackdown. 8. *Catopogon carinatus*, Goldf.; *U*. Greensand, Wilts. Some of the secondary species of Cidaris have the ambulacral pores widely separated (= Rhabdocidaris); in others the rows of pores are doubled (= Diplocidaris). The genus Hemicidaris (fig. 8, 4), distinguished by the large spine-bearing tubercles on the lower part of the ambulacra, ranges from the trias to the chalk-marl. Diadema, with smooth, solid spines (= Hemidiadema), appear in the lias, and continue to the chalk, where the modern type, with annulated, hollow spines, appears. Echinopis also occurs in the lias; and Acroechinum, a genus characteristic of the oolites, and distinguished from Selenia by its perforated tubercles. Acrocidaris and Heliocidaris, with Glyptus, and several other sub-genera of Echinus, are also peculiar to the oolites. Selenia (fig. 8, 5), with its ornamental disk, is characteristically cretaceous. Arbacia and Temnocephalus appear first in the eocene. The Cassidulidae commence in the oolites, with Pygaaster (fig. 7, 1) and Holocyclus, and abound in the cretaceous system. Galerites (fig. 7, 3), Discoides, Pyrura, and Cassidulus are peculiar to the chalk. The Clypeastridae are represented in the oolites by numerous species of Echinolampas and Nucleolites (or Clypeus); the latter genus attains a large size. The sub-genus Cato- pycus (fig. 8, 8) is peculiar to the cretaceous series. Cono- clypeus occurs in the chalk and tertiaries. Clypeaster flourished most in the miocene age; many large species are found in the south of Europe, Madeira, and the West Indies. Numerous genera, remarkable for their flattened form, and popularly known as "cake-urchins," are peculiar to the tertiaries and existing seas. Leuota and Scutellina are eocene; Seastella (fig. 7, 4) is miocene. Mellita and Echinarchinus are both fossil and recent. The heart-shaped urchins (Spatangidae), are only remotely represented in the oolites by Disaster (fig. 8, 6); they are numerous in the chalk, to which Micraster, Hemipemnites (fig. 8, 7), Archiacia, Holaster, and Ananchytes (fig. 7, 2), are peculiar. Hemaster is cretaceous and tertiary. Spatangus, Equatorius, Brisum, Amphidatus, and Schizaster are tertiary and recent forms.

The shell of the Echinodermata has the same intimate structure in all the orders and families, and in every part of the skeleton, whether "test," or "spine," or "tooth." The smallest plates resemble bits of perforated card-board, and the largest and most solid are formed of a repetition of similar laminae. In a few membranous structures, minute spicula, curved, bi-hamate, or anchor-shaped, are met with. They are always composed of carbonate of lime; but owing to their porosity, fossil examples are commonly impregnated with earth, or pyrites, or silica, and form bad subjects for microscopic investigation. Without, however, losing their organic structure, the fossil Echinoderms exhibit a cleavage like that of calcareous spar, by which the smallest ossicle of star-fish or Crinoid may be recognised: this peculiarity is most strikingly obvious in the great spines of the Cidaris (fig. 8, 3), or the enlarged column of the "pear Encrinite" (fig. 6, 7). Examples of the latter may be seen which had been crushed when recent, and before the sparry structure was superinduced.

Order 6.—Holothuroidea.

(Sea-Cucumbers, Trepang.)

Char.—Body vermiform; integument flexible, with scattered reticulate calcareous corpuscles, or beset with small anchor-shaped spicula.

The Holothurioide order presents scarcely any examples likely to be met with in a fossil state, except the genus Pseudos, of whose imbricated shield a fragment has been found by Mr Richmond in the northern drift of Bute. Count Münster has figured the microscopic plates, apparently of a Holothuria, from the chalk of Warnminster; and the anchor of a Synapta from a still older formation,—the Articulata upper oolite of Bavaria. Microscopic observers will doubtless meet with many such detached plates and spines when searching for Polycystinea and other Rhizopods in the oolitic and cretaceous strata; but it is scarcely probable that the order has dated far back in time.

Province II.—Articulata.

In the great division of invertebrate animals called Articulata the brain is in the form of a ring encircling the gullet. A double ganglion above the tube supplies the chief organs of sense. The ganglia below the tube are connected with two chords which extend along the ventral surface of the abdomen, and are in most species united at certain distances by double ganglia, which are connected with the nerves supplying the body segments and their appendages. The body presents a corresponding symmetrical form. The skeleton is external, and consists of articulated segments of a more or less annular form. The articulated limbs, in the species possessing them, have a like condition of the hard parts, in the form of a sheath which incloses the muscles. The jaws, when present, are lateral, and move from side to side.

The worm, the lobster, the scorpion, and the beetle exemplify this province.

The articulate division of the animal kingdom, most universally distributed and numerically abundant at the present day, is least perfectly represented amongst the relics of the former world. Their chitinous integuments, often hardened with earthy salts, are quite as capable of preservation as the shells of the Mollusca, and remains of them are met with in all aqueous deposits; but that manifold, complex organization, which in the recent state fits them so admirably for generic and specific comparisons, is fatal to their entire preservation, and the fossil examples are often so fragmentary as to admit of little more than the determination of their class and family.

The most ancient fossiliferous rocks bear imprints which have been regarded as the tracks and burrows of marine worms. With these are found Crustacea of the lowest division, and of a group which is wholly extinct. A little later appear the Phyllopods, Copepods, and other existing orders of Entomostraca. Only a few obscure forms, doubtfully referred to the higher division Malacostraca, have been found in the carboniferous and Permian systems. The secondary strata contain abundant remains of Isopods, and of lobsters and hermit-crabs. True crabs (Brachyura) abound in the oldest tertiaries. Air-breathing insects and Arachnida existed even in the palaeozoic age; the "sombre shades" of the carboniferous forests were not "uncheered by the hum of insects;" nor were the insects blind, like those which now inhabit the vast caverns of Kentucky and Carniola. The Articulata which come latest are the Cirripedes, whose lowest family appears in the lias; while the Balanidae are only found in the tertiaries.

The number of fossil Articulata catalogued and described forms but a very small proportion of those which have probably existed. Bronn enumerates 1551 fossil insects: 131 Arachnida, 894 Crustacea, and 292 Anellida. Darwin describes 69 fossil Cirripedes, 12 of which are living species.

Class I.—Anellida.

(Worms, Tube-Worms, Neridae.)

Char.—Body soft, symmetrical, vermiform, annulated, with suckers, or setae, or setigerous tube-feet; blood of a red colour in most.

1 Beitrage, heft 6, 1843. The peculiar markings on the surface of the old Cambrian slate rocks, conjectured to afford the earliest indications of the existence of marine worms, are not without suspicion as to their origin. The so-called "Nereites" bear considerable resemblance to other equally ancient impressions which have been described as Zoophytes, under the name of *Protoxirgularia* (fig. 3, 1). No such doubt attaches to the worm-tracks which abound in the thin-bedded sandy strata of the forest-marble; and the "Cololites" of the lithographic limestone are most probably the castings of worms. Long calcareous tubes occur in the upper Silurian and carboniferous strata, which have received the name of *Serpulites*. The *Microcosmus* of the carboniferous period is now regarded as an Annelide; and in all the later formations, tubicular Annelides, especially of the genera *Serpula*, *Spirorbis*, and *Vermilia* abound. Some of these, although attached and gregarious, are so regular in their growth as to have been usually called *Vermuti*, but are now placed in the genus *Vermiculatia*. *Spirogyrus*, and some other shell-excavators, are indicated in the terataries. Amongst the problematic fossils of the palaeozoic strata, two are supposed to be *onellidoides*—viz., the *Tentaculites* (fig. 10, 7), which was apparently free, and almost always regular in its growth, so as more to resemble one of the gregarious Pteropods; and the *Cormulite* (fig. 10, 8), which is attached when young, singly or in groups, to Silurian shells and corals; the structure of its shell is vesicular, and the cavity resembles a series of inverted cones. The unattached and gregarious *Ditrupa* appears in the upper chalk, and abounds in the London clay and crag.

**Class II.—CIRRIPEDEA.**

*(Barnacles, Acorn-Shells.)*

Char.—Body chitinous or chitino-testaceous, subarticulated, mostly symmetrical, with aborted antennae and eyes; thorax attached to the sternal surface of the carapace, with six pairs of multijarticulate, biramous, setigerous limbs; metamorphosis resulting in a permanent parasitic attachment of the fully-developed female to some foreign body.

The fossil Cirripedes belong chiefly to the sessile division, and consist of the ordinary forms of the still-existing *Balanidae*. They are rare in the eocene tertiary, but more abundant afterwards. The *Balanus porcatus* attains a great size in the shelly beds of northern drift; its large basal plate, when detached, is a puzzling fossil, and has caused some mistakes. A *Corona* has been found in the middle division of the crag which has afforded so many cetaceous bones. Remains of pendunculated Cirripedes occur in older deposits, but are mostly scarce and fragmentary. A species of *Pollicipes* is found adhering to drift-wood, perforated by bivalves, in the lias; another occurs in the Oxford clay, attached in groups to drift-wood, and the shells of Ammonites, which probably floated in the sea after death. The chalk affords many species of *Pollicipes* and *Scalpellum*, a species of the anomalous genus *Verruca*, and the only extinct genus of Cirripedes—*Loricula* (fig. 10, 6). This remarkable fossil is found attached to Ammonites, and exhibits only one side in any of the examples hitherto found. In this unsymmetrical development and the imbrication of its valves it more resembles *Ferrucca* than any other Cirriped. "During the deposition of the great cretaceous system, the *Lepidulae* arrived at their culminating point: there were then three genera, and at least thirty-two species," whereas at the present day the Philippine Archipelago, which is the richest marine province, affords but five species.

whilst the marine *Cytheridea* assist with their multitudinous atoms in building up the chalk. Amongst the Phyllodops, the gregarious *Eustheria* covers the slabs of Wealden and of Keuper with crowds of bivalve shells which have been commonly mistaken for *Cyclades* and *Posidoniomyer*. The globose *Entomocochlis* (fig. 9, 2) is found in the carboniferous limestone; *Lepidella* (fig. 9, 1) in the Silurian rocks of the north; and *Beyrichia* (fig. 9, 3), which is characteristically Silurian, may be distinguished from the young forms of Trilobites by the unsymmetrical shape of its separated valves. Other palaeozoic Phyllodops (*Ceratiocaris* and *Hymenocaris*) related to the recent *Nebatia*, and having a conspicuous tail, occur in the upper and lower Silurian strata; the genus *Leptocoles* (M'C.) was founded on the tall-spines of these Crustacea. *Dithyrocaris* (fig. 9, 4), which resembles the recent *Apus* in the horizontal compression of its carapace, is found in the carboniferous limestone. The lower coal measures also contain in their nodules of clay-ironstone frequent examples of *Bellinurus* (fig. 9, 6), a small Pocillopod, differing from the recent king-crab (*Limulus*) in the moveable condition of the body-segments. But the most extraordinary of the palaeozoic Crustacea are the *Eurypterus*, *Himantopterus*, and *Pterygotus* (fig. 9, 5), from the upper Silurian and old red sandstone, of which some far surpassed the largest living lobster or king-crab in size. They have been considered an extinct family, related to the *Limuli*; or as the repre-

---

1 This figure (by Mr Salter), as well as several others, are taken from the *Siluria* of Sir R. Murchison, P.G.S. Articulata, representatives of the larval condition of the stalk-eyed *Malacostraca*. But the following structures show an affinity to the *Ostracoda*. Their carapace is comparatively small, with compound eyes on the antero-lateral margins; the body segments are eleven or twelve in number, without appendages, and terminated by a pointed or bilobed tail. *Eurypteris* has eight feet; the others have three pairs of limbs,—viz., the chelate antennae, the foot-jaws, and the natatory feet, with their fin-like palettes, which spring from the under side of their cephalo-thorax. The surface of the body and limbs often presents a peculiar imbricated sculpture, which caused them at one time to be regarded as fishes by Agassiz. The *Pterygotus problematicus* is supposed to have attained a length of seven feet, and some of the others were a yard long. Crustacea of this magnitude may have formed tracks on the sea-bed, like those on the Potsdam sandstone of America, called "Protichnites," subsequently to be described.

The great family of Trilobites is entirely confined to the palaeozoic age; none are found even in the upper coal measures or Permian system. Above 400 species have been described, and grouped in 50 genera. Of these, 46 are Silurian, 22 Devonian, and 4 carboniferous. According to Broom, 13 genera are peculiarly lower Silurian, 3 upper Silurian, 1 Devonian, and 3 carboniferous.

**Order Trilobites.**

Char.—Trunk segments trilobed; sessile compound eyes in most; limbs aborted.

The skeleton of the Trilobite consists of the cephalic shield, a variable number of trunk-rings or segments, and the pygidium or tail composed of a number of joints more or less archylosed. In some species a labrum (or "hypostome") has been discovered, but no indications of antennae or limbs have ever been detected; still there can be no doubt they enjoyed such locomotive power as even the limpet and chiton exhibit when requisite. Variations in the length of the cephalic and caudal spines (*e.g.*, in *Asaphus cundatus* and *longi-cundatus*), and in the prominence of the head-lobes, have been considered indications of difference of sex. One of the oldest and simplest forms is the minute *Agnostus* (fig. 9, 11); it is usually found in little shoals, with only the cephalic shield preserved, as if it were the larval form of some large Trilobite. According to the observations of M. Barrande, the *Sao* passes through twenty stages of growth, being first a simple disc, and ultimately having seventeen free thoracic segments and two caudal joints; the additional segments are developed between the thorax and abdomen. The *Trinucleus* (fig. 9, 10) with its ornamental border, and *Illenus* (fig. 9, 7), in which the trilobation is less conspicuous than in most genera, are characteristic of the lower Silurian strata. Two others from the Wenlock limestone have long been celebrated,—viz., *Calymene* (fig. 9, 9), or the "Dudley Trilobite," so compactly rolled up; and *Asaphus* (or *Placops*) *cundatus* (fig. 9, 8), in which the lenses of the large eyes are frequently well preserved, and visible without a glass. Each eye has at least 400 facets, and in the great *Asaphus tyrannus* each is computed to have 6000. In one species (*Asaphus Koralewskii*) the eyes are supported on peduncles. The largest Trilobite is *Asaphus gigas*; some of the fragments indicate a creature eighteen inches long.

**Sub-Class 2.—MALACOSTRACA.**

Char.—Body divided into thorax and abdomen, with seven segments in each.

The Isopods are represented in the upper oolite by *Archaeoniscus Brodiei*, which is gregarious, in large numbers in the slabs of Purbeck limestone; and in the Permian system by the *Prosoponiscus* (or *Palaeocrangon*). The problematic *Pygocephalus*, and the "*Apus dubius*" Articulata, both from the carboniferous strata, are doubtfully referred to the *Stomopoda*, and, with the exception of the *Gito-crangon* of Richter, are the oldest of the known stalk-eyed Decapods.

Macrourous Crustacea are of constant occurrence throughout the oolites and cretaceous strata. One of the most remarkable forms, *Eryon* (fig. 10, 3), is found in the lias (with the closely-allied *Tropifer* and *Coleia*) and in the Oxford clay. The small lobsters of the genus *Glyphhea*, in the oolites, and *Megeria*, in the Speeton clay and greensand, are commonly the nucleus of hard nodules of phosphate of lime. The larger species of the chalk form the genus *Enoplocydia*. The Oxfordian oolite of Solenhofen, with its finely-laminated lithographic slates, opens like a book filled with compressed and wonderfully-preserved shrimps and lobsters. One of them, remarkable for its long and slender arms (*Megachirus*, fig. 10, 4) is also found in the Oxford clay of Wiltshire. One of the most remarkable repositories of fossil Crustacea is the Isle of Sheppy, where the "London clay" has afforded countless examples of the higher organized division, including 9 Brachiura, 3 Anomura, and 5 macrourous species. The island of Hainan, on the coast of China, abounds with fossil crabs of the genus *Macrothelana*, which are sold in the drug-market of Shanghai. Others are found in the miocene of Malta, and of Perim Island in the Red Sea. The reputed instances of secondary *Brachyura* are open to doubt; in England we have only the little *Eugyn Martini* (or *Reussia*) from the gault, for the *Podopilumus* (M'C.) is probably from some foreign tertiary deposit. Pairs of chelate claws occur in the upper chalk which are referred to the hermit-crab (*Mesostylus Fayasi*). Small Crustaceans, resembling in form the living *Corystes*, abound in the gault (fig. 10, 2), but they are known to be anomurous by the small size and dorsal position of the posterior legs, and by the little plates intercalated between the last joints of the tail, as seen also in the *Dromilites* (fig. 10, 1) from the London clay.

**Class IV.—INSECTA.**

Char.—Body chitinous, articulated, with articulated and uncinate limbs; head provided with jointed antennae; respiratory system tracheal.

The fossil insects hitherto examined have afforded no new types or forms of unusual interest. The oldest known, those from the lower coal measures, resemble the *Curculionidae* and *Blattidae* or *Locustidae* of the pre- The lias limestones have afforded a greater variety to the persevering skill of Mr Buckland: species of the genera Berosus, Elater, Gyrinus, Laccophilus, and Melolontha, and undetermined genera of the families Carabidae, Buprestidae, Chrysomelidae, and Telephoridae; Panorpa-like insects of the genus Orthophlebia; dragonflies, Nepidae and Cicadidae, Cicada, and the dipterous genus Asilus. Next in age is the insect depositary of the Stonesfield slate, which affords the large wing-covers of Buprestis Bucklandi, species of Primus and Coccinella, and the great neuropteran Hemerobioides. The Purbeck limestone has supplied, in addition, species of Cerylon and Calymbetes, Cypion, Heloporus, and Limnus; and examples of Staphylinidae, Cantharidae, Harpidae, Hydrophilidae, and Tenebrionidae, Libellula and Phryganus, Achaeta and Blatta, Aphid, Cercopis, and other Homoptera, and ten dipterous genera. In the newer pleiocene freshwater formations the recent Copris lunaris has been detected, and the elytra of Donacia and Harpalus. The principal foreign sources of fossil insects have been the lithographic slates of Solenhofen, and the tertiary deposits of Aix in Provence, and Eningen, near Constance, on the Rhine. Remains of species of Tinea and Subinae are said to have been found in the lower Jura, and of a diurnal Lepidopteran in the Molasse. Numerous examples of insects in true amber have been obtained, and much more abundantly in "gum animi," a more modern fossil resin. These are all unknown to entomologists, and are probably extinct, since no department of recent natural history has been so closely worked, although the fossil insects have been comparatively neglected. It has been suggested by Mr Westwood that the lias insects have a sub-alpine character, and may have been brought down by torrents from some higher region. But no attempt has been made to show whether these or any other group of fossil insects most nearly resemble those of any particular zoological province of the present day.

Much has been said of the "inclusial limestone" of Auvergne, supposed to be built up of the fossilized cases of caddis-worms (Phryganoidae); but the only entomologist who has visited the country and examined the formation entertains doubts of the correctness of this interpretation.

Of the Myriapoda, 17 fossil species have been found, commencing in the oolitic system. And of the Arachnida, 131 species are catalogued; the earliest and most interesting of these is the fossil scorpion (Cycloptilum senior) of the Bohemian coal measures (figured in Buckland's Bridgewater Treatise). Fossil spiders are found in the Solenhofen slates and in the tertiary marls of Aix.

Province III.—MOLLUSCA.

Remains of the Testacea, or shell-bearing molluscous animals, are the most common of all fossils, and afford the most complete series of "medals," or characteristic signs for the identification of strata. The duration of types and species, as a general rule, is inversely proportional to rank and intelligence. The most highly organized fossils have the smallest range, and mark with greatest exactitude the age of the deposit from whence they have been derived. But the evidence afforded by shells, if less precise, is more easily and constantly obtained, and holds good over larger tracts of country.

Class I.—BRACHIOPODA.

The lamp-shells (Brachiopoda), more than any other group, have suffered with the lapse of time. Of 1300 known species, only 75 are living; and of the 34 genera, the larger part (21) are extinct. The number of generic forms is greatest in the Devonian period and least in the upper oolites, after which a second set of new types gradually appears. The preponderance of fossil Brachiopoda is contrasted with the scarcity of the recent shells even more strongly by the abundance of individuals than by the number of species; for the living shells mostly inhabit deep water and rocky situations inaccessible to the dredger, and are seldom obtained in large numbers.

The genus Terebratula, as now restricted to shells with a short internal loop, musters above 100 fossil species, of which only one survives (T. vitrea), an inhabitant of the Lusitanian province.

The Waldheimias, or Terebratula, with long loops, are widely distributed in our present seas, although only 9 in number, being found on the coast of Spitzbergen and Labrador, at Cape Horn, and most abundantly in New South Wales and New Zealand: there are 60 fossil species dating from the trias. The Terebratella commenced in the lias, and occur in small numbers throughout the cretaceous and tertiary periods, and are the only lamp-shells which attain their climax in recent seas. Five species of Argiope occur in the greensand, chalk, and tertiaries. The allied genus Thecidium is represented by one species in the carboniferous and one in the triassic system, becomes comparatively common in the secondary period, and dwindles again to a single species in the newer tertiary; this species survives within still narrower limits in the Mediterranean sea. The sub-genus Terebratulina is represented by 20 species in the secondary and tertiary formations. T. striata of the chalk is so like the recent T. capit serpentinus as to be with difficulty distinguished from it. Several extinct sub-genera occur in the cretaceous strata, of which the most remarkable are Trigonosomus (fig. 11, t) and Lyra, shaped like a violin. The genus Stri-

Fig. 11.

Brachiopoda.

1. Trigonosomus Pallasi, Wood. 2. U. Cretonensis, Clapy. 2. Stringocephalus Bartoli, Dehr., Devonian, Effel. 3. Spirifer striatus Garwood, Brittain. 4. Cyrtia trapezoidalis D. Miller, Devonian, Drury. 5. Athysa Roewy, Devonian, Carboniferous Ireland. 6. Uncinata gryphoides Schlo., Devonian, Belgium. 7. Athysa reticulata, L. U. Silurian, Malvern. 8. Pentamerus levis Carvood & Salop.

gocephalus (fig. 11, 3) is peculiar to the Devonian strata, and has a large internal loop, and a very prominent cardinal process, forked at the end, and fitting over the central plate of the opposite valve.

The shell of Terebratula and some of its allies (Argiope, Thecidium, Cyrtia, and Spiriferina) is dotted with minute quincuncial perforations, sometimes visible to the naked eye, as in T. linna, but usually requiring a lens of low power.

---

1 For the characters of this province and of its classes, see article MOLLUSCA. 2 They are smallest in T. carnea, which Dr Carpenter has figured by mistake for T. linna in the Trans. of the Palaeontographical Soc. The lamp-shells with sharp beaks and plaited valves have been separated from the *Terebratula* under the name *Rhynchonella* (Fisch.). Their shells do not exhibit the punctate structure under a magnifying-glass, and they have no internal skeleton to support their arms, which in the recent species are coiled up spirally, and directed towards the concavity of the smaller valve, like the spires of the extinct *Atypa* (fig. 11, 7). Of the 3 living species of *Rhynchonella*, one is found throughout the Arctic seas, a second in New Zealand, and the third at the Feejeees (?). The fossil species exceed 250, and are found in all parts of the world; those from the palaeozoic strata may prove distinct from the rest, since the permian species are known to be provided with large internal processes (*Camarophoria*, King). Casts of these shells are frequently impressed with the narrow and angular pallio-vascular impressions. The extinct genus *Atypa* differs from *Rhynchonella* solely in having calcareous spires, which are preserved in many instances, and may be cleared to some extent by the application of acid. The foramen is separated from the hinge-line by a deltidium; and the interior of the valve is marked by ovarian and vascular spaces exactly as in *Rhynchonella*. The lower Silurian rock contains another genus, *Porombonites* (Pander), as yet imperfectly understood, but having the valves marked externally by impressed dots, which are not perforations. The genus *Pentamerus* occurs in all the strata below the carboniferous limestone, and is remarkable for its great internal partitions, causing the shell to split readily across the middle; and giving rise to deep incisions in those casts of the interior which are so common in the Caradoc sandstone (fig. 11, 8).

The extinct family *Spiriferidae* are characterized by the possession of internal calcareous spires extending from the centre of the shell outwards (fig. 11, 3). These spires, like the shell itself, are frequently silicified, and may be disengaged from the matrix by the action of acid. At other times the shell is imbedded in soft marl, removable by careful washing, so as to show the calcareous lamina of the spire fringed with hair-like processes, formerly the support of cirri. In the genus *Spirifer*, the shell has a long straight hinge-line, and the flattened area of the larger valve has a deltoid byssal notch. The typical species are characteristic of the palaeozoic strata, and have a shell-structure like *Rhynchonella*. The liassic species (*Spiriferina*, D'Orb.), have punctate shells, and the byssal opening is closed (at least in the adult) by a thin arched plate or "pseudo-deltidium." In the sub-genus *Cyrtia* (fig. 11, 4), the hinge-area is ultimately as long as it is wide, and the deltidium is perforated in the centre by a byssal tube; some of the species have a punctate shell. The genus *Athyrus* (Dalman), not always easily distinguished from *Terebratula*, has usually a smooth and rounded shell, ornamented with concentric lamellae or wing-like expansions (fig. 11, 5); the beak is truncated by a round foramen; the hinge-area is obsolete; and the spires are as in *Spirifer*, with the addition of some further complications near the hinge. There are 25 species, mostly from the Devonian and carboniferous rocks. The species of *Retzia* (King) are still more like plaited *Terebratula*, but have lateral spires; they range from the Silurian strata to the trias. *Uncites graphus* (fig. 11, 6), a peculiar Devonian fossil, has a prominent beak, perforated in the young shell by a minute apical foramen; the hinge-area is filled up by a deeply concave deltidium, on each side of which (but only in some specimens) there is a lateral pouch formed by an inflection of the margin of both valves.

The family *Orthidae* consists of shells with a straight hinge-line, bordered by a flat, narrow area, with a central notch in each valve; the ventral valve is furnished with articulating hinge-teeth, and the dorsal valve has short processes for the support of the oral arms, which appear to have been horizontally spiral (as in *Atypa*). Between the oral processes there is a central projection for the attachment of the cardinal muscles. Internal moulds of the *Orthis* (fig. 12, t) exhibit on the ventral side the single attachment of the adductor muscles in the centre, and on each side of it the cardinal muscles; these are surrounded by the punctate ovarian spaces and impressions of the large pallial sinuses. The genus *Orthis* includes 100 species, ranging upwards to the Permian, but it is most abundant in the Silurian rocks. Some of the lower Silurian species have a round foramen as the "pseudo-deltidium," and are called *Orthisina* (D'Orb.). Other species in the upper palaeozoic rocks have the beak twisted or deformed, probably owing to the attachment of the shell when young (= *Streptorhynchus*, King). In *Strophomena*, Rafin (= *Leptena*, Dalman), there is a minute byssal foramen when young, of which no trace exists in the adult; and the deltoid notch is also closed, except the space required to receive the divided cardinal process of the dorsal valve. The oral processes appear to be shifted to the centre of the valve. The shell, when young, is plano-convex, but when it has attained a certain size the valves are bent over to one side or the other, and more or less suddenly. The pallial impressions are the same as in *Orthis*.

The genus *Davidsonia*, peculiar to the Devonian limestones, resembles an *Orthis* attached, like *Thecidium*, by the ventral valve to corals, and sometimes taking the markings of the body on which it grows, like the oyster and *Anomia*. The pallial impressions are like those of *Orthis*, and the form of the spiral arms is indicated by prominences which almost fill up the interior of the shell in aged examples. Some indications have been obtained of slender calcareous spires for the support of the arms in this genus; and also in *Koninckia*, a small shell from the trias of St Cassian, in which there are always spiral grooves in the interior of the valves crossed by the impressions of the pallial sinuses.

The anomalous fossil called *Calcicola sandalina* (Lam.) is also peculiar to the Devonian limestones. In shape it resembles *Cyrtia*, but has no hinge, and neither foramen nor internal processes, except a row of small projections.

---

1 The term *deltidium*, applied by Von Buch to this foramen, has, by misconception of his meaning, become constantly used for the plates which partially close it. along the hinge-line, and two small lateral groups of ridges in the smaller valve. The interior is punctato-striate, but has no recognisable muscular markings.

The Productidae are altogether palaeozoic fossils, and most abundant in the carboniferous limestones. Their valves are concavo-convex, the hinge-line straight, and the interior marked with distinct impressions of the muscles for opening and closing the valves, and simple vascular spaces. There are 60 species of Producta found in the upper palaeozoic rocks, and having a very wide range in North and South America, and from Spitzbergen to Tibet and Tasmania. Some of them are extremely variable in form; many are armed with long tubular spines, and others completely clothed with short, hair-like processes; they have no hinge-teeth, and the hinge-area is extremely narrow, except in the sub-genus Autosteges of the Russian zechstein. Producta proboscidea has its convex valve prolonged into a tube, as if for the constant supply of respiratory currents. The Permian genus Strophalosoma has its valves articulated by hinge-teeth, and covered with long and slender hollow spines; the shell is attached when young by the umbo of the large valve. Chonectes is distinguished from Producta by a row of spines along the hinge-margin of the convex valve; it also has a narrow hinge-area with a covered notch, and small hinge-teeth. There are 25 species in the Silurian and carboniferous, usually of small size, and finely striated.

Crania is one of the oldest living types, ranging upwards from the lower Silurian. One of the earliest species appears to have been unattached, and another to have had hinge-teeth. Crania ignabergensis, of the chalk of Sweden, has the valves externally alike, being attached only when very young. The internal markings of C. antiqua, and other fossil species, are remarkably grotesque. Lower valves of this genus and Thecidium are not uncommon, attached to the tests of sea-urchins, in the chalk; but upper valves are scarce, either detached or in situ.

The Discinidae are also ancient fossils, few in number, but appearing in every period. Some of the palaeozoic Discina (= Orbiculoides, D'Orb.) cannot be generically distinguished from the recent species by any characters with which we are as yet acquainted; but others (= Trematis, Sharpe) are ornamented with quincuncial punctures, and the casts exhibit indications of diverging internal plates, which imply very considerable difference in the organization of the animal. The genus Siphonotreta (Verneuil), peculiar to the Silurian formations, is covered with moniliform tubular spines.

Lingula, which has given its name to the older fossiliferous rock, is another form occurring unchanged in strata of every period. Only 34 species are known, and none of them are very common. The latest British Lingula is found in the coralline crag (older pliocene) of Suffolk: the nearest living species is as far off as the Philippines. L. Davisii, of the "Lingula flags" in North Wales, has a pedicle groove in the ventral valve, by which the posterior adductor (or cardinal muscle) must have been divided into two elements, as in the genus Obolus; externally it has all the appearance of an ordinary existing shell. Obolus, Eichw. (= Ungula, Pander) is so abundant in the lower Silurian sandstones of Sweden and Russia as to have given its name to the "obolite grit." In England it occurs only in the upper Silurian of Dudley. The shell is horny in texture, and often stained blue, like the Lingula, by the presence of phosphate of iron. In shape it is regularly oval, and differs from Lingula in the character of the internal muscular impressions.

Class II.—Lamellibranchiata.

(Bivalve Shells.)

More than a third part of the known fossil shells are ordinary bivalves (Conchyfera, Dh.) They amount to nearly 6000, while the recent species scarcely exceed half that number. Nevertheless, it is a group which attains its maximum in the present seas. The genera are seven times more numerous in the newer tertiary than in the oldest geological system; and the number of species found in the entire Silurian series is less than 100, while the chalk contains 500, and the miocene 800. Out of 150 genera, 35 have become extinct, besides numerous sub-genera. The families Cyprinidae, Astartidae, and Anatidae have passed their maximum; the Trigoniidae are nearly extinct; and the Hippitidae have no living representatives.

The monomyary bivalves, and others with an open mantle, attain a degree of importance at an early period; and with them some of the burrowing families (Mycetidae and Anatinidae); while the highest organized siphonated shells (e.g., Veneridae and Tellinidae), unknown in the older rocks, are most abundant now.

The family Ostreidae, distinguished from the Pectens and Anomiae by resting on the left valve, contains two fossil forms. Of these, Exogyra resembles an oyster with spiral umbones, directed backwards, or to the left hand; it is an attached shell, characteristic of the cretaceous strata. The genus Gryphaea (fig. 14, 1) abounds in the oolites, and is


gregarious, but unattached, the umbo of the larger valve being curved inwards like a claw. A single Ostrea occurs in the carboniferous limestone, after which the species become abundant, and are with difficulty distinguishable from the smooth and plaited, or "cocks-comb," oyster of the present day.

Several curious modifications of Anomia and Placuna have been obtained in a fossil state. Limanomia (Bouchard) has ears like Lima, and is attached to shells and corals of the Devonian age. Placunopsis (M. and L.), found in the oolites, has a transverse ligamental groove, which, like the umbo of the upper valve, is some way within the margin of the shell. And Carolia (Cantr.), a tertiary form of Placuna, has a byssal plug passing through a foramen like that of Anomia when young, but closed in the adult.

Fossil Pectinidae are very numerous. Some of them in the carboniferous limestone (e.g., P. Sonneratii) cannot be distinguished generically from the living Pectens, and retain diverging bands of colour. But the greater part of these old species are somewhat aviculoid in form (fig. 13, 1), and their hinge-area is grooved with cartilage-furrows like those of *Area*. The most beautiful forms occur in the chalk and greensand, and resemble the recent scallop (*Janira*, Schum.) in the inequality of their valves, but are further characterized by the possession of articulating hinge-teeth like *Spondylus*. These constitute the genus *Neithia* (fig. 14, 2). *Plicatula* exist in the trias and oolites, along with shells referred dubiously to *Hinnites* and *Spondylus*. True *Hinnites* (a sub-genus of *Pecten*) are characteristic of the miocene. *Spondylus* appear in the greensand and chalk. Some of them (like the so-called "Plagiostoma spinosum") are unattached; others resemble the recent deep-water *S. Gussonii*, and have been called "Dianchore." The inner layer, including the hinge of these shells, is seldom preserved. *Lima duplicata*, and some other oolitic species, have two ranges of little hinge-teeth, but not like those of the recent species of *Lima*. The large and smooth or striated Limas of the oolites have been called *Plagiostoma*, a name originally given by Llwyd.

The pearl-oysters (*Arculidae*) are also very abundant fossils; but owing to the frequent repetition of similar forms, it is difficult to determine the genera with any degree of certainty by the aid of external characters alone. The Silurian species mostly belong to the genus *Pterinea* (Goldfuss), and are broadly winged, and have the hinge-area striated lengthwise, and a few diverging hinge-teeth. *Ambonocyba* (Hall) resembles *Inoceramus*, and ranges from the Silurian to the carboniferous strata (fig. 13, 3). The Silurian genus *Cardiola* is ridged like a cockle; and *Posidonomya*, which is found in all the palaeozoic rocks, is very thin and concentrically furrowed (fig. 13, 2). Many other genera have been proposed whose characters are even more imperfectly understood. *Monotis* (*Satinarius*), one of the common shells of the trias, has no anterior ear. *Pteroperma* (Lyceyt), an oolitic form, has a winged shell, with numerous small anterior teeth and long posterior laminae. The genus *Gerrillia* (fig. 14, 4), ranging from the carboniferous limestone to the chalk, consists of elongated shells, with several cartilage-pits in the ligamental area. *Bakewellia*, found in the Permian, has an anterior muscular impression like *Area*. The recent genus *Perna* commenced in the lias or preceding formation, and exhibits great variety of shape. *Pulvinites Adansonii* (fig. 14, 3) appears to have been a Perna with a byssal foramen like

*Anomia*; and *Inoceramus* (fig. 14, 5), characteristic of the cretaceous strata and oolites, differs from *Perna* chiefly in form, the larger valve being sometimes completely involute, and resembling a *Nautilus*. The genus *Pinna*, which appears to belong to this family, although provided with two adductor muscles, occurs fossil in the Devonian and all subsequent strata. Some of the oolitic species, distinguished by the name *Trichites*, are inequivalve and irregular, and attain a thickness of more than an inch, resembling mineral masses of fibrous carbonate of lime.

Amongst the *Mytilidae* are many Silurian species distinguished by their large, round, anterior muscular scar (*Modiolopsis*, Hall), and others which have a straight hinge-line and plaited valves (*Orthonotus*, Conrad). *Myalina* has the cartilage-groove repeated (fig. 13, 4), and is found in the upper palaeozoic rocks. Sometimes the anterior adductor is supported on a shelf, as in the recent *Septifera* and *Dreissenia*. True *Mytili* and *Modioles* abound in the oolitic strata. *Dreissenia*, now confined to the rivers of the Aralo-Caspian region, or only naturalized in Western Europe, was represented by many species, and some of large size, in the eocene of Hampshire and miocene of Vienna.

Fossil *Arcadae* are far more numerous than the recent shells, and mostly belong to the division *Cucullacea*, of which a single species survives in the Coral Sea. The palaeozoic *Arks* have anterior teeth like *Area*, and posterior teeth like *Cucullaria*, and differ from both in the reduction of the hinge-area to a narrow tract corresponding with the posterior half only in the recent shells. The casts of Ark-like shells in the Silurian rocks are further distinguished by a deep furrow behind the front muscular impression. These constitute the genus *Ctenodonta* (Salter), which has hinge-teeth like *Nucula*, and a prominent external ligament (fig. 13, 5). Some of the oolitic *Arks*, with a byssal sinus, and the posterior teeth very long and parallel, form a sub-genus called *Macrodon* (fig. 14, 6). Others, with prominent umbones, teeth like *Nucula*, and a striated ligamental area, form the genus *Isoarea* of Münster (fig. 14, 7). Above 200 species of *Nucula* and *Leda* are known only as fossils, and range through all the rock systems. The palaeozoic species are anomalous in form, and when better understood, will certainly be considered distinct as genera. *Yoldia* is a newer tertiary form characteristic of high northern latitudes; and *Solenella* occurs fossil in Patagonia and New Zealand. The problematic genus *Solemya* is supposed to have existed in the carboniferous period. *Pectunculi* appear first in the cretaceous strata, being less ancient than *Limopsis*, which occurs in the Bath oolite. A member of the latter genus found in the Belgian eocene has the ligamental area entirely behind the cartilage-pit, and is called *Nucinella* by D'Orbigny. The "Stalagmum" of Conrad (= *Myopara*, Lea) is identical with *Crenella* (T. Br.), a sub-genus of *Modiola*, found in the cretaceous and tertiary strata.

The *Trigoniidae* are represented in the lower Silurian strata by *Lyrodema* (fig. 13, 6), a shell with several radiating hinge-teeth, striated transversely; and in the upper palaeozoics by *Axinus* (fig. 13, 7) and several other imperfectly-known genera. The trias contains true *Trigonia* associated with the genus *Myophoria* (fig. 14, 8), which has the umbones turned forwards, and a posterior hinge-tooth. The only member of this family which has yet been found in tertiary strata is the little genus *Verticordia* (Wood) of the crag. No *Trigonia* have been met with, although 100 species are known in the secondary rocks, and two are still living on the coasts of South Australia.

Fresh-water mussels (*Unionidae*), of large size and various form, occur in the Wealden formation, and are not generically distinguishable from recent shells; but those of the coal measures and older rocks are extremely problematic, and may even belong to marine genera.

Of the genus *Chama* there is one species in the upper greensand and chalk of England, and another in the London clay. Elsewhere they are more abundant, amounting to thirty species. Closely allied to *Chama* is the *Diceras* (Lam.), of which the remarkable casts attracted attention at an early period (fig. 14, 1). They are found in the coral rag of France and Germany, and resemble the horns of some animal. The shell is attached by the umbo of either valve, indifferently, like some of the recent Chamas. The posterior adductor muscle is supported on a prominent ridge (as in *Pachydesma*, *Megalodon*, and the recent *Cardilia*), which causes a spiral furrow in each horn of the cast. The shells which succeed *Diceras*, in the lower cretaceous strata, have the right valve usually much smaller than the left; and in one instance (fig. 14, 2) it is like the operculum of a spiral univalve. The only British species of this group is *Requienia lonsdaliae*, found in the ironstone of Bowood. In France, and also in Texas, another form occurs, with the attached valve simple and conical, like a Hippurite. The ligamental groove is straight, and the umbo of the free valve marginal.

These shells are so intimately allied to the *Hippuritidae*, that *Requienia* has been frequently included with them in the apocryphal order "Rudista." The members of the Hippurite group are attached and gregarious, like oysters, often occurring in great numbers, and filling large tracts of rock. Their valves are different in structure and sculpturing, and are articulated by two prominent teeth above and one below; the cartilage is internal, but there is a conspicuous ligamental furrow outside. There are nearly 100 species characteristic of the cretaceous strata, and especially of the lower chalk, or "hippurite limestone." Only one species (*Radiolites mortoni*) is found in England; the rest are from the West Indies, Southern Europe, Algeria, and the East. The form which approaches nearest to *Chama* is the little genus *Caprotina* (fig. 15, 7), whose upper valve has a marginal umbo, but is in other respects like a miniature *Radiolite*. *Caprina* (D'Orb.) has the free valve perforated by canals which open in the inner margin, and in *Caprinella* the outer lamina of both valves possesses this structure. One valve is sometimes spiral (fig. 15, 6), and partitioned off internally by numerous septa, like the water-Spondylus, but so regularly as to resemble the chambered shell of a *Nautilus*. In the *Radiolite* (fig. 15, 5), both valves are conical, and the umbo of the free valve (marginal in the very young shell) becomes central in the adult. The structure of the hinge is modified by Mollusca, the absence of any spirality in the valves, but is essentially the same as in *Caprotina* and *Diceras*; the prominent teeth of the upper valve support curved plates for the attachment of the adductor muscles, which become continuously more undercut in the course of their growth. In *Hippurites*, the anterior muscular plate projects horizontally, the posterior vertically, like a third tooth, for which it has been mistaken. In this genus there are two longitudinal inflexions of the outer shell-wall beside the ligamental furrow, one corresponding to the posterior muscular plate, the other (or third) apparently a siphonal inflexion like that in *Trigonia* and *Leda* (fig. 15, 4).

The cockle-shells (*Cardiidae*), as they have a world-wide distribution now, had a corresponding range in time, and are found in all strata from the Silurian upwards. The commonest fossil type of *Cardium* is ribbed concentrically on the sides, and radiately on the posterior slope, a style of ornament almost unique amongst the 200 recent species. The Caspian cockles, distinguished by a sinus in the pallial line, appear to have inhabited the Aralo-Caspian region almost from the middle tertiary period; the hinge-teeth are reduced to one (*Monodacna*) or two (*Didacna*) in each valve, and are sometimes quite wanting even in the young shell (*Adacna*, Eichw.). *Lithocardium aviculare* (fig. 16, 7) is a characteristic shell of the Paris basin, and appears to have spun a byssus, like the fry of some recent cockles; it also resembles the oriental *Tridacna*, of which a species is found in the miocene of Poland. The genus *Conocardium* (fig. 13, 8) of the upper Silurian and carboniferous systems is remarkable for the prismatic cellular structure of its shell, and the truncation of the posterior (?) side of the valves, which are furnished in some species with a slender siphonal process.

The *Lucinidae*, allied to the cockles in their hinge-structure, are also plentiful in the fossil state, and have as wide a range. They are usually recognisable, even when in the condition of internal casts, by their circular form and the oblique ridge on their disk. Casts of *Lucina* also exhibit the peculiar narrow outline of the anterior adductor detached from the pallial line. *Cryptodon*, *Diplodonta*, *Keltia*, and *Pythina* are found in the eocene tertiary. *Corbis*, under the sub-generic form of *Sphara*, commences in the trias; another modification, found in the oolites and chalk (*Unicardina*, D'Orb.), is edentulous; and *Tancredia* (Lyell), a compressed triangular shell, with a dentition like *Corbis*, is frequent in the lias and oolite.

The fresh-water *Cycladidae* are represented in the Wealden and eocene by many species of *Cyrena*, mostly of small size. The recent *Corbicula fluminalis* of eastern rivers is a common fossil of the pliocene tertiary in England and Sicily.

The *Cyprinidae* and *Astartidae* are more abundant as fossil shells, and had a wider range of old than at the present day. Nearly 100 species of *Cyprina* have been catalogued, commencing in the trias; the dentition of the older species is, however, somewhat peculiar. The *Isocardiae* are almost as numerous, and have the same range, but many of the fossil Isocardia-looking shells are really related to the *Anatinidae*. A yet higher antiquity has been assigned to *Cypricardia*, a genus now very scarce and difficult to obtain, on account of its habit. The palaeozoic *Pleurophorus* (fig. 13, 9) is distinguished by the prominent ridge behind the anterior muscular impression; and *Megalodon* (J. Shy), by the plate supporting the posterior adductor. This genus is represented in the oolites by *Pachyrima* (fig. 16, 1), and in the tertiaries and modern seas by *Cardilia*.

The genus *Astarte*, now limited to a dozen species in the North Atlantic and Arctic seas, has an almost worldwide geological distribution, and counts 200 species in Mollusca. D'Orbigny's catalogue, commencing with the lias period. *Crassatella*, now almost a southern form, is common in the cretaceous and tertiary strata of Europe. Closely allied to *Astarte* is the extinct genus *Opis* (fig. 16, 3), of which there are 42 species in the secondary series; and *Cardinia* (fig. 16, 2), characteristic of the lias and oolites.


**Secondary and Tertiary Bivalves.**

1. *Pachyrhizus septiferum*, Bur.; *Ceratina*, Meuse. 2. *Cardinia hybrida*, Sib.; *Liss*, Gloucester. 3. *Opis tumidula*, Sib.; *Glyptodon*, Bagnac. 4. *Pecten scaphiformis*, Dray.; *Liss*, Saxony. 5. *Sowerbya crassa*, D'Orb.; *Oxfordina*, Ardenne. 6. *Goniomya scripta*, Sib.; *Kelvedon rock*, Willa. 7. *Lithocardium angulare*, Lam.; *Escom*, Paris. 8. *Grateloupia irregularis*, Bart.; *Miosse*, Bordeaux. 9. *Teedina persimilis*, Lam.; *Eocene*, Belgium.

The so-called Unios of the coal measures (*Anthracosia*, King) are probably members of this group. One hundred species of *Cardita* (including *Venericardia*) are found in the secondary and tertiary strata; of the 50 recent forms, one only is Arctic, and this occurs in the glacial deposits of England. The allied genus *Myoconcha* is characteristic of the older secondary rocks, and *Hippopodium* of the lias.

The *Veneridae* are pre-eminently characteristic of the tertiary and present period. Some obscure species of Venus are found in the oolites; *Cytheraea* occur in the greensands; *Artemis*, *Trigona*, *Lucinopsis*, *Venerupis*, and *Tapes* in the middle tertiary; *Petricola* in the eocene. The only extinct form is *Grateloupia* (fig. 16, 8), which differs but little from *Trigona*.

The Mactras and Tellens are also comparatively modern groups; most of the supposed oolitic species belong to *Lucinidae*, except *Sowerbya* (fig. 16, 5), which has a palatal sinus, and is found in the oolites of Malton and Portland. *Psammobius* and *Mesodon* occur in the greensand; *Donax* and *Syndosmya* in the eocene; *Gastrana* and *Lutraria* in the miocene. *Lutraria rugosa*, still living on the coast of Portugal, is fossil in the raised beaches of Sussex.

The oldest forms of razor-fish (*Solenidae*) are those with the transverse internal rib (*Solecurtus*), which occur in the neocomian, whilst true Solens and *Glycimeris* appear first in the eocene strata. The genus *Mya*, as now restricted to the species resembling *M. arenaria*, are only met with in the newer tertiary. *Corbula* ranges upwards from the lower oolites; *Nucara* appears in the upper greensand; and *Thetis* (= *Poromya*, Forbes) in the neocomian.

Above 100 species of *Panopea* (a genus essentially like *Mya*) have been obtained from oolitic and tertiary strata in all parts of the world. They are with difficulty distinguished from those equally numerous forms of *Anatinidae* which have been associated with *Pholadomya* on account of the tenacity of their finely-granulated valves; they constitute the genus *Myacites* (Bromi), and occur in all the palaeozoic and secondary rocks; some of the oolitic and cretaceous species are distinguished by V-shaped furrows (fig. 16, Mollusca, 6). Still more numerous are the fossil forms of *Pholadomya*, which range upwards from the lias, but are reduced to a single species now living in the Caribbean seas. Shells with the umbones fissured like *Anatina* also occur in the oolites. *Pandora* first appears in the older tertiary. Amongst the extinct genera referred to this family are the Silurian *Grammysia* (fig. 13, 10), with valves folded transversely; the carboniferous *Edmondia* (fig. 13, 11), with large oblique cartilage plates; and *Cardiomorpha*, shaped like *Isocardia*; and the oolitic *Ceromya* (Ag.), which also resembles the heart-cockle in form. *Ceromya* is an oolitic *Anatina*, with the posterior end of the valves much attenuated.

The genus *Gastrochaena* appears in the lower oolites; and casts of its burrows are frequently preserved after the decomposition of the coral in which they were made. *Clavarella* dates from the upper greensand, and *Aspergillum* from the miocene. *Saxicara* is found in the newer tertiary and raised beaches of Northern Europe; and the great species commonly called *Panopea* *Norvegica* is a characteristic fossil of the newer pliocene of Britain and Greenland.

The Pholades and ship-worms appear first in the oolitic strata. Forms resembling the recent *Martesia striata* have been discovered in fossil wood of the lias and Speeton clay. *Jouannetia* (Desm.) was first known as a miocene fossil; and *Pholas* occurs in the older tertiary. Extinct species of *Teredo* are found in the silicified wood of the greensand of Blackdown and in the fossil palm-fruits of Brabant and Sheepy. The drift-wood of the London clay is usually perforated by the ship-worm, and also by an extinct form (*Teredina*, fig. 16, 9), which resembles *Martesia* in possessing an umbonal shield: when adult, it not only closes the anterior pedal opening, but also cements its valves to the shelly lining of its burrow, like an *Aspergillum*. Specimens have been obtained in which the whole interior of the valves and tube had been excessively thickened towards the close of life by successive layers of shell.

**Class III.—GASTEROPODA.**

Fossil univalves—the remains of spiral and limpet-like shells—are not wanting in any but the very oldest fossiliferous rocks ("lingula flags"). From the lower Silurian, where less than 100 species, referable to scarcely more than 10 genera, are found, they increase in number and variety slowly and regularly up to the newer tertiaries, which have afforded ten times as many genera and twenty times as many species. The total number of fossil marine univalves is less than 6000; the recent exceed 8000; and although we may expect to discover more new fossil species than recent, yet it is evident the group of univalves has only now attained its maximum development.

Between the extinct and living air-breathers the numerical discrepancy is still greater. About 300 land-snails, and half as many fresh-water *Pulmonifera*, are enumerated in the fossil catalogues; but the greater part of these are recent species, and the whole bears no proportion to the number of living land-snails, which exceed 4000. That many more have formerly existed is indicated by the fact, that the fossil land-snails of the older tertiaries of Europe are entirely different from their living successors, and most nearly represented at the present time in the West Indies and Brazil. The generic forms peculiar to oceanic islands (remains of old continents) are more numerous than those of the mainland, as if this order had once been more important. But the circumstances favourable to their petrification must have been of such rare occurrence as to preclude the probability of attaining more than the scantiest information concerning them.

From the large and proportional number of living Gas- Mollusca. teropods, and the great amount of information which has been obtained of late years respecting their structure and habits, it might be expected that the affinities of the fossil univalves would be easily worked out, and their indications fully interpreted. Such, however, is not the case. Univalve shells present no internal markings, easily accessible owners crawling over the bottom, for it can scarcely be Mollusca.

insisted that all were necessarily floaters on account of their organization. The species of *Bellerophon* are numerous in all the palaeozoic rocks, and some of the smaller kinds appear to have been gregarious: those with disconnected whirls have been called *Cyrtolites* (Conrad.) The *Bellerophina* of D'Orbigny (fig. 18, ii) is a minute shell found in the gault. The other division (*Firoloidea*) consists of Molusks in which the shell is wanting or rudimentary, and small compared with the bulk of the animal. A single species of the genus *Carinaria* has been found in the middle tertiary of Turin.

**Strombidae.**—The Strombs, with their massive shells, nevertheless, resemble the fragile Heteropods in some respects. They have the same lingual dentition, and the same carnivorous habits; and though living on the sea-bed, they rather leap than glide, having a narrow sole and a deeply-divided opercular-groove lobe. Characteristic of the warmer zones of existing seas, they are only found fossilized in the newer tertiary strata of countries south of Britain; but there is a group of little shells related to the recent *S. ficusellus* in the older tertiaries of London, Paris, and America, to which Agassiz has given the name *Rimella*.

The allied genus of scorpion-shells (*Pterocera*), now peculiar to eastern seas, has been described as occurring fossil in the secondary strata of Europe; but the extinct species appear to be more nearly related to *Aporrhais*. This genus, now confined to the western shores of Europe, occurs in all the tertiaries, and is represented in the secondary rocks by many remarkable forms. Some have been separated under the name *Aloria*; and to this group the so-called *Pterocera Bentleyi* may perhaps be referred (fig. 18, 2). *Rostellaria* and *Serapis* (or *Terebellum*), now peculiar to the Red or eastern seas, are conspicuous fossils of the European eocene, at which time their range extended to America. Some of the ancient Rostellarias have the outer lip enormously expanded, as in the *R. ampla* (*Hippocrena*) of the London clay. In the oolites and chalk there are slender fusiform shells (*Spiniger*, D'Orb., fig. 18, i) with spines on the sides of the whorls, as in some recent *Rimella*.

**Muricidae.**—The great family of whelks, by far the most

like those of bivalves, and exhibiting the essential characters of the soft parts; and their external forms are often so overlaid with ornament, and disguised by mimetic characters, as to mislead upon a first examination. Shells of any family may be limpet-shaped, or turritated, or discoidal, plain or ornamented. It is more desirable to ascertain whether they have been nacreous or porcellaneous; whether the apex (or nucleus) presents any peculiarities; and if operculated, whether the operculum was few-whirled or multispiral.

The earlier describers of fossil univalves unhesitatingly recognised many familiar recent genera, even in the older rocks. But their "Melemas" were marine shells; the supposed *Buccinum* had no notch; the *Solaria* were pearly; the *Neritas* assumed, when adult, the irregular aperture of *Pileopsis*; the Naticas had non-spiral opercula; and the *Maclurea* was figured upside down.

The more closely palaeozoic univalves are examined, so much the more do they appear to differ from ordinary recent types; and the search for allied forms has to be conducted amongst the rare and minute and least understood of recent shells.

**Nucleobranchiata.**—Those fossil univalves, which in their symmetry resemble the *Nautilus*, but are unfurnished with air-chambers, have been compared to the recent *Heteropoda* (or *Nucleobranchiata*, Bl.), and especially to that division typified by the tiny *Atlanta*, in which the animal can withdraw itself completely into its shell, and close the aperture with an operculum. The genus *Porcellia*, characteristic of the carboniferous age, has a discoidal shell, with a spiral nucleus projecting, as in *Atlanta*, from the right side; the whorls are exposed, and marked with a narrow band along the back, ending in a deep slit (fig. 17, 6). Another genus (*Bellerophon*) resembles the recent *Oxygyrus* in its more globose form, with a similar narrow umbilicus on either side (fig. 17, 7); sometimes the shell is thin and the aperture expanded, like a trumpet, whilst other species are globular and solid; the former may have been tenanted by large animals living at the surface of the open sea, the latter seem to have been more adapted to protect their

---

**Fig. 17.**

Palaeozoic Univalves.

1. *Lepidoceras* Lebedev, Lér.; Carboniferous, Tournay. 2. *Macrechilus* Schlotheim, D'Archi; Depression, Eifel. 3. *Scaphoconcha expansilabrum*, Söhr.; Devonian, Nassau. 4. *Eumicrotremus* Lebedev, Lér.; Wesser, Liassic, May Hill. 5. *Archibuccinum angustatum*, Söhr.; Devonian, Eifel. 6. *Parcellia* Pansel, Lér.; Carboniferous, Tournay. 7. *Bellerophon bi-carinatus*, Lér.; Carboniferous, Tournay. 8. *Tubima armata*, Barr.; U. Siberian, Bohemia. 9. *Maculina Peach*, Söhr.; L. Silurian, Staffordshire. 10. *Columbella quadriloculata*, Söhr.; Carboniferous, Lanark.

---

**Fig. 18.**

Secondary Univalves.

1. *Spiniger*, sp.; Oxford Clay, Chippenham. 2. *Aloria* Bentley, M. and L.; Great Oolite, Collyweston. 3. *Purpurina Morisoni*, Barr.; Great Oolite; Mitchinhampton. 4. *Neritina Brunsdeniana*, Thurn.; Coven, Poland. 5. *Crenulina* Sowerby, Lér.; Great Oolite; Mitchinhampton. 6. *Trochotoma elongata*, D'Orb.; Great Oolite, Mitchinhampton. 7. *Neritina Maimata*, Barr.; Oxfordian, Ardennes. 8. *Plicolus pilatus*, Söhr.; Great Oolite, Ameliff. 9. *Cassula incrassata*, J. Söhr.; L. Greensand, Blackdown. 10. *Acteolina concava*, Debl.; Lias, Normandy. 11. *Bellerophina salinata*, Söhr.; Great, Folkstone.

important group of living sea-shells, is scarcely of higher antiquity than the eocene tertiary. The *Purpurina* of the oolites (fig. 18, 3), and *Columbellina* of the chalk, are Mollusca. extinct genera somewhat resembling *Purpura* and *Colum- bella*. But since the so-called "cones" of the oolites have proved to be *Tornatellae*, it may not be unreasonable to dis- trust these other presumed affinities. The huge uni- valve of the chalk, which Sowerby called a *Dolium*, has been described as a *Pterocera* by D'Orbigny. In the tertiaries siphonated univalves abound, and are mostly re- ferable with certainty to recent genera. The only marked change consists in the comparative abundance of some scarce existing forms, and the absence or rarity of many now most conspicuous. Moreover, the geographical dis- tribution of the genera has undergone a great change since the close of the eocene period. This change is most no- ticeable in the cold-temperate zone, and is evidently the result of altered climate. The northern seas must ever have been inclement, and the tropical seas always tropical; but the latitude of England being most liable to vicissitudes of climate, might be expected to show the greatest variety, and the most complete and rapid alterations of organic

![Fig. 19. Tertiary Univalves.]

1. Nautilus (Aturia) de nie, Ebr.; *Eocene*, Britain. 2. Nautilus (Aturia) front view of a region. 3. Conorbis dormitor, Sol.; *Eocene*, Britain. 4. Boreomia lineata, T. Edw.; *M. Eocene*, Nant. 5. Volutilithes loculator, Sol.; *Eocene*, Britain. 6. Natica (Deshayesia) (recently *Natica*, gen. n.), *Eocene*, N. Italy. 7. Turritella (Prototurritella), Brongniart, *Eocene*, Bordeaux. 8. Nerita (Velutina) peruviana, Gau.; *Eocene*, France. 9. Helix (Lychmus) Matheroni, Rey.; *Eocene*, S. France. 10. Fernandesia tricarinata, M. Br.; *Miocene*, Hockenheim. 11. Volvaria bulloides, Lam.; *Eocene*, Grignon. 12. Vaginella depressa, Basil.; *Miocene*, Bordeaux.

Life. In the London clay are found many species of *Cla- vella*, *Typhis*, *Mitra*, *Pseudoliva*, *Oliva*, and *Ancillaria*; and some extinct forms (*Lestoidea* and *Strepsidura*) re- lated to *Fusus*. The middle tertiary, wanting in England, but largely developed in Central and Southern Europe, also contains many genera belonging now to warmer latitudes, and many species still living in the south. In the newer tertiaries of Europe these southern forms disappear, and are gradually replaced by others of an opposite charac- ter (*Trophon*, *Neptunia*, and *Trichotropis*), now inhabiting the Arctic and boreal coasts. The entire number of fossil *Muricidae* amounts to 1000, or about half as many as the recent. The older tertiaries of England also contain species of *Triton*, *Cassidaria*, *Cancellaria*, and *Pyrula*, shells (now foreign to our seas), which have formerly been included in this family. As regards bulk, there are no fossil species of *Fusus*, *Triton*, and *Cassis* (or *Strombus* and *Voluta*) to com- pare with those of the present day.

Conidae.—The Cones and Pleurotomas appear first in the chalk, and are abundant in the eocene, accompanied by an intermediate form (*Conorbis*, fig. 19, 3), and another extinct sub-genus (*Boronia*, fig. 19, 4), in which the column is plaited, as in *Mitra*. The genus *Terebra* is commoner in the miocene.

Volutidae.—The Volutes also appear as cretaceous fossils in Europe and Southern India; they are very abundant in the London clay, and one occurs in the English crag. The ancient species are mostly distinguished by their spires being acute, as in *Mitra* (fig. 19, 5), a peculiarity only found in one very rare living (?) species, dredged from a bed of dead shells in 132 fathoms water (792 feet) off the Cape. The crag Volute resembles the Magellanic form. *Cymba olla*, the only living European Volute, is a fossil in the *pliocene* of Majorca.

Cypraeidae.—The Cowries form another group of sub- tropical shells once common in the temperate zone. Seve- ral large species are found in the London clay, most nearly related to the southern *Cypraea*; whilst the crag contains only members of the sub-genus *Trivia*, one of which still lives on our coast.

The round-mouthed shells (*Holostomata*), whether ani- mal-feeders or vegetarians, make a conspicuous figure amongst the fossils of an earlier period than that in which the last group began to flourish. The carnivorous *Naticidae* and *Pyramidellidae* are represented in the palaeozoic strata by *Naticopsis*, *Loxoneema* (fig. 17, 1), and *Macrochilus* (fig. 17, 2). The violet-snail (*Janthina*), so unlike any other existing shell-fish, seems related to the Silurian *Scalites*, *Rophistoma*, and *Holopecia*. Shells like *Scalaria* and *So- larium* occur in the trias and oolites associated with *Chem- nitia* (?) of extraordinary size, and species of *Euclia* and *Niso*. These families of shells and the *Cerithiidae* are more abundant fossil than recent, the known numbers being 1500 extinct and 900 living forms. *Solaria*, with discon- nected whorls and pyramidal opercula (*Bifrontia*, Db.), are common in the eocene tertiary, and a single living species (*B. zanclaea*) has been discovered by M'Andrew.

Amongst the tertiary Naticas are many with an oblique aperture and peculiar perforation (*Globulus*, J. Shy., = *Am- pullina*, Bl.), and others with prominences on the pillar (*Deshayesia*, fig. 19, 6). The *Neritacea* of the oolites are re- markable for the spiral ridges (like the "worm" of a screw) winding round their interior, and giving rise to the variety of singular patterns seen in sections (fig. 18, 4). A simi- lar structure exists in the recent "telescope-shell" (*Tere- brolia*). The fresh-water univalves of the Wealden and older tertiaries differ but little from their recent congeners of the genera *Paludina*, *Potamides*, *Melania*, and *Meta- nopsis*. Fossil *Turritella* are of doubtful occurrence before the tertiary; the Silurian species have the peristome com- plete (*Holopella*, M'C.); another form (*Proto*, fig. 19, 7) is characteristic of the miocene.

The bonnet-limpets (*Calyptraeidae*) are common in the old rocks, which also contain a few species of *Chiton* and shells like *Dentalium*. Fossil *Trochidae* are very numerous, but hitherto many *Littorinidae* have doubtless been included with them. Perhaps no true *Turbo* is known from strata be- fore the cretaceous. The *Euomphali* (fig. 17, 4), which char- acterize the older rocks, have multispiral calcareous oper- cula, like the recent *Cyclostrema* (= *Adecorbis*). The genus *Maclurea* (fig. 17, 9), which has been regarded as a "left- handed" *Euomphalus*, is probably very different; it has a thick shelly operculum, sinistrally spiral, and furnished with an internal process, as the Nerites are; the spire is sunk and concealed, whilst the whorls are exposed on the flattened un- der-side; it occurs in the older Silurian rocks of Scotland and North America. One common feature of the palaeozoic spiral shells is their tendency to become irregular towards the conclusion of their growth; in *Serpulina* (= *Phanerotinus*, Shy.), the whorls are all disunited; in *Scoliotoma* (fig. 17, 3) and *Catantostoma* the aperture is expanded. Some small oolitic shells have a thickened peristome (*Crosso- stoma*, fig. 18, 5), like the recent *Lieta*, which commences in the older tertiary. A large proportion of the troch- iform fossil shells have their whorls, whether round or angu- lar, marked by a peculiar band, terminating in a deep slit Mollusca at the aperture; most of these were solid nacreous shells belonging to the genus Pleurotomaria, of which but a single species survives; others in their slenderness resemble Turritella, and have been named Murchisonia (fig. 17, 3). The carboniferous shell called Polytremaria has a row of holes in place of a slit; and the Silurian Tubina (fig. 17, 8) has three rows of tubular spines. The Cirrus of the inferior oolite is a reversed shell with one row of similar ornaments; and Trochotoma (fig. 18, 6) has a perforation near the margin of the aperture, which is carried onward as the shell grows. Scissurella, which is always diminutive and not pearly, makes its first appearance only in the newer tertiary. Halictis occurs in the miocene of Malta. The Neritidae appear in the oolites; besides true Nerites, there are Neritoides (fig. 18, 7), with a channeled outer lip; Pileolus, which is perfectly limpet-like above (fig. 18, 8); and Neritopsis, with its angular columellar notch most distinctly marked. Key-hole limpets (Fissurellidae) occur as early as the carboniferous period, but are very scarce at first, and never become numerous. The oolitic Rimula is a minute shell supposed to be related to a very rare living species. Ordinary limpets (Patellidae) of unequivocal form are found in the Bath oolite, but are afterwards less plentiful, and almost disappear from the tertiarys; M. D'Orbigny regarded them as generically distinct, but employed for them a name (Helcion, Mont.) synonymous with Patella.

Pulmonifera.—The existence of air-breathing snails in the palaeozoic rocks is indicated only by the somewhat problematic Dendropupa, discovered by Sir C. Lyell in a coal tree of Nova Scotia. The Purbeck limestone contains a modern-looking Physa; and other species of extraordinary size are found in the older tertiary of France, and also in Central India, where the genus does not exist at the present day. The fresh-water eocene of the Isle of Wight and Paris has afforded many species of Limona and Planorbis; a Glandina rivaling in size the G. truncata of South Carolina; a Cyclotoma, with a sculptured operculum like the Cyclotis Jamaicensis; and an elongated species of the section Megalomastoma, which is now living in both East and West Indies. At Hordle has been found the little Helix labyrinthica, still living in Texas; and in the south of France occur representatives of the Brazilian genera Megaspira and Anastoma (fig. 19, 9). In the miocene is found another genus (Ferussima, fig. 19, 10) resembling the lamp-snail, but supposed to be operculated. The Pulmonifera of the English pliocene are in a few instances extinct, at least in England; nearly all are still living here, but more or less abundant now than they were in the times of the mastodon and elephant. The extinct land-snails of the Atlantic islands Madeira and Porto Santo are associated with remains of many recent species occurring in numbers which have relatively altered, telling the same tale of gradual changes, affecting some species prejudicially, but favourable to the increase of others. The fossil land-snails of St Helena were supposed by Mr Darwin to have become finally extinct only in the last century, owing to the destruction of the native woods by the instrumentality of goats and swine.

Tectibranchiata.—The families typified by Tornatella, Ringicula, and Bulla played a more important part in the secondary and tertiary periods, but their affinities have been seldom understood. The cone-like Aeteonina appeared in the carboniferous rocks, and attained a remarkable development in the lias (fig. 18, 10). They were succeeded by the Aeteonellidae, with a plated columella, in the cretaceous strata; and by Volvaria (fig. 19, 11) in the eocene. The diminutive Ringicula of our seas were preceded by large species of the same genus in the tertiarys, and by Cinula (fig. 18, 9), Globiconcha, and Tylostoma, in the cretaceous strata. The genus Varigera has varices recurring twice in each whorl, like Eulima; and Pterodonta is winged like Strombus.

Pteropoda.—The fragile shells of Hyalea and Cladorda are found in the newer tertiary of Italy, with Vaginella (fig. 19, 12), a form allied to Cucurria. But the occurrence of Pteropoda in the older rocks is attended with considerable obscurity. The shells called Theca are slender and conical; Pterotheca has a wing-like expansion; and Conularia (fig. 17, 10) is a four-sided sheath, with the apex partitioned off, as in the recent Cucurria. If really pteropodous, these shells are the giants of the order.

Class IV.—CEPHALOPODA.

Order I.—TETHYRHACHINATA.

(Sauropod Cephalopoda.)

Of the lower group of Cephalopods, possessing chambered shells similar to the pearly Nautilus, there are 1400 extinct species, belonging to above 30 genera, while 5 or 6 species alone exist in modern seas. These fossils resemble the Nautilus, and differ from the dibranchiate Spirula in the structure of their shell, which is composed of two layers, the outer porcellaneous, the inner pearly; whereas the Spirula—an internal shell—is entirely nacreous. They also agree with the Nautilus in the relative capacity of their last chamber, which seems obviously large enough to contain the whole animal. Moreover, it appears, from the position of the siphuncle and the form of the aperture, that these shells were revolute spiral, or coiled over the back of the animal, and not involute like the Spirula. No traces of fossil septa or horny claws have been found associated with them, nor any indications of dense muscular tissue, even in the same matrix which has preserved so completely the mummy cuttle-fish. By their form and size they were ill adapted for rapid locomotion, and must have depended for safety on the shelter afforded by their solid shell. The discoidal Ammonites attained a diameter approaching 3 feet, and the straight-shelled Orthocerata were sometimes not less than 6 feet in length. These latter must have lived habitually in a position nearly vertical; whilst the discoidal genera would creep over the sea-bed with their air-chambers above them, like a snail-shell reversed. The Ammonites appear to have been provided with an operculum, more secure than the "hood" of the Nautilus, but, like it, composed of two elements united by a straight suture. These opercula, frequently mistaken for bivalve shells, have a porous structure altogether peculiar, and are frequently sculptured on their outer convex surface; whilst their concavity exhibits only lines of growth (fig. 21, 7). Special forms are associated, in all localities, with particular species of ammonite; and their size is adapted exactly to the specimens in which they are found. Calcareous mandibles occur in all the secondary strata, but not (hitherto) in such numbers or circumstances as to imply that they belonged to any other genus beside the true Nautilus. They are of two forms: those corresponding to the upper mandible (fig. 21, 8) have been called Rhyncholites (Paleocephalus and Rhynchostethus of D'Orbigny); whilst the lower mandibles constitute the "genus" Couchorhynchaus of De Blainville (fig. 21, 9). The arms of the extinct Tetrabranchia may have been organized like those of the Nautilus, but were probably less numerous in the genera with slender shells, and in those early forms with a small many-lobed aperture. The length of the body-chamber is greatest when its diameter is least; and the prominent spines which ornament the exterior are partitioned off internally by a nacreous lamina, indicating considerable motion of the animal in its shell. When the outer shell of the fossil is removed by decomposition or the hammer, the margins of the internal septa (or parti- The chambered shells may be divided into two principal groups: those with simple sutures, like the recent Nautilus; and the Ammonitidae, in which the margins of the septa are lobed and foliaceous. In the former the siphuncle is central or internal (i.e., at the margin next the spire); in the latter it is external (i.e., at the back of the shell, but ventral as regards the animal). There are, however, Nautili with lobed sutures (Aturia, Bronn, fig. 19, 1); and some with an external siphuncle (Cryptoceras, D'Orb.). And on the other hand, the sutures of the Ammonite are at first very slightly lobed, and become progressively more complex; so that specimens of the same species have been referred to three genera—Goniatiites, Ceratites, and Ammonites—according to their age.

With the exception of Goniatiites, the Ammonitidae are peculiar to, and co-extensive with, the secondary strata; while the Nautilidae, with the exception of Nautilus and Aturia, are confined to the palaeozoic rocks. But the so-called palaeozoic Nautilidae exhibit peculiarities suggesting very wide differences. It has been proposed to associate the greater part of them with the Orthocerata as a distinct family, but at present the data are defective. Like the Ammonitidae, their shells assume almost every conceivable form and curvature, and the genera founded on these characters are very ill defined.

The simplest form of Orthoceras is like a Nautilus unrolled; and Lituites (fig. 20, 2) is the same with the apex spiral. Some of the carboniferous Nautili have a square back, and the whorls either compact or open in the centre (fig. 20, 1); whilst the last chamber is more or less disunited. The species with the whorls quite disunited constitute the genus Trigonoceras, M'C. (=Nautiloceras, D'Orb.). The Silurian genus Trochoeceras, Barr, is a spiral Nautilus. Clymenia, a characteristic Devonian fossil, has angular sutures and an internal siphuncle; it may perhaps be coiled up ventrally like the Spirula. The tertiary shell called Nautilus zig-zag (Aturia, Br., fig. 19, 1, 2), which is so widely distributed in Europe, America, and India, has a siphuncle nearly marginal when young, but gradually becoming more central in the adult: it has no special relation to Clymenia.

Those species of Orthocerata in which the aperture is contracted form the genus Apioceras, Fischer (=Poterioceras, M'C.), or when also curved, the Oncoceras of Hall. In Barrande's genus Accoceras (fig. 20, 9) the shell is flask-shaped, the chambered and siphunculated apex being apparently deciduous; the aperture is contracted, and the air-chambers occupy only the dorsal half of the shell. In Phragmoceras (fig. 20, 7), the shell is slightly curved to the ventral side, and the aperture is remarkably contracted, the opening for the respiratory funnel being nearly distinct from the cephalic aperture. In Cyrtoceras the curvature is dorsal.

In some other members of this family the siphuncle attains a remarkable size or extraordinary complexity.

In Camaroceras (fig. 20, 4), the siphuncle is lateral, quite simple and equal to half the diameter of the shell. Casts of these great siphuncles were called "Hyolithes" by Eichwald; they frequently contain small shells of Orthoceras, Bellerophon, and other genera. In some species the siphuncle is strengthened internally by repeated layers of shell, or partitioned off by a succession of funnel-shaped diaphragms; these constitute the genus Endoceras of Hall. The same author has given the name Discosaurus to a fossil which is evidently the siphuncle of some very delicate and perishable chambered shell (fig. 20, 6). In those Orthocerata with siphuncles most nearly resembling the

---

1 In the carboniferous species of Actinoceras (e.g., A. giganteum), these foramina form a cross on the ventral side of the siphuncle. Mollusca, having the sutures lobed but not foliated. They seldom exceed 6 inches in diameter, and are usually very much smaller.

The whirls are most frequently concealed to some extent, and often marked by cross furrows or "periodic mouths."

The Ceratites are distinguished by having the lobes of the sutures serrated, while the intervening "saddles" (or curves directed towards the aperture) are simple. They are found in the trias of Europe, Tibet, and South America; and again very rarely in the cretaceous strata of France and Syria,—a circumstance quite anomalous in the history of the geological distribution of life. Many Ammonites, perhaps all, are like Ceratites when young.

The species of Ammonite exceed 300; and their range is co-extensive with that of the secondary rocks. They are found throughout Europe, and at the Cape, in Kamtschatka, Tibet, and S. India. They are absent from a large area of the United States, but are found in the cretaceous strata of New Jersey, Missouri, and the West Indian Islands; also in Chile and Bogota. The sections into which, for the sake of convenience, this extremely natural group has been broken up, are very ill defined, and have no pretension to be considered sub-generic. The group (called Cassiani) characterizing the triassic period, is remarkable for many-lobed and elaborately-foliated sutures,—a circumstance more important, because it is the oldest group, and associated with Ceratites and the last-surviving Goniatites and Orthocerata. They abound in the "alpine limestone" of St Cassian, and Hallstatt in Austria. A second group (Arietes), having the back keeled, with a furrow on each side of the keel, as in the great Ammonites called Pachylandi and Coneycerai, mark the lias period; they are less plentiful in the oolites, and are represented in the greensands by the "Cristati," which are keeled, but not furrowed, and develop a "beak," or process, from the keel when adult. The "Arietes" pass by many intermediate forms into the "Falciferi" (e.g., A. serpentinus), also characteristic of the upper lias, and these are represented by a few quoit-shaped species (Disci), with sharp backs, in the oolites.

Ammonites with serrated keels (Amalthei), exemplified by A. spinatus and margaritatus, abound in the middle and upper lias, and again in the oolites (e.g., A. cordatus and excavatus). They are succeeded by the "Rothomagenses" in the chalk,—thick Ammonites with a line of tubercles in the place of the keel.

Ammonites with channelled backs (Colliciati) are represented in the lias (A. anguliferus), inferior oolites (A. Parkinsoni), and middle oolite (A. anceps), and in the cretaceous strata by numerous species (e.g., A. serratus, Mollusca, laurus, and fulcatus), remarkable for their elegance.

Of the species with backs more or less squared, armatus and capricornis occur in the lias, athleta and peroramatus in the Oxfordian. But the oolitic forms which have the back square, and ornamented with two rows of spines when young, like Goweri, Duncani (fig. 21, 2), and Jason, become rounded and unarmed in their old age.

Round-backed Ammonites abound in the lias and oolites. The snake-like Annulatus, the spine-bearing coronatus, and fimbriatus with its ornamental fringes, have been regarded as types of small groups. A more important division (ligata) is distinguished by nearly smooth whirls, constrictions recurring at regular intervals. These are seen in A. taenius, and others related to Heterophyllus; in many neocomian Ammonites, and in A. planulatus of the lower chalk.

These constrictions, often accompanied by a prominent rib, undoubtedly indicate periods of rest, when the Ammonite ceased for awhile to grow. They may be traced in species belonging to other groups, as well, e.g., in biplextus and triplicatus, as in the ligata; but most frequently all indications are obliterated by subsequent growth. It has been a question whether the lateral processes of Ammonites Duncani (fig. 21, 2), are formed and removed periodically, or whether they are peculiar to the adults, and mark the close of their outward growth. The first conclusion is more probable from analogy; and they are commonly found with small and apparently young shells, but not (any more than the lateral spines of the living Argonaut) in those of adult size and condition.

It was remarked by the elder Sowerby that Ammonites were most beautiful when of middle growth, the ornamental characters being less developed in the young, and lost in the adult. The ribs and spines, and even the keel or furrow of the back disappear, in many instances, from the body-whirl of the full-grown shell.

Varieties of form, such as marked the palaeozoic Nautilidae, are met with in the Ammonitidae, chiefly towards the close of their reign. The Baculite (fig. 21, 4), with its straight shell, is characteristic of the upper chalk; and the Turrilite, which is spiral, and usually a left-handed spiral, abounds in the lowest beds of the same formation. In Hamites the shell is straight, returning upon itself after a certain space, and forming a simple or complex hook. In Ptychoceras these limbs of the hook-like shell are in close contact. The Toroceras is curved like a bow; in Crioceras the discoidal whirls are separate; and in Scaphites (including Ancyloceras) the shell, at first compact like an Ammonite, or open-whirled like Crioceras, lengthens out finally, and returns upon itself like the crozier of the Hamite. Helicoceras, again, connects the last with the Turrilite by its elevated spire terminating in a prolonged crozier.

Of these forms, Ancyloceras alone is found in the oolites; all the rest are cretaceous; and most abound in the alpine districts of the south of France.

Order 2.—Dibranchiata.

(Cuttle-fishes.)

Of the two great divisions of cephalopodous Mollusca, that which is represented at the present day by the pearly Nautilus was developed in the greatest profusion and variety in the palaeozoic and secondary periods; whilst the more active and intelligent cuttle-fishes and squids have not been (certainly) found in rocks older than the lias, and scarcely above 100 are found in the whole secondary and tertiary series, while twice as many have been obtained in existing seas.

The Sepiidae are represented in the middle and upper oolites by the genus Coccotuthus (fig. 22, 6), whose strong and granulated bone is furnished with broader lateral expansions than the recent cuttle-fishes. In the older ter- Mollusca, taries of London and Paris, many species of Sepia appear to have existed, but only the solid macro (fig. 22, 5) of the shell is usually preserved. In the miocene tertiary of Malta, a diminutive cuttle-bone is not rare; and at Turin a remarkable form (Spirulirostra, fig. 22, 7) has been discovered, in which the apex is provided with a chambered and siphonated cavity like the shell of the Spirula. Two other genera, Beloptera (fig. 22, 8) and Belemnitis, very imperfectly known by rare and fragmentary examples, occur in the eocene tertiary.

Remains of the Calamaries (Loligidae) are often found in the fine-grained and laminated argillaceous limestones of the lias and Oxford clays, as at Lyme Regis, and Böll of Solenhofen. Some of these are slender, like the pens of the recent Ommastrephes, and furnished with a small conical appendix, as in that genus; whilst others are broad, and pointed at each end (Beloteuthis). The most common form has the shaft wide and longer than the wings, and is truncated posteriorly. It has a nacreous lining, and is usually accompanied by a large and well-preserved ink-bag (fig. 22, 4). These were called Belemnosepia by Agassiz and Buckland, who supposed them to belong to the same animal with the Belemnite. They have also been called Loligo-sepia and Loliginites; but the name Geotethis, given by Count Münster, appears least objectionable. One species (Mastigophorus latifrons) is of frequent occurrence in the Oxford clay near Chippenham, which retains not only the horny (chitinous) pen and ink-bag, but also the muscular sheath, the rhombic terminal fins, and at least the bases of the arms, with the minute hooks, and traces of the mandibles. Horny claws, like those of the uncinated Calamary (Onychodenthis), have been observed arranged in double series in the lias of Watchett, and they sometimes occur in great numbers in the coprolitic remains of the Enaliosauri. The most remarkable examples of this kind are preserved in the lithographic limestones of Solenhofen, and show that the extinct Calamary had ten nearly equal arms, the tentacles, in their retracted condition, being undistinguishable from the rest—each furnished with 20 to 30 pairs of formidable hooks. What further evidence was needed respecting the nature of this creature has been supplied by the Chippenham fossils, which in all probability are identical in genus, if not in species, with the Anacothetis, described by Münster. One of these extraordinary fossils—the mummy of a cuttle-fish more ancient than the chalk formation and the upper oolites—is represented in fig. 22, 2, reduced to one-sixth from the original in the British Museum. Nine of the arms are preserved, the sclerotic plates of the eyes, the bases of the large lateral fins, the small ink-bag, and the conical shell. This shell, which is chambered internally, like the phragmocone of the Belemnite (fig. 22, 1), has an outer sheath of fibrous structure, one-fourth of an inch thick at the apex, and furnished with two converging ridges on its dorsal side; the external surface, however, is horny (or chitinous), like the pen of the Calamary. These chambered shells occur in great numbers, and are so like the phragmocones of the associated Belemnites, both in structure and proportions, that it is difficult to avoid believing they are in some way related; but hitherto they have only been noticed in the laminated Oxford clay of Wits and the equivalent lithographic shales of Solenhofen.

The Belemnite, on the other hand, is found in all the oolitic and cretaceous strata, from the lowest lias to the upper chalk. In its ordinary imperfect state, it is a cylinder pointed at one end, and truncated or excavated by a funnel-shaped cavity (alveolus) at the other, and has a radiating fibrous structure, with less distinct concentric laminae of growth. But even this "guard," which corresponds simply to the macro of the cuttle-bone, exhibits such remarkable modifications of form, that nearly 100 species have been founded upon no higher evidence. In some Belemnites of half an inch diameter, the guard is scarcely an inch longer than the phragmocone, whilst in others it attains a length of ten inches, and is tubular, as in B. acanthus. Some are fusiform, others laterally compressed; some have a longitudinal groove extending from the apex along the upper or under side, and in others the apex is furrowed laterally as well. The Belemnites of the chalk have been called Belemnitelle (D'Orb.), because they have a slit in the ventral side of the alveolar border of the guard; their external surface also exhibits more distinct traces of vascular impressions.

Specimens of Belemnite have been discovered in which the guard had been broken during the lifetime of the animal; but the broken portions being held together by the investing organized integuments, had been re-united by the deposition of new layers of the fibrous structure peculiar to the guard. Several examples of Belemnites, with the apex injured and healed during life, are preserved in the British Museum. In all perfect Belemnites, the alveolus is occupied by a phragmocone, with tender nacreous walls and septa, and terminating in a minute globular apex, and perforated by a ventral sipuncule (fig. 22, 1). The last chamber is rarely preserved, and appears to have thinned off into a mere horny sheath, with sometimes two pearly bands like knife-blades on the dorsal side. It must have been sufficiently capacious to contain all the viscera. The ink-bag has been very rarely found, and is even smaller than in the last genus, as if in relation to the more greatly developed shell.

The Conotethis (fig. 22, 3) of the Gault has an oblique phragmocone, with a very thin shell, and seems to have been attached to a slender style, like the funnel-shaped appendix of the gladius in the recent sagittated Calamary.

Mr Dana has described, under the name Helicurus fugiensis, a belemnitoid fossil from the "slate" rock of Cape Horn. It is half an inch in diameter, has a thick fibrous guard, and the slender phragmocone terminates in a fusiform spiral nucleus.

Subjoined is a table of the extinct genera of the molluscan province:

For the drawings and most of the facts relating to invertebrate fossils, the writer is indebted to his experienced colleague in charge of that department of the British Museum, Mr S. P. Woodward, F.G.S. It has been shown that every type of invertebrate animal Vertebrata is represented in the superimposed stratified deposits called Cambrian and lower Silurian.

An important work, embodying the labours of the accomplished naturalist and acute observer, Dr Christian H. Pander, has recently been published by the Russian government, descriptive of the fossil fishes of the Silurian formations of that empire. Of some hundred fossils described and beautifully figured in this work, and referred to different genera and species of fishes, from lower Silurian rocks, the writer, after the closest comparison and consideration of the evidence, is disposed to regard only those referred by Pander to the genera *Ctenognathus*, *Corydolodus*, and *Gnathodus*, as having any probable claims to vertebrate rank; and to this admission must be appended the remark, that the parts referred to jaws and teeth may be but remains of the dentated claws of *Crustacea*. With regard to the fossils called "Conodonts," on which the main part of M. Pander's evidence of lower Silurian fishes rests, the following remarks, penned after microscopic examination of specimens kindly submitted to the writer by Sir Roderick Murchison, are applicable to them.

Minute, glistening, slender, conical bodies, hollow at the base, pointed at the end, more or less bent, with sharp opposite margins, might well be lingual teeth of Gastropods, acetalular hooklets of Cephalopods, or teeth of cartilaginous fishes. Against the latter determination is the minute size of the "Conodont" bodies. Their basal cavity doubtless contained a formative pulp, but the proof that the product of such pulp was "dentine" is wanting; the observed structure of the booklet presents concentric conical lamellae of a dense structureless substance, containing minute nuclei or cells.

In some specimens the base is abruptly produced and divided from the body of the booklet by a constriction,—a form unknown in the teeth of any fishes, but presented by certain lingual teeth of Gastropods,—e.g., the lateral teeth of *Sparcella*. In other Conodonts the elongated base is denticulate or serrate, as in the lateral teeth of *Bucephum* and *Chrysodonum*. It is improbable, however, that they belong to any conchiferous toothed Mollusk, the shells of such being wanting in the deposit where the Conodonts are most abundant.

The more minute hooklets have a yellowish, transparent, horny appearance; the larger, perhaps older ones, present a harder whitish appearance. Their analysis by Pander yielded "carbonate of lime," carbonic acid being evolved by application of dilute nitric acid, and oxalic acid producing an obvious precipitate. Some English analysts have believed that the Conodonts yielded a trace of phosphate of lime.

The detached condition of the hooklets, and the integrity of the thin border of the basal pulp-cavity, indicate that they have not been broken away from any of those kinds of attachment to a bone which the minute villiform teeth of osseous fishes would show signs of. The Conodonts have been supported upon a soft substance, such as the skin of a Mollusk or worm, the mucous membrane of a mouth, or throat, or the covering of a proboscis; but to select the teeth of cyclostomous or plagiostomous fishes as the exclusive illustration of the above condition, is to take a partial and limited view of the subject.

In comparing the Conodonts with the teeth of fishes, they present, as Dr Pander recognises, most resemblance with the conical, pointed, horny teeth of Myxinoidea and Lampreys in that class; and the absence of any other hard part in the strata containing the Conodonts tallies with the condition of the cyclostomous skeleton; but not more than it does with the like soft condition of anneliduous worms and naked Mollusks. But the teeth of all known Cyclostomes are... Vertebrata much less slender and are less varied in form than in the Conodonts. Certain lingual plates of Myxinoidea are serrate, but not with a main denticule of much greater length,—such as shown in the form of the Conodont called *Maehatroodus* by Pander. Most cyclostomous teeth are simple, thick cones, with a subcircular base; and every known tooth of a cyclostomous fish is much larger than any of the forms of *Conodon*, which rarely equal half a line in length. This minuteness of size, with the peculiarities of form, supports a reference of the Conodonts rather to some soft invertebrate genus. Certain parts of small Crustacea,—e.g., the pygidium or tail of some minute *Entomostraca*,—resemble in shape the more simple Conodonts; but when we perceive that these bodies occur in thousands, detached, with entire bases, and that any part of the carapace, or shell of an Entomostracan or other Crustacean, has been rarely detected in the lower Silurian Conodont beds, it is highly improbable that they can have belonged to an organism protected by a substance as susceptible of preservation as their own substance. Much more likely is it that the body to which the minute hooklets were attached was as soluble and perishable as the soft pulp upon which the Conodont was sheathed. The writer finds no form of spine, denticule, or hooklet in any Echinoderm, and especially in any soft-bodied one, to match the Conodonts; and concludes that they have most analogy with the spines, or hooklets, or denticles of naked Mollusks or Annelides. The formal publication of these minute ambiguous bodies of the oldest fossiliferous rocks, as evidences of fishes, is much to be deprecated.

**Order I.—Plagiostomi.**

*(Sharks, Rays.)*

Char.—Endo-skeleton cartilaginous or partially ossified; exo-skeleton placoid; gills fixed with five or more gill-apertures; no swim-bladder; scapular arch detached from the head; ventrals abdominal; intestine with spiral valve.

The earliest good evidence which has been obtained of a vertebrate animal in the earth's crust is a spine, of the nature of the dorsal spine of the dog-fish (*Acanthias*), and of the dorsal spines of the extinct *Acanthodii*. It occurs in the most recent deposits of the Silurian period, in the formation called "Ludlow rock." Its discovery is due to Murchison; its determination to Agassiz, who assigns it to a genus of plagiostomous cartilaginous fishes called *Onchus*.

The *Onchus* spines from the Ludlow bone-beds are compressed, slightly curved, less than two inches in length, with no trace at their base of the joint characteristic of the dorsal spines of the "shark-fishes" (Ganoids of the family *Siluridae*), or "file-fishes" (*Balistidae*). The sides of the spine are finely grooved lengthwise, with rounded ribs between the grooves. They are referred to two species—*Onchus Murchisoni*, and *O. semistrigatus*. Sir P. Egerton has lately figured another species from the argillaceous beds near Ludlow, which is more curved, and is armed along the posterior edge; the longitudinal ribs are fine and numerous, but are constricted at intervals, as in the genus *Ctenacanthus*, and become subtuberculate at the base. He deems them significant of a distinct genus of shark-like fishes.

With the dorsal spines of *Onchus* are found petrified portions of skin, tubercular and prickly, like the shagreen of shark's skin, and referred to a genus called *Sphagodus*; also coprolitic bodies of phosphate and carbonate of lime, including recognisable parts of the small Mollusks and Cricetoides which inhabited the sea-bottom in company with the Onchus-fish. No vertebrae, or other parts of the endoskeleton of a fish, have been discovered, unless the fragments of a calcified bar, with tooth-like processes, called *Plectrodus*, be truly jaws with teeth. They resemble, however, parts of the pioneer-claws of Crustaceans, as well as of the jaws and teeth of fishes, and do not indicate that class so satisfactorily as the *Onchus* spines and *Sphagodus* shagreen. Yet the denticles are confluent with an outer ridge of the bone, according to the "pleurodont" type, and consist of separated large teeth, with minute serial teeth in the interspaces; and the large teeth are grooved longitudinally.

If the Plectrodonts be jaws with anchylosed teeth, they belong to an order distinct from the Plagiostomi. If they should belong to any of the fishes indicated by the dorsal spines and shagreen skin, a combination of characters would be exemplified not known in other formations or in any existing fishes.

No detached teeth unequivocally referable to a plagiostomous genus, nor any true ganoid scale of a fish, have yet been found in the formations that have revealed these earliest known evidences of vertebrate animals. What, then, it may be asked, were the conditions under which so immense an extent, as well as amount, of sediment was deposited,—including chambered Cephalopods, Gastropods, Lamellibranchs, Brachiopods, various and large trilobite Crustaceans, with Crinoids, Polypes, and Protozoa,—that precluded the preservation of the fossilizable parts of fishes, if that class of vertebrate animals had existed in numbers, and under the variety of forms, comparable to those that people the ocean at the present day? Bonitos now pursue flying-fishes through the upper regions of an ocean as deep as any known part of the Silurian seas of which the deposits afford an idea of greatest depth. If fishes of cognate habits with the present deep-sea fishes, under whatever difference of form such Silurian fishes may have been manifested, had really existed, we might reasonably expect to find the remains of some of the countless generations that succeeded each other during that vast and indefinite period, sufficing for the gradual deposition of sedimentary beds of thousands of feet in depth or vertical thickness.

The evidences of plagiostomous fishes afforded by fossil spines will be here pursued. In most of the existing cartilaginous fishes of this order the defensive spine which stands erect in front of the dorsal fin is smooth; such is the case in the dog-fishes (*Spinacidae*) in which each dorsal fin is fringed with a spine. In the Port-Jackson sharks (*Cestraciontidae*) the spine in front of each dorsal is bony, and is armed along its hinder or concave border with bent spines. The fin is connected with this border, and its movements are regulated by the elevation or depression of the spine during the peculiar rotary action of the body of the shark. This action of the spine in raising and depressing the fin resembles, Dr Buckland has remarked, that of the moveable or jointed mast, raising and lowering backwards the sail of a barge. But their more obvious use, in the small Plagiostomes possessing such spines, is as defensive weapons against the larger and stronger voracious fishes.

Certain bony fishes are similarly armed,—e.g., sticklebacks (*Gasterosteus*), sheat-fishes (*Siluridae*), trigger-fishes (*Balistes*), and some species of snipe-fishes (*Fistularidae*). In the latter family the *Centriscus humerosus* (fig. 23) shows a dorsal spine, denticulated behind, as in the Cestracionts, but the base of the spine in bony fishes is peculiarly modified for... Vertebrata, articulation with another bone. In the Plagiostomes the base of the spine is hollow, becomes thin and smooth when the body of the spine is sculptured, and is in the recent fish implanted in the flesh.

The following genera of plagiostomous fishes have been founded on the fossil spines, or "ichthyodorulites," which have been discovered in the "Devonian" or "Old Red Sandstone series." Onchus (represented by O. semistriatus, O. heteroeyrus), Dimeracanthus, Haplacanthus, Narceodes, Naudus, Byssacanthus, Cosmocanthus, Homacanthus (fig. 24), Ctenacanthus, Ptycanthus, Climatius, Parexus, Odontacanthus, and Pleuracanthus.

The genus Homacanthus is founded on small compressed spines, with fine recurved teeth on the back edge, and longitudinal striæ on the sides. Specimens of Homacanthus arcuatus (fig. 24) have been found in Devonian formations near St Petersburg.

The carboniferous series of formations includes the mountain limestone, millstone grit, and the coal measures (see fig. 1). In this series the genus Onchus is still represented by the O. subcatus, O. rectus, and O. subulatus; and the genus Homacanthus, by H. macroodus and H. microodus, from the carboniferous limestone of Armagh. Ptycanthus, Ctenacanthus, and Pleuracanthus are also forms common to the Devonian and carboniferous periods. The spine of the latter genus is denticulated along both margins, a structure which is presented, in existing Plagiostomes, only by species of the ray family: Pleuracanthus, therefore, as Agassiz concludes, may offer the earliest example of the flat form of cartilaginous fish, which is represented by the sting-rays (Trigon, Myliobates) in the present seas. The ichthyodorulite (ichtyos, a fish; dora, a spear; lithos, a stone) here selected to illustrate this fossil, is a portion of the spine of the Pleuracanthus levisimus (fig. 25), from the carboniferous beds near Dudley. The other plagiostomous genera based upon fossil spines from the coal formations are—Oracanthus, Gyracanthus, Nemacanthus, Cosmocanthus, Leptacanthus, Homacanthus, Trygichthys, Asteroptychius, Phycogonemus, Sphenacanthus, Platycanthus, Diploacanthus, Eristamacanthus, Orthacanthus, Cladacanthus, Lepracanthus.

Immediately above the coal-measures lie a variable series of sands and clays of different colours, including the coal plants: above this, a marl-slate in thin layers, containing scanty evidences of fishes; but these are more abundant and instructive in the supercumbent magnesian limestone, in which formation near Belfast ichthyodorulites of the genus Gyropristis (Ag.), have been found. Above this are the penean red sandstones, in which, at Westoe, have been found fossil spines closely allied to, if not identical with, the Gyracanthus for-

The foregoing formations constitute the Vertebrata, uppermost of the palaeozoic series called "Permian," from the Russian province in which these strata are most extensively developed. Their relative position is known by the term "magnesian limestone" in the "Table of Strata," fig. 1.

The superimposed strata, marked "new red sandstone," includes also a varied series of red and white sands, marls, and conglomerates, forming collectively the system called "triassic." The ichthyodorulites of this system are referable to the genera Nemacanthus, Leiacanthus, and Hybodus. In the "lias," which is the oldest or lowest of the great "oolitic" system, the dorsal spines of the genus Hybodus (fig. 26), are the largest and most abundant; this genus, however, is represented by detached teeth in the keuper and muschelkalk members of the "trias." The lias formations give evidence that the dorsal spines and fins of Hybodus were two in number; and the genus is shown, both by the structure of the spine and the form of the teeth, to have had its nearest affinities with the Cestracion amongst existing Plagiostomes. Hybodus continued to be represented by successive and varying specific forms up to, and including, the cretaceous period. Hybodus is therefore a genus of cartilaginous fishes eminently characteristic of the secondary or mezozoic period in palaeontology, and ranges through every formation of that period. The specimen selected for the illustration of the dorsal spine of Hybodus is that of the H. subcarinatus, from the Wealden of Tilgate Forest.

Large fossil spines, longitudinally grooved, have been found associated with the teeth of the extinct cestraciont genus of the chalk called Ptychodus.

In the tertiary formations, the fossil spines present for the most part the generic characters of those of existing Plagiostomes,—e.g., Spinax, Trigon, and Myliobates; but one form, found in the eocene beds near Paris, is the type of the extinct genus Aulacanthus of Agassiz.

The teeth of the plagiostomous fishes,—viz., sharks (Squalide), rays (Raidere), and Cestractious, are described in the article ODONTOLOGY (vol. xvi., pp. 422-424). From the account and figures there given, it will be seen that the teeth are very numerous, and that, being attached only by ligament to the membrane of the mouth, they must soon fall off; in the decomposition of the dead fish, become scattered abroad by the movements of the body through the action of the waters, and sink into the sediment.

FAMILY I.—CESTRACONTIDAE.

(Port-Jackson Shark.)

The existing genus which has thrown most light upon the fossil teeth which have thus become imbedded in the oceanic deposits of the palaeozoic and mezozoic periods, is the Cestracion, now restricted to the Australian and Chinese seas, where it is represented by two or three species, and suggests the idea of a form verging towards extinction. It formerly flourished under a great number of varied generic or family modifications, represented by species, some of which attained dimensions far exceeding the largest known living Cestracions. The dentition of these fishes is adapted to the prehension and mastication of crustaceous and testaceous animals; they are of a harmless, timid character; and have the before-described denticulate dorsal spines given to them as defensive weapons. Figure Vertebrata. 27 gives a side view of the upper and lower jaws of the "Port-Jackson shark," showing the oblique disposition of the large crushing teeth, which cover like a pavement the working borders of the mouth. Figure 28, p. 424, article Odontology, gives a view of the crowns of the teeth of the upper jaw of the same species. From their mode of attachment, these teeth would become detached from the jaws of the dead fish, and dispersed in the way above described; and it is by such detached fossil teeth that we first get dental evidence of the Cestraciont family in former periods of the earth's history.

The teeth of the Hybodonts are conical, but broader and less sharp than those of true sharks. The enamel is strongly marked by longitudinal grooves and folds. One cone is larger than the rest, and called the "principal;" the others are secondary. In one genus (Cladodus, Ag.), the secondary cones go on enlarging as they recede from the "principal;" and teeth of this genus, referred by Eichwald to the Hybodus longiconus, have been discovered in the old red sandstone in the vicinity of Petersberg.

In the Orodus, the cones are more compressed, trenchant, and distinct from the body of the tooth than in Hybodus; but they present a principal and secondary cones. Fig. 28 is a tooth of the Orodus cineratus (Ag.), from the carboniferous beds near Bristol. The O. porosus and O. compressus are from deposits of similar age near Armagh.

If fig. 29 be compared with fig. 28, p. 424, vol. xvi., it would seem as if the several teeth of each oblique row in Cestracion had been welded into a single dental mass in Cochliodus, the proportions and direction of the rows being closely analogous. Whether in Cochliodus there were any small anterior prehensile teeth, is hypothetical; the large crushing dental plates must have been admirably adapted to crack and bruise the shells of molluscs and Crustaceans. The Cochliodus contortus (Ag.) (fig. 29) has been found in the carboniferous formations near Bristol and Armagh, and the genus is peculiar to that geological period.

Teeth referable to the genus Hybodus occur in all the secondary rocks from the trias to the chalk inclusive.

A form of tooth which more closely resembles the crushing-teeth of Cestracion, is that on which the genus Aerodus is founded, and which also ranges from triassic strata to the upper chalk of Maestricht. The species here selected (fig. 30) is the Aerodus nobilis, from the lias of Lyme Regis. The upper figure shows the grinding surface, which, from its finely and transversely striated character and dark colour, has suggested to the quarrymen Vertebrata, the name of "fossil leeches." The older fossilists regarded these teeth as petrified Vermes; but the structure, as shown by the microscope, is closely similar to that of the teeth of Cestraciont. Portions of the jaw of the Aerodus have been discovered which show that these teeth were arranged, as in Cestraciont, in oblique rows, with at least seven teeth in each row. Aerodus lateralis is a muschelkalk fossil, A. hirudo a Wealden, and A. transversus a cretaceous fossil. No tooth referable to the genus has been found in any tertiary stratum.

The genus Ptychodus is founded on teeth usually of large size, and of a more or less square form (fig. 31). The crown is deeper than the root, which is obtuse and truncate. The enamelled summit of the crown is granulate at the margin, and raised in the middle into an obtuse eminence, disposed in large transverse, parallel, sometimes wavy and rather sharp ridges. With teeth of this form are sometimes found others of smaller size, with more convex rounded crowns, doubtless forming the extremes of the multilateral pavement which, as in modern sharks and rays, covered the broad jaws of the Ptychodonts. Large dorsal spines have been found so associated with the above-described teeth as to indicate the affinity of the Ptychodus to the Cestraciont family of sharks. All the Ptychodus latissimus specimens and species referable to this genus have been found in the cretaceous strata.

Family II.—Squalidae.

(Sharks.)

The well-marked, saw-shaped tooth (fig. 32), so closely resembles the lower jaw-teeth of the sharks, called "gri-suts" by the French (Notidanus, Cuv.), as to be referred to that genus by Agassiz. Such teeth nevertheless occur in strata of oolitic age (Notidanus Minsteri, Ag., fig. 32). Other species,—e.g., N. prictinatus,—are found in the chalk of Kent; and N. serratissimus, in the eocene clay at Sheppy.

The tooth (fig. 33) on which Agassiz has founded the genus Corax indicates, by its close resemblance to those of Carcharias, its relationship with the true sharks (Squalidae). Most of the species of Corax, including C. falcatus, are cretaceous; a few are tertiary: all are extinct.

Another form of shark's tooth, deeply notched at one margin, and with the rest of the border finely denticulate, resembles more that of the "Topes" or gray sharks (Galeus, Cuv.), and is referred by Agassiz to the genus Galeocerdo. The species are found in both the cretaceous and tertiary formations; Galeocerdo adunca (fig. 34) is from the miocene of Europe and America. In the same tertiary series are found the teeth of the Hemipristis serra, Ag. (fig. 35).

Odontaspis (Ag.), presents a form of tooth most like that in the blue sharks (Lamna) of the present seas. Species of Odontaspis occur in the cretaceous and tertiary beds. The O. Hopei (fig. 36) is from the London clay of Sheppy. It indicates a destructive and formidable species of shark.

---

1 See Owen' Odontography vol. i. p. 54, pls. 14 and 15. Vertebrata. With these fossil teeth of sharks are found, though sparingly, in both the cretaceous and tertiary beds, petrified bodies of vertebrata, showing by their extreme shortness in comparison with their breadth, by their bi-concavity,

and the fissures on the external surface (as shown on the upper figure of cut 37), that they belonged to a shark closely allied to the Porbeagle, (Lamna, Cuv.)

Family III.—Raide.

(Rays.)

Fossil evidences of this peculiar family of cartilaginous fishes have been discovered in oolitic, cretaceous, and tertiary formations, and consist of defensive spines, dermal tubercles, and teeth, but chiefly the latter. The most peculiar and distinctive modifications of the dental system, presented by the eagle-rays (Myliobatidae) are unequivocally shown by fossils of the tertiary formations, and have not been found in earlier strata.

The form, structure, and arrangement of the dental plates of the existing Myliobates are described and illustrated in the article ODONTOLOGY, p.423, figs. 26 and 27. To this genus, as now restricted, certain fossils from the London clay of Sheppay (Myliobates totiopicus, Ag., fig. 38) belong.

In Zygobatus (fig. 39), the middle series of teeth is less broad; and a still narrower series is interposed between it and the small lateral teeth. Existing rays showing this modification are found in Brazilian seas; fossil teeth of this genus, e.g., Zygobatus Woodwardi, Ag. (fig. 39), occur in the tertiary crag (probably miocene) of Norfolk, and in the miocene mollasse of Switzerland.

When the teeth form broad transverse undivided plates, as in fig. 40, they characterize the genus Zetobatus. Fossils of this genus occur in the English eocenes and the Swiss mollasse.

In the "crag" of Norfolk and Suffolk, and in marine pliocene beds, fossils have been found which closely resemble the osseous and spinigerous plates that beset the skin of the ray, thence called "thornback" (fig. 41), and which indicate the existence of a species allied to the Raia clavata.

Thus we obtain evidence of fishes of the plagiostomous Vertebrata order in the marine deposits of every formation from the upper Silurian beds to the present period. But none of the palaeozoic fossils are referable to any existing genus. A few only of the mezoic Plagiostomes, and those chiefly from the chalk are so determinable. Most of the secondary Plagiostomes belong or are allied to a family (Cestraciontidae), now nearly extinct; the evidence of the generic forms of Plagiostomes characteristic of the present time become common only in the tertiary periods. No fossil species is the same with any existing one.

Order II.—HOLOCEPHALI.

(Chimaeroid Fishes.)

Char.—Jaws bony, traversed and encased by dental plates; endo-skeleton cartilaginous; exo-skeleton as placoid granules; most of the fins with a strong spine for the first ray; ventrals abdominal; gills laminated, attached by their margins; a single external gill aperture.

To judge from the paucity of existing representatives of this order of cartilaginous fishes, it would seem, like the Cestracions, to be verging towards extinction. One genus (Chimera, Linn.) is founded on a single known species of the northern seas called "king of the herrings" (Chimera monstrosa); another genus (Callorhynchus of Gronovius) is represented by two known species in the Australian and Chinese seas. The only parts of chimaeroid fishes likely to be fossilized are the jaws and spines. The bony and dental substances are so combined in the more or less beak-shaped jaws, that they characterize the order, and are never found separate. It is chiefly on such fossil mandibles, and portions of them, that the evidence of the Holoccephali in former geological periods rests. These singular fishes ranged, under different generic and specific modifications, from the bottom of the oolitic series to the present period.

Genus Ischichonax, Egerton.—Of this genus, I. Johnsoni is from the lias of Dorsetshire; I. Egertoni from the Kimmeridge of Shotover; and I. Townshendi, a magnificent species, from the Portland stone. Two species (I. Agassizii and I. brevirostris) are from the cretaceous beds; at which period the genus appears to have perished.

Genus Ganodus, Egerton (including Ganodus and Psittacodus of Agassiz).—This genus is exclusively represented by species from the oolitic slate of Stonesfield,—e.g., G. Bucklandi, G. Colei, G. Oscenii.

Genus Edaphodus, Egerton (including Edaphodon and Passalodon of Buckland).—The large E. Sedgwickii is from the greensand near Cambridge; the still larger E. gigas from the chalk of Kent and Sussex. The ichthyodorulite called Psittacodus Mantelli by Agassiz may be the dorsal spine of this species. Three species, including the E. Bucklandi, are found in the eocene of Bagshot and Bracklesham; and one species (E. helveticus) is from the mollasse of Switzerland.

The genus Elasmodus, Egerton, is exclusively represented by species,—e.g., E. Hunteri, from the London clay of Sheppy.

Order III.—GANOIDEI.

Char.—Endo-skeleton in some osseous, in some cartilaginous, in some partly osseous and partly cartilaginous; exo-skeleton formed by enamelled bones; fins usually with the first ray a strong spine. Sub-Order I.—PLACOGANOIDEI.

(Char.—Endo-skeleton cartilaginous, or retaining the notochord; head and more or less of the trunk protected by large ganoid, often reticulated, plates; heterocercal.

The last term signifies a form and structure of tail illustrated by fig. 42, and to be seen in the sharks, dog-fishes, half of the trunk are defended by ganoid plates,—i.e., plates Vertébrata, composed of a hard bone coated with enamel; those of the trunk forming a buckler composed of a back-plate (fig. 43) and breast-plate (fig. 44), articulated together at the sides. The rest of the trunk was defended by small ganoid scales, flexible, like scale-armour, and bore a small dorsal fin (fig. 43, d'), and a terminal heterocercal fin, very rarely displayed

Fig. 42. Heterocercal Tail (Lepidosteus osseus).

and sturgeons of the present day: it results from a prolongation of the vertebral column into the upper lobe dn, producing an unsymmetrical form of the caudal fin, which is contrasted with the symmetrical form of the same fin presented by most fishes of the present day, and illustrated by the skeleton of the perch, in art. Ichthyology, fig. 10, p. 208, vol. xii., in which the vertebral column terminates at the middle of the base of the caudal fin. There are a few exceptional intermediate forms and structures of this fin.

The fossil remains of the singular fishes of the extinct order Placoganoidae were first discovered about 1813, in formations of the "old red" or Devonian age in Russia, and are preserved in museums at St Petersburg and Dorpat. The relation of these specimens to the class of fishes was first announced by Professor Asmus, and shortly after, the generic names Asterolepis and Bothriolepis were invented by Professor Eichwald, to express certain modifications of the external surface of portions of the ganoid plates, subsequently recognised as constituting the buckler of the fore-part of the extinct fishes. In September 1840 Mr Hugh Miller submitted to the geological section of the British Association, at Glasgow, the first discovered specimens affording a recognisable idea of the form of one of these "old red" fishes, and for this form Professor Agassiz assigned the generic name Pterichthys (pteron, a wing, ichthys, a fish). Although, therefore, the term Asterolepis had been attached to a fragment of the cuirass of this fish a few months previously, yet, as no recognisable generic characters were associated with such name, and as Asterolepis has been applied also to another genus,—the Heterostius of Asmus,—the example of British palaeontologists will be here followed, in retaining the name Pterichthys for the genus. "Of all the organisms of the system," wrote the lamented Hugh Miller in his work on the Old Red Sandstone, "one of the most extraordinary, and the one in which Lamarck would have most delighted, is the Pterichthys, or winged fish, an ichtyolite which the writer had the pleasure of introducing to the acquaintance of geologists nearly three years ago (1840), but which he first laid open to the light about seven years earlier" (1833).

Genus Pterichthys (fig. 43).—The head and the anterior in fossil specimens. The pectoral spines, c, are formed of ganoid material, like the buckler. The armour of the head, or helmet, appears to have been articulated by a moveable joint to the trunk-buckler. One of the few existing ganoid fishes (Lepidosteus) is remarkable for the degree in which the head moves upon the trunk. The component dermal plates of the helmet correspond in some measure with the position of the cranial bones in osseous fishes, but not sufficiently to sanction the application to them of corresponding names. They are indicated by figures in the cut 43: 2 is the front terminal or rostral plate; it is followed in the median line by four other plates in the following order:—4, premedian; 6, median; 8, postmedian; 10, nuchal; 3 is the marginal, and 7 the

---

1 This term is used by other German writers as the equivalent of the Placoides of Agassiz, which answers to the Plagiostomes of Cuvier.

2 Ibid., t. vii., p. 78, communicated March 13th, 1840. Dr Fleming had recognized certain fossil scales as those of fishes in the "Old Red" of Fifeshire, in 1827. Vertebrata: postmarginal; 5 is the prelateral, and 9 the postlateral.

The dorsal shield of the trunk-cuirass is composed of two mid-plates and two on each side. 12 is the "dorsomedian," 14 the post-dorsomedian; 11 is the dorsolateral, 13 the post-dorsolateral.

The ventral shield (fig. 44) consists of one mid-plate and three or four side-plates: 15 is the pre-ventrolateral, 19 the ventrolateral, 21 the post-ventrolateral; the small supplementary plate marked 17 is sometimes confluent with 19; 16 is the centromedian plate; its margins are bevelled off and overlapped by the lateral plates.

In the first-formed imperfect specimens of Pterichthys, the ventral shield was deemed the dorsal one. Sir P. Egerton has the merit of having first discerned its true position.

The pectoral spines are long and slender, and consist of two principal segments, both defended by finely tuberculated ganoid plates, like those of the head and trunk. From their form, they would seem to have served to aid the fish in shuffling along the sandy bottom or bed, if left dry at low-water. The fins attached to the flexible part of the body indicate a certain power of swimming, though not with any great rapidity. The jaws are small, and possess confluent denticles.

The type-species is the Pterichthys Milleri; others have been based upon proportions of the cuirass, of the pectorals, and the tail; all are from the "old red," and the great majority have been found in the Devonian strata of Ross-shire, Caithness, and other Scotch localities.

Genus Cephalaspis (hephale, head, aspis, buckler).—In this genus the posterior angles of the shield-shaped helmet are produced backwards in a pointed form, giving to the head the form of a "saddler's knife;" in other respects the genus closely resembles Pterichthys.

Mr D. Page has recently acquired specimens of Cephalaspis from Lanarkshire tile-stones, forming the base of the Devonian system, which show a dorsal fin, pectoral fins, and a large heterocercal fin, besides a well-marked capsule of the eye-ball.

Genus Cocosteus (kokkos, berry, osteon, bone).—If a heterocercal fin were added in outline to the restoration of the fish of this genus, given in fig. 45, a correct idea would be given of the "old red" fish, which, in the progress of its reconstruction, has suggested so many strange notions of its nature and affinities.

The helmet and cuirass are firmly united, and there is no trace of the jointed appendages, like pectoral fins, which characterize Pterichthys. The unprotected part of the trunk shows an ossification of the neural and haemal spines, and of their appendages, the rays of a "dorsal" and "anal" fin; and by the analogy of Cephalaspis, the tail was most probably terminated by an unequal-lobed fin. The lower jaw is composed of two rami, loosely connected at the symphysis; so that, when displaced, as in fig. 46, 24, and as commonly in crushed fossil specimens, they gave the notion of the fish being provided with laterally-working jaws, like those of the lobster. But, in reality, the jaw worked vertically upon a fixed upper jaw; both jaws being provided with from ten to twelve teeth on each side, anchylosed to the bone; each ramus of the jaw consisting of a premandibular (figs. 45 and 46, 24) and a postmandibular (fig. 45, 18) element.

An under-view of the cephalothoracic buckler of Cocosteus, according to Dr Pander's restoration, is given in fig. 46, showing the sutures of most of the cephalic plates, and the external surface of the plates of the plastron. 9, Rosstral plate; 7, premedian; 5, median; 8, prelateral; 6, lateral; 16, postmandibular; 24, premandibular; 15, precentromedian; behind the lozenge-shaped centromedian, and on each side, are (22) the pre-ventrolateral and (23) the post-ventrolateral. The same figures mark the above plates in the side view (fig. 45), with the addition of (12) the dorsomedian and (14) the post-dorsomedian.

The blank space between the neural (n) and haemal (h) spines of the fossil endo-skeleton indicates the position of the soft "notochord" (c), which has been dissolved away. The cylindrical gelatinous body, so called,—in Latin ekorda dorsalis,—pre-exists to the formation of the bony bodies of the vertebrae in all vertebrate animals; and the development of these bodies seems never to have gone beyond this embryonal phase in any palaeozoic fish; such fishes are accordingly termed "notochordal," retaining the notochord.

There are but two genera of existing fishes which manifest, when full grown, such a structure, associated with Vertebrata, ossified peripheral elements of the vertebrae, viz., the Protopterus of certain rivers of Africa, and the Lepidosiren of certain rivers of South America. Those fishes alone would, if fossilized, present the appearance of the vertebral column shown in fig. 45, and which characterizes all the oolitic fossil ganoid fishes (see figs. 54 and 55). It is a strong illustration of a principle of "progression," this persistence in palaeozoic and most mesozoic fishes of an embryonic vertebral character, transitory in nearly all existing fishes.

The external "ganoid" surface of the buckler plates of Coccosteus is ornamented with small hemispherical tubercles; whence the generic name, signifying "berry-bone." The similarity of this ornamentation to that of the plates of the buckler in some Tortoises, led to the belief, when the coccostean plates were first found, of their being evidence of the chelonian genus Trionyx in Devonian beds. Passing notions also got into print of the crustaceous affinities of Coccosteus; whence the trivial name of the type-species decipiens, or the "deceiving" Coccosteus.

Strange as seem the forms and structure of the placogonoid fishes of the "old red" period, there are not wanting existing species which throw much truer light on their nature than any existing Chelonia or Crustacea. The singular little family of "trunk-fishes" (Ostracionidae) shows Vertebrata species in which the body is inclosed in a more or less quadrangular cuirass, composed of sutured-articulated ganoid plates, which are usually tuberculated on the external surface, and with the angles prolonged into spines in some species, like those of the helmet of Cephalaspis. The caudal part of the trunk protrudes from the back opening of the cuirass, as in Coccosteus and Pterichthys, and ossification of the endo-skeleton is incomplete. The species of this family are for the most part natives of seas of tropical or warm temperate latitudes.

In another family of existing fishes, called "Siluroids," there are species in which the broad cranial bones, connate with dermal ossifications, form a helmet to the head, whilst one or two dermal spine-bearing bones combine to form the part called "buckler" by Cuvier. In the genus Doras, the lateral line is armed with bony ganoid plates; and in Callichtys, these biserial plates are developed so as to incase the whole body. But generally, as in Pimelodus, the hinder muscular part of the trunk is undefended, as in Coccosteus. The ganoid plates of the head and back shields are fretted with rows or ridges of confluent tubercles, radiating from the centre to the circumference of the plate, whilst the inner surface is smooth, as in Coccosteus (fig. 46); and, moreover, the dorsal plate in existing Siluroids sends down a median ridge from its inner surface, like that from the "dorsal-median" plate in Coccosteus. The point of resemblance to be mainly noticed, however, is the contrast furnished by the powerful armature of the head and back with the unprotected nakedness of the posterior portions of the creature—a point specially noticeable in Coccosteus, and apparent also, though in a lesser degree, in some of the other genera of the old red, such as the Pterichthyes and Asterolepis. "From the snout of the Coccosteus down to the posterior termination of the dorsal plate the creature was cased in strong armour, the plates of which remain as freshly preserved in the ancient rocks of the country as those of the Pimelodi of the Ganges on the shelves of the Elgin museum; but from the pointed termination of the plate immediately over the dorsal fin to the tail, comprising more than one-half the entire length of the animal, all seems to have been exposed, without the protection of even a scale; and there survives in the better specimens only the internal skeleton of the fish and the ray-bones of the fins. It was armed, like a French dragoon, with a strong helmet and a short cuirass; and so we find its remains in the state in which those of some of the soldiers of Napoleon's old guard, that had been committed unstripped to the earth, may be dug up in the future on the fatal field of Borodino, or along the banks of the Danube or the Wap. The cuirass lies still attached to the helmet, but we only find the naked skeleton attached to the cuirass. The Pterichthys to its strong helmet and cuirass added a posterior armature of comparatively feeble scales, as if, while its upper parts were shielded with plate-armour, a lighter covering of ring or scale armour sufficed for the less vital parts beneath. In the Asterolepis the arrangement was somewhat similar, save that the plated cuirass was wanting. It was a strongly-armed warrior in slight scale-armour; for the disproportion between the strength of the plated head-piece and that of the scaly coat was still greater than in the Pterichthys. The occipital star-covered plates are, in some of the larger specimens, fully three-quarters of an inch in thickness, whereas the thickness of the delicately-fretted scales rarely exceeds a line.

"Why this disproportion between the strength of the armature in different parts of the same fish should have obtained, as in Pterichthys and Asterolepis, or why, while one portion of the animal was strongly armed, another portion..." Vertebrata should have been left, as in Coccoctes, wholly exposed, cannot of course be determined by the mere geologist. His rocks present him with but the fact of the disproportion, without accounting for it. But the natural history of existing fish, in which, as in the Pimelodii, there may be detected a similar peculiarity of armature, may perhaps throw some light on the mystery. In Hamilton's Fishes of the Ganges, the habitats of the various Indian species of Pimelodii, whether brackish estuaries, ponds, or rivers, are described, but not their characteristic instincts. Of the Silurus, however, a genus of the same great family, I read elsewhere that some of the species, such as the Silurus Glanis, being unwieldy in their motions, do not pursue their prey, which consists of small fishes, but lie concealed among the mud, and seize on the chance stragglers that come in their way. And of the Pimelodus guttatus, a little strongly-armed fish with a naked body, I was informed by Mr Duff, on the authority of the gentleman who had presented the specimens to the Museum, that it burrowed in the holes of muddy banks, from which it shot out its armed head, and arrested, as they passed, the minute animals on which it preyed. The animal world is full of such compensatory defences; there is a half-suit of armour given to shield half the body, and a wise instinct to protect the rest. Now it seems not improbable that the half-armed Coccoctes, a heavy fish, indifferently furnished with fins, may have burrowed, like the recent Silurus Glanis or Pimelodus guttatus, in a thick mud, of the existence of which in vast quantity, during the times of the old red sandstone, the dark Caithness flagstones, the ferruginous breccia of Strathpeffer, and the gray stratified clays of Cromarty, Moray, and Banff unequivocally testify; and that it may have thus not only succeeded in capturing many of its light-winged contemporaries, which it would have vainly pursued in open sea, but may have been enabled also to present to its enemies, when assailed in its turn, only its armed portions, and to protect its unarmed parts in its burrow.

Sub-Order 2.—LEPIDOGANOIDEI.

Family L—Dipteridae.

This family includes a few heterocercal fishes with a double anal as well as dorsal fin. The head is large and flattened; the teeth subequal; the scales perforated by small foramina; the notochord persistent.

In the genus Dipterus (fig. 47), the two dorsals, d1, d2, are opposite the two analts, a1, a2; the ventrals, v, are in advance of the first anal and first dorsal. The Dipterus macrolepidotus is characterized by the large size of its scales. Its remains are found in the old red sandstone of many localities of Scotland and England.

In the allied genus Diplopterus the vertical fins are opposite, but the dorsals are wider apart, and the teeth are larger and fewer. Four species have been recognised in the "old red" of Gamrie, Orkney, and Lethenbar, Ross-shire. Two species occur in the carboniferous series.

In the genus Osteolepis the vertical fins are alternate in position, the first dorsal being near the middle of the back. All the species of this genus are from the "old red."

Family II.—Acanthodii.

The species of this family are characterized by their very small scales; they are heterocercal and notochordal. There is a strong spine in front of each fin. The head is large; the orbits approximate; the mouth wide, and opening obliquely upwards, so that they have somewhat the aspect of the Uranoscopii. The principal genera are from the old red sandstone, and are as follows:—Cheiracanthus, with a single dorsal situated in front of the anal; Cheiropleis, in which the dorsal is situated behind the anal; and Diplocanthus (fig. 48), in which there are two dorsals.

The Diplocanthus striatus is found in the "old red" of Cromarty. In fig. 48, as in the other figures, p is the pectoral fin, d the dorsal, v the ventral, a the anal, and c the caudal. In this species the upper lobe of the caudal is much prolonged. The fin-spines in the Acanthodii were, like those of the recent dog-fish (Spinax), simply imbedded in the flesh, with their base, as it were, unfinished, not provided, as in the Silurids and other modern bony fishes, with a joint-structure. Some species of Acanthodii existed in the seas of the carboniferous period.

Family III.—Coelacanthi.

The species of this family are characterized by the holowness of the rays or spines; whence the name. The caudal fin has a peculiar structure, the vertebral column being continued into and beyond its middle part, supporting a kind of slender appendage between the two normal lobes. The species of the genus are most abundant in the Devonian and carboniferous formations; but some occur in oolitic and even cretaceous beds; but all became extinct before the tertiary epoch.

Glyptolepis had a heterocercal tail, with rounded scales, smooth externally, and with radiating compartments internally. The G. microlepidotus, of which a magnified view of some scales is given in fig. 49, occurs in the old red sandstone of Scotland and England.

Family IV.—Holoptychiidei.

The type-genera of this family were first recognised and characterized by the fossil scales, under the name Holoptychius (Ag.), and by the fossil teeth, under the name Rhizodus (Ow.). They include species which have left their remains in the "old red" and the coal measures. They are nearly allied to the Coelacanthians, having, like them, but partially ossified bones and spines, the interior of which retained their primitive gristly state, and appear hollow in the fossils. The head was defended by large externally sculptured and tuberculate ganoid plates. The teeth consist of two kinds,—small

---

1 Hugh Miller, Rambles of a Geologist, p. 288. Vertebrata: serial teeth, and large laniary teeth,—at long intervals; both kinds showing the "labyrinthic" structure at their base, which is ankylosed to the jaw-bone. Large fossil teeth, with the more complex "dendritic" disposition of the tissues, characterize a genus (Dendrodus), most probably of the Holoptychian family.

The generic term Rhizodus is now retained for the Holoptychians of the coal measures which have more robust and obtuse serial teeth, and longer, sharper, and more slender laniaries, exemplified by the R. Hibberti. Species of true Holoptychius,—e.g., H. giganteus (Ag.), H. nobilissimus (Ag.), occur in the old red sandstone. A most noble specimen of the latter species, 2 feet 6 inches in length, discovered in the old red sandstone at Clashbinnie, near Perth, is now in the palaeontological series of the British Museum. It is chiefly remarkable for the size and bold sculpturing of the ganoid scales (fig. 50).

"The amount of design exhibited in the scales of some of the more ancient Ganoids,—design obvious enough to be clearly read,—is very extraordinary. A single scale of Holoptychius nobilissimus, fast locked up in its red sandstone rocks, laid by, as it were, for ever, will be seen, if we but set ourselves to unravel its texture, to form such an instance of nice adaptation of means to an end as might of itself be sufficient to confound the atheist. Let me attempt placing one of these scales before the reader in its character as a flat counter of bone, of a nearly circular form, an inch and a half in diameter, and an eighth part of an inch in thickness; and then ask him to bethink himself of the various means by which he would impart to it the greatest possible degree of strength. The human skull consists of two tables of solid bone, an inner and an outer, with a spongy cellular substance interposed between them, termed the diplos; and such is the effect of this arrangement, that the blow which would fracture a continuous wall of bone has its force broken by the spongy intermediate layer, and merely injures the outer table, leaving not unfrequently the inner one, which more especially protects the brain, wholly unharmed. Now, such also was the arrangement in the scale of the Holoptychius nobilissimus. It consisted of its two well-marked tables of solid bone, corresponding in their dermal character, the outer to the cuticle, the inner to the true skin, and the intermediate cellular layer to the rete mucosum; but bearing an unmistakable analogy also, as a mechanical contrivance, to the two plates and the diplos of the human skull. To the strengthening principle of the two tables, however, there were two other principles added. Cromwell, when commissioning for a new helmet, his old one being, as he expresses it, "ill set," ordered his friend to send him a "fluted pot;" i.e., a helmet ridged and furrowed on the surface, and suited to break, by its protuberant lines, the force of a blow, so that the vibrations of the stroke would reach the body of the metal deadened and flat. Now, the outer table of the scale of the Holoptychius was a "fluted pot." The alternate ridges and furrows which ornamented its surface served a purpose exactly similar with that of the flutes and fillets of Cromwell's helmet. The inner table was strengthened on a different but not less effective principle. Now, the inner table of the scale of the Holoptychius was composed on this principle, of various layers or coats, arranged the one over the other, so that the fibres of each lay at right angles with the fibres of the others in immediate contact with it. In the inner table of one scale I reckon nine of these alternating, variously-disposed layers; so that any application of violence, which, in the lath-splitter, would run lengthwise along the grain of four Vertebrae, of them, would be checked by the cross grain in five. In other words, the line of the tear in five of the layers was ranged at right angles with the line of the tear in four. There were thus in a single scale, in order to secure the greatest possible amount of strength,—and who can say what other purposes may have been secured besides?—three distinct principles embodied, the principle of the two tables and diplos of the human skull, the principle of those of the variously-arranged coats of the human stomach, and the principle of Oliver Cromwell's "fluted pot." There have been elaborate treatises written on those ornate flooring-tiles of the classical and middle ages, that are occasionally dug up by the antiquary amid monastic ruins, or on the sites of old Roman stations; but did any of them ever tell a story half so instructive or so strange as that told by the incalculably more ancient ganoid tiles of the palaeozoic and secondary period?"

Such are some of the forms and structures of fishes that swam in the seas from which were deposited the sediment that has hardened into the "old red sandstones" of Great Britain, Russia, and other parts of the world. And in this process of consolidation the carcasses of the fishes entombed in the primordial mud have had their share. For, just as a plaster-cast boiled in oil derives greater density and durability from that addition, so the oily and other azotized and ammoniacal principles of the decomposing fish operated upon the immediately surrounding sand so as to make it harder and more compact than the sediment not reached by the animal principles. Accordingly it has happened that in the course of the upheaval and disturbance of old red strata, parts of it, broken up and exposed to the action of torrents, have been reduced to detritus, and washed away, with the exception of certain nodules, generally of a flattened elliptic form, which are harder than the surrounding sandstone. Such nodules form the bed of many a mountain stream in "old red sandstone" districts of Scotland. If one of these nodules be cleft by a smart and well-applied stroke of the hammer, the cause of its superior density will be seen in a more or less perfect specimen of the fossilized remains of some animal, most commonly a fish.

But the placognoid and ganoid, heterocercal and notochordal, fishes of the Devonian epoch existed in such vast shoals in certain favourable inlets, that the whole mass of the sedimentary deposits has been affected by the decomposing remains of successive generations of those fishes. The Devonian flagstones of Caithness are an instance. They owe their peculiar and valuable qualities of density, tenacity, and durability wholly to the dead fishes that rotted in their primitive constituent mud. In no other part of the world, perhaps, can the builder set a large flagstone on its edge with assurance of its holding long together in that position. A great proportion of the county of Caithness formed, before its upheaval, the bottom of what may truly be termed a "piscina mirabilis." Yet there are minds, one at least, who, cognisant of the wonderful structures of the extinct Devonian fishes—of the evidence of design and adaptation in their structures—of the altered nature of the sediment surrounding them, and its dependence on the admixture of the decomposing and dissolved soft parts of the old fish—would deliberately reject the conclusions which healthy human reason must, as its Creator has constituted it, draw from such evidences of His operations. There are now individuals, one at least, who prefer to try to make it be believed that God had recently, and at once, called into being all these phenomena; that the fossil bones, scales, and teeth, had never served their purpose,—had never been recent,—were never truly developed, but created fossil; that the

---

1 See art. ODONTOLOGY, vol. xvi., p. 412, fig. 12. 2 Ibid., p. 413, fig. 14. 3 Owen's Odontology, vol. ii., p. 75, pls. 35 and 36. 4 Hugh Miller, Remarks of a Geologist, p. 284. 5 See Osteolites, by P. H. Gosse, 1858. Vertebrata, creatures they simulate never actually existed; that the superior hardness of the inclosing matrix was equally due to primary creation, not to any secondary cause; that the geological evidences of superposition, successive stratification, and upheaval were, equally with the palaeontological evidences, an elaborate design to deceive and not instruct! Surely, on such hypothesis, the workmanship must be that of the father of lies, not of the Author of Truth, and the imaginer of such hypothesis must be a Manichean at heart.

The sub-order of Placognathoids, so richly represented in the Devonian epoch, disappears in the carboniferous one; the Ganoids, with rhomboid scales, increase in number. These are characteristically represented by the genus *Paleoniscus*, species of which range throughout the carboniferous and Permian beds. *Paleoniscus* (fig. 51) is characterized by the heterocercal tail and moderate-sized fins, the dorsal, D, being single, and opposite the interval between the anal, A, and ventral, V, fins; each fin has an anterior spine.


The fore-part of the head is defended by strong ganoid plates, of a beautiful vertebrate polish; the trunk-scales are usually granulate externally.

The sub-order *Pycnodontes* is represented in the carboniferous strata by the heterocercal genus *Platysomus* (fig. 53), and by the species *P. parvulus*, which has been found in that formation at Leeds; but most of the pycnodont fishes belong to the mezozoic period.

*Pygopterus*, *Acrolepis*, *Eurygnathus*, *Elomichtys*, *Plectolepis*, *Graptolepis*, *Orognathus*, *Pododus*, *Acantododes*, and *Diplopterus* are carboniferous genera of Ganoids, with rhomboid scales. *Cetacanthus*, *Isodus*, *Phyllolepis*, *Hoplopteryx*, *Uronemus*, *Colonodus*, *Centrodus*, *Asterolepis*, *Psammomystes*, and *Osteopax*, are genera of Ganoids with rounded scales, represented by species in carboniferous strata.

Of the above-named genera, *Acrolepis*, *Pygopterus*, *Paleoniscus*, *Platysomus*, and *Cetacanthus*, continue to be represented in Permian beds; in which also are found species of the ganoid genera *Dorypterus*, *Holacanthodus*, and *Globodus*, if the teeth on which the latter is based be not those of *Platysomus*.

The formations of the mezozoic or secondary periods give evidence of the full development of the ganoid order. In the lowest or "triassic" division this order is still represented by heterocercal and notochordal species belonging to some of the genera of the Permian period, as, e.g., *Cetacanthus*, *Amblypterurus*, and *Paleoniscus*. The genus *Placodus*, a supposed pycnodont fish of the muschelkalk, has been shown to be a conchivorous Saurian. In the oolitic division the heterocercal Ganoids are almost completely superseded by homocercal genera, which now, for the first time, appear on the stage of life; but the ossification of the endo-skeleton is still incomplete. In the cretaceous series the Teleostian, or well-ossified bony fishes, are numerous; and here also first are seen fishes with the flexible "cycloid" or "ctenoid" scales, and of genera which continue to be represented by living species.

Of 33 genera of fishes in the lias, 4 only were represented in older strata, while the rest extend into the upper oolitic beds. Of these, 19 genera are Ganoids with rhomboid scales, and two (*Leptolepis* and *Gyrotes*) have rounded scales. The sub-order *Sturionii*, represented by the sturgeons of the present seas, makes its first appearance in the lias, under the generic form of *Chondrostetus*, which has recently received a full description and illustration in a memoir communicated by Sir P. Egerton to the Royal Society of London. In this it is shown "that *Chondrostetus*, though essentially sturionian, yet evidences a transitional form between the sturgeons and more typical Ganoids; that its food was similar to that of the existing members of the family, but that it was procured in a tranquil sea, rather than in the tumultuous waters frequented by sturgeons at the present time."

---

1 Owen, in Phil. Trans. 1838, p. 169. 2 Proceedings of the Royal Society, April 20, 1838. The name of this group of ganoid fishes refers to the form of the greater proportion of the teeth, especially those attached to the palate and hind alveolar part of the lower jaw: the few anterior teeth are small and sub-prehensile; but the whole dentition bespeaks fishes adapted to feed on small testaceous and crustaceous animals.

The Pycnodonts were for the most part deep-bodied fishes, symmetrically compressed from side to side. They were notochordal; a few of the earlier forms (Platycephalus, e.g., fig. 53) were heterocercal, but the majority of the sub-orders were homocercal.

In the lias, most beautiful fossil fishes of this group are found, which were referred by Bronn to the genus Tetragonolepis, and by Agassiz to the lepidoid sub-order. Sir P. Egerton has, however, shown that the dentition is truly "pycnodont," having a very close resemblance to that of Microdon, but with the masticatory apparatus smaller in proportion to the size of the fish. The scales, moreover, instead of being articulated by interlocking pegs and sockets, as in fig. 52, are joined in a peculiar way, which Sir P. Egerton describes as follows:

Each scale bears upon its inner anterior margin a thick solid bony rib, extending upwards beyond the margin of the scale, and sliced off obliquely above and below, on opposite sides, for forming splices with the corresponding processes of the adjoining scales. These splices are so closely adjusted, that without a magnifying power, or an accidental dislocation, they are not perceptible. When in situ, and seen internally, these continuous lines decussate with the true vertebral apophyses, and cause the regular lozenge-shaped pattern so characteristic of the pycnodont family.

The Pycnodonts so characterized are further distinguished from the closely-resembling lepidoid genus Dapedius, by having the small anterior teeth conical and single-pointed, instead of being bifurcate; and although this character is subject to occasional variations, nevertheless, on taking a comprehensive view of all the dapediid species, it seems to have been sufficiently constant to warrant the continuance of the separation of the group into the unicuspid and bicuspid species. And Sir P. Egerton has accordingly proposed to apply the generic terms Echmedus (from εχμος, a point, and οδον, a tooth), for the unicuspid and pycnocodont species, formerly termed Tetragonolepis, and to continue the name Dapedius for the bicuspid and unequivocally lepidoid homocercal deep-bodied Ganoids, many beautiful species of which are found in the lias.

Genus Pycnodus (fig. 54).—The type-genus of this sub-

order is characterized by the large size of the round flat Vertebrae, crowned teeth, which cover the broad jaws as by a pavement of from three to five rows; at the fore-part of the jaws are two or more trenchant incisive teeth both above and below.

The oblique inner processes of the scales, and which in some species appear as distinct dermal ossicles decussating the neural spines, are limited in Pycnodus to the space between the occiput and the dorsal fin (fig. 54).

This species of Pycnodus abound in the oolitic formations above the lias: the one figured (P. rhombus) is from a calcareous deposit, so charged with animal remains as to be fetid, at Torre d'Orlando, near Naples. Species of Pycnodus (P. cretaceus, e.g.) occur in the chalk of Kent; and one species (P. toliapicus) has left its remains in the eocene clay of Sheppey. Some teeth from German miocene have been referred to this genus; but at this period, if not at the earlier tertiary one, Pycnodus became extinct.

Sub-Order 4.—LEPIDOSTEIDÆ, Müller.

Char.—Ganoids with rhomboid scales and conical teeth; a single dorsal.

Family L.—SAUROIDEI, Agassiz.

Char.—Some of the teeth much larger than the rest, and laniariform.

Genus Caturus.—In this genus the jaws are armed with close-set, large, conical teeth; the scales are delicate; all the species are homocercal and notochordal (fig. 55). The dorsal, d, is opposite the ventral, v, all the fins are of moderate size. One species of Caturus (C. Bucklandi) is from the lias; but the majority, like C. furcatus, are from the lithographic slates of Solenhofen. The most recent known species (C. similis) is from the chalk of Kent.

Pauchycormus, Saurostomus, Sauropis, Thryssonotus, and Eugnathus, are among the well-marked genera of the Sauroid family. It is represented at the present day by the North American genus Lepidosteus; but in this fish not only is the notochord converted into bony vertebral bodies, but these are united by ball-and-socket joints.

Sub-Order 5.—CYCLOGANOIDEI.

Family L.—LEPTOLEPIDÆ.

The Ganoids of this family are homocercal, and have rounded scales. In the type-genus (Leptolepis, fig. 56), the scales are extremely thin, yet a fine layer of ganoin may be discovered in them. The teeth are minute and en brosse, with two of larger size in front of the mouth. It has not been determined whether the notochord is ossified; but traces of distinct vertebral bodies appear to the writer to be discernible in some specimens. Species of Leptolepis range from the lias to the calcareous slates of Eichstätt. Genus Macropoma.—Fine specimens of homocerebral ganoid fishes, with rounded scales, sculptured externally, as in fig. 57, have been discovered in the chalk formations of Kent and Sussex. They have been referred by Agassiz to the genus called Macropoma, significative of the large size of the gill-cover, and to the cœlacanthal family; but traces of vertebral bodies are apparent in some specimens. Casts of the "interior" of the alimentary canal, showing impressions of a broad spiral valve, are preserved in certain specimens in the British Museum. One species (M. Egertonii) is from the Speeton clay; the other (M. Mantelli) from the chalk.

In the tertiary division the ganoid order rapidly diminishes, and its place is taken by fishes with better ossified internal skeletons, and with thinner, more flexible, and usually soluble scales. The gills are supported on bony arches, and are protected by branchiostegal rays, and by an operculum or gill-cover. The aortic bulb is provided with but two valves; and the optic nerves decussate. For this group, including the majority of existing fishes, and of those which made their appearance during the tertiary period, Müller proposed the name "Teleosteï," which almost corresponds with the "osseous fishes" of Cuvier. After the full and accurate illustration of this great group in the article Ichthyology, little is required to be added by way of illustration from extinct species.

Order IV.—ACANTHOPTERI

Sub-Order 1.—CTENOIDEI

Char.—Endo-skeleton ossified; exo-skeleton as ctenoid scales (fig. 58); fins with one or more of the first rays unjointed or inflexible spines; ventrals in most beneath or in advance of the pectorals; swim-bladder without air-duct.

Of this order may be given two genera, both of which are now extinct. One (Semionophorus) belongs to the chetodont family; the other (Smerdis) to the Percoids.

The genus Semionophorus, Ag. (fig. 59), is represented exclusively by extinct species peculiar to the tertiary deposits at Monte Bolca. It is characterized by the extreme height or prolongation of the anterior part of the dorsal fin, D, and for the correlative elongation of the slender pointed ventral fins. The anal fin, A, is much shorter than the dorsal. Owing to the soluble nature of the scales, and to the well-ossified skeleton, the fossils of this, as of most other tertiary fishes, are exemplified by the vertebral column and skull more than by the skin.

Sub-Order 2.—CYCLOIDEI.

Genus Smerdis.—The species composing this genus are of small size, and are wholly extinct; they likewise are chiefly met with in the tertiary ichthyolite beds of Monte Bolca; but some (e.g., the Smerdis minutus, fig. 60) are from cocoon deposits in France. In all the species the first suborbital or lacrymal bone is strongly dentate, as is also the preoperculum; but this has no spine at the angle. The operculum terminates behind by a rounded prominence. There are two dorsals. The scales are minute, but are occasionally preserved. Those of Smerdis minutus present, under a magnifier, the structure shown in fig. 61.

In the article Ichthyology, figures are given of the two kinds of existing sword-fish (Xiphias, fig. 112; and Histiothorax, fig. 113); in the former the sword-like prolongation of the confluent premaxillaries is flattened, in the latter it is rounded.

---

1 See art. Ichthyology, pp. 209–232. 2 Ibid., p. 273. Fossil remains of a rounded rapier-like sword, but much longer and more slender than in the existing *Histiothorax*, have been found in the eocene clay at Sheppy and Bracklesham. They are referred to an extinct genus of the xiphoid family by Agassiz, called *Cantorhynchus*, or "hollowbeak." The most perfect specimen hitherto found is figured in fig. 62, of half the natural size. It forms part of the instructive collection of Captain Le Hon at Brussels. The upper transverse section shows the single cavity at the middle of the rostrum; and the lower section shows the double or divided cavity near its base.

**Order V.—ANACANTHINI.**

*Char.*—Endo-skeleton ossified; exo-skeleton in some as cycloid, in others as ctenoid scales; fins supported by flexible or jointed rays; ventrals beneath the pectorals, or none; swim-bladder without air-duct.

**Family.—PLEURONECTIDE.**

*(Flat-Fishes.)*

In this family the symmetrical form is lost, and both eyes are on one side of the head. Species of still existing genera of this much-modified family have been found in tertiary deposits. The little turbot (*Rhombus minimus*, *fig.* 63) occurs in the tertiary deposits of Monte Bolca. An equally extinct species of sole (*Solea antiqua*) has been found in tertiary marls near Ulm.

Fossil fishes of the cod, mullet, carp, salmon, and herring genera, are found in the tertiary formations, but are distinct from all known species.

The Ganoids are reduced to the genera *Lepidosteus* and *Acipenser*, but may have been represented by the palates with crushing teeth, to which the names *Pisodon* and *Phyllodus* have been given. With respect to the fishes of the tertiary period, "they are so nearly related," says Agassiz, "to existing forms, that it is often difficult, considering the enormous number (above 8000) of living species, and the imperfect state of preservation of the fossils, to determine exactly their specific relations. In general I may say that I have not yet found a single species which was perfectly identical with any marine existing fish, except the little species *Mallotus villosus*, which is found in the nodules of clay of unknown geological age in Greenland."

For a list of the genera and species of fossil fishes known at the date of publication, reference may be made to the article *Fossil Ichthyology*. Since that was written, it has been determined that the "schists of Glaris" and of "Monte Bolca," belong to the eocene tertiary period.

---

1 See Owen's *Odontography*, p. 133, pl. 47, fig. 3. 2 Ibid., p. 139, pl. 47, figs. 1 and 2. 3 See for the explanation of these terms in the system of Agassiz the article *Fossil Ichthyology*, *Enzyk. Bru.*, vol. xii. Ichnology, on the score of the defect of negative evidence, to which attack conclusions from the known genetic history of air-breathing animals are open. Many creatures living on land may never be carried out to sea; but marine deposits may be expected to yield adequate evidence for general conclusions as to the character of the vertebrate animals that swarmed in the seas precipitating such deposits.

One other conclusion may be drawn from a general retrospect of the mutations in the forms of the fishes at different epochs of the earth's history,—viz., that those species, such as the nutritious cod, the savoury herring, the rich-flavoured salmon, and the succulent turbot, have greatly predominated at the period immediately preceding and accompanying the advent of man; and that they have superseded species which, to judge by the bony Garpike (Lepidotes), were much less fitted to afford mankind a sapid and wholesome food.

**ICHOLOGY.**

In entering upon the genetic history of the class of reptiles, we have to inquire, as in that of fishes, in what period of the earth's history the class was introduced, and under what forms; at what period it attained its plenary development, in regard to the size, grade of structure, number and diversities of its representatives; and the relations which the existing members of the class bear to its past condition. Fifteen years ago, the oldest known reptilian remains were those of the so-called "Thuringian Monitor," from the Permian copper-slates of Germany. Five years ago, the batrachian Apateon, or *Archegosaurus* of the Bavarian coal-field, represented the known commencement of reptilian existence. In October 1851 the following notice appeared in the *Elgin Courant* for the 10th of that month:

"Geological Discovery.—A fossil has been obtained from the old red sandstone at Spynie, near Elgin, which serves to establish the fact that air-breathing Vertebrata of the order *Sauria* existed during the deposition of the Devonian system of rocks, which hitherto had only been surmised from the discovery of impressions of foot-prints on the surface of the strata. This fossil, or rather impression, was brought to a gentleman in Elgin, who now possesses it, by a quarryman, who with praiseworthy care preserved the fragments, without tampering with the impression to any injurious extent, so that its character is unmistakable. It presents the figure of a reptile about four inches in length, showing the head containing fragments of conical teeth, part of the neck, the back-bone and ribs, the pelvic bones and hind legs, with part of the tail (the rest of the tail, which seems to have been long, being still unexposed). The appearance of this small specimen is very striking,—the animal matter in its decomposition having stained the matrix to a dark ochre, while the rest of the stone is a pale gray, almost white. No doubt this interesting relic will receive a more scientific description than we can pretend to give, but we cannot resist the impulse to announce the occurrence in this district of an object so well calculated to forward the investigations of geologists."

An accurate drawing of this fossil was transmitted to the writer by its possessor, Mr Patrick Duff of Elgin, together with the newspaper containing his announcement of the discovery; and the writer's opinion, written on the inspection of Ichnology, that drawing in October 1851, was to the effect that it was a true saurian reptile. The fossil itself was subsequently transmitted to London, and submitted to different palaeontological authorities. To the present writer was kindly accorded the opportunities of having the requisite drawings made for the illustration of the fossil in his *History of British Fossil Reptiles*; the results of a minute and careful comparison of the impression were briefly given in the *Literary Gazette* of December 20; and the name *Leptopleuron*, significative of its most conspicuous lacertian character, was proposed for it. Other palaeontologists regarded the fossil as a batrachian reptile, and Dr Mantell assigned to it the name of *Telerpeton*; but all concurred that it belonged to an air-breathing vertebrate animal. The following is the summary of the grounds on which the writer's conclusions as to the higher affinities of the little Elgin reptile, in its class, were based:

"Vertebrate air-breathing Life in the Old Red Sandstone."

"Our readers may recollect that the geological world was startled by the announcement, in the *Elgin Courant* of October 10th, of the discovery of a fossil reptile in the 'old red' at Spynie, near Elgin. The specimen has been submitted to the examination of Professor Owen, from whom we have received the following notice of its nature and affinities:

"Royal College of Surgeons, Dec. 15, 1851.

"Mr Duff, the proprietor of the very remarkable fossil recently discovered in a sandstone of the Devonian system of rocks at Elgin, transmitted me a drawing of it, with the request that I would undertake its examination; to which having gladly acceded, the specimen itself was brought to me by a friend of Mr Duff's. It is the impression, in two pieces of a gray variety of the old red sandstone, of a long and slender four-footed vertebrate animal, four inches and a half in length, clearly belonging, by the form, proportions, and positions of the scapular and pelvic arches, and the supposed limbs, to the reptilian class. The impression retained its dimensions, the cavities in the matrix which contained it remain, stained by a deposit of an ochreous tint. The impressions are so well defined as to show that there were twenty-six vertebrae between the skull and sacrum, two sacral vertebrae, and thirteen caudal vertebrae, before the tail disappears by dipping into an unexposed part of the matrix. Impressions of twenty-one pairs of ribs are preserved, all very slender, short where they commence near the head, but rapidly gaining length as they are placed further back. The cervical and anterior ribs are expanded, but not bifurcate, at their vertebral end; and all the ribs articulate close to the bodies of the vertebrae. In the crocodilian reptiles the anterior ribs are bifurcate, and the posterior ones, with a simple head, arrange themselves along diaphyses. The distinctive characters of the batrachian skeleton are the double occipital condyle; ribs wanting, or very short and weak; a single cervical vertebra; and rib-shaped ilium. The first character cannot be determined, the occipital articulation not being preserved in the fossil. Instead of the second character, the fossil shows ribs of varied length, and most of them much longer than in the salamanders, newts, or any known Batrachian. With regard to the third character, the impression in the matrix clearly shows two sacral vertebrae and a short subquadrate pelvis.

"Both the humerus and the femur show the lacertian sigmoid shape, and near equality of length, which distinguish them alike from the crocodilian and batrachian orders; they are likewise, as in lizards, relatively longer than in the newts and salamanders. In the perfect impression of the head may be seen the hollow bases of some large, slightly-compressed, conical teeth, which also tell for the saurian and against the batrachian nature of this ancient reptile. I propose to call it *Leptopleuron lacertinum*. Many particulars of minor import, bearing upon the more immediate affinities of this most rare and interesting fossil, have been noted, and will be given, with the figures, in my *History of British Fossil Reptiles*, for which work Mr Duff has kindly consented to place the..." Admitting the saurian affinities of the *Leptopleuron*, it may have been, like the little *Amblyrhynchus* of the Gallopagos Islands, a lizard of marine habits. As no part of the osseous texture is preserved, and the impressions of the vertebrae are only those of the neural arches, we cannot be assured that the bodies of the vertebrae may not have been unossified, and in the same embryonal stage as in some reptiles, next to be noticed, from the coal formations. It is most probable, however, that the *Leptopleuron* was more nearly allied to the thecodont than to the labyrinthodont Reptilia.

Very recently the remains of a true saurian reptile, protected by pitted bony plates like those of crocodiles, but with a broad coracoid like that of *Cetiosaurus* or *Polyptychodon*, have been discovered in the same formation and locality. The matrix is a fine-grained whitish sandstone, with a cement of carbonate of lime. It belongs, according to Sedgwick and Murchison, to the uppermost beds of the "old red" or Devonian system. As yet, however, no characteristic Devonian or old red fossils of any class have been discovered associated with the foregoing evidences of reptiles, which, according to the determination of strata by characteristic fossils, would belong to the secondary or mezozoic period. In the same sandstone, in the quarry at Cumnockstone, near Elgin, a continuous series of thirty-four impressions have been observed by Captain Binkendonk, and attributed by him to the foot-prints of a quadruped. The impressions are in pairs, forming two parallel rows, the hind one being one inch in diameter, and larger than the fore one in the proportion of three to four. The stride of the animal, on this hypothesis, must have been about four inches.

There are several circumstances under which impressions made on a part of the earth's surface, soft enough to admit them, may be preserved after the impressing body has perished. When a shell sinks into sand or mud, which in course of time becomes hardened into stone, and when the shell is removed by any solvent that may have filtered through the matrix, its place may become occupied by crystalline or other mineral matter, and the evidence of the shell be thus preserved by a cast, for which the cavity made by the shell has served as a mould. If the shell has sunk with its animal within it, the plastic matrix may enter the dwelling-chamber as far as the retracted soft parts will permit; and as these slowly melt away, their place may become occupied by crystallized deposits of any silicious, calcareous, or other crystallizable matter that may have been held in solution by water percolating the matrix, and such crystalline deposit may receive and retain some colour from the soft parts of which it thus becomes a cast.

Evidences of soft-bodied animals, such as *Actinia* and *Medusa*, and of the excremental droppings of higher animals, have been thus preserved. Fossil remains, as they are called, of soft plants, such as sea-weeds, reeds, calamites, and the like, are usually casts in matrix made naturally after the plant itself has wholly perished.

Even where the impressing force or body has been removed directly or shortly after it has made the pressure, evidence of it may be preserved. The hailstone, the ripple wave, the rain-drop, even the wind that bore it along and drove it slanting on the sand, have been registered in casts of the cavities which they originally made on the soft seabed; and the evidence of these and other meteoric actions, so written on imperishable stone, have come down to us from times incalculably remote. Every form of animal life, that, writhing, crawling, walking, running, hopping, or leaping, could leave a track, depression, or foot-print, behind it, might thereby leave similar lasting evidence of its existence, and also to some extent of its nature.

The interpretation of such evidences of ancient life has much exercised the sagacity of naturalists since Dr. Duncan, in 1828, first inferred the existence of tortoises at the period of the deposition of certain sandstones in Dumfriesshire, from the impressions left on those sandstones, and the casts afterwards formed in those impressions. The faculty of interpreting has been still more racked by similar evidences of more extraordinary foot-prints, probably of large brachian reptiles, first noticed in 1834 at Hildburghausen in Saxony, in sandstones of the same geological age as those in Scotland.

The vast number and variety of such impressions, due either to physical or meteoric forces, to dead organic bodies, parts or products, or to the transitory actions of living beings, have at length raised up a distinct branch of paleontological research, to which the term "Ichnology" has been given.

In this class of evidences the impressions called "protichnites" (fig. 64), left upon the "Potsdam sandstones" of the older Silurian age in Canada, are the most ancient; but the foot-prints of birds surpass all others in regard to their number, distinctness, and variety of sorts.

But how, it may be asked, are such foot-prints preserved? A common mode may be witnessed daily on those shores where the tide runs high, and the sea-bottom is well adapted to receive and retain the impressions made upon it at low-water.

Dr. Gould of Boston, U.S., first called the attention of naturalists to this interesting operation on the shores of the Bay of Fundy, where the tide is said to rise in some places 70 feet in height. The particles deposited by that immense tidal wave are derived from the destruction of previously existing rocks, and consist of silicious (flinty) and micaceous (tacky) particles, cemented together by calcareous (limy) or argillaceous (clayey) paste, containing salts of soda, especially the muriate (common salt), and coloured with various shades of the oxide of iron, of which the red oxide predominates. The perfection of the surface for receiving and retaining an impression depends much upon the micaceous element. Vast are the numbers of wading and sea birds that course to and fro over the extensive tract of plastic red surface left dry by the far retreat of the tide in the Bay of Fundy. During the period that elapses between one spring tide and the next, the highest part of the tidal deposit is exposed long enough to receive and retain many impressions; even during the hours of hot sunshine, to which, in the summer months, this so-trodden tract is left exposed, the layer last deposited becomes baked hard and dry, and before the returning tidal wave, turbid with the same comminuted materials of a second stratum, has power to break up the preceding one, the impressions left on that stratum have received the deposit. A cast is thus taken of the mould previously made, and the sediment superimposed by each succeeding tide, tends more and more surely to fix it in its place. Then, let ages pass away, and the petrifying influences consolidate the sand layers into a fissile rock: it will split in the way it was formed, and the cleavage will expose the old moulds on one surface and the casts on the other.

Another condition for fixing the impressions on a sandy shore is the following:—When an extensive level tract is... Ichnology left dry by the retreating tide, as at the estuary of the small rivers entering the Bay of Morecombe, on the Lancashire coast, those rivers occasionally overflow the sands at low-water, and deposit in the foot-prints made previous to such overflow the fine mud which sudden heavy rains have brought down from the surrounding hills. Again, those sudden "freshets," as they are locally called, sometimes as quickly subside, and the thin layer of argillaceous mud is left dry on the sand before the returning tide. Such layer of mud readily receives and retains the foot-prints of the many birds that course over the flat expanse; and as the tide returns, it deposits in such foot-prints a layer of the fine sand which the rising waters hold in suspension.

The best-defined foot-prints in the new red sandstone quarries at Stourton, on the Cheshire coast, are found where strata of sandstone are separated by a thin layer of argillaceous stone, which, when exposed, soon breaks up and crumbles away. This layer has, however, received the impressions when it was plastic, and the superincumbent deposit of sandstone retains those impressions in relief upon its under surface. The observations which have just been recorded, of the circumstances that produce an interposition of a thin layer of claystone between thicker beds of sandstone, and which circumstances the writer has witnessed in the Bay of Morecombe, explain the formation and the preservation of the best "ichnites" of the labyrinthodont and other reptiles in the new red sandstone of Stourton.

There is a third condition under which impressions, and casts of impressions, on a sandy beach may be preserved. On a dry windy day clouds of fine sand are drifted along the surface exposed at low-water, are spread lightly over all its little inequalities, and fill up every impression that may have been made on it since it was left bare by the retreating waves. On the return of the tide, the fine sand filling the impressions is moistened, and more wet fine sand is added to it; and a cast is thus fixed in the moulds, to be more and more firmly fixed by each deposition from successive tidal waves.

Thus may be witnessed the actual conditions and the circumstances daily occurring that tend to preserve foot-prints and other impressions made on the sea-shore, and which have operated in past time to similarly preserve the impressions then made on tracts alternately exposed and covered by the tidal wave. The merit of having first discerned the nature and cause of the numerous small hemispheric pits and tubercular casts in relief on the surface of certain sandstone slabs, is due to John Cunningham, Esq., F.G.S., architect, of Liverpool. Since that light was thrown on their nature, they have been recognised under various modifications, as impressions of soft rain, of the big-dropped thunder-showers, of rain driven obliquely by the gale, and making impressions with the side of the cup highest opposite the point whence the wind blew, of frozen rain or hail, &c. Whenever a stratum is proved to be a "sedimentary" one,—i.e., to be due to the precipitation of its constituent particles from water, in which they had been previously suspended,—we have evidence of some expanse of water—proof, in fact, of the existence of that element, with all its properties of condensation by cold, and expansion and vaporization by heat and exposure. Evaporation makes the raw material of rain. No wonder, then, that impressions of rain-drops should be seen on the oldest sedimentary rocks. Conditions are co-ordinated in meteoric as in organic phenomena; one being given, the rest may be deduced.

The oldest rocks in which rain-drop impressions have been observed are those of the Cambrian age at Longmynd, Wales, described and figured by Mr Salter. Many of the ichnologic flags of the same formation are covered with ripple, or current marks. They show borings of worms, and a trace of a trilobite (Palaeopyge) nearly allied to the Dikelocephalus—the oldest known trilobite of America (Lower Silurian or Cambrian at St Croix, Minnesota).

It is in "Potsdam sandstones" of the same geological antiquity that the impressions have been discovered which the writer has interpreted to be those of a large entomotaceous Crustacean, in evidence of which the following sample, applicable to a single species, may be given, in illustration of the ichnologist's mode of work.

**Protichnites tepus-natatus** (fig. 64).

The subject so named consists of a series of well-defined impressions, continued in regular succession along an extent of 4 feet; and traceable, with an inferior degree of definition, along a further extent of upwards of 2 feet.


**Protichnites tepus-natatus (Cambrian).**

In the extent of 4 feet there are thirty successive groups of foot-prints on each side of a median furrow, which is alternately deep and shallow along pretty regular spaces of about 2½ inches in extent. The number of prints is not the same in each group; where they are best marked, as in fig. 64, I, II, we see 3 prints in one group, α, α', α", 2 prints in the next, β, β', and 2 in the third, ε, ε', which is followed by a repetition of the group of 3 prints, α, α', α", making the numbers in the three successive groups 3, 2, 2; the three groups of impressions being recognisably repeated in succession along the whole series of tracks on both sides of the median groove.

The principal foot-prints are disposed in pairs, placed with different degrees of obliquity, in each of the three groups, towards the median track; the innermost print in the second, II, and third, C, pairs, which are best marked, being usually rather more than half the size of the outer print, β' and ε'.

The two foot-prints of the same pair are a little further apart from each other, in the three succeeding pairs, as at α', α", β', ε', especially in the second and third groups of each set; the two forming the pair α', ε', again approximating in the next series, and the pairs β', ε', diverging in the same direction and degree; and this alternate approximation and divergence is repeated throughout the entire series of the present tracks.

But what strikes the ichnologist, heretofore conversant chiefly with the foot-prints of bipeds or quadrupeds, is the occurrence in Ichnology, the present series of the third impression \(a\), which complicates the most approximated pair \(A\), being placed in front and a little to the inner side of the hindmost impression, \(a'\), of that pair. The superadded impression \(a\) is about the same size as the innermost in each pair, the average diameter of that impression being 5 lines.

Taking this view of the impressions, it appears that whilst the innermost in each pair, \(a', b, c\), are of equal size, the outermost, \(a'', b', c'\), progressively increase in size, from the most approximated to the most divergent of the three pairs; that of the first \(a''\), being narrow in proportion to its length, that of the second \(b''\), as broad as long, and the outermost \(c'', c'\) of the third pair being oblong, but larger than that in the next pair. In some places where the most approximated pair of impressions, \(a', a''\), are deeply marked, they are complicated by a fourth shallow and very small pit, \(a'''\), midway between the third, \(a\), and the outermost, \(a''\), of the pair of impressions.

There are no clear or unequivocal marks of toes or nails on any of the impressions which form the lateral pairs or triplets. Their margins are not sharply defined, but are rounded off, and sink gradually to the deepest part, which is a little behind the middle of the depression. There is a slight variation in the form and depth of the answerable impressions, but not such as to prevent their correspondence being readily appreciable through the whole of the extent here described; and it is evident that almost of each of the three pairs here described as first, A, second, B, and third, C, may be identified with the corresponding innermost impression on the opposite side, and with the same impression of the same pair in the three preceding and the three succeeding pairs.

The impressions selected for fig. 64 clearly demonstrate that the animal, progressing in an undulating course, made at each action of its locomotive members, answering to the single step of the biped and the double step of the quadruped, not fewer than, in Protichnites 7-motus, fourteen impressions, seven on the right and seven on the left; and in Protichnites 8-motus, sixteen impressions, eight on the right and eight on the left; these seven and eight impressions respectively being arranged in three groups, viz., Protichnites 7-motus, three, two, and two; in Protichnites 8-motus, three, two, and three,—the groups being reversed, in successive series, so similarly as to regularly and admit of no doubt that they were made by repeated applications of the same impressing instruments, capable of being moved so far in advance as to clear the previous impressions and make a series of new ones at the same distance from them as the sets of impressions in the series are from each other. What then was the nature of these instruments? To this three replies may be given, or hypotheses suggested—They were made either, first, as in the case of quadrupedal impressions, each by his own limb, which would give seven and eight pairs of limbs to the two species respectively; or, secondly, certain pairs of the limbs were bifurcate, as in some insects and crustaceans, another pair or pair being trifurcate at their extremities; or, each group of impressions was made by a single so subdivided limb, in which case we have evidence of a remarkably broad and short, and, as regards ambulation, very leggy quadruped; or, thirdly, three pairs of limbs were bifurcate, and the supplementary pits were made by small superadded limbs, as in some crustaceans; or, fourthly, a single broad fin-like member, divided at its impressing border into seven or into eight obtuse points, so arranged as to leave the definite pattern described, must have made the series of three groups by successive applications to the sand.

The latter hypothesis appears to be the least probable,—first, as being most remote from any known analogy; and, secondly, because there are occasional varieties in the groups of locomotion which would hardly accord with impressions left by one definitely subdivided instrument or member. Thus in the group of impressions marked 1 L in fig. 64, the outer impression, \(c'\), is single, but in the preceding set it is divided; whilst the impressions \(a', a''\) are conjoined, and the innermost \(a'''\) is separate in 1 L. The same variety occurs in the outer pair, \(c', c''\), in Protichnites 8-motus.

Yet, with respect to the hypothesis that each impression was made by its own independent limb, there is much difficulty in conceiving how seven or eight pairs of jointed limbs could be aggregated in so short a space of the sides of one animal. So that the most probable hypothesis is, that the creatures which have left these tracks and impressions on the most ancient of known sea-shores belonged to an articulate and probably crustaceous genus, either with three pairs of limbs employed in locomotion, and several divided to accord with the number of prints in each of the three groups, or bifurcated merely, the supplementary and usually smaller impressions being made by a small and simple fourth, or fourth and fifth pair of extremities.

The great entomostracous king-crab (Lambrus), which has the small anterior pair of limbs near the middle line, and the next four reptilian lateral pairs of limbs bifurcate at the free extremity, the last pair of internal limbs with four lamelliform appendages, and a long and slender hard tail, comes nearest to my idea of the kind of animal which has left the impressions on the Pottedan sandstone.

The shape of the pits, so clearly shown on the incrustated slabs, impressed by Protichnites 8-motus, accords best with the hard, sub-ovoid, and irregular terminations of a crustaceous ambulatory limb, such as may be seen in the blunted legs of a large Pulmonaria or Bisopae; and it is evident that the animal of the Pottedan sandstone moved directly forwards after the manner of the Macrourus and Xiphosaurus, and not sideways, like the brachyurous Crustaceans.

The appearances in the slab impressed by the Protichnites multimotus favour the view of the median track having been formed by a caudal appendage, rather than by a prominent part of the under surface of the trunk.

The imagination is baffled in the attempt to realize the extent of time past since the period when the creatures were in being that moved upon the same shores of that most ancient Silurian sea; and we know that, with the exception of the microscopical forms of life, all the higher species of animals came into being at a period geologically very recent in comparison with the Silurian epoch.

The deviations from the living exemplars of animal types usually become greater as we descend into the depths of time past; of this the Archaeosaur and Ichthyosaur are instances in the reptilian class, and the Perichthys and Cocceosaurus in that of fishes. If the vertebrate type has undergone such inconceivable modifications during the Secondary and Devonian periods, what may not have been the modifications of the articulate type during a period probably more remote from the secondary period than this is from the present time. In all probability no living form of animal bears any resemblance to that which the Pottedan footprints indicate as to afford an exact illustration of the mode and number of the intramuscular, and of the mode of locomotion, of the silurian Protichnites.

Since the foregoing interpretation of the silurian Ichnites of North America was published, similar impressions have been observed in rocks of the like high antiquity in Scotland, as at Binks, Eskdale, which have received the name of Protichnites Scoticus.

Class L.—REPTILIA.

Sub-Class I.—AMPHIBIA.

(Batrachia.)

Genus Batrachopus (Batrachopus primarius, King).

In 1844 the scientific world was startled with the announcement by Dr King of Greensburg, Pennsylvania, of the fact that he had discovered unquestionable fossil footmarks of reptiles in the sandstone of the coal measures, in Westmoreland county, near that town. No reptilian footprints or saurian remains had previously been found lower in the series than the new red sandstone. Dr King states the impressions to be "near 800 feet beneath the topmost stratum of the coal formation."

Sir C. Lyell, in Silliman's Journal, July 1846, describes his visit to Greensburg, where he examined these footmarks, and sustained Dr King's observation and description of them. He considered them to be allied to the labyrinthodont foot-prints which have been referred to the genus Cherotherium. He says—"They consist, as before stated, of the tracks of a large reptilian quadruped, in a sandstone in the middle of the carboniferous series, a fact full of novelty and interest; for here in Pennsylvania, for the first time, we meet with evidence of the existence of air-breathing quadrupeds capable of roaming in those forests where the Sigillaria, Lepidodendron, Caulopteris, Calamites, ferns, and other plants flourished."

These foot-marks were first observed standing out in relief from the lower surface of slabs of sandstone resting on thin layers of fine unctuous clay, which also exhibited the cracks due to shrinking and expansion. Now these cracks, where they traversed the foot-prints, had produced distortion in them, for the mud must have been soft when the animal walked over it and left the impressions; whereas, when it afterwards dried up and shrunk, it would be too hard to receive such indentations, and could only affect them in the way of subsequent dislocation.

---

1 Harkness and Salter "On the Lowest Rocks of Eskdale," Quarterly Journal of the Geological Society, vol. xii., pp. 238, 243, fig. 2. No less than twenty-three footprints, the greater part so arranged as to imply that they were made successively by the same animal, were observed by Dr King in the same quarry.

Everywhere there was a double row of tracks, and in each row they occur in pairs, each pair consisting of a hind and fore foot, and each being at nearly equal distances from the next pair. The hind footprint is about one-third larger than the fore footprint; it has five toes, but the front only four; some of them exhibit a stunted rudiment of the innermost toe or "pollex," which is the underdeveloped one. The outermost toe in the hind footprint is shorter and rather thicker than the rest, and stands out like a thumb on the wrong side of the hand.

With this general resemblance to the footprints of Labyrinthodon, and also the red and red-brown shades of Europe, there are well-marked distinctions. In the first place, the right and left series of impressions are wider apart, indicative of a broader-bodied animal. The front print in Labyrinthodon has only four well-developed toes instead of five, as in Labyrinthodon; it is also proportionably larger, —the fore foot in Labyrinthodon being less than half the size of the hind foot. The distance between the fore and hind print of each pair, and of one such pair from the next on the same side, is nearly the same in Labyrinthodon and Labyrinthodon.

Genus Sauropus, Rogers.—Very similar footprints were discovered and described by Mr Isaac Lee in a formation of red shales, at the base of the coal measures at Pottsville, 78 miles N.E. of Philadelphia. These are of older date than the preceding, inasmuch as a thickness of 1700 feet of strata intervenes between the footprints at Greensfield and these older Pottsville impressions.

Professor H. D. Rogers, in 1851, announced his discovery in the same red shales, between the Devonian and carboniferous series, of three species of four-footed animals, which he deems to have been rather saurian than batrachian, seeing that each foot was five-toed; one species, the largest of the three, presented a diameter for each foot-print of about two inches, and showed the fore and hind feet to be nearly equal in dimensions. It exhibits a length of about nine inches, and a breadth between the right and left footprints of nearly four inches. The impressions of the hind feet are but little in advance of the fore feet. With these footprints associated shrinkage cracks, caused as are caused by the sun's heat upon mud, and rain-drop plittings, with the signs of the trickling of water on a wet beach,—all confirming the conclusions derived from the foot-prints, that the quadrupeds belonged to air-breathers, and not to a class of animals living in and breathing water.

ORDER I.—GANOCEPHALA.

Genus Apateon, Von M.; Archegosaurus, Goldf.—Certain fossils, discovered in the spheroidalitic clay-slate


Fig. 65.

Apateon or Archegosaurus (Carboniferous).

ribs, with the indications of stunted swimming limbs, impressed the writer with the conviction of the near alliance of the Archegosaurus with the Proteus and other perennibranchiate reptiles.

This conclusion of the affinity of Archegosaurus to existing types of the reptilian class is confirmed by subsequently-discovered specimens, some of which have been acquired by the British Museum, others have been described and figured by H. von Meyer in his Palaeontographica (Bd. vi., 2nd Liefl. 1857); more especially by his discovery of the embryonal condition of the vertebral column,—i.e., of the persistence of the notochord, and the restriction of ossification to the arches and peripheral vertebral elements. In this structure, the old carboniferous reptile resembled the existing Lepidosternum.

---

1 Arceus, a chest; argyros, beginning, emery, lizard. 2 Mainz, October 1843. "In dem Brandschleiter von Münster-Appel im Rhein-Rhein habe ich in vorigen Jahre einen Salamander aufgefunden. Gehört dieser Schleier der Kalksteinformation? In diesem falle wäre der Fund auch in anderen Hinsicht interessant." (Leonhard und Brunn, Neues Jahrbuch für Mineralogie, &c., 1844, p. 49.) 3 Ob das—Apateon pedetris—ein Salamander-artiges Geschöpf war, ist keineswegs ausgemacht." (Ibid.) 4 Archegosaurus, Fossile-Saurier aus dem Stein Kohlengebirge des Ubergangs der Ichthyoden zu den Lacerten und Krokodilen bilden?" p. 3. (Beiträge zur vorzeitlichen Fauna des Steinkohlengebirges, 4to, 1847.) 5 Reptilien aus der Steinkohlen Formation in Deutschland, Sechster Band, p. 61. Repilika. and affords farther ground for regarding that remarkable existing animal as one which obliterates the line of demarcation between the fishes and the reptiles.

Coincident with this non-ossified state of the basis of the vertebral bodies of the trunk (fig. 65, s), is the absence of the ossified occipital condyles which characterize the skull in better developed Barraclia. The fore part of the notochord has extended into the basi-sphenoid region, and its capsule has connected by ligament to the broad flat ossification of expansions of the same capsule, forming the basi-sphenoid or basi-sphenoid plate. In fig. 63 are represented the chief modifications of the vertebrae, as shown in the neck, thorax, abdomen, sacrum, and tail. The vertebrae of the trunk in the fully-developed, full-sized animal present the following stage of ossification:

The neurospines (fig. 65, n) coalesce at top to form the arch from the summit of which was developed a compressed, sub-quadrate, moderately high spine, with the truncate or slightly convex summit expanded in the fore and aft direction so as to touch the contiguous spines in the back; the spines are distinct and separate. The sides of the base of the neural arch are thickened and extended outwards into diaphyses, having a concave articular surface for the attachment of the rib. The neural arch is slightly produced at each angle into a synapophysis looking upwards and a little forward; the hinder part was much produced backwards, supporting two thirds of the neural spine, and each angle developed into a zygapophysis, with a surface of opposite aspect to the anterior one. In the capsule of the notochord three bony plates were developed, one on the ventral surface, and one on each side, at or near the back part of the diaphysis. These bony plates may be termed cortical parts of the centrum, in the same sense in which that term is applied to the element which is called "body of the atlas" in man and Mammalia, and "sub-vertebral wedge-bone" at the fore part of the neck in Enaliosaurus.

As such neural or inferior cortical elements co-exist with seemingly complete centrums in the Ichthyosaurus, thus affording ground for regarding them as being distinct from a true centrum, the term "hypopophysial" has been proposed for such independent inferior cortical elements in and from the notochordal capsule; and by that term may be signified the sub-notochordal plates in Archeosaurus, which co-exist with proper hypopphyses (a) in the tail. In the trunk they are flat, subsquare, oblong bodies, with the angles rounded off; in the tail they bend upwards by the extension of the ossification from the under to the side parts of the notochordal capsule; sometimes touching the lateral cortical plates. These serve to strengthen the notochord and support the intervertebral nerve in its outward passage. The ribs (pi) are short, almost straight, expanded and flattened at the ends, and much slender at the middle. They are developed throughout the trunk, and along part of the tail, occurring together with the basial arches, as in the Mosasaur. The basial arches (b), which are at first open at their base, become closed by extension of ossification inwards from each produced angle, converting the notch into a foramen. This forms a wide oval, the apex being produced into a long spine; but towards the end of the tail the spine becomes shortened, and the basial arch reduced to a mere flattened ring.

The size of the canal for the protection of the caudal blood-vessels indicates the powerful muscular actions of that part, as the produced spines from both neural and basial arches bespeak the provision made for muscular attachments, and the vertical development of the caudal swimming organ.

The skull of the Archeosaurus appears to have retained much of its primary cartilage in a state of ossification to have been chiefly active at the surfaces where, as in the combined dermo-neural ossifications of the skull in the sturgeons and salamandrid fishes,—e.g., Polypterus, Acipens, Lepidotes, these ossifications have started from centres more numerous than those of the true vertebral system in the skull of saurian reptiles. This gives the character of the present extinct order of Barraclia.

The skull is much flattened or depressed, triangular, with rounded angles, and the front one more or less produced according to the species; and in some species according to the age of the individual. The base is concave; the sides nearly straight, or slightly concave. The basi-occipital appears to have retained its primordial soft, unossified state.

Of the ex-occipitals, in a distinctly ossified state, no clear view has yet been had. The super-occipital (fig. 65, 4) is represented, as in the salamandrid fishes, by a pair of flat bones, more probably developed in the epithelial membrane and integument than in the cartilaginous osteocranium.

The pair of bones external to these, and forming the prominent angles of the occipital region, represent the "par-occipitals."

The lower peripheral surface of the basi-sphenoidal cartilage is ossified with a concave border towards the notochord behind, to the capsule of which it seems to have been attached. The basi-sphenoids were double-boned cartilaginous, and the protocerium there unaltered, as it was apparently in the Ichthyosaurus region.

The lateral ossifications above representing the "parietal" (7), form a pair of oblong flat bones, with the "foramen parietale" in the mid-suture. External to these, and wedged between them, in the super- and par-occipitals, are the pair of bones answering to the "mastoids" (8). They give attachment externally and below to the tympanic (28), and to a subsidiary bony plate, holding the position of that development of the mastoid which roofs over the temporal fossa in the Chelonia; it may be termed "super-squamosal" (the bone between S and 27 in fig. 65). The frontal bones (13), divided by a mid-suture, like the parietals, increase in length, and are continued far in advance of the orbits. The bone (12) which occupies the position of the post-frontal in Chelonia is divided into two centres, one articulating with the mastoid (8), the other, which is external to it, with the pre-frontal. This other bone may be termed the "post-orbital," as proposed by Von Meyer. The post-frontal extends forward above the orbit to meet the pre-frontal, separating the frontal (11) from the orbit, as in the sturgeon (Acipenser), Polypetres, and Lepidotes, and also in some Chelonia.

The pre-frontal extends far forward, terminating in a point between the nasal (15) and lacrymal. The nasals (13), divided by the median suture, extend to the external nostrils, their prolongation varying with the species and age of the individual.

Thus far the ossification of the superciliary of the skull of Archeosaurus closely conforms to that of the salamandroid ganoid fishes above mentioned; and the homologous bones are determinable without doubt.

The lacrymal bone obviously answers to the from large suborbital scale-bones of Ichthyosaurus; its large size and forward extension in Archeosaurus is a mark of that affinity.

The upper jaw consists of pre-maxillary (22), maxillary (21), and palatine bones. The pre-maxillaries are divided by a median suture, as in Lepidotes and Crocodilus, and are short bones, the breadth exceeding the length in A. latreutes, and also in the young of A. Decheni; but in the old animal opposite proportions prevail. In A. Decheni each pre-maxillary contains eight teeth; in A. latreutes not less than eleven. The maxillary (21) which extends from the pre-maxillary to beneath and beyond the orbit, presents a great length, varied according to species and age; it is of small vertical extent, and terminates in a point, which reaches the tympanic. Anteriorly it unites with the pre-maxillary, and enters into the formation of the anterior part of the orbit; medially it unites above with the lacrymal and suborbital, and below forms the outer boundary of the choanal aperture, joining the former anteriorly, and the palatine posteriorly. The palatine is a long narrow bone, rather expanded at both extremities; it forms anteriorly the hinder border of the choanal aperture, and medially throughout a great part of its extent the outer boundary of the great palatal vacuity. It supports a row of teeth, of which one or two at the fore part are of large size.

Between the orbit and the maxillary extends an oblong flat bone (26), forming the lower or outer border of the orbit, uniting with the pre-frontal and lacrymal anteriorly, with the maxillary below, and with the tympanic (9) another bone behind. In this position, and in its connection, it agrees with the condition of the crocodile, and also with the primitive condition of bones of fishes. These are unossified purely mucro-dermal bones, and may not be the homologues of the endo-skeletal malar bone of saurians, birds, and mammals.

To which of the bones, therefore, suborbital or malar, the one in question of the Archeosaurus answers, may be doubtful. The writer inclines to view it as a dermal ossification, and to conclude that, as in the higher Barraclia, the true malar and the zygomatic arch are not developed. Admitting the doubt on this point, the bone (26) may be termed the "suborbital."

With regard to the next bone (27), the same question, whether it answers to the squamosal in the crocodile, or whether it is a dermal ossification, applies.

If a homology with a determinate endo-skeletal bone in the crocodile and higher vertebrates were to be predicated, it would be as the "temporal." Essentially it indicates the tendency to excessive dermal ossification of the skull, like that which extends into the superficial temporal fascia from the squamosal and mastoid in the Chelonia; this separate ossification in Archeosaurus roofs over the temporal fossa. It appears like a posterior repetition of the supernumerary surface-bones called post-orbital and paramastoid, and, like them, corresponds in position with the posterior suborbital scale-bones in Asia and Lepidotes.

The hinder angles of the skull are formed by the tympanic; in

---

1 "Principal Forms of the Skeleton," Orr's Circle of the Sciences, p. 187, fig. 11. young individuals the tympanic does not extend backward beyond the par-occipital, but as age advances it projects further backward. It appears to abut internally against the pterygoid.

The two roots of the mandible were originally united at a short symphysis, not exceeding the breadth or depth of the jaw at that point; the depth gradually augmenting to near the articular end, but never exceeds a sixth, and is usually only an eighth of the length of the jaw, no definite coronoid process being developed; the upper and lower borders are nearly straight as far as the deepest part. The lower border behind this part rises rather abruptly to an angle, which is just below the articular pit. The angular element (30) presents a convexity answering to the point of ossification whence some faint ridges radiate upon its outer surface. The dentary (31), if it does not form the articular surface, lies very near it, and each ramus appears to be composed of these two bones. The anterior develops into coronoid rising. Neither articular nor splenial element has been clearly demonstrated. If an articular element has existed, it was probably very small.

From fishes the lower jaw of *Archegosaurus* differs in the great length or forward extension of the angular piece (30); but it resembles the piscine type in the simplicity of its composition. The angular piece is, however, longer in the Ganoids, e.g., *Asius*, *Polypterus*, *Lepidotes*, than in other fishes; and in *Lepidotes* its proportions are almost those of the *Archegosaurus*. It offers similar proportions to those in *Archegosaurus*, in the mandible of the *Azelot* and *Protox* (fig. 65).

The teeth have the simple conical pointed shape. They are implanted in the premaxillaries, maxillary, mandibular, and vomerine bone, and in a single row in each.

In the short premaxillaries there are from 8 (*A. Decani*) to 12 (*A. latiremis*); they are rather larger than the maxillary teeth. These follow in an unbroken series to beneath and beyond the orbit, and are about 30 in number; but their interspaces are such as would lodge double that number in the same extent of alveolar border.

The vomerine teeth are in a single row, parallel with and near to the maxillary row; one or two behind the choanae are much larger than the rest, which resemble the maxillary teeth in size. The mandibular teeth extend backward to the coronoid rising, and decrease in size from the front towards the hinder end.

Each tooth is implanted in a socket, and in a shallow cup-shaped socket, with a slightly raised border, to which the circumference of the tooth becomes anchylosed. The tooth is loosened by absorption and shed to make way for a successor. These are developed on the inner, hinder, and fore part of the base of the old tooth.

The teeth are usually shed alternately. They consist of osteodentine, dentine, and cement. The first substance occupies the centre; the last covers the surfaces of the tooth, but is introduced into its substance by many concentric folds, extending along the basal half. These folds are indicated by fine longitudinal, straight striæ along that half of the crown. The section of the tooth at that part (see fig. 65, tooth-section) gives the same structure which is shown in the section of a modern mammalian tooth.

The same principle of dental composition is exemplified in the teeth of most of the ganoid fishes of the carboniferous and Devonian systems, and is carried out to a great and beautiful degree of complication in the "old red" Dendrodonites.

The repetition of the same principle of dental structure in one of the earliest genera of Reptilia, associated with the defect of ossification of the endo-skeleton and the excess of ossification in the exo-skeleton of the head and maps, decisively illustrate the true affinities and low position in the reptilian class of the so-called *Archegosaurus*.

Resting upon and protected by the throat-plate in the middle line, there is a long slender bone, which must belong to the median series of the hyoid system, either basi- or uro-hyal; it is most probably homologous with the uro-hyal of Amphibia and other Perennibranchiates.

That two pairs of long slender bones projected outward and backward from the median series, is shown by more than one specimen of *Archegosaurus* (vol. xiii., fig. 10). The anterior pair is the longest; these are situated as if they had been attached, one to each side of the broad "throat-plate," which may have represented a basi-hyal. The anterior pair are homologous with the corresponding longer pair of appendages to the broad basi-hyal of Amphibia, and are ceratohyals. The shorter posterior pair answer to the branchiostegals in *Archegosaurus* and other Perennibranchiates. There is no such pair in the hyoidan arch of any known Saurotia.

Just external to the ends of the above lateral elements of the hyoid apparatus feebly traces of archæal series of bony nuclei were detected by Goldfuss, and interpreted by him as remains of partially ossified branchial arches. In this determination the writer agrees with Goldfuss. In all those specimens possessing them they present the outline of two or three arches in dots, or slightly curved series of dots or points. In the small relative size of these indications of branchial arches, the *Archegosaurus* agrees with the Amphibia.

No doubt, in the fully-grown *Archegosaurus*, the lungs would be equal to the performance of the required amount of respiration; but the retention of such traces of the embryonal water-breathing system in the adult leads to the inference that the animal must have affected a watery medium of existence for as great a proportion of its time as is observed to be the case in the existing perennibranchiate turtles; in which, notwithstanding the degree of development of the gills, the great proportion of the respiratory functions seems to be performed by the gills.

The additional marks of affinity to fishes which the *Archegosaurus* presents in its persistent motorduct, its cartilaginous basi-occipital, its dermal ossifications on the head, and its minute body-scales (fig. 65, scales), remove it further from the saurian reptiles, and exhibit it more strongly in the light of an obscure form between the Batrachians and the Ganoids.

Throat-Plates.—The under surface of the body between the head and trunk, is defended by broad bony plates, three in number. One is median and symmetrical, of an elongate lozenge shape, with the angles rounded off; slightly convex externally, a little produced at the middle of the hinder half into something like a low quadrate-bone. The under surface is ornamented by radiating furrows, except at so much of the marginal parts as is overlapped by the lateral pieces, and by the scapular arch. The latter three plates are attached to the anterior half of the sides of the median one, are shaped like beetles' elytra, and converge forwards. Their centre of ossification is towards their outer and back part, from which the external ridges and grooves radiate towards the inner border.

Von Meyer compares these dermal shields to the ento- and epi-sternal elements of the plastrum of Chelonia; their truer homology seems to the writer to be with the median and lateral large throat-plates or scales of *Megalichthys* and *Sudis gigas*. The ento-sternal element is the only endo-skeletal piece uncombined with a dermal ossification in most Chelonia. The epi-sternal, like the hyoid and hypophract, supports the abdominal ribs, with superadded dermal ossifications in Chelonia.

The scapulae (51) are instructively exhibited in the very young specimen of the *Archegosaurus* figured in t. xiv., fig. 4, of Von Meyer's treatise. The coracoids being doubtless wholly cartilaginous at that stage, are not discernible in the specimen referred to. The upper slender end of the scapula is opposite the side of the vertebral column, about the fifth neurapophysis from the head, and it curves gently downward and forward, expanding at its humeral end.

This expansion is more sudden in the fully-developed animal, giving the bone the shape of a rudder, and the direction of the broadest part is forward. At least in the specimen (the great majority) in which the skeleton is seen from above, the slender dorsal end of the scapula is seen overlying, or near the hinder border of, the lateral throat-plate, and it extends outward and backward to its expanded humeral end.

The coracoids (52) are a pair of flat reniform plates, with the convex border turned forward, the concave one backward; they seem to have overlapped the smooth margins of the posterior half of the median throat-plate. It is most probable that, as in Amphibia, a portion of the broad coracoid remained in the cartilaginous state, and that the full reniform plate answers to the ossified part of that coracoid which it resembles in shape and relative position.

The position of the slender scapulae, styiform and rib-like, as in the Perennibranchiates, is distinctly shown in t. xviii., figs. 1 and 2, of M. von Meyer's treatise. The coracoids, as in Amphibia, form the chief part of the articular cavity for the humerus.

The perennibranchiate affinities of *Archegosaurus* are shown as clearly by the scapular as by the hyoidian arch. The fore-limb does not exceed half the length of the head. The humerus (53) is a short thick bone, slightly constricted at the middle, expanded and rounded at both ends, the proximal one being the largest. For some time the bone is hollow and open at each end; when ossification finally closes the terminal apertures, it shows that the ends were connected to the coracoids and to the fore arm by interposed ligaments—just as in true turtles, by a synovial joint. Of the two bones of the fore arm, the ulna is a little longer and larger than the radius (54). Both bones present the simplest primitive form, gently constricted in the middle, with the proximal... Reptilia—ends a little concave, the distal ones a little convex. The space between the antibrachium and the metacarpus plainly bespeaks the mass of cartilage representing, as in Amphibian, the carpal segment (56) in *Archegosaurus*. No trace of a carpal bone is found save in the largest and oldest examples, in which five or six small roundish ossicles are aggregated near the ulnar side of the carpus. Four digits are present; and considering the pollex to be, as usual, wanting, the second digit, answering to the median of pentadactyle feet, is the largest, and includes at least four phalanges (58); those, with the metacarpals (57), are long, slender, terminally expanded, and truncate. They obviously supported a longish, narrow, pointed pad. The outermost or middle finger was the shortest, and has the shortest metacarpal and first phalanx.

It is true that in *Myosaurus* the fore limbs are relatively as short as in *Archegosaurus*, and the coltlike crocodile recalls the arrest of development of the same limbs in the marsupial Potorous; but in *Archegosaurus*, not only is the small size of the fore limbs, but also their type of structure, especially that of their scapular arch, closely in accordance with that in the Perennibranchiates, as shown in the tridactyle fore limbs of the *Proteus anguinus*, of which a figure is added to that of the *Archegosaurus* in fig. 65.

The ilium (62), like the scapula, is expanded at its articular or femoral end. It is less long and slender; one border is straight, the other concave, by the expansion toward that border of the femoral end. Two shorter bones on each side complete the pelvis below. One is of a simple form, straight, thicker in proportion to its length than in the ilium.

The other bone is shown, with its fellow, in t. xiii., fig. 6, and xviii., figs. 8 and 9, of Von Meyer's treatise. That author compares the pair of bones to the *Apateon* in shape; they may be the pubic bones. On this hypothesis, they are restored to their true position at 64 (pubis) in fig. 65. The femur (65) is slightly expanded, and truncate at both ends; it is not longer than the ilium. The tibia (66) and fibulae are separate bones, like those of the fore arm; the margins, which are turned toward each other, are most concave. They are rather more than half the length of the femur.

The foot-bones are separated by a fibro-cartilaginous tarsal mass (68) from those of the leg. The form of the phalanges, expanded and truncate at both ends, bespeak their simple ligamentous joints, and show them supported, like the fore limb, a fin or limb adapted simply for swimming. The argument for the saurian affinities of *Archegosaurus*, based by V. Meyer on the short fore-limbs of *Myosaurus*, already invalidated by the difference of structure, is controverted by the fact, that the hind limbs of *Archegosaurus*, like those of the Perennibranchiates, are not only as simple in structure, but also as short, as the fore limbs.

In 1852 Sir Charles Lyell and Mr Dawson, in the course of their investigations of the coal strata of Nova Scotia, remarkable for the erect fossil trees in certain parts, discovered in the hollow of the trunk of one of these trees (*Sigillaria*, 2 feet in diameter), which was wholly converted into coal, some small bones, which Professor Wyman of Boston surmised to have belonged to a batrachian reptile. By the professor's advice they were brought to England and submitted to the writer, who has described and figured them as batrachian, under the name *Dendrerpeton Acadianum*, and with close affinities, from the plicated structure of the teeth, the sculpturing of some broad cranial plates, and the structure and proportions of certain limb-bones, to the genus *Apateon* or *Archegosaurus*.

Genus Raniceps.—About the centre of the great carboniferous basin of Ohio, United States, at the mouth of the "yellow creek," is a seam of coal 8 feet in thickness, the lower four inches of which is "cannel coal." In this has been found the skull, part of the vertebral column, scapular arch, and fore limbs of a reptile referred by Professor Wyman to the batrachian sub-class, under the name of *Raniceps*. Two closely-allied fossils, also referred to *Batrachia*, have been found in the same formation and locality.

Order II.—Labirinthodonta.

Genus Baphetes, Ow.

Sp. Baphetes planiceps.—In January 1854 the writer communicated to the Geological Society of London a description of part of a fossil cranium of an animal, from the Pictou coal, Nova Scotia, measuring 7 inches across the orbit. From the characters then specified, the fossil was determined to be the fore part of the skull of a saurid Batrachian of the extinct family of the Labyrinthodonts. It agreed with them in the number, size, and disposition of the teeth; in the proportions and mode of connection of the premaxillaries, maxillaries, nasals, prefrontals and frontals; and in the resultant peculiarly broad and depressed character of the skull. The traces of the nostrils were less definite and satisfactory than the remains of the orbits; but the latter were decisive against a piscine nature. The fossil also presents the same well-marked external sculpturing as in the Labyrinthodonts; and amongst the genera that have been established in that family, the form of the end of the muzzle, or upper jaw, in the Pictou coal specimen, best accorded with that in the *Capitosaurus* and *Metopias* of Von Meyer and Burmeister. But the orbits had been evidently larger and of a different form than in the reptiles so called; and, for the convenience of distinction and reference, the writer proposed to name the fossil *Baphetes planiceps* (Bärro, I dip or dive), in reference to the depth of its position and the shape of its head.

Being thus introduced at the carboniferous period to the labyrinthodont order, which attained its full development in the triassic period, we shall proceed to notice the more decisive evidences and typical illustrations of that extinct order of *Batrachia*.

The name of this sub-class is from the Greek word *batrachos*, signifying a frog; and it is represented in the present animal population of England by a few diminutive species of frogs, toads, and newts, or water-salamanders. But, at the period of the deposition of the new red sandstone, in the present counties of Warwick and Cheshire, the shores of the ancient sea, which were then formed by that sandy deposit, were trodden by reptiles having the essential bony characters of the *Batrachia*, but combining these with other bony characters of crocodiles, lizards, and ganoid fishes; and exhibiting all under a bulk which, as made manifest by the fossils and foot-prints, rivalled that of the largest reptiles of the present day. The form of the largest labyrinthodonts, if we may judge by the great breadth and flatness of the skull, most, however, have more resemblance to that of the toad or land-salamander.

The Batrachians had no fixed type of external form like the higher orders of reptiles, but some, as the broad and flat-bodied toads and frogs, most resemble the Cheloniens, especially the soft-skinned mud-tortoises (*Trionyx*); other Batrachians, as the *Cecilia*, resemble Ophidiens; a third group, as the newts and salamanders, represent the Lacertians; and among the perennibranchiate reptiles there are species (*Siren*) which combine with external gills the mutilated condition of the apodal fishes.

Thus it will be perceived that, even if the entire skeleton of one of the new red sandstone Batrachians had been obtained, there is no fixed or characteristic general outward form in the batrachian order whereby its affinity to that group could have been determined. The common characters by which the Batrachians, so diversified in their reptilian ancestors, are associated into one group or sub-class of reptiles, besides being taken from the condition of the circulating and generative systems, and other perishable parts, are, however, fortunately as strongly manifested in modifications of the skeleton, and principally in the skull. This is joined to the atlas by the medium of two tubercles, developed exclusively from the occipitals; the bony palate is formed chiefly by two broad and flat bones, called "vomerine" by Cuvier, and generally supporting teeth. It is only in the Batrachians among reptiles that examples are found of two or more rows of teeth on the same bone, especially on the lower jaw (*Cecilia*, *Sirens*). With regard to vertebral characters, no such absolute batrachian modifications can be adduced as those above mentioned from the anatomy of the cranium. Some Batrachians, as is well known, have the vertebrae united by ball-and-socket joints, as in most recent reptiles; others by biconcave joints, as in a few recent and most extinct Saurians. Some species have ribs, others want those appendages; the possession of ribs, therefore, even if longer than those of the *Cecilia*, by a fossil reptile combining all the essential batrachian characters of the skull, would not be sufficient ground for pronouncing such reptile to be a Saurian.

---

1 Quarterly Journal of the Geological Society, vol. ix., 1853. 2 American Journal of Science and Arts, March 1857. Much less could its saurian nature be pronounced from the circumstance of its possessing large conical striated teeth,—as the ordinary characters of size, form, number, and even presence or absence of teeth, varies much in existing Batrachians, the location of teeth on the vomerine bones being the only dental character in which they differ from all other orders of reptiles.

The writer's first acquaintance with the remarkable fossils under consideration was founded on the examination, in 1840, of portions of teeth found in the red sandstone of Coton End quarry, Warwickshire. The external characters of these teeth corresponded with those which had previously been discovered by Professor Jaeger in the German Keuper formation in Württemberg, and on which the genus *Mastodonsaurus* had been founded (fig. 66).

The results of a microscopic examination of the teeth of the *Mastodonsaurus* from the German Keuper, and of those from the new red sandstone of Warwickshire have been detailed in the *Art. Odontology* (vol. xvi., p. 412). They proved that the teeth from both localities possessed in common a very remarkable and complicated structure (fig. 67), to the principle of which,—viz., the convergence of numerous inflected cells of the external layer of cement towards the pulp cavity, a very slight approach was made in the fang of the tooth of the *Ichthyosaurus*, and that a closer approximation to the labyrinthic structure in question was made by the teeth of several species of fishes, as by those of *Anarhichas*, while the teeth of existing Batrachians were simple like those of most Saurians.

Thus, inasmuch as the extinct animals in question manifested in the intimate structure of their teeth an affinity to fishes, it might be expected that, if they generally belonged to the class of reptiles, the rest of their structure would manifestly bear the characters of the lowest order,—viz., the Batrachia, the existing members of which pass, though not by the dental character alluded to, yet by so many other remarkable degradations of structure, towards fishes. Now it has happened that, in the same formation in Württemberg from which the labyrinthic teeth of the so-called *Mastodonsaurus* had been derived, a fragment of the posterior portion of the skull has been obtained, showing the development of a separate condyle on each ex-occipital bone; whence Professor Jaeger, recognizing the identity of this structure with the batrachian character above mentioned, founded upon the fossil a new genus of Reptilia which he called *Salamandrodoides giganteus*. Subsequent discoveries, however, satisfied the present that the bi-condyloid fragment, which originally constituted the genus *Salamandrodoides*, belonged to the same reptile as the teeth on which he had founded the genus *Mastodonsaurus*. The following fossils, from the new red sandstone of Warwickshire, gave additional proof of the batrachian nature of the genus to which those fossils belong, with the establishment of five distinct species, one of which is most probably identical with the *Mastodonsaurus salamandrodoides* of Professor Jaeger. In reference to the generic denomination *Mastodonsaurus*, it unavoidably recalls the idea of the mammalian genus *Mastodon*, or else a mammilloid form of tooth, while in all the teeth of reptiles so called are originally, and most of them are permanently, of a mammilloid or not of a mammilloid form; secondly, because the second element of the word *saurus*, indicates the genus to belong to the saurian and not to the batrachian order of reptiles. For these reasons, the writer has proposed to designate the genus in question *Labyrinthodon*, in allusion to the peculiar and characteristic structure of the teeth (fig. 67).

**Fig. 67.**

Transverse section of a tooth of the *Labyrinthodon* (magn.)

The specimens from British localities are referable to five species,—viz., 1. *Labyrinthodon salamandrodoides*; 2. *L. leptognathus*; 3. *L. pachymastus*; 4. *L. centricornis*; and 5. *L. scutulatus*; and we shall here briefly notice the characters exhibited by the bones assignable to the second, third, and fifth species.

*Labyrinthodon leptognathus.*—The remains of this species consist of fragments of the upper and lower jaws, two vertebrae, and a sternum. They were found in the new red sandstone quarries at Coton End, in Warwickshire.

A dorsal vertebra from Coton End presents further evidence of the batrachian nature of the *Labyrinthodon*. It has concave articular cavities at the extremities of the body,—a condition now known among existing reptiles only in the Geckos, and in the lower or perennibranchiate division of Batrachians. It is a common structure in extinct Saurians, but the depth of the vertebral articular cavities in the *Labyrinthodon* exceeds that in the amphibian Crocodilians and in most Plesiosauria. The body of the vertebra is elongate and sub-compressed, with a smooth but not regularly curved lateral surface, terminating anteriorly in a slightly projected longitudinal median ridge; and it exhibits the same exceptional modifications of the reptilian type as the vertebrae of existing Batrachians, in having the superior arch or neurapophysis anchored with the centrum. From each side of the base of the neural arch a thick and strong transverse process extends obliquely outwards and upwards.

A symmetrical bone, resembling the episternum of the *Ichthyosaurus* was associated with the preceding remains. It consists of a stem or middle, which gradually thicken to the upper end, where cross pieces are given off at right angles to the stem, and support on each a pretty deep and wide groove indicating strongly the presence of clavicles, and thus pointing out another distinction from crocodiles, in which clavicles are wanting. Most Batrachians possess these bones.

The modifications of the jaws, and more especially those of the bony palate of the *Labyrinthodon leptognathus*, prove the fossil to have been essentially Batrachian, but with affinities to the higher Sauria, leading, in the form of skull and the sculpturing of the cranial bones, to the crocodilian group, in the collocation of the larger fangs at the anterior extremities of the jaws to the Plesiosauria, and in one part of the dental structure, in the form of the episternum, and the bi-concave vertebra, to the *Ichthyosaurus*. Another marked peculiarity in this fossil is the anchylosis of the bases of the teeth to distinct and shallow sockets, by which it seems to resemble the Sphenacodonts, instead of the fish. From the absence of any trace of deposition at the inner side of the base of the functional teeth, of alveoli of reserve for the successional teeth, it may be concluded that the teeth were reproduced, as in the lower Batrachians and in many fishes, in the soft mucous membrane which covered the alveolar margin, and that they subsequently became fixed to the bone by anchylosis, as in the pike and Lophius. This anatomical fact militates strongly against the idea that the *Labyrinthodon* is a Saurian. Labyrinthodon pachygnathus.—The remains of this species, which have been obtained, consist of portions of the lower and upper jaws, an anterior frontal bone, a fractured hammer, an ilium with a great part of the acetabulum, the head of a femur, and two ungual phalanges. A portion, nine and a half inches long, of a right ramus of a lower jaw, in addition to the characters common to it and the fragment of the lower jaw of the L. leptognathus, in the structure of the angular and dentary pieces, shows that the outer wall of the alveolar process is not higher than the inner, as in frogs and toads, the salamanders and monotremes, in all of which the base of the teeth is anchylosed to the inner side of an external alveolar plate. The smaller serial teeth are about forty in number, and gradually diminish in size as they approach both ends, but chiefly so towards the anterior part of the jaw. The sockets are close together, and the anterior ones larger. The greatest ivory teeth were probably those in each symphysis, and the length of the largest is considered to have been one and a half inches. The base of each tooth is anchylosed to the bottom of its socket, as in somecodon and sarolell fishes; but the Labyrinthodon possesses a still more ichthyic character in the continuation, preserved in this specimen, of a row of small teeth anterior and external to the two or three larger tusks. The premaxillary bone presents the same peculiar modification as in the higher organized Batrachia, the palatal process of the premaxillary extending beyond the outer plate both externally and, though in a less degree, internally, where it forms part of the boundary of the anterior palatal foramen, whence the outer plate rises in the form of a shelf, projecting from the longitudinal suture in the upper part of the palatal process; it is here broken off near the margin, and the fractured surface gives the breadth of the base of the outer plate, stamping the fossil with a Batrachian character conspicuous above all the saurian modifications by which the essential nature of the fossil appears at first sight to be masked.

In the pre-frontal bone there are indications of crocodilian structure. Its superior surface is slightly convex and pitted with irregular impressions; and from its posterior and outer part it extends downwards a broad and slightly concave process, which appears to be the anterior boundary of the orbit. This process presents near its upper margin a deep pit, from which a groove is continued forwards; and in the corresponding orbital plate of the crocodile there is a similar but smaller formation.

From the remains of the cranium of the L. pachygnathus, it is evident that the facial or maxillary part of the skull was formed in the main after the crocodilian type, but with well-marked batrachian modifications in the premaxillary and inferior maxillary bones. The most important fact which they show is, that this sarolell Batrachian had subterminal nostrils, leading to a wide and shallow nasal cavity, separated by a broad and almost continuous palatal floor from the cavity of the mouth; indicating, with their horizontal position, that their posterior apertures were placed far behind the anterior or external nostrils; whereas in the air-breathing Batrachia the nasal meatus is short and vertical, and the internal apertures pierce the anterior part of the skull. It may be inferred, therefore, that the apparatus for breathing by inspiration must have been similar in the Labyrinthodon as in the crocodile and hence still further, that the skeleton of the Labyrinthodon will be found to be provided with well-developed costal ribs, and not, as in most of the existing Batrachians, with merely rudimentary styles. Since the essential condition of this defective state of the ribs of Batrachians is well known to be their fish-like mode of generation and necessary distension of the abdomen, it is probable that the generative economy of these fossil reptiles, in which the more complete ribs would prevent the excessive enlargement of the ovaria and oviducts, may have been similar to that of saurian reptiles.

A fragment of a vertebra of L. pachygnathus presents analogous characters to the vertebra of the L. leptognathus previously noticed.

Of the few bones of the extremities which have come under the writer's inspection, one presents all the characteristics of the corresponding part of the humerus of a toad or frog, viz., the convex, somewhat transversely extended articular end, the internal longitudinal depression, and the well-developed deltoid ridge. The length of the fragment is two inches, and the breadth is thirteen lines. The ridges are moderately thick and compact, with a central medullary cavity. In its structure, as well as in its general form, the present bone agrees with the batrachian, and differs from the crocodilian type.

In the right ilium, about 6 inches in length, and in the acetabulum, there is a combination of crocodilian and batrachian characters. The acetabular cavity is bounded on its upper part by a produced and sharp ridge, as in the frog, and not emarginate at its anterior part, as in the crocodile.

As the fragment of the ilium was discovered in the same block as the two fragments of the cranium and the portion of the lower jaw, it is probable that they may have belonged to the same animal; and if so, as the portions of the head correspond in size with those of the head of a crocodile six or seven feet in length, but the acetabular cavity with that of a crocodile 25 feet in length, then the hinder extremities of the Labyrinthodon must have been of disproportionate magnitude compared with those of existing Saurians, but of approximate magnitude with some of the living anourous Batrachians. That such a reptile, of a size equal to that of the species whose remains have just been described, existed at the period of the formation of the new red sandstone is abundantly manifested by the remains of those similar impressions to which the term Chirotherium has been applied. Other impressions, as those of the Chirotherium Hercules, correspond in size with the remains of the Labyrinthodon salamandroides, which have been discovered at Guy's Cliff. The head of a femur from the same quarry in which the ilium was found is shown to correspond in size with the articular cavity of the acetabulum. The two toe-bones, or terminal phalanges, resemble those of Batrachians in presenting no trace of a nail, and from their size they may be referred to the hind feet of the L. pachygnathus.

An entire skull of the largest species discovered in the new red sandstones of Wurttemberg; a lower jaw of the same species found in the same formation in Warwickshire; some vertebræ, and a few fragments of bones of the limbs, have recently, with the indication of size and shape of the trunk of the animal yielded by the series of distinctive foot-prints, as the basis of the restoration of the Labyrinthodon salamandroides, at the Crystal Palace. It is to be understood, however, that, with the exception of the head, the form of the animal is necessarily more or less conjectural.

Labyrinthodon cucullatus.—The remains to which this specific designation has been applied compose a closely and irregularly aggregated group of bones imbedded in sandstone, and manifestly belonging to the same skeleton; they consist of four vertebrae, portions of ribs, a humerus, a femur, two tibii, one end of a large flat bone, and several small ossaceous dermal scales. The mass was discovered in the new red sandstone at Leamington, and was transmitted to the writer in the summer of 1840.

The vertebrae present biconcave articular surfaces similar to those of the other species. In two of them the surfaces slope in a parallel direction obliquely from the axis of the vertebrae, as in the dorsal vertebrae of the frog, indicating a habitual inflexion of the spine, analogous to that in the humped back of the frog. The neural arches are anchylosed to the vertebral body. The spinous process rises from the whole length of the middle line of the neural arch, and its chief peculiarity is the expansion of its elongated summit into a horizontally-flattened plate, sculptured irregularly on the upper surface. A similar flattening of the summit of the elongated spine is exhibited in the large atlas of the toad.

The body of the vertebra agrees with that of the L. leptognathus. The humerus is an inch long, regularly convex at the proximal extremity, and expanded at both extremities, and constructed for the middle of the body of a large bird; the short and flattened bone is bent at a subacute angle with the distal extremity, and resembles most nearly the anchylosed radius and ulna of the Batrachia.

The femur wants both the extremities; its shaft is subtriangular and slightly bent, and its walls are thin and compact, including a large medullary cavity. The tibiae are as long, but thicker and stronger than the femur. They had lost their articular extremities, but exhibited that remarkable compression of their distal portion which characterizes the corresponding bone in the Batrachia; they likewise have the longitudinal impression along the middle of the flattened surface. Were more of the skeleton of the above-defined species of Labyrinthodon known, they might present differences of sufficient value. Such differences in the forms and proportions of the skull, and in the relative position of the orbits, of specimens that have been discovered separately in the triassic sandstones of Germany, have been so interpreted.

In the Labyrinthodon (Mantodonaurus) Jaegeri—the largest of the species—the skull is triangular, the two condyles projecting from the middle of the base; the sides are straight, and converge to the obtuse apex. The orbits are oval, narrowest anteriorly, and are situated nearly midway between the fore and back part of the skull. The nostrils are very small, and are as wide apart as the orbits.

Labyrinthodon (Trematosaurus) Brunnii, Von Meyer.—The name Trematosaurus was given by Braun to a labyrinthodont reptile, in reference to the partial formation at that time deemed to be peculiar to the genus. It seems hard to be convinced that all the fragments of the genus was founded on an unusually perfect skull discovered in the rich fossiliferous hunter-sandstein of Bernburg. It is about one foot long, and, relatively to its basal breadth, it is longer and narrower than in L. Jaegeri, the sides converging at a more acute angle. The orbits are elliptical, situated in the middle of the skull, and Reptilia, wider apart than in L. Jogarei; the nostrils are relatively nearer together, their interspace being only half that in the L. Jogarei.

Labyrinthodon (Motopius) diagnosticus, H. von Meyer.—In this species the skull is broader in proportion to its length than in the foregoing; the sides are convex as they converge to the obtuse muzzle. The orbits are small, of a wide elliptical form, situated in the anterior third of the skull; they are twice as wide apart as are the nostrils. The parietal foramen is near the occipital ridge. The remains of this species are from the upper beds of the Keuper sandstone in Würtemberg.

The Labyrinthodon (Osteocephalus) arenaceus, Münster, is distinguished by a much broader and almost truncate muzzle. The orbits are elliptic, and situated almost wholly in the hinder third of the cranium; their interspace is the same as that between the nostrils, which are relatively as large as in L. Brunnii.

The name Zygosaurus appears to have been applied with better grounds, by Eichwald, to a labyrinthodont reptile from the Permian cupiferous beds at Oremberg. It has the parabolic skull of L. Jogarei and L. diagnosticus; the orbits large, and divided by an interval less than their own diameter. The temporal fossae are relatively larger, and bounded by stronger zygomatic arches, and seem not to have been roofed over by bone. The dentition is strictly labyrinthodont.

Osteocephalus Valenciennes is a genus and species founded by Von Meyer on a portion of a lower jaw, containing fifty teeth lodged in rather a deep groove, but apparently presenting the labyrinthine structure. The specimen is from the bunter sandstone of Soulitz-les-Bains.

Xestorhyncha Perini.—By this name M. von Meyer would indicate certain flat cranial bones, sculptured like those of Labyrinthodon, but with a peculiarly polished ganoid-like surface, from the muschelkalk of Luxemburg.

In all the foregoing forms of Labyrinthodonts, represented by complete crania, with the exception perhaps of Zygosaurus, the supplemental osseous plates roofing over the temporal fossae are present, as in Archeosaurus, viz., the "post-orbital" and the "super-squamosal" bones. In all of them the occipital condyles are distinct, forming a pair; and in all the vomer is divided and bears teeth. The structure and disposition of the entire dental system is strictly labyrinthodont.

The relation of these remarkable reptiles to the saurian order has been advocated to be one of close and true affinity, chiefly on the character of the extent of ossification of the skull, and of the outward sculpturing of the cranial bones. But the true nature of some of these bones appears to have been overlooked, and the glance of research for analogous structures has been too exclusively upward. If directed downward from the Labyrinthodonts to the Archeosaurus and certain ganoid fishes, it suggests other conclusions.

The conformity of pattern in the dermal, semidermal, or neurodermal bones of the outwardly well-ossified skull of Polypterus, Lepidosteus, Sturio, and other salamandroid-ganoid fishes, with well-developed lung-like air-bladders, and of the same skull-bones in the Archeosaurus and the Labyrinthodonts; the persistence of the notochord (chorda dorsalis) in Archeosaurus, as in Sturio; the persistence of the notochord and branchial arches in Archeosaurus, as in Lepidosiren; the absence of occipital condyle or condyles in Archeosaurus, as in Lepidosiren; the presence of labyrinthine teeth in Archeosaurus, as in Lepidosteus and Labyrinthodon; the large median and lateral throat-plates in Archeosaurus, as in Megalichthys, and in the modern Arapaima and Lepidosteus—all these characters point to one great natural group, peculiar for the extensive gradations of development, linking and blending together fishes and reptiles within the limits of such group. The salamandroid (or so-called "sauroid") Ganoids—Lepidosteus and Polypterus—are the most piscine, the true Labyrinthodonts are the most reptilian, of the group. The Lepidostiren and Archeosaurus are intermediate gradations, one having more of the piscine, the other more of the reptilian characters. The Archeosaurus conducts the march of development from the fish proper to the labyrinthodont type; the Lepidostiren conducts it to the perennibranchiate batrachian type. Both illustrate the artificiality of the supposed class distinction between fishes and reptiles, and the naturality of the "Haemacrymes," or cold-blooded Vertebrata, as the one natural and truly definite group. There is nothing in the known structure of the so-called Archeosaurus or Mastodonosaurus that truly indicates a belonging to the saurian or crocodilian order of reptiles. The exterior ossifications of the skull and the canine-shaped labyrinthine teeth are both examples of the salamandroid modification of the ganoid type of fishes.

The small proportion of the fore limb of the Mystriosaurus in nowise illustrates this alleged saurian affinity; for though it be as short as in Archeosaurus, it is as perfectly constructed as in the crocodile, whereas the short fore limb of Archeosaurus is constructed after the simple type of that of the Proteus and Siren. But the futility of this argument of the sauroid affinities is made manifest by the proportions of the hind limb of Archeosaurus. As in Proteus and Amphibuna, it is as stunted as the fore limb; whereas in Mystriosaurus, as in other Teleosaurus, the hind limbs are relatively larger and stronger than in the existing crocodiles. M. von Meyer leaves the hind limb out of sight in his advocacy of the saurian nature of the so-called Archeosaurus. One regrets that Von Meyer's original name Apateon, though proposed to express his scepticism of the alleged nature of the fossil submitted to him in 1844 by Dr Gergens, was not retained by Professor Goldfuss. It is still more to be regretted that a compound name should in any case be adopted or constructed, where the proof of the affinity it may be meant to indicate is not perfect. Archeosaurus, like Mastodonosaurus, will become at length mere arbitrary terms; but until then, they will really recall or express little more than the mistaken views of the inventors of those names in respect of the true affinities of the remarkable extinct piscine reptiles to which they have been applied.

Fig. 68 gives a reduced view of a portion of new red sandstone, with three pairs of foot-prints in relief.

Consecutive impressions of such prints have been traced for many steps in succession in quarries of that formation in Warwickshire, Cheshire, and also in Lancashire, more especially at a quarry of a whitish quartzose sandstone at Storton Hill, a few miles from Liverpool. The foot-marks are partly concave and partly in relief; the former are seen upon the upper surface of the sandstone slabs, but those in relief are only upon the lower surfaces, being in fact natural casts, formed on the subjacent foot-prints as in moulds. The impressions of the hind foot are generally 8 inches in length and 5 inches in width; near each large footstep, and at a regular distance—about an inch and a half—before it, a smaller print of the fore foot, 4 inches long and 3 inches wide, occurs. The footsteps follow each other in pairs, each pair in the same line, at intervals of about 14 inches from pair to pair. The large as well as the small steps show the thumb-like toe alternately on the right and left side, each step making a print of five toes.

Foot-prints of corresponding form, but of smaller size, have been discovered in the quarry at Storton Hill, imprinted on five thin beds of clay, lying one upon another in the same quarry, and separated by beds of sandstone. From the lower surface of the sandstone layers the solid casts of each impression project in high relief, and afford models of the feet, toes, and claws of the animals which trod on the clay.

Similar foot-prints were first observed in Saxony, at the village of Hessburg, near Hillburghausen, in several quarries of a gray quartzose sandstone, alternating with beds of red sandstone, and of the same geological age as the sandstones.

This generic term has been applied to another fossil by Eichwald. of England that had been trodden by the same strange animal. The German geologist who first described them proposed the name of *Cheirotherium* (cheir, the hand; therion, beast) for the great unknown animal that had left the foot-prints, in consequence of the resemblance, both of the fore and hind feet, to the impression of a human hand; and Dr Kaup conjectured that the animal might be a large species of the opossum kind. The discovery, however, of fossil skulls, jaws, teeth, and a few other bones, in the sandstones exhibiting the foot-prints in question, has rendered it more probable that both the foot-prints and the fossils are evidences of the same kind of huge extinct batrachian reptiles.

All the labyrinthodont remains from the Warwick and Leamington sandstones agree in their essentially batrachian nature with those from the German keuper, and other new red or triassic strata. The impressions of the *Cheirotherium* resemble those of the foot-prints of a Batrachian; but are not identical with those of any known Batrachian or other reptile. They show a papillose integument like that on the sole of certain Geckos, and which may be another mark of sauroid departure from the modern batrachian type. In the attempt to solve the difficult problem of the nature of the animal which has impressed the new red sandstone with the cheirotherian foot-prints, we cannot overlook the fact, that we have in the Labyrinthodons also batrachian reptiles, differing as remarkably from all known Batrachia, and from all other reptiles in the structure of their teeth; both the footsteps and the fossils are, moreover, peculiar to the new red sandstone; the different size of the foot-prints referred to different species of *Cheirotheria* correspond with the different size of ascertained species of *Labyrinthodon*; and the present facts best support the hypothesis, that the foot-prints called "cheirotherian," are those of labyrinthodont reptiles.

---

**Sub-class 2.—SAURIA.**

**Order I.—THECODONTIA.**

**Genus PROTOREOSAURUS, Von Meyer.**

Sp. *Protoresaurus Speneri*, Von M.—The first fossil Saurian on record is that which records the circumstance by its generic name, and honours its describer by the specific one. The slab of "copper-slate" from the Permian beds of Eisenach in Thuringia, displaying, either in fossils or impressions, the skull, vertebral column, and bones of the fore foot of the reptile in question, was figured and described by Spener, a physician at Berlin, in 1710. The original specimen is now in the museum of the Royal College of Surgeons, London. It was obtained from a copper-mine near Eisenach, at a depth of 100 feet from the surface.

A second specimen, showing the two fore limbs, a hind limb, and part of the trunk, was described by Link in 1718. Cuvier gives copies of portions of two other specimens in his *Osseens Fossiles*.

The healthy, honest mind of Speerer is shown by the conclusions which he formed from the state of preservation of his specimen—"omnia, enim, minutissima etiam apophyses, spinæ;" &c.,—and from its association with equally well-preserved remains of fishes, and even of the delicate leaves of plants, against the notions of those fossils merely simulating, and never having been, the living organisms which they represented—nations which were then advocated under the sounding phrase of "plastic force," as they have lately been under that of "prochronism." Speerer's only doubt was, whether the reptile had been a crocodile or a lizard; but he inclined to the former view, on account of the proportions of the head to the trunk. He then enters upon speculations as to how a crocodile could have come into Germany; and shows the usual effect of a mind biased by a hypothetical diluvial catastrophe, not demonstrated by observation and inductive research, and to the extent of such bias benumbed in the exercise of the faculty for the acquisition of natural truth.

The seven cervical vertebrae are proportionally larger than in any known recent or fossil terrestrial or aquatic Saurian; they resemble in this respect the cervical vertebrae of Pterodactyls; the tail is long, and its vertebrae differ from those of all other known reptiles, recent or fossil, in having the spinous processes bifurcate, diverging in the direction of the axis of the body.

The muscular power of the neck is indicated by traces of bone tendons. The dorsal vertebrae exceed eighteen in number, and have higher spines than in the modern Monitor; the dorsal ribs are long, and longitudinally impressed. The hind limb is much longer than the fore limb, and the leg is longer, in proportion to the thigh and foot, than in the Monitors. The teeth are sharp-pointed, slender; there appear to be at least twenty in both upper and lower jaws in Spener's specimen.

The writer concludes, from the length and strength of the tail and the peculiar provision for muscular attachments in that part, and from the proportions of the hind limbs, that the *Protoresaurus* was of aquatic habits, and that the strength of its neck and head, and the sharpness of its teeth, enabled it to seize and overcome the struggles of the active fishes of the waters which deposited the old Thuringian copper-slates.

**Genus THECODONTOSAURUS.**

Sp. *Thcodontosaurus antiquus*.—In 1836 certain reptilian remains from the "dolomitic conglomerate" at Reiland, near Bristol, were described by Messrs Riley and Stutch-

---

1 *Miscellanea Berolinensis*, 4to, i., p. 99, figs. 24 and 25. 2 Ed. 8vo, 1836, pl. ccxxxvii., figs. 1 and 2. 3 A character first pointed out in the writer's "Report on British Fossil Reptiles," Trans. of Brit. Assoc., 1841, p. 155. The matrix has been referred to the Permian period; it is now thought by some good observers to be not older than the Triassic.

The teeth in these reptilian fossils are lodged in distinct sockets in *Thecodontosaurus*; they are arranged in a close-set series, slightly decreasing in size towards the posterior part of the jaw; each ramus of the lower jaw contained twenty-one teeth. These are conical, rather slender, compressed and acutely pointed, with their anterior and posterior finely serrated edges, the serratures being directed towards the apex of the tooth; the outer surface is more convex than the inner, and the apex is slightly recurved; the base of the crown contracts a little to form the fang, which is subcylindrical.

**Genus Palaeosaurus.** Riley and Buttenbury.—In the same formation as contained the jaw and teeth of the *Thecodontosaurus* two other teeth were separately discovered, differing from the preceding and from each other; the crown of one of these teeth measuring 9 lines in length and 5 lines in breadth. It is compressed, pointed with opposite concave and serrated margins, but its breadth as compared with its length is so much greater than in the *Thecodontosaurus*, that upon it has been founded the genus *Palaeosaurus*, and it is distinguished by the specific name of *pliognathus*, from the cusped tooth, which is referred to the genus under the name of *Palaeosaurus cylindrodon*. The position of the tooth of the *Palaeosaurus cylindrodon* which has been preserved shows that the crown is subcompressed and traversed by two opposite finely-serrated ridges, as in the *Thecodontosaurus*; its length is 6 lines, its breadth at the base 2 lines.

The vertebrae associated with the two kinds of teeth above described are biconcave, with the middle of the body more constricted, and terminal articular cavities rather deeper than in *Telesaurus*; but they are chiefly remarkable for the depth of the spinal canal at the middle of each vertebra, where it sinks into the substance of the centrum; thus the canal is wider, vertically, at the middle than at the two ends of the vertebra. An analogous structure, but less marked, obtains in the dorsal vertebrae of the *Rhycho- saurus* in the new red sandstones of Shropshire.

Besides deviating from existing lizards in the thecodont dentition and biconcave vertebrae, the Saurians of the dolomitic conglomerate also differ in having some of their ribs articulated by a head and tubercle to two surfaces of the vertebrae, as at the anterior part of the chest in crocodiles and Dinosaurs. The shaft of the rib was traversed, as in the Protorosaur and Rhychosaurus, by a deep longitudinal groove. Some fragmentary bones indicate obscurely that the pectoral arch deviated from the crocodilian, and approached the lacertian or eosaurolium type, in the presence of a clavicle and in the brevity and anteroposterior width of the coracoid. The humerus appears to have been but little curved, and the length of the femur and to have been, like that of the *Rhychosaurus*, unusually expanded at the two extremities. The femur is chiefly remarkable for a third process or trochanter, just above the middle of the shaft, which shows a medullary cavity. The distal condyles are flattened, the outer one being the larger; there is a deep depression between them posteriorly, and a very light one anteriorly.

The tibia, fibula, and metatarsal bones manifest, like the femur, the fitness of the Saurians for progression on land. The ungual phalanges are sub-compressed, curved downwards, pointed, and impressed on each side with the usual curved canal.

The general conclusions which may be drawn from the knowledge at present possessed of the osteology of the *Thecodontosaurus* and *Palaeosaurus* are, in the first place, that the dentition, biconcave vertebrae, double-jointed ribs, and proportionate size of the bones of the extremities, they are allied to the *Telesaurus*; but that they combine a dinosaurian femur, a lacertian form of tooth, and structure of the pectoral and probably pelvic arch with these crocodilian characters, having distinctive modifications, as the mosasauriform spinal canal, in which, however, the almost contemporary *Rhychosaurus* participates. It would be interesting to ascertain whether the caudal vertebrae are characterized, as in the Taurignian Protorosaur, by double diverging spinous processes.

**Genus Belodon, Von Meyer.**

Sp. *Belodon Plieningeri.*—The reptile from the upper white keuper sandstone of Württemberg, described by Plieninger, agrees in its essential characters so closely with the thecodont Saurians of the Bristol conglomerate as to add Reptilia to the probability of both belonging to the same lower mesozoic period.

Three vertebrae are modified to afford adequate attachment to the iliac bones in *Belodon*, and this additional evidence of affinity to *Dinosaurus* may have characterized also the English Thecodonts.

**Genus Cladyodon, Ow.**

Sp. *Cladyodon Lloydii.*—In the Memoir on the Triassic Red Sandstones of Warwick, by Murchison and Strickland, published in 1840, in the 2d series of the *Geological Transactions*, vol. v., a tooth, which is an extremely rare fossil in those English formations, was figured in pl. xxviii., fig. 6.

Having had the opportunity of studying the original specimen and fragments of some others of seemingly the same species from the new red sandstones of Warwick and Leamington, the writer recognised the affinity of the reptile with those teeth to the thecodont reptiles of the Bristol conglomerate, and indicated what appeared a generic modification of form by the term *Cladyodon*.

The writer has subsequently received other specimens of the teeth characterizing this genus, which may be described as being two-edged, sub-compressed; the sides more or less convex; the edges more or less sharp, and frequently finely serrate; the crown slightly bent sideways, the inner side towards the mouth-cavity. The teeth are sometimes lanceet-shaped, through convergence of the edges towards point; sometimes through one edge being convex and the other concave, the crown is slightly curved or sickle-shaped; sometimes through use, the points are blunted. The root is very thin, and shows traces of very fine slight longitudinal striations, forming wrinkles. The dentine is disposed in concentric layers; it is not labyrinthic; the base of tooth shows a conical pulp-cavity.

These teeth indicate a Saurian about 12 feet in length.

The writer cannot discern any generic, or even good specific distinctions, between the teeth from the Warwickshire keuper, on which in 1840 he founded the genus *Cladyodon*, and those from the Württemberg keuper, on which M. von Meyer in 1844 founded the genus *Belodon*. Both are nearly allied to *Palaeosaurus*.

**Genus Bathygnathus, Leidy.**

Sp. *Bathygnathus borralis*, Leidy.—Allied to the *Cladyodon* and *Belodon* by the shape of the teeth is the Saurian from the new red sandstone of Prince Edward's Island, North America, the generic and specific characters of which have been deduced by Dr Leidy from a portion of lower jaw, containing seven teeth, but with interspaces from which others have been lost. The depth of the dentary bone is 5 inches; a peculiarity which suggested the generic name (*bathys*, deep, *gnathos*, jaw). The precise mode of implantation of the teeth is not described.

The fossil was discovered at a depth of 21 feet from the surface, in a red sandstone supposed to be of the same age as that of Connecticut, so remarkable for the various and singular foot-marks, referable, some to reptiles, and others to large birds.

**Order II.—Cryptodontia.**

**Genus Rhychosaurus, Ow.**

Sp. *Rhychosaurus articeps*, Ow.—The fossils in which the above order, genus, and species of reptile have been based are from the new red sandstone (trias) of Shropshire. They occur at the Grinsill quarries, near Shrewsbury, in a fine-grained sandstone, and also in a coarse burlandstone; in the latter the writer found imbedded some vertebrae, portions of the lower jaw, a nearly entire skull,

---

1 Geologische Transactien, 2d series, vol. v., p. 344. 2 Wirzsch. naturf. Jahressette, viii., Jahrg. 1857, p. 389. Jaeger's *Phycoctonum* appears to have been founded on casts of the sockets of the teeth of *Belodon*. 3 Report of the British Association, "Brit. Fossil Reptiles," 1841, p. 155. (See fuller descriptions, with figures, in Odontography, pl. 62, A, fig. 4, a, k.) 4 Journal of the Academy of Sciences, Philadelphia, vol. ii., p. 327, pl. xxxiii. 5 Transactions of the Cambridge Philosophical Society, vol. vii., part iii., 1842, p. 355, pl. 5 and 6. fragments of the pelvis and of two femora: in the fine-grained sandstone, vertebrae, ribs, and some bones of the scapular and pelvic arches are imbedded. The bones present a very brittle and compact texture; the exposed surface is usually smooth, or very finely striated, and of a light blue colour. The sandstones containing these bones occasionally exhibit impressions of footsteps which resemble those figured in the Memoir by Messrs Murchison and Strickland (Geol. Trans., 2d series, vol. v., pl. xxviii. fig. 1), but differ in the more distinct marks of the claws, the less distinct impression of a web, the more diminutive size of the innermost toe, and an impression corresponding with the hinder part of the foot, which reminds one of a hind toe pointing backwards, and which, like the hind toe of some birds, only touched the ground with its point. The foot-prints are likewise more equal in size, and likewise in their intervals, than those figured by Messrs Murchison and Strickland: they measure from the extremity of the outermost or fifth toe to that of the innermost or first rudimental toe, about one inch and a half. They are the only footprints that have as yet been detected in the new red sandstone quarries at Grimsill. As the fossil bones have always been found nearly in the same bed as that impressed by the footsteps above described, they probably belong to the same animal.

Vertebræ.—Both articular surfaces of the centrum are concave, and are deeper than in the biconcave vertebrae of the extinct Crocodilians; the texture of the centrum is compact throughout.

The neural arch is anchored with the centrum, without trace of suture, as in most lizards; it immediately expands and sends outwards from each angle of its base a broad triangular process with a flat articular surface; the two anterior surfaces look directly upwards, the posterior ones downwards; the latter are continued backwards beyond the posterior extremity of the centrum; the teardrop-like diminution of the rib is situated immediately beneath the anterior oblique process. So far the vertebrae of the Rhynchosaurus, always excepting their biconcave structure, resemble the vertebrae of most recent lizards. In the modification next to be noticed, they show one of the vertebral characters of the Dinosaurs. A broad obtuse ridge rises from the upper convex surface of the posterior articular process and arches forwards along the neural arch above the anterior articular process, and gradually subsides anterior to its base: the upper part of this arched angular ridge forms, with that of the opposite side, a platform, from the middle line of which the spinous process is developed. Nothing of this kind is present in existing lizards; the sides of the neural arch immediately converge from the anterior extremity to the base of the spine, without the intervention of an angular ridge formed by the sides of a raised platform. The base of the spinous process is broadest behind, and commences there by two roots or ridges, one from the upper and back part of each posterior articular process. The anterior margin of the spinous process is thin and trenchant; the height of the spine does not exceed the antero-posterior diameter of its base; it is obliquely rounded off. The spinal canal sinks into the middle part of the centrum and rises to the base of the spine, so that its vertical diameter is twice as great at the middle as at the two extremities: this modification resembles in a certain degree that of the vertebrae of the Plesiosaurus from the Bristol conglomerate.

The skull presents the form of a four-sided pyramid, compressed laterally, and with the upper face arching down in a graceful curve to the apex, which is formed by the termination of the muzzle. The very narrow cranium, wide temporal fossae on each side, separated merely by the parietal and the mastoid bones, and laterally by strong compressed zygomatics; the long tympanic pedicle, descending freely and vertically from the point of union of the posterior transverse and zygomatic arches, and terminating in a convex pulley for the articular concavity of the lower jaw; the large and complete orbits, and the short, compressed, and bent down maxillae, all combine to prove the fossil to belong to the Iacertian division of the saurian order. The mode of articulation of the skull with the spine cannot be determined in the present specimen, but the lateral compression and the depth of the skull, the great vertical breadth of the superior maxillary bone, the small relative size of the temporal spaces, the vertical breadth of the lower jaw, prove that it does not belong to a reptile of the batrachian order. The shortness of the muzzle, and its compressed form, equally remove it from the Crocodilians. No Chelonian has the tympanic pedicle so long, so narrow, or so freely suspended to the pterygoid and lateral angles of the cranium.

The general aspect of the skull differs, however, from that of existing Lacertians, and resembles that of a bird or turtle, which resemblance is increased by the apparent absence of teeth. The pre-maxillary bones, moreover, are double, as in crocodiles and Cheloniens, but with this exception, all the essential characters of the structure of the skull are those of the lizard.

The rami of the lower jaw are remarkable, as in Batrachomoeus, for their great depth, but not the least trace of a tooth is discernible in the alveolar border of the dentary element.

The cranium has been preserved with the mouth in the naturally closed state, and the upper and lower jaws in close contact. In this state we may suppose that they were originally buried in the sand matrix, which afterwards hardened around them; and since lizards, owing to the unlimited reproduction of their teeth, do not become edentulous by age, we must conclude that the state in which the Rhynchosaurus was buried, with its lower jaw in undisturbed articulation with the head, accorded with its natural condition, while living, so far as the less perishable hard parts of its mastiodary organs were concerned. Nevertheless, since a view of the inner side of the alveolar border of the jaws has not been obtained, we cannot be quite assured of the actual edentulous character of this very singular Saurian. The indications of a dental system are much more obscure in the Rhynchosaurus than in any existing Lacertian; the dentations of the upper jaw are absolutely feeble than in the chameleon, and no traces thereof can be discovered on the lower jaw, where they are strongest in the chameleon. The absence of the prominent process in the Rhynchosaurus—which is conspicuously developed in all existing lizards—corresponds with the unarmed condition of the jaw, and the resemblance of the Rhynchosaurus in this respect to the Chelys ferrea, would seem to indicate that the correspondence extended to the toothless condition of the jaws. The resemblance of the mouth to the compressed beak of certain sea-birds, the bending down of the curved and elongated pre-maxillaries, so as to be opposed to the deep symphysial extremity of the lower jaw, are further indications that the ancient Rhynchosaurus may have had its jaws encased by a bony sheath, as in birds and turtles.

There are few genera of extinct reptiles of which it is more desirable to obtain the means of determining the precise modifications of the locomotive extremities than the Rhynchosaurus. The fortunate preservation of this skull has brought to light modifications of lacertine structure leading towards Chelonia and birds which before were unknown; the vertebrae likewise exhibits very interesting deviations from the lacertian type. The entire reconstruction of the skeleton of the Rhynchosaurus may be ultimately accomplished, if due interest be taken in the collection and preservation of the fossils of the Grimsall quarries.

Genus OUDENODON, Bain.

Sp. Oudenodon Bainii.—The fossils on which the above genus and species are founded are from a bluish argillito-ferruginous limestone in South Africa, and form part of a collection transmitted to the British Museum by A. G. Bain, Esq.

One portion of the fossil skull includes all that part in advance of the temporal fossa; the fore part of the temporal ridges, at the upper and back part of this fragment, curve as they diverge from each other to the back part of the orbit. The upper interorbital part of the cranium is nearly flat, with the orbital margins slightly raised, and terminating anteriorly in a low, horizontal, shelf-like projection, the least breadth of the interorbital space is 1 inch. A slight depression divides the interorbital from the supranasal tubercles. The nasal bones form an almost flat rhomboid surface, from the contracted fore part of which the broad premaxillary part of the upper jaw inclines downward and forward at an open angle. This part is traversed by a low obtuse median ridge, and terminates below in a trenchant edentulous border.

The nostrils are small, oval, and separated from each other by the broad junction of the ascending branch of the premaxillary with them.

The maxillary bone presents the chief peculiarity, being traversed obliquely by a strong angular ridge, commencing a little anterior to the orbit, and terminating at the alveolar border, not far from the maxillo-premaxillary suture. The alveolar border gently curves to this termination, and shows no trace of a tooth or alveolus.

The compound structure of the lower jaw is shown at the fractured back part, where an upper (supranasal?) element, thick and rounded above, is received into an outer and lower element, thin above, and thick and bent below, forming a groove for the reception of the upper element. On the outer side of the jaw, about the middle of the part preserved, there is a longitudinal depression or narrow vacuity, above which there is a low ridge. The symphysis is thick, long, and bent up in the form of a beak, terminating by an edentulous sub-trenchant border; its fore and outer part is traversed by a low median ridge.

The length of this portion of the skull is 6 inches; its breadth across the maxillary ridges is 2 inches 10 lines; the extent of the symphysial of the lower jaw is 2 inches 6 lines. Besides the evidence from the teeth, turtles, which removes the Oudenodon, like the Rhaebosaurus, from the chelonian order, Mr Bain, in a letter announcing the discovery of the fossils in South Africa, mentions the association of other bones with skulls, which gives additional proof of the saurian nature of the edentulous reptiles. He writes: "There were many skulls entirely without teeth, which we at first thought had belonged to Chelonians or turtles; but afterwards, finding that the animals had distinct narrow ribs, which Chelonians have not, we put them down also for something new, and named them 'Oudenodon,' or toothless animals."

**Order III.—Dicynodontia.**

**Genus Dicynodon, Ow.**—In 1844 Mr Andrew G. Bain, who had been employed in the construction of military roads in the colony of the Cape of Good Hope, discovered, in the tract of country extending northwards from the county of Albany, about 450 miles east of Cape Town, several nodules or lumps of a kind of sandstone, which, when broken, displayed in most instances evidences of fossil bones, and usually of a skull with two large projecting teeth. Accordingly these evidences of ancient animal life in South Africa were first notified to English geologists by Mr Bain under the name of "Bidentals;" and the specimens transmitted by him were submitted to the writer for examination. The results of the comparisons thereupon instituted went to show that there had formerly existed in South Africa, and from geological evidence, probably, in a great lake or inland sea, since converted into dry land, a race of reptilian animals presenting in the construction of their skull characters of the crocodile, the tortoise, and the lizard, coupled with the presence of a pair of huge sharp-pointed tusks, growing downwards, one from each side of the upper jaw, like the tusks of the mammalian horse or walrus. No other kind of teeth were developed in these singular animals: the lower jaw was armed, as in the tortoise, by a trenchant sheath of horn, some bones of the back, or vertebræ, by the hollowness of the co-adapted articular surfaces, indicate these reptiles to have been good swimmers, and probably to have habitually existed in water; but the construction of the bony passages of the nostrils proves that they must have come to the surface to breathe air.

Some extinct plants allied to the Lepidodendron, with other fossils, render it probable that the sandstones containing the dicynodont reptiles were of the same geological age as those that have revealed the remains of the Rhynchosaurus and Labyrinthodonts in Europe.

The generic name *Dicynodon* is from the Greek words signifying "two tusks or canine teeth." Three species of this genus have been demonstrated from the fossils transmitted by Mr Bain.

**Sp. Dicynodon lacerticeps, Ow.**—This species is founded on a skull 6 inches in length, of which a reduced figure is given in cut 69, in which C shows the canine tusks.

**Sp. Dicynodon tetanideus, Ow.**—In this species the skull, and the facial part more particularly, is shorter than in *D. lacerticeps*.

**Sp. Dicynodon strioliceps, Ow.**—The shortening of the jaws and blunting of the muzzle are carried to an extreme in this species, in which the nostrils are situated almost beneath the orbits.

**Sp. Dicynodon tigriceps, Ow.**—In this species the length of the skull is 20 inches, its breadth across the widest part of the zygomatic arches being 18 inches. It differs from the *D. lacerticeps* not only in size, but in the relatively larger capacity of the temporal fossa, and smaller size of the orbits. These cavities in *D. lacerti-

---

1 The Eastern Province Monthly Magazine, Graham's Town, September 1856. 2 Trans. Geological Society, 2d series, vol. vii. (dis., two, knedoes, canine-teeth). 3 Trans. Geol. Soc., 2d series, vol. vii., p. 233. 4 From εις, εις, the sea, εικοσια, a lizard. the upper one is formed by the mastoid (fig. 70, 8) and the post- frontal (12); the lower one by the malar (27), and squamosal (28); the latter answering to the true zygomatic arch in Mammals. The squamosal abuts by its hinder expanded end against the almost vertical tympanic pedicle, which gives attachment to the lower jaw. This shows the reptilian compound structure; 29 marks the surangular element, 30 the angular and 32 the dentary. In the side-view of the skull in fig. 70, 22 is the premaxillary, 21 the maxillary, 24 the nasal, thereby below being the nostril, to the prefrontal—between which and 21 is the lacrymal, 11 the frontal above the orbit. The premaxillary teeth and correspond- ing premandibular ones are unusually long, strong, and sharp; there are two similar teeth in each maxillary; the remaining serial teeth are smaller, but equally acute. There are no teeth on the palate.

The almost entire and undisturbed vertebral column, from the meschelkalk of Bayreuth, figured by Von Meyer in pl. 23 of his work on meschelkalk dinosaurs, and attributed by him to Notho- saurus mirabilis, gives the earliest indication of that modification of the trunk-bones which reaches its maximum in the Plesiosaurus (fig. 72), in which it was first detected by the sagacity of Cony- bear.

Twenty of the anterior vertebrae of this series, in Nothosaurus, which begins with the atlas, have the whole or part of the rib-pit situated on the centrum, as in the first vertebra in fig. 70; the pit is wholly there on fourteen vertebrae; it begins to ascend upon the neural arch in the fifteenth, as in the second vertebra, given in fig. 70, and is wholly placed there on the twenty-first vertebra.

According, therefore, to the characters by which the writer has proposed to distinguish the cervical from the dorsal vertebrae, Nothosaurus has twenty of the former. In the specimen referred to, nineteen consecutive vertebrae show the rib-pit supported

wholly on an outstanding dispophysis from the neural arch, as in the third vertebra in fig. 70; these are to be reckoned therefore as dorsal vertebrae. In the cervical vertebrae the rib-pit is large, vertically reniform, not divided by a groove; its circumference slightly projects in Nothosaurus.

There is no clear evidence of any of the cervical ribs being ter- minally expanded and hatchet-shaped, as in Plesiosaurus; those of the back are vertically longer than in Plesiosaurus, and more convex.

In the sacral vertebrae, fourth in fig. 70, the rib-pits again begin to sink upon the centrum.

There are two distinct sacral vertebra in Nothosaurus. They are known by their long, straight, terminally-bent and convergent pleurophyses, the first of which overlaps a little the second. To the convergent ends of these ribs, the ilium (fig. 70, 62, p) was doubtless ligamentously affixed. In the first caudal vertebra the par- and suprapophyses stand out much farther than in the sacrum, but rapidly shorten in the second and third caudals. The convergent process in each supports a short stiffly straight rib- let, as in the fifth figured vertebra (fig. 70); the anterior and suc- ceeding caudals support humeral arches and spines, after the dis- appearance of the pleurophyses. The humeral arch disappears in about the eighth vertebra from the head, and finally the neural arch. The terminal centraums are subcylindric and subcom- pressed. Both Nothosaurus and Plesiosaurus had abdominal ribs, of which the median piece (fig. 70, 68) was symmetrical, the two rays diverging at a very open angle, and terminating in a point or a fork; the side-pieces (p) seem not to have been so nume- rous as in Plesiosaurus.

The scapula (fig. 70, 54) is a short and strong bone, its blade appearing as a short and narrow sub-compressed process extending from the subquadrate, thick, and expanded end which affords the articular surfaces for the coracoid, clavicle, and humerus.

The clavicle, which is an exogenous process in Plesiosaurus, is here united by a strong oblique suture to the scapula. It expands into, or sends off from its outer part, a broad, flat, obtuse process, near the suture; then contracts and bends inwards to the epister- num, to which it is articulated also by suture.

The coracoid (fig. 70, 52) sends forward a broad and short flat- tened process, separated by a narrow notch from the scapular part of its head; it then contracts and soon expands into a broad, flat, sub-triangular plate, the broad and straight border of which arti- culates with that of the opposite coracoid.

A wide, unfissured intersternum separates the coracoid from the epi- sternum; their combination in the direction of this interval gives the peculiar longitudinal or fore-and-aft extent to the coracoids of the Plesiosaur, in which these bones unite with the episternum.

The pelvic arch presents a closer correspondence with that in the Plesiosaurus (fig. 72).

---

**Genus PISTOSAURUS**, Von Meyer.

Sp. **Pistosaurus longiceps**.—In this genus the facial part of the skull contracts abruptly in front of the orbits; so that, viewed from above, it resembles a long-necked bottle; the orbits are situated in the posterior half of the skull, and the nostrils are lateral. From the meschelkalk of Bayreuth.

**Genus CONCHIOSAURUS**, Von Meyer.

Sp. **Conchiosaurus circinatus**.—The facial part of the skull is less projecting than in Pistosaurus, and the nostrils are terminal. The teeth are twelve in number on each side, are subequal, with widish intervals with pyriform crown. From the meschelkalk at Laineck, near Bayreuth.

**Genus SIMOSAURUS**, Von Meyer.

Sp. **Simosaurus Guillardii**.—The fossils, chiefly cranial, on which this genus is founded, occur in the dolomitic meschelkalk near Ludwigsdorf, and in the meschelkalk of Lunéville. The skull presents the large temporal fossa, the divided nuchals, and the general depressed form and composition of that of Nothosaurus and Plesiosaurus. But its facial part is much shorter; the muzzle is neither prolonged nor terminally expanded, but forms the obtuse end of the short depressed face, of which the premaxillary part is the narrowest. The nostrils, consequently, although distant from the orbit by half the breadth of the latter, are yet nearer the fore end of the skull than in the abovementioned archosauromorph genera. The nostrils are relatively nearer to each other, the intervening bony tract due to the premaxillaries chiefly being, relatively to the breadth of the skull, much narrower in Simosaurus than in Notho- or Pterosaurus.

The profile of the skull rises from the intermaxillary to the interorbital regions much more than in the Nothosaurus, and the depths of the skull, behind the orbit, is greater in proportion to its length. The post-frontals are most clearly produced backwards, along the upper border of the zygoma to the mastoid. The maxillaries are extended, and connected with the post-frontals, which terminate freely and obliquely a little below the expanded hinder part of the middle lobe, without being met by or joining a squamosal.

Most complete and extensive is the ossification of the roof of the mouth in this genus. The pterygoids are expanded into one broad unbroken imperforate flat expanse of bone, from about one-third of the distance from the snout to the occipital condyle; they are united by a median suture, and underlap the whole of the sphenoid. The teeth, compared with Nothosaurus, are few and large, and are subequal, save one or two at the fore and hind extremity of the series. The crown expands a little above the fang, is conical, sub-bifurcated, and impressed by a few coarse longitudinal ridges; some are obtuse, others acute; but all are shorter and thicker than in Notho- or Pterosaurus.

The vertebrae have flat or very slightly concave articular surfaces on the body; the neural arch articulates therewith by suture. In these characters, and in their general proportions, they resemble those of Notho- and Pterosaurus. It is significant of some difference in respect of the arrangement of the vertebrae in the same column, that although specimens from the tail, and from different parts of the back, have been obtained, no cervical vertebra with any probability belonging to this genus has yet been found. The caudal centra present two well-defined, rather prominent, hypaxial processes. The cervical is like the contraction of the body reminded Cuvier of that of the Ichthyosaurus, but its expanded median part was differently shaped. The pubis, like that of the Plesiosaurus, resembles to a certain degree the pubis in Chelonia. The few bones of the limbs which have been found still resemble, as do those of Plesiosaurus, the corresponding bones of marine Chelonia.

Accordingly, there have been entered in palaeontological catalogues an Ichthyosaurus Lamelliferus (De la Beche), a Plesiosaurus Lamelliferus (Münster), and a Chelonia Lamelliferus (Gray and Keferstein); but all these are parts of one and the same genus of Enchelosaurian, or "Cheloniens des environs de Luméville" of Cuvier, the "Simosaurus" of H. von Meyer.

Genus PLACODUS.—The cranial structure in this genus of muschelkalk reptile is closely similar to that in Simosaurus, but its proportions are different: it is as broad as long; the greatest breadth being behind, whence the sides converge to an obtuse muzzle; the entire figure viewed from above being that of a right-angled triangle, with the corners rounded off. The temporal fossae are the widest, and zygomatic arches the strongest, in the whole class Reptilia; the lower jaw presents a like excessive development of the coronoid processes. These developments, for great size and power of action of the biting and grinding muscles, relate to a most extraordinary form and size of the teeth, which resemble paving-stones, and were evidently adapted to crack and bruise shells and crusts of marine Invertebrata.

The teeth of the upper jaw consist of an external or maxillary series, and an internal or palatal series. The maxillary series are supported in a marginal row of alveoli by the premaxillary and maxillary bones; the palatal series are implanted in the palatine and pterygoid bones. The maxillo-premaxillary teeth are five in number on each side, two implanted in the premaxillary, and three in the maxillary. The premaxillary teeth are subequal, smaller than the maxillary teeth; their crowns are subhemispheric in P. laticeps, but in P. gigas they present bent, pointed, prehensile character. In P. laticeps the first maxillary tooth has a full oval crown, 44 lines by 41 diameter; the second much larger, 51 lines by 44 lines in diameter; the third is subcircular, 8 lines in diameter, on the right side. The palatal series begins on the inner side of this tooth, and consists of two teeth on each side. The first tooth has a full elliptical crown, 10 lines by 8; the second tooth, developed in the broad pyriform bone, presents a full oval shape, 1 inch 9 lines by 1 inch 3 lines in diameter. In Placodus gigas and P. laticeps the palatal teeth, three in number on each side, are all of large size, slightly increasing from before backwards; they are situated close together, forming on each side a series a little curved with the convexity outwards, and the interspace between the two series is very narrow. The first tooth is triangular, the second and third are quadrangular; each with the angles rounded, and the transverse diameter exceeding the form and att or longitudinal one. The maxillary teeth are much smaller than the palatal ones, have a rounded or subquadrate crown, are four in number, and of subequal dimensions. The premaxillary teeth, three in number on each side, are more remote and distinct from the maxillary teeth than in Placodus rostratus and P. laticeps; their crowns are more elongated and conical than in P. laticeps; the prehensile power of the prolonged premaxillary part of the jaw being obviously greater in Placodus gigas than in P. laticeps or P. rostratus. The size of the last tooth in P. laticeps surpasses that of any of the teeth in the previously discovered species. In proportion to the entire skull, it is the largest grinding tooth in the animal kingdom, the elephant's molar not excepted.

All these teeth are implanted by short simple bases in distinct hollow sockets, subject to the same law of displacement and succession as in other reptiles. By some it may be deemed requisite to separate generically the Placodi with two teeth from those with three teeth in each palatal series; but the Placodus rostratus offers a transitional condition in the small relative size of the first two palatal teeth, and in the rounded form of all the teeth, from the P. Andrias to the P. laticeps.

We cannot contemplate the extreme and peculiar modification of form of the teeth in the genus Placodus without a recognition of the adaptation for the pounding and crushing of hard substances, and so convinced that the association of the fossils with shell-clad Mollusca in such multitudes as to have suggested special denotations to the strata containing Placodus (e.g., muschelkalk, terrestriellenkalk, &c.), is indicative of the class whence the Placodi derived their chief subsistence.

No doubt the most numerous examples of similarly-shaped teeth for a like purpose are afforded by the class of fishes, as, e.g., by the extinct Pycnodonts, and by the wolf-fish (Anarhichas lupus) and the Cetacean of the existing seas. But the reptilian class is not without its instances at the present day of teeth shaped like paving-stones, of which certain Australian lizards exhibit the peculiarity in so marked a degree that the generic name Opelosus has been given to express that peculiarity. Amongst extant reptiles, also, a species of lizard from the tertiary deposits of the Limagne in France presents round obtuse teeth, of which the last, in the lower jaw, is suddenly and considerably larger than the rest.

Nothosaurus, Simosaurus, and Pterosaurus present the same evidences of lacertian affinities in the division of the nostrils by the median extension of the premaxillary backwards to the nasals, the same thecodont dentition, and the same circumscription of the orbits and temporal fossae as in Placodus; there is also a general family likeness in the upward aspect of these apertures, accompanying an almost entire depression of the skull. The nostrils, though varying greatly in length in these genera, present the same obtuseness, and the alveolar border of the teeth the same smooth outward convexity which we observe in the Placodus. The peculiar confluence of the elements of the upper and lower zygomatic arches,—i.e., of the post-frontal and malar,—forming the broad wall of bone behind the orbit, is continued still further backwards in the Simosaurus.

In Pterosaurus the elongated post-frontal, malar, and squamosal are united together in one deep zygomatic arch, which has the mastoid and tympanic for its hinder abutment.

It is remarkable that hitherto no vertebrae or other bones of the trunk or limbs have been found so associated with the teeth of Plesiosaurus, as to leave any grounds for believing them belonging to the same species. Usually, after the indication of a reptile by detached teeth, the next step in its reconstruction is based upon detached vertebrae. The twelve or more evidences of Plesiosaurus, afforded by bone as well as teeth, are all portions of the skull.

It is possible that some of the singularly modified vertebrae from the muschelkalk, next to be described, may belong to the Placodus; and the same surmise suggests itself in reference to some of the limb-bones from the muschelkalk that cannot be assigned to other known saurian genera.

The consideration of the dentition of Placodus to the crushing of very hard kinds of food, its close analogy to the dentition of certain fishes known to subsist by breaking the shells of whelks and other shell-clad Mollusca, and the characteristic abundance of fossil shells in the strata to which the remains of Placodus are peculiar, concur in producing the belief that the species of this genus were reptiles frequenting the sea-shore, and probably good swimmers. But as at present we have got no further than the head and teeth in the reconstruction of this mesozoic form of molluscivorous reptile, the present notice will conclude with a remark suggested by the disposition and form of the teeth. In all the species, under the rather wide range of specific varieties of the dentitions, there are two rows of the crushing teeth in the upper jaw, and only one row in the lower jaw, on each side of the mouth; and the lower row plays upon both upper rows, with its strongest (middle) line of force directed against their interspaces. Thus the crushing force between produces a part between the two planes or points of resistance above, on the same principle on which we break a stick across the knees; only here the fulcrum is at the intermediate point, the moving powers at the two parts grasped by the hands. It is obvious that a portion of shell pressed between two opposite flat surfaces might resist the strongest bite, but subjected to alternate points of pressure its fracture would be facilitated.

**Genus TANYSTROPHUS.**

Sp. Tanytropheus compicus, H. von Meyer.—Certain long,


slender, hollow bones, from the German muschelkalk strata, were referred by Count Münster to the class Reptilia, under the name of *Mecosaurus*, under the impression that they were bones of the limbs. H. von Meyer subsequently, in more perfect specimens, observing that each slightly expanded extremity of the long bone was terminated by a symmetrical oval concave articular surface, surmounted by a pair of symmetrical lateral incurved plates, resembling confluent neurapophyses, with articular surfaces, and with their sometimes confluent bases arching over a neural canal (as in the left-hand figure in cut 71), recognised their vertebral character; and, adopting the determination of their reptilian nature, but repudiating the idea of their being limb-bones, he discarded Münster's name, and substituted for it that of *Tanytropheus*, indicative of their peculiar proportions avertebrally.

Although the articular ends are for the most part symmetrical, the long intervening part is not so. It is sometimes, usually broader at the base below than above, sometimes more flattened on one side than on the other, giving an irregular, vertically oval, or triangular cross section. A low median ridge is not uncommon on the lower surface towards the ends of the vertebrae; and similar less regular ridges project from the sides of the otherwise smooth outer surface. The centrum is excavated by a canal, resembling a medullary one, but more probably filled, in the recent state, as in the long nasal style of the frog, with unossified cartilage. The walls of this cavity are compact, and in thickness about one-sixth of the diameter of the bone. The terminal neural arches support each a low median ridge or rudimental spine, which soon subside. The trace of neural spines in like manner disappears, etc., is continuous by two distinct slender canals which traverse for a certain extent the entire extent of the thicker upper wall of the cavity of the vertebral body. A single large vascular canal opens on the wider surface midway between the two ends of the body. There is no trace of transverse processes, rib-surfaces, or neurapophyses; this, and the absence of the continuous neural canal, indicate these singular vertebrae to belong to the tail. From the long caudal-vertebral style of anomalous Batrachia the vertebrae of *Tanytropheus* differ in having distinct articular surfaces at both ends. The difference of shape and size in the few that have been found also indicates that there were more than two such vertebrae in the tail of the extraordinary animal to which they have belonged.

Caudal vertebrae of the normal proportions and structure, from muschelkalk in the same localities, with *Tanytropheus* have been referred to *Nikrosaurus*. It is possible, however, that one or other of the remarkable genera *Sternosaurus*, *Parodon*, etc., may have possessed the peculiar structure in the tail, or some part of it, which the tanytrophean vertebrae indicate. The first four vertebrae of the neck or trunk of the *Plesiosaurus* teloecoria are those which most resemble in their proportions the vertebrae above described; but none of the fistularian vertebrae have the articular concavity, and the neurapophyses, at both ends; the first presents them at the fore end, and the last at the hind end, and the modifications of both these finished articular ends pretty closely correspond with those of *Tanytropheus*; but the second and third vertebrae of *Plesiosaurus* are united with the first and fourth by natural surfaces with deeply-interlocking flattened processes.

**Genus SPHENOSAURUS.**

Sp. Sphenosaurus Sternbergi, Von M.—The fossil vertebrae on which this genus is founded are imbedded in a sandstone, most likely the hunter, from Bohemia or the south of Germany. Of the twenty-three vertebrae so preserved in nearly their natural position, and with their under surface exposed, five belong to the tail, the rest to the trunk. Of these, two are sacral, two lumbar, the rest are dorsal or thoracic, with long and slender ribs connected with them. The neural arch appears to have been naturally united to the centrum with large neurapophyses. The articular end of the centrum is vertical to its axis; both are slightly concave. Between each centrum is a transversely concave depressed osmeterium, homologous with the cervical wedge-bone or hypopophys in *Elasmosaurus*. This is the chief peculiarity in *Sphenosaurus*, and recalls a character in the vertebral column of *Archeosaurus*.

**Genus PLEISOAURUS.**

The discovery of this genus forms one of the most important additions that geology has made to comparative anatomy. Baron Cuvier deemed the structure of the Plesiosaurus "to have been the most singular, and its characters the most anomalous that had been discovered amid the ruins of a former world." "To the head of a lizard it united the teeth of a crocodile, a neck of enormous length, resembling the body of a serpent, a trunk and tail having the proportions of an ordinary quadruped, the ribs of a chameleon, and the paddles of a whale." "Such," writes Dr. Buckland, "are the strange combinations that form the structure in the Plesiosaurus, a species remarkable of which after imperfect records for thousands of years amidst the wreck of millions of extinct inhabitants of the ancient earth, are at length recalled to light by the researches of the geologist, and submitted to our examination, in nearly as perfect a state as the bones of species that are now existing upon the earth."

The first remains of this animal were discovered in the lias of Lyme Regis about the year 1823, and formed the subject of the paper by the Rev. Mr Conybeare (afterwards dean of Llandaff), and Mr (afterwards Sir Henry) De la Beche, in which the genus was established, and named *Plesiosaurus* (in approximate to the Saurians'), from the Greek words *pleion* and *sauros*, signifying "more" and "lizard," to be called "lizard," because the authors saw that it was more nearly allied to the lizard than was the *Ichthyosaurus* from the same formation.

The entire and undisturbed skeletons of several individuals, of different species, have since been discovered, fully confirming the sagacious restorations by the original discoverers of the *Plesiosaurus*.

**Vertebral Column.**—The vertebral bodies have their terminal articular surfaces either flat or slightly concave, or with the middle of such cavity a little convex. In general the bodies present two pits and holes at their under part. The cervical vertebrae consist of centrum, neural arch, and neurapophyses. The latter are wanting in the first vertebra; but both this and the second have the hypapophyses.

The cervical ones are short, and expand at their free end, so as to have suggested the term "hatched-bones" to their first discoverers. They articulate by a simple head to a shallow pit, which is rarely supported on a process, from the side of the centrum; but is commonly bisected by a longitudinal groove, a rudimental indication of the upper and lower processes which sustain the cervical ribs in Crocodiles.

The body of the atlas articulates with a large hypapophysis below, with the neurapophysis above, with the body of the axis behind, and with part of the occipital condyle in front; all the articulations save the last become, in *Plesiosaurus* pectenopterus, and probably with age in other species, obliterated by calcification. The hypapophysis forms the hinder two-thirds, the neurapophysis comprises the anterior and lateral parts, and the centrum forms the middle or bottom of the cup for the occipital condyle. The second hypapophysis is lodged in the inferior interspace between the bodies of the atlas and axis; it becomes anchylosed to these and to the first hypapophysis. The first neurapophysis, or rudimental rib, is developed from the centrum of the axis.

As the cervical vertebra approach the dorsal, the lower part of the costal pit becomes smaller, the upper part larger, until it forms the whole surface, gradually rising from the centrum to the neurapophysis.

The dorsal region is arbitrarily commenced by this vertebra, in which the costal surface begins to be supported on a diaphysis, which progressively increases in length in the second and third dorsal, continues as a transverse process to near the end of the trunk; and on the vertebra above or between the iliac bones, it subsides to the level of the neurapophysis. In the caudal vertebrae the costal surface gradually descends from the neurapophysis upon the side of the centrum; it is never divided by the longitudinal groove which, in most *Plesiosaurus*, indents that surface in

---

1 Previous to the writer's Memoir on *Plesiosaurus* in the Philosophical Transactions (1858), all palaeontologists had referred the genus to the pycnodont order of fishes. The cervical vertebrae. The neural arches remain long unanchylosed with the centrum in the Plesiosaurus, and appear to be always little in some species. The pleurophyseal plates in length, and lose in terminal breadth in the hinder cervicals; and become long and slender ribs in the dorsal region, curving outwards and downwards so as to encompass the upper two-thirds of the thoracic abdominal cavity. They decrease in length and curvature as they approach the tail, where they are reduced to short straight pieces, as in the neck, but are not terminally expanded; they cease to be developed near the end of the tail. The hemapophyses in the abdominal region are subdivided, and with the basial spine or median piece, form a kind of "plastron" of transversely-extended, slightly-bent, median and lateral, overlapping bone bars, occupying the subabdominal spaces between the coracoids and humerals. In the tail the hemapophyses are short and straight, and remain re-united both with the centrum above and with each other below. The basial spine is not developed in this region. This modification has been expressed by the statement that there were no chevron-bones in the Plesiosaur. The tail is much shorter in the Plesio- than in the Ichthyosaurus.

The skull is depressed; its length is rather more than thrice its breadth; but the proportions somewhat vary in different species. The cranial part, or that behind the orbits, is quadrato; hence it contracts laterally to near the maxillo-premaxillary suture, where it confines either parallel or with a slight swelling before round into the obtuse anterior termination.

The orbits are at or near the middle of the skull, estimating the length of this by that of the lower jaw, they are in advance of the middle in Plesiosaurus Haemibatis. The orbits are rather subtriangular than round, being somewhat squared off behind, straight above, and contracted anteriorly. No trace of sclerotic plates has yet been discerned in any specimen. The temporal fossae are large subquadrate apertures. The nostrils, which are a little in advance of the orbits, are scarcely larger than the parietal foramen. Below them, upon the palate, are two similar-sized apertures, probably the palatal nostrils.

The lower jaw presents an angular, surangular, splenial, and dentary element, in each ramus; the dentary elements being confluent at the expanded symphysis. There is no vacuity between the angular and surangular or any other element of the jaw. The coronoid process is developed, as in Placodus, from the surangular, but rises only a little higher than in crocodiles. The alveoli are distinct cavities, and there is a groove along the inner border in both jaws.

When the successional teeth first project in that groove they give the appearance of a double row of teeth. All the teeth are sharp-pointed, long and slender, circular in cross section, with fine longitudinal ridges on the enamel; the anterior teeth are the longest.

---

**Fig. 72.**

Plesiosaurus (Lias).

---

The scapula is a strong triradiate bone, the longest ray being formed by the acromial or clavicular process which arches forward and inward to abut against the sternum or epiplecoid.

The proper body of the scapula is short and straight, somewhat flattened; the thick articular end, which forms the shortest ray, is subequally divided by the articular surface for the coracoid, and that for the head of the humerus.

The coracoids are chiefly remarkable for their excessive expansion in the direction of the axis of the trunk, extending from the abdominal ribs forward, so as to receive the entostrernum, which is wedged in the thoracic interstices. The median borders meet and unite from an anterior determination by their degree of curvature or convexity, which is always slight.

The coracoids unite anteriorly with the clavicles as well as with the episternum; laterally they articulate with the scapula, combining to form the glenoid cavity for the humerus.

The episternum has the same general form as the median pieces of the abdominal ribs, being, like those pieces, a modified humeral spine, only more advanced in position; the lateral wings or prolongations are broader and flatter; the median process is short; a long narrow ridge projects from the middle of the internal surface. The humerus is moderately thin and long bone with a conical head, sub-cylindrical at its proximal end, becoming flattened and gradually expanded to its distal end, where it is divided into two indistinct surfaces for the radius and ulna. The shaft in most species is slightly curved backwards, or the hind border is concave, whilst the front one is straight. The radius and ulna are about half the length of the humerus; the former is straight, the latter curved or reniform, with the concavity towards the radius; both are flattened; the radius is a little contracted towards its carpal end, and in some species is longer than the ulna. The carpus consists of a double series of stout rounded discs, the largest at the radial side of the wrist; the ulnar side hinder to have contained more unossified matter. The metacarpals, in number, are elongate, slender, slightly expanded at the two ends, flattened, and sometimes a little bent. The phalanges of the five digits have a similar form, but are smaller, and progressively decrease in size; the expansion of the two ends, which are truncate, makes the sides or margins concave. The first or radial digit has generally three phalanges, the second from five to seven, the third eight or nine, the fourth eight, the fifth five or six phalanges. All are flattened; the terminal ones are nailless; and the whole were obviously included, like the paddle of the porpoise and turtle, in a common sheath of integument. The pelvic arch consists of a short but strong and straight narrow moveable ilium, and of a broad and flat pubis and ischium; the former subquadrate or subcircular, the latter triangular; the fore-and-aft expanse of both bones nearly equalling that of the coracoids. All concur in the formation of the hip-joint. The ischium is prolonged again, and together near their medial borders, leaving a wide elliptical vacuity, or "fossa ovalis," between this junction and their outer acetabular one. The pelvic paddle is usually of equal length with the pectoral one, but in Plesiosaurus it is longer. The bones closely correspond, in number, arrangement, and form, with those of the fore limb. The femur has the hind margin less concave, and so appears more straight. The fibula, in its reniform shape, agrees with its homologue the ulna. The tarsal bones are also smaller than the tibia and fibula. Of existing reptiles, the lizards, and amongst these the old world Monitor (Plesiosaurus Fitz.), by reason of the peculiarities in front of the orbit, most resemble the Plesiosaur in the structure of the skull. The division of the nostrils, the vacuities in the occipital region between the exoccipitals and tympanics, the parietal foramen, the zygomatic extension of the post-frontal, the palato-maxillary, and pterygo-sphenoid vacuities in the bony palate, are all lacertian characters, as contrasted with crocodilian ones.

But the antorbital vacuities between the nasal, pre-frontal, and maxillary bones are the sole external nostrils in the Plesiosaurus; the zygomatics are absent against the dorsal part of the tympanic, as in lizards. A much greater extent of the mouth than is confined in lizards, and the palato-maxillary and pterygo-sphenoid fissures are reduced to small size. The teeth, finally, are implanted in distinct sockets. That the Plesiosaur had the "head of a lizard" is an emphatic mode of expressing the amount of resemblance.

---

1 This is omitted in most of the published restorations of the Plesiosaurus. Reptilia.

blance in their cranial conformation. The crocodilian affinities, however, are not confined to the teeth, but extend to the structure of the skull itself.

In the simple mode of articulation of the ribs, the lacertian affinity is again strongly manifested; but to this vertebral character such affinity is limited; all others exemplify the ordinal distinction of the Plesiosaurus from known existing reptiles. The shape of the joints of the centra; the number of vertebrae between the head and tail, especially of those of the neck; the slight indication of the sacral vertebrae; the non-confluence of the caudal hemapophyses with each other, are all "plesiosauroid." In the size and number of abdominal ribs and sternum may perhaps be discerned a first step in that series of development of the hemapophyses of the trunk which reaches its maximum in the plastron of the Chelonia.

The connotation of the clavicle with the scapula is common to the Chelonia with the Plesiosaurus; the expansion of the coracoids—extreme in Plesiosaurus—is greater in Chelonia than in Crocodilia, but is still greater in the Lacertilia. The form and proportions of the pelvis and ilium, as compared with the ilium, in the pelvic arch of the Plesiosaurus, find the nearest approach in the pelvis of marine Chelonia; and no other existing reptile now offers so near, although it be so remote, a resemblance to the structure of the paddles of the Plesiosaurus. Amongst the many figurative illustrations of the nature of the Plesiosaurus in which popular writers have indulged, that which compares it to a snake threaded through the trunk of a turtle is the most striking; but the number of vertebrae in the Plesiosaurus is no true indication of affinity with the ophidian order of reptiles.

The reptilian skull from formations underlying the lias, to which that of Plesiosaurus bears the nearest resemblance, is the skull of the Plesiosaurus of the German muschelkalk.1 The nostrils have a similar position and diminutive size in Plesiosaurus, but are somewhat more in advance of the orbit, and the premaxillaries enter into the formation of their boundary: the premaxillary muzzle and the temporal fossa are also somewhat longer and narrower.

The post-frontals and mastoids more clearly combine with maxillae and squamosals in forming the zygomatic arch, which is of greater depth in Plesiosaurus; the parietal foramen is larger; there is no trace of a median parietal crest. On the palate, besides the vacuities between the pterygoids and premaxillaries, and the small foramina between the palatines, premaxillaries, and maxillaries, there is in Plesiosaurus a single median foramen in advance of the latter foramina, between the pointed anterior ends of the pterygoids and the premaxillaries. In Nothosaurus the pterygoids extend back, underlapping the basi-sphenoid, as far as the basi-occipital, the median suture uniting them being well marked to their terminations; and there is no appearance of vacuities like the pterygo-sphenoid ones in Plesiosaurus.

The tympanics are relatively longer, and extend farther back in Plesiosaurus than in Plesiosaurus. There is no trace of lacrymals in Plesiosaurus; and its maxillaries are relatively larger than in Plesiosaurus. In Plesiosaurus there are 18 teeth on each side the upper jaw, including the 5 premaxillary teeth; in Plesiosaurus there are from 30 to 40 teeth on each side. In Plesiosaurus the teeth are relatively larger, and present a more oval transverse section: the anterior teeth are proportionally larger than the posterior ones than they are in Plesiosaurus. The disproportion is still greater in Nothosaurus, in some species of which the teeth behind the premaxillary and symphysial terminal expansions of the jaws suddenly become —e.g., in Nothosaurus mirabilis (fig. 70)—very small, and form a straight, numerous, and close-set single series along the maxillary and corresponding part of the mandibular bone.

Both Nothosaurus and Plesiosaurus had many neck-vertebrae; and the transition from these to the dorsal series was effected, as in Plesiosaurus, by the ascent of the rib-surfaces from the centrum to the neurapophysis; but the surface, well divided between the two elements, projected further outwards than in most Plesiosaurus.

In both Notho- and Plesiosaurus the pelvic vertebra develops a comb-like process (para- and diapophysis), but of relatively larger, vertically placed, standing well out, and from near the fore part of the side of the vertebra. This process, with the coalesced riblet, indicates a stronger ilium, and a firmer base of attachment of the hind limb to the trunk, than in Plesiosaurus. Both this structure, and the greater length of the bones of the fore arm and leg show that the muschelkalk predecessors of the classic Plesiosaur were better organized for occasional progression on dry land. More than twenty species of Plesiosaurus have been described by, or are known to, the writer: their remains occur in the oolitic, Wealden, and cretaceous formations, ranging from the lias upwards to the chalk, inclusive. A comparison of remains of various Plesiosaur has led to a conviction, that specific distinctions are accompanied with well-marked differences in the structure and proportions of answerable vertebrae, but are not shown in small differences of number in the cervical, dorsal, or caudal vertebrae.

When any region of the vertebral column presents an unusual excess of development in a genus, such region is more liable to variation, within certain limits, than in genera where its proportions are more normal. The differences of the number of cervical and dorsal vertebrae, ranging between 29 and 31 in the Plesiosaurus Hawkiensis, e.g.,—as noted in the description of that species in the writer's Report on British Fossil Reptiles, 1838—indicate the range of variety observed in the same region of which, at that time, the vertebral columns of different individuals could be compared.

General LIGUARDI, Dr.—M. von Meyer regards the number of cervical vertebrae and the length of neck as characters of prime importance in the classification of Reptilia, and founds therein his order called Macrotracheliens, in which he includes Simosaurus, Plesiosaurus, and Nothosaurus, with Plesiosaurus. No doubt the number of vertebrae in the same skeleton bears a certain relation to ordinal groups: the Ophioid find a common character therein; yet it is not their essential character, for the snake-like form, dependent on multiplied vertebrae, characterizes equally certain Batrachians (Cordylus) and fishes (Maroma). Certain regions of the vertebral column are the seats of greatly varied forms in some natural group of Reptilia. We have long-necked and short-necked lizards; but do not therefore confound them with numerous caudal vertebrae, as in Microcosmus, from those with few or none. The extinct Dolicosaurus of the Kentish chalk, with its procanean vertebrae, cannot be ordinarily separated, by reason of its more numerous cervical vertebrae, from other shorter-necked procanean lizards. At little can we separate the short-necked and big-headed amphibian Plesiosaur from the Macrotracheliens with which it has its most intimate and true affinities.

There is much reason, indeed, to suspect that some of the muschelkalk Saurians, which are as closely allied to Nothosaurus as Plesiosaurus is to Plesiosaurus, may have presented analogous modifications in the number and proportions of the cervical vertebrae. It hardly seems likely to be possible to articulate the broad and short-necked skull of the Simosaurus, with its proportionally large teeth, without inferring that such a head must have been supported by a shorter and more powerful neck than that which bore the long and slender head of the Nothosaurus or Plesiosaurus. The like inference is more strongly impressed upon the mind by the skull of the Plesiosaurus, still shorter and broader than that of Simosaurus, and with vastly larger teeth, of a shape indicative of their adaptation to crushing molluscous or crustacean shells.

Neither the proportions and armature of the skull of Plesiosaurus, nor the mode of obtaining the food indicated by its cranial and dental characters, permit the supposition that its head was supported by other than comparatively short and strong neck. Yet the combination of the skull, its proportions, cavities, and other like-giving anatomical characters, all bespeak the close essential relationship of Plesiosaurus to Simosaurus and other so-called "macrotracheliens" reptiles of the muschelkalk beds. The writer continues, therefore, to regard the fin-like modification of the limbs as a better ordinal character than the number of vertebrae in any particular region of the spine. But whilst retaining the term Enaliosauria for the large extinct natory group of saurian reptiles, the essential distinctness of the groups Sauropterygia and Ichthyopterygia, typified by the Ichthyosaurus and Plesiosaurus respectively, should be borne in mind.

Sp. Plesiosaurus brevicyclorhynx, Owen.—The generic characters of Plesiosaurus are given by the teeth and the cervical vertebrae. As compared with those of Plesiosaurus, the teeth are thicker in proportion to their length, are subtriangular in transverse section, with one side flattened, and bounded by lateral prominent ridges from

---

1 Von Meyer, Muschelkalk Saurier. of the neck and the equality of the width of the back of the head with the front of the chest, impressing the observer of the fossil skeleton with a conviction that the ancient animal must have resembled the whale tribe and the fishes, in the absence of any intervening constriction or "neck."

This close approximation in the Ichthyosaurus to the form of the most strictly aquatic vertebrate animals of the existing creation, is accompanied by an important modification of the surfaces forming the joints of the tail fin, each of which surfaces are hollow, leading to the inference that they were originally connected together by an elastic bag or "capsule" filled with fluid,—a structure which prevails in the class of fishes, in the Labyrinthodonts and a few extinct aquatic reptiles, in the existing perennialbranchiate Batrachia, but not in any of the whale or porpoise tribe.

With the above modifications of the head, trunk, and limbs, in relation to swimming, there co-exist corresponding modifications of the tail. The bones of this part are much more numerous than in the Plesiosaurus, and the entire tail is consequently longer; but it does not show any of those modifications that characterize the very supple tail of the fish. The number of caudal vertebrae of the Ichthyosaurus gradually decrease in size to the end of the tail, where they assume a compressed form, or are flattened from side to side, and thus the tail, instead of being short and broad as in fishes, is lengthened out as in crocodiles.

The very frequent occurrence of a fracture of the tail, about one-fourth of the way from its extremity, in well preserved and entire fossil skeletons, is owing to that proportion of the end of the tail having supported a tail fin. The only evidence which the fossil skeleton of the Ichthyosaurus gives of this is the depression or horizontally-flattened form of the bone supporting the fin. This is inferred therefore from the corresponding condition of the Plesiosaurus being flattened in the vertical direction, or from side to side, that it possessed a tegumentary tail fin expanded in the vertical direction.

The shape of a fin composed of such perishable material is of course conjectural, as is the outline in fig. 74. Thus, in the construction of the principal swimming organ of the Ichthyosaurus we may trace, as in other parts of its structure, a combination of mammalian (beast-like), saurian (lizard-like), and piscine (fish-like) peculiarities. In the great length and gradual diminution of the tail we perceive its saurian character; in the tegumentary nature of the fin, and the manner by which it was supported, we see some part in the mammalian whales and porpoises is shown; whilst its vertical position makes it closely resemble the tail fin of the fish.

The horizontal tail fin of the whale tribe is essentially connected with their necessities as warm-blooded animals breathing atmospheric air; without this means of displacing a mass of water in the vertical direction, the head of the whale could not be brought with the required rapidity to the surface to breathe; but the Ichthyosaurus, not being warm-blooded or quick breathers, would not need to bring their head to the surface so frequently or so rapidly as the whale; and moreover a compensation for the want of hori-

---

1 The anatomical reader is referred to the writer's "Report on British Fossil Reptiles," Trans. Brit. Assoc., 1839, and to the Annals and Magazine of Natural History, 1858, p. 388. the auditory passages; on the upper surface the parietal foramen and the temporal fossa; on the lateral surfaces the orbits and nostrils, the plane of the aperture in both being vertical; on the inferior surface the palato-maxillary, the pterygo-sphenoid, and the pterygo-malar vacuities. The occipito-parietal vacuities are larger than in Crocodiles, smaller than in Lacertilia; they are bounded internally by the basi-, ex-, and super-occipitals, externally by the parietal and mastoid. The auditory apertures are bounded by the tympanic and squamosal. The tympanic takes a greater share in the formation of the "meatus auditorius" in reptiles than in crocodiles; it is restricted to that which it takes in Ichthyosaurus.

The orbit is most remarkable in the Ichthyosaurus, amongst reptiles, because its large proportional size and its posterior position; in the former character it resembles that in the lizards, in the latter that in the crocodiles. It is formed by the pre- and post-frontals above, by the lacrymal in front, by the post-orbital behind, and by the peculiar long and slender malar bar below. In crocodiles and in most lizards the frontal enters into the formation of the orbits, and in lizards the maxillary also. The nostril is a longish triangular aperture, with the narrow base behind; it is bounded by the lacrymal, nasal, maxillary, and pre-maxillary. It is proportionally larger than in the Plesiosaurus, and is distant from the orbit about half its own long diameter. Like the orbit, the plane of its outlet is vertical.

The pterygo-palatine vacuities are very long and narrow, broadest behind, where they are bounded, as in lizards, by the anterior cavities of the basi-sphenoid, and gradually narrowing to a point close to the palatine nostrils. These are smaller than in most lizards, and are inclosed by the palatines, ecto-ptyergoid, maxillary, and pre-maxillary. The pterygo-maxillary fissures are the lower outlets of the temporal fossa; their sudden posterior breadth, due to the emargination of the pterygoid, relates to the passage of the muscles for attachment to the lower jaw. The parietal fossa is bounded by both parietals and frontal; its presence is a mark of labyrinthodont and lacertian affinities; its formation is like that in iguanids and Rhampholeon.

The cerebral fossa is bounded above by the parietal internally, by the mastoid and post-frontal externally; they are of an oval form, with the great end forward. In their relative size and backward position they are more crocodilian than lacertian.

In the Ichthyosaurus communis there are seventeen sclerotic plates forming the fore part of the eyeball. In a well-preserved example, the papillary or corneal vacuity, as bounded by these plates, is of a full oval form, 14 inch in long diameter, the length of the plates (or breadth of the frame) being from 8 to 10 lines. In the skull the long diameter of the orbit is 4 inches. The deep position of the sclerotic circle in this cavity showed how they had sunk, by pressure on the external side of the eyeball became collapsed by escape of the humours in decomposition.

Wherever the antecedent forms an extinct genus of any class are known, the characters of such genus should be compared with those of its predecessors rather than with its successors or with existing forms, in order to gain an insight into its true affinities.

We derive a truer conception of the affinities of the Ichthyosaurus by comparison with the Labyrinthodonts and other triassic reptiles, as we do of the Plesiosaurus by comparison with the mosasaurids and Crocodilians. It is commonly said that the Ichthyosaurus resembles more the lizard, in some of such characters, and in a less degree the crocodiles, in others. Such a character. The truer expression would be that the lizards, which are the predominant type of Sauropsida at the present day, have retained more of the ontological type of the triassic and colitic reptiles, and that the crocodiles deviate further from them or exhibit a more modified or specialized structure. The posterior position of the nostrils, the small size and position of the palato-ptyergoid foramen, are marks of affinity to Plesiosaurus, in common with which genus the cranial structure of the Ichthyosaurus exhibits a majority of lacertian characters.

In comparing the jaws of the Ichthyosaurus communis with those of the gangatic Gharial an equal degree of strength and of alveolar border for teeth result from two very different proportions in which the maxillary and premaxillary bones are combined together to form the upper jaw. The proportionation of the snout has evidently no relation to this difference, and we are accordingly led to look for some other explanation of the disproportionate development of the premaxillaries in the Ichthyosaurus. It appears to me to give additional proof of the collective tendency of the affinities of the Ichthyosaurus to the lacertian type of structure. The backward or antorbital position of the nostrils, like that in whales, is related to their marine existence. But in the Lacertilians in which the nostrils extend to the fore part of the head, their anterior boundaries are formed by the premaxillaries; it appears, therefore, to be in conformity with the lacertian affinities of the Ichthyosaurus that the premaxillaries should still enter in the same relation with the nostrils, although this involves an extent of anterior development proportionate to the length of the jaws, the forward production of which sharp-toothed instruments fitted them, as in the modern dolphins, for the prehension of agile fish.

That the Ichthyosaurus occasionally sought the shores, crawled on the strand, and basked in the sun, may be inferred from the bony stricture connected with their fore fins, which does not exist in any porpoise, dolphin, grampus, and whale; and for want of which, chiefly, these warm-blooded, air-breathing, marine animals are so helpless when left high and dry on the sands. The structure in question in the Ichthyosaurus is a strong osseous arch, inverted and spanning across beneath the chest from one shoulder-joint to the other; and what is most remarkable in the structure of this "scapular" arch is, that it closely resembles, in the number, shape, and disposition of its bones, the same part in the singular aquatic mammalian quadruped of Australia, called Antechinus, and Platypus, or duck-mole. The Ichthyosaurus, when on the shore either for feeding or procreation, would lie or crawl prone, or with its belly resting or dragging on the ground.

Another extraordinary feature of the head was the enormous magnitude of the eye; and from the quantity of light admitted by the expanded pupil, it must have possessed great powers of vision, especially in the dark. It is not uncommon to find in front of the orbit in fossil skulls, a circular series of petrified thin bone plates, ranged round a central aperture, where the pupil of the eye was placed. The eyes of many fishes are defended by a bony covering consisting of two pieces; but a compound circle of overlying plates is now found only in the eyes of turtles, tortoises, lizards, and birds. This curious apparatus of bony plates would aid in protecting the eyeball from the effects of the water when the Ichthyosaurus rose to the surface, and from the pressure of the dense element when it dived to great depths; and they show, writes Dr Blackland (Bridgewater Treatise), "that the enormous eye of which they formed the front, was an optical instrument of varied and prodigious power, enabling the Ichthyosaurus to descry its prey at great or little distances, in the obscurity of night, and in the depths of the sea."

Of no extinct species are the materials for a complete and exact restoration more abundant and satisfactory than of the Ichthyosaurus; they plainly show that its general external figures must have been that of a huge predatory abdominal fish, with a longer tail and a smaller tall-fin; scaleless, moreover, and covered by a smooth or finely wrinkled skin, analogous to that of the whale and shark.

The mouth was wide, and the jaws long, and armed with numerous pointed teeth, indicative of a predatory and carnivorous nature in all the species; but these differed from one another in regard to the relative strength of the jaws, and the relative size and length of the teeth.

Masses of masticated bones and scales of extinct fishes that lived in the same seas and at the same period as the Ichthyosaurus, have been found under the ribs of fossil specimens, in the situation where the stomach of the animal was placed; smaller, lighter, and more digested masses, containing also fish-bones and scales have been found, bearing the impression of the projections of the internal surface of the stomach of the great predatory sea-lizard. One of these Coprolites is figured beneath the skeleton in fig. 74.

In tracing the evidences of creative power from the earlier to the later formations of the earth's crust, remains of the Ichthyosaurus are first found in the lower lias, and occur more or less abundantly through all the superincumbent secondary strata up to, and inclusive of, the chalk formations. They are most numerous in the lias and oolite, and the largest and most characteristic species have been found in these formations.

More than thirty species of Ichthyosaurus are known to the writer, many of which have been described or defined.

ORDER V.—DINOSAURIA.

Char.—Cervical and anterior dorsal vertebrae, with par- and diapophyses, articulating with bifurcate ribs; dorsal vertebrae, with a neural platform, sacral vertebrae exceeding two in number; body supported on four strong ungulate-like limbs.

The well-defined vertebrae, large and hollow limb-bones, and trochanteric femora of the thecodont reptiles of the British conglomerate, together with the structure of the sacral vertebra in the allied Boloros, indicate the beginning, in the triassic period, of an order of reptiles which attained, in its full development and typical characteristics in the colitic period.

Genus SCILLIDOMARUS, OW.—By this name is indicated a Saurian with large and hollow limb-bones, with a femur having the third inner trochanter, and with metacarpal and phalangial bones, Genus Megalosaurus, Bkld.—The true dinosaurs have been established by the discovery of the sacrum, which consists of five vertebrae, interlocked by the alternating position of neural arch and centrum. The articular surfaces of the free vertebrae are nearly flat; the neural arch develops a platform which in the anterior dorsals supports very long and strong spines. The dental characters are described and figured in the article ONTOLOGY, vol. ii., p. 435, figs. 48.

The oldest known bed from which any remains of Megalosaurus have been obtained are the lower colites at Selby Hill and Chippenham Norton, Gloucestershire. Abundant and characteristic remains occur in the Stonesfield slate, Oxfordshire. Teeth of the Megalosaurus have been found in the Corribash and Bath colites; both teeth and bones are common in the Wealden strata and Purbeck limestones. Some of these fossils indicate a reptile of at least 30 feet in length.

Genus Hylaeosaurus, Mlll.—Remains of the Dinosaurian so-called have hitherto been found only in Wealden strata, as at Tilgate, Sussex, and Bath. The most instructive evidence is that which was exposed by the quarrymen of the Wealden stone at Tilgate, and obtained and described by Mantell in 1832. It consisted of a block of stone measuring 41 feet by 27 feet (fig. 76), and included the following parts of the skeleton in almost natural juxtaposition:—10. Anterior vertebrae, the first supporting part of the base of the skull; several ribs, 4, 4; some enormous dermal bony spines, 5, 6, which supported a strong defensive crest along the back; two coracoids, 7, 7; scapulae, 8, 8; and some detached vertebrae and fragments of bone. The sacrum was dinosaurian, and included five vertebrae.

The teeth were relatively small, close-set, thecodont in implantation, with subcylindrical fang and a subcompressed slightly expanded and incurved crown, with the borders straight, and converging to the blunt apex. They indicate rather a mixed or vegetable diet than a carnivorous one. The skin was defended by subcircular bony scales. The length of the Hylaeosaurus may have been 25 feet.

Genus Iguanodon, Mlll.—Remains of these large herbivorous reptiles have been found in Wealden and ammonian (greenland) strata. Femora, 4 feet in length, showing the third inner trochanter have been discovered. The sacrum included five, and in old animals six vertebrae; the claw-bones are broad, flat, and oblong. There were no well-developed toes on the hind foot, and singular large tridactyle impressions, discovered by Eccles in the Wealden at Hastings have been conjectured to have been made by an iguanodon. The characteristic dentition of this genus is described in the article ONTOLOGY, p. 435, figs. 42, 43, and 44. All traces of dinosaurian reptiles disappears in the lower cretaceous beds.

ORDER VI.—PTEROsaURIA.

The species of this order of reptiles are extinct, and peculiar to the mesozoic period. Their chief characteristic is the development of the pectoral limbs into organs of flight (fig. 76). This is due to an elongation of the antibrachial bones, and more especially to the still greater length of the metacarpal and phalangeal bones of the fifth or innermost digit (fig. 76, 5); the last phalanx of which terminates in a point. The other fingers were of more ordinary length and size, and terminated by claws. The number of phalanges is progressive from the first (fig. 76, 1) to the fourth (4), which is a reptilian character. The whole osseous system is modified in accordance with the possession of wings; the bones are light, hollow, most of them permeated by air-cells, with thin compact outer walls. The scapula and coracoid are long and narrow, but strong. The vertebrae of the neck are few, but large and strong, for the support of a large head with long jaws, armed with sharp-pointed teeth. The skull was lightened by large vacuities, of which one (o, fig. 76) is interposed between the nostril n and the orbit l. The

Fig. 76. Fossil Skeleton of Pterodactylus crassirostris: A, Sketch of Living Pterodactyl.

vertebrae of the back are small, and grow less to the tail. Those of the sacrum are few and small, and the pelvis and weak hind limbs bespeak a creature unable to stand and walk like a bird. The body must have been dragged along the ground like that of a bat. But the Pterosauria may have been good swimmers as well as flyers. The vertebral bodies unite by ball-and-socket joints, the cup being anterior, and in them we have the earliest manifestation of the "procellari" type of vertebra. The atlas consists of a discoid centrum, and of two slender neuropophyses; the centrum of the axis is ten times longer than that of the atlas, with which it ultimately coalesces; it sends off from its under and back part a pair of processes above which is a transversely extended convexity articulating with the third cervical vertebra. In the vertebra there is a large pneumatic foramen at the middle of the side. The neural arch is confluent with the centrum. Dentition thecodont.

Sp. Dimorphodon macronyx, Bkld.—The Pterodactyls are distributed into sub-genera, according to well-marked modifications of the jaws and teeth. In the oldest known species, from the lias, the teeth are of two kinds; a few at the fore part of the jaws are long, large, sharp-pointed, with a tall elliptical base; behind them is a close-set row of short, compressed, very small lance-shaped-teeth. In a specimen of Dimorphodon macronyx, from the lower lias of Lytton Regis, the skull was 8 inches long, and the expansion of wing about 4 feet. There is no evidence of this species having had a long tail.

Genus Rhamphorhynchus, Von Meyer.—In this genus the fore part of each jaw is without teeth, and may have been encased by a horny beak, but behind the edentulous production there are four or five large and long teeth, followed by several smaller ones. The tail is long, stiff, and slender.

The Rhamphorhynchus longipinnis, R. Gemmingeri, and R. Münsteri belong to this genus. All are from the lithographic (middle oolitic) slates of Bavaria.

Genus Pterodactylus, Cuv.—The jaws are provided with teeth to their extremities; all the teeth are long, slender, sharp-pointed, set well apart. The tail is very short.

P. longirostris, Ok.—About 10 inches in length; from lithographic slate at Pappenheim. P. crassirostris, Goldf.—About 1 foot long; same locality (fig. 76.) P. Kochii, Wagn.—8 inches long; from the lithographic slates of Kehlheim. P. medius, Mnat.—10 inches long; from the lithographic slates at Meulenhardt. P. Reptilia grandis, Cuv.—14 inches long; from lithographic slates of Solenhofen. Two small and probably immature Pterodactyloidea, showing the short jaws characteristic of such immaturity, have been entered as species under the names of P. brevisirostris and P. Meyeri. The latter shows the circle of sclerotic eye-plates.

The fragmentary remains of Pterodactyloidea from British oolites,—e.g., Stonesfield slates, usually referred to as Pterodactylus Bucklandi,—indicate species about the size of a swan.

The evidences of Pterodactyloidea from the Wealden strata indicate species about 16 inches in length of body. Those (P. Fittoni, and P. Selwyni, Ow.) from the greensand formation, near Cambridge, with neck-vertebra 2 inches long, and humeri measuring 3 inches across the proximal joint, had a probable expanse of wing of from 18 to 20 feet. The P. Cuculli, Ow., and P. compressirostris, Ow., from the chalk of Kent, attained dimensions very little inferior to those of the greensand Pterodactyloidea.

More evidence is yet needed for the establishment of the pterosaurian genus, on the alleged character of but two phalanges in the wing-finger, and for which the term "Ornithopterus" has been proposed by Von Meyer.

With regard to the range of this remarkable order of flying reptiles in geological time, the oldest well-known Pterodactyloid is the Dimorphodon macrocephalus, of the lower lias; but bones of Pterodactyloidea have been discovered in the coeval lias of Württemberg. The next in point of age is the Rhamphorhynchus Bambusensis, from the "Posidonienschiefer" of Hanx in Bavaria, answering to the alum shale of the Whitby lias; then follows the P. Bucklandi from the Stonesfield oolite. Above this come the first-defined and numerous species of Pterodactyloidea from the lithographic slates of the middle oolitic system in Germany and from Gürin, on the Rhone. The Pterodactyloidea of the Wealden are as yet known to us by only a few isolated bone fragments. The largest known species are those from the greensand of Cambridgeshire. Finally, the Pterodactyloidea of the middle chalk of Kent, almost as remarkable for their great size, constitute the last forms of flying reptile known in the history of the crust of this earth.

ORDER VII.—CROCODILIA.

Char.—Teeth in a single row, implanted in distinct sockets, external nostril single and terminal or sub-terminal. Anterior trunk vertebrae with par- and di-aphophyses, and bifurcate ribs; sacral vertebrae two, each supporting its own neural arch. Skin protected by bony, usually plated plates.

Sub-order 1.—AMPHICHELIA.

Crocodiles closely resembling in general form the long and slender-jawed kind of the Ganges called Gavial, existed from the time of the deposition of the lower lias.

Their teeth were similarly long, slender, and sharp, adapted for the prehension of fishes, and their skull was modified for more efficient progress in water by the vertebral surfaces being slightly concave behind, the limbs being relatively larger and stronger, and by the orbits forming no prominent obstruction to progress through water. From the nature of the deposits containing the remains of the so-modified crocodiles, they were marine. The fossil crocodile from the Whitby lias, described and figured in the Philosopical Transactions, 1758, p. 688, is the type of these amphiichelian species. They have been grouped under the following generic heads:—Telosaurus, Steneosaurus, Myriostoma, Macrognathus, Macrognathus, to which must be added Plesiosuchus, Pelagiosaurus, Melosaurus, Sarcosuchus, Geosuchus, Polyptychodon.

Species of the above genera range from the lias to the chalk inclusive.

Steneosaurus of the Wealden is characterized by the compressed crown and trenchant margins of the teeth; Geosuchus, of the Purbeck beds, by some of the dermal scales having the same peg-and-pit interlocking as in the scales of the gasoid fish in fig. 52; Polyptychodon of the greensand and chalk, by the well-defined numerous longitudinal ridges of the enamel; from the size of some of these teeth, this crocodile, like the Pterodactyloidea of the same period, appears to have been the largest of its group; it surpassed all other Amphicheelians in size.

Sub-order 2.—OPISTHOCHELIA.

The small group of Crocodilia so called is an artificial one, based upon more or less of the anterior trunk vertebrae being united by ball-and-socket joints, but having the ball in front, instead of at the back, as in modern crocodiles, behind. Cuvier first pointed out this similarity in a Crocodilian from the Oxfordian beds at Honflur, and the Kimmundgian at Havre. The writer has described similar opisthochelian vertebrae from the great oolite at Chipping Norton, from the upper lias of Whitby, and, but of much larger size, from the Wealden formations of Sussex and the Isle of Wight. These specimens probably belong, as suggested by the writer in 1841, to the fore part of the same vertebral column as the vertebra, flat at the fore part, and slightly hollow behind, on which he founded the genus Cetiosaurus. The smaller opisthochelian vertebrae described by Cuvier have been referred by Von Meyer to a genus called Stegosuchus.

In our species from the Wealden, dorsal vertebrae measuring 8 inches across are only 4 inches in length, and caudal vertebrae nearly 7 inches across are less than 4 inches in length. These characterize the species called Cetiosaurus brevis.

Caudal vertebrae measuring 7 inches across and 5½ inches in length, from the lower oolite at Chipping Norton, and the great oolite at Easton, represent the species called Cetiosaurus medius.

Caudal vertebrae from the Portland stone at Garlington, Oxfordshire, measuring 7 inches 9 lines across and 7 inches in length, are referred to the Cetiosaurus longus. The latter must have been the most gigantic whale-like of Crocodilians, unless it were equaled in bulk by the Polyptychodon of the chalk.

Sub-order 3.—PROCHELIA.

Crocodilians with cup-and-ball vertebrae, like those of living species, first make their appearance in the greensand of North America (Crocodylus basiformis and C. bauri). In Europe their remains are first found in the tertiary strata. Such remains from the lacustrine clay of Meudon have been referred to C. polyptychodon, C. colorognathus, C. Bequeireli. In the calcareous grottoes of Argentum and Castelbury we find remains of the Rauisaurus and Dodecath. In the coastal recesses London clay at Shepphy Island the entire skull and characteristic parts of the skeleton of C. polyptychodon and C. Cheongsoidea occur. In the somewhat later coeval beds at Bracklesham occur the remains of the gavial-like C. Dixonii. In the Horndale beds have been found the C. Hastingsiensis, with short and broad jaws; and also a true alligator (C. Hamiltoniensis). It is remarkable that forms of procrelian Crocodilia, now geographically restricted—the gavial to Asia, and the alligator to America—should have been associated with true crocodiles, and represented by species which lived, during nearly the same geological period, in rivers flowing over what now forms the south coast of England.

Many species of procrelian Crocodilia have been founded on fossils from miocene and pliocene tertiaries. One of these, of the gavial sub-genus (C. crassidens), from the Sewalik tertiary, was of gigantic dimensions.

ORDER VIII.—LACERTILIA.

Char.—Vertebræ prococlean, with a single transverse process on each side, and with single-headed ribs; sacral vertebrae not exceeding two.

Small vertebrae of this type have been found in the Wealden of Sussex. They are more abundant, and are associated with other generic characteristic parts of the species, in the cretaceous strata. On such evidence have been based the Rauisaurus subulatus, the Ossiniasaurus crassidens, and the Dolichosaurus longicollis. The last-named species is remarkable for the length and slenderness of its trunk and neck, indicative of a tendency to the ophidian form. But the most remarkable and extreme modification of the lacertian type in the cretaceous period is that presented by the huge species, of which the remains 5 feet long, was discovered in the upper chalk of St. Peter's Mount, near Maestricht, in 1790. The vertebrae are gently concave in front, and convex behind; there are thirty-four between the head and the base of the tail; a sacrum seems to have been wanting. The caudal vertebrae have long neural and hemal spines, both of which arches coalesce with the centrum, and formed the basis of a powerful swimming tail. The teeth are anchylosed to eminences along the alveolar border of the jaw, according to the acrodont type. There is a row of small teeth on each pterygoid bone. For this genus of huge marine lizard the name Mosasaurus has been proposed. Besides the M. Hofmanni of Maestricht, there ORDER IX.—Ophidia.

(Serpents.)

The earliest evidence of an ophidian reptile has been obtained from the eocene clay at Sheepy; it consists of vertebrae indicating a serpent of 12 feet in length, the Palaeophis typhlops. Still larger, more numerous, and better preserved vertebrae have been obtained from the eocene beds at Bracklesham, on which the Palaeophis typhlops and P. porosus have been founded. These remains indicate a boa-constrictor-like snake, of about 20 feet in length. Ophidian vertebrae of much smaller size, from the newer eocene at Horsham, refer the species called Falderia rhombifera and P. porosus. Fossil vertebrae from the tertiary formation near Salonica have been referred to a serpent, probably poisonous, under the name of Laophis. A species of true viper has been discovered in the miocene deposits at Sansans, South of France. Three fossil Ophidians from the Oeningen slate have been referred to Coluber arenarius, C. Kargii, and C. Ouenii.

ORDER X.—Chelonia.

(Tortoises and Turtles.)

Reference has already been made to the impressions in sandstones of triassic age in Dumfriesshire, referred by Dr Duncan to tortoises. These impressions have been finely illustrated in the great work by Sir William Jardine on the foot-prints at Corncockle Muir. The earliest proof of chelonian life which the writer has obtained has been afforded by the skull of the Chelone planiceps, from the Portland stone; and by the carapace and plastron of the extinct and ingenuous modified emydian genera Pleurosternum and Plesiochelys (fig. 77). In the first genus the plastron retains its central vertebrity; in the second genus an additional pair of bones is interposed between the hyoartes (ha) and hypoplastrons (pa). In the specimen figured (fig. 77), the plastron, and the under surface of the marginal pieces (2 to 12) of the carapace, of Pleurosternum emeryiornatum (Parbeck).

Fig. 77.

Pleurosternum emeryiornatum (Parbeck).

of the marginal pieces (2 to 12) of the carapace, of Pleurosternum emeryiornatum are shown. This fine Chelonian is now in the British Museum.

True marine turtles (Chelone Camperii, C. obsoleta, C. pulchri-

---

1 Monograph of the Fossil Chelonian Reptiles of the Wealden and Purbeck Limestones, 4to, 1853, Palaeontographical Society. 2 The upper end of the femur from Sheepy, in t. xxix. of Monograph of Fossil Reptiles of the London Clay, Palaeontographical Society, 1850, belongs to this species.

VOL. XVII. Table of Geological Distribution of Reptilia.

| Class II.—AVES. | |-----------------| | Long before any evidence of birds from actual or recognizable fossil remains is obtained in tracing the progress of life from the oldest fossiliferous deposits upwards, we meet with indications of their existence impressed in sandstones of the triassic or liassic period. | | These earliest evidences of the class are by foot-prints in some former tidal shore, preserved in one or other of the ways explained in the section "Ichthyology." The fossil bones of birds have not been found save in strata of much later date than the impressed sandstones; and they are much more rare than the remains of mammals, reptiles, and fishes, in any formations except the most recent, again limited locally, e.g., New Zealand. | | Sir C. Lyell has well remarked, that "the powers of flight possessed by most birds would insure them against perishing by numerous calamities to which quadrupeds are exposed during floods." The same writer further argues, that "if they chance to be drowned, or to die when swimming on water, it will scarcely ever happen that they will be submerged so as to become preserved in sedimentary deposits." It is true that the carcass of a floating bird may not sink where it has died, but be carried far along the stream; ultimately, however, if not devoured, its bones will subside when the soft parts have rotted, and both the compressibility of the osseous tissue, and the facts made known by the ornithologists of the present day near Cambridge, on the London clay, at Shepperton, and of the Museum, Wisconsin, show that they can be preserved in the fossil state. The length of time during which the carcass of a bird may float, doubtless exposes it the more to be devoured, and so tends to make more scarce the fossil remains of birds in sedimentary strata. | | Certain it is that the major part of the remains of extinct birds that have as yet been found are those of birds that were deprived of the power of flight, and were organized to live on land. | | The existence of birds at the triassic period in geology, or at the time of the formation of sandstones which are certainly intermediate between the lias and the coal, is indicated by abundant evidences of foot-prints impressed upon those sandstones which extend through a great part of the valley of the Connecticut River, in Connecticut and Massachusetts, New England. | | The foot-prints of birds are peculiar, and more readily distinguishable from those of most other animals. Birds tread on the toes only; these are articulated to a single metatarsal bone at right angles equally to it, and they diverge more from each other, and are less connected with each other, than in other animals, except as regards the web-footed order of birds. | | Not more than three toes are directed forward; the fourth when it exists, is directed backward, is shorter, usually rises higher from the metatarsal, and takes less share in sustaining the superincumbent weight. No two toes of the same foot in any bird have the same number of joints. There is a constant numerical progression in the number of phalanges (toe-joints), from the innermost to the outermost toe. When the back toe exists, it is the innermost of the four toes, and it has two phalanges, the next has three, the third or middle of the front toes has four, and the outermost has five phalanges. When the back toe is wanting, as in some waders, and most wingless birds, the toes have three, four, and five phalanges respectively. When the number of toes is reduced to two, as in the ostrich, their phalanges are respectively four and five in number, thus showing these to answer to the two outermost toes in tridactyl and tetradactyl birds. | | The same numerical progression characterizes the two phalanges in most lizards from the innermost to the fourth; but a fifth toe exists in them which has one phalanx less than the fourth toe. It is the fifth toe which is wanting in every bird. In some Gallinaceae, one or two (Pavo bicalcaratus) spurs are superadded to the metatarsus; but this peculiar weapon is not the stunted homologue of a toe. Dr Deane and Mr Marsh of Greenfield, United States, first noticed, in 1835, impressions resembling the feet of birds in the sandstone rocks near that town. Dr Hitchcock, president of Amherst College, United States, whose attention was called to these impressions, first made public the fact, and submitted to a scientific ordeal his interpretations of these impressions as having belonged to the feet of living birds, and he gave them the name of Ornithodactylus. | | It was a startling announcement, and a conclusion that must have had strong evidence to support it, since one of the kinds of the tracks had been made by a pair of feet, each leaving a print 20 inches in length. Under the term Ornithodactylus giganteus, however, Dr Hitchcock did not shrink from announcing to the geological world the fact of the existence, during the period of the deposition of the red sandstones of the valley of the Connecticut, of a bird which must have been at least four times larger than the ostrich. The impressions succeeded each other at regular intervals; they were of two kinds, but differing only in a right and left foot, alternating with each other. The left foot was little to a side, and the right a little to the right of the mid-line between the series of tracks. Each foot-print (fig. 79, b and r) exhibits three toes, diverging as they extend forwards. The distance between the tips of the inside and outside toes of the same foot was 12 inches. Each toe was terminated by a short strong claw projecting from the mid toe a little on the inner side of its axis, from the other two toes a little on the outer side of theirs. The end of the metatarsal bone to which these toes were articulated rested on a two-lobed cushion which sloped upwards behind. The inner toe (r) showed distinctly two phalangeal divisions, the middle toe three, the outer toe (b) four. And since, in living birds, the penultimate and ungual phalanges usually leave no trace, a single impression, the inference was just, that the toes of this large foot had been characterized by the presence of a steadily-increasing number of phalanges, from the inner to the outer one, as in birds. And, as in birds also, the toe with the greatest number of joints was not... |

---

1 Principles of Geology, ed. 1847, p. 721. 2 Save in the Swift. the longest; it measured, e.g., 124 inches, the middle toe in the same base-line measured 16 inches, the outer toe 12 inches. Some of the impressions of this huge tridactylous footstep were so well preserved as to demonstrate the papillose and striated character of the integument covering the cushions on the under side of the foot. Such a structure is very similar to that in the ostrich. The average extent of stride, as shown by the distance between the impressions, was between 3 and 4 feet; the same limb was therefore carried but each step from 6 to 7 feet forward in the ordinary rate of progression.

These foot-prints, although the largest that have been observed on the Connecticut sandstones, are the most numerous. The gigantic Brontosaurus, as Principal Hitchcock proposes to term the species, "must have been," he writes, "the giant rulers of the valley. Their gregarious character appears from the fact, that at some localities we find parallel rows of tracks a few feet distance from one another."

The strata of red sandstone, with the above-described impressions, occupy an area more than 150 miles in length, and from 5 to 10 miles in breadth. "Having examined this series of rocks in many places, I feel satisfied that they were formed in shallow water, and for the most part near the shores and banks of the beds, from three to ten miles above the level of the water and laid dry, while a newer series, composed of similar sediment, was forming." The tracks have been found in more than twenty places, scattered through an extent of nearly 80 miles from N. to S., and they are repeated through a succession of beds attaining at some points a thickness of more than 1000 feet, which may have been thousands of years in forming.

One of the evidences of birds from the Cambridge greensand, transmitted to the writer by their discoverer, Mr Barret, is the lower half of the trifid metatarsal, showing the outer toe-joint much higher than the other two, and projecting backwards above the middle joint; it indicates a bird about the size of a woodcock.

In a conglomerate of the pliocene age, exposed near the town of Mondon, near Paris, the leg and thigh bones (tibia and femur) of a bird (Gastornis Parisitanus) have been discovered; they indicate a genus now extinct. They belonged to a species as large as an ostrich, but more robust, and with affinities to wading and aquatic birds. In the eocene clay of Sheepy fossil remains of birds have been found, indicating a small vulture (Lithornis culminaris); also a bird, probably of the king-fisher family (Halcyonidae), and a species of the sea-gull family. In the same formation at Highgate remains of a species of the heron family have been found.

The fossil bones of birds from the gypsum quarries at Montmartre were referred or appropriated by Cuvier to eleven distinct species. Great ornithites have been obtained from the Hordwell fresh-water deposits. The most ancient example of a passerine bird is the Protornis Gloriosissimus, founded on an almost entire skeleton discovered in the schistose rock of Olaris, referable to the older division of the eocene tertiary series. This skeleton is about the size of a lark, and in some respects similar to that bird.

Comparisons of the ornitholites of the eocene teritories show that the following ordinal modifications of the class of birds were at that period represented, the raptorial, or birds of prey, by species of the size of our ospreys, buzzards, and smaller falcons, and most probably also by an owl; the insectivorous, or tree-perching birds, by species seemingly allied to the nuthatch and the lark; the omnivorous or anisodactyls, by species as large as the cuckoo and kingfisher; the aquatic, by species as large as the heron and curlew; by a species as large as, but with thicker legs than, an ostrich; the gallinaceous, by a curlew of the size of the ibis, and by species allied to Scopulæ, Fringa, and Pelidna, of the size of our woodcocks, lapwings, and sandpipers; and the notatorial, by species allied to the cormorant, but one of them of larger size, though less than a pelican; also by a species akin to the divers (Merganer).

The remains of birds become more abundant and varied as we approach the present time; especially in the miocene strata, so richly developed in France, although wanting in Britain. One of the most singularly-modified forms of beak is shown by the flamingo.

The fossil skull of a species of the genus (Phoenicopterus) has been found in the miocene fresh-water deposits of the plateau of Grignon, near Grignon-le-Château; and the entire articulated bone of a species of eagle (Aquila) or osprey (Pandion) in the same deposits at Chaptalou, Allier; and the humerus of a bird allied to and as large as the albatross, in the molasse coquillère marine at Armagne. Remains of a vulture, most probably a Cathartes, have been found in the miocene lacustrine deposits of Cantal. Indications of all the other orders of birds, save the great Cursores or Struthionidae, have also been discovered in miocene strata—those of wading birds being the most numerous. Fossil eggs of birds occur in miocene deposits in Auvergne; and impressions of feathers have been discovered in the pliocene calcareous marls at Montbeloeix. In pliocene brick-earth deposits in Essex has been found a fossil metatarsal of a swan, as large as, and not distinctly divisible from, those of the living species. In the gravel at Lawford a fossil humerus like that of a wild goose. But most of the ornitholites of this recent tertiary period have been discovered in ossiferous caverns. They belong to birds closely resembling the falcon, wood-pigeon, lark, thrush, teal, and a small wader. The writer has received information of skeletons of birds found deeply imbedded in stratified clay at Aberdeen and Peterhead. The most remarkable additions to the present class have been obtained from the superficial deposits, turbaries, and caves in New Zealand. This island is remarkable for the absence of aboriginal species of land-mammals, and for the presence of a small bird with very rudimental wings, and the keelless sternum and loose plumage of the Struthious order; out of a peculiar group called Apteryx, the leg-bones are short, and the humerus from four to a very small five inches. Birds resembling the Apteryx in the shape of the sternum and bony structure of the pelvis and hind limbs, some retaining also the small back toe, others apparently without it, formerly existed in New Zealand under different specific forms ranging in height from 3 feet to 10 feet. They have been referred by the writer to the genera Dinornis and Palopteryx. The gigantic species are interesting as exhibiting birds equal to the formation of tridactyl impressions as large as those of the Connecticut sandstones called Ornithichnites (Brontosaurus) gigas (fig. 79, r, b). In that cut is given a figure of the leg-bones of Dinornis giganteus, in which the tibia (r) measures upwards of a yard in length. In the entire skeleton (d) of another species, the metatarsus is as thick, but only half as long, as in the D. giganteus; the framework of the leg is the most massive of any in the class of birds; the toes almost rival those of the elephant; whence the name Dinornis elephantopus, given to this species. Several other species of these

---

1 Lyell, Manual of Elementary Geology, etc., 1855, p. 348. 2 Owen, "On the Affinities of Gastornis Parisitanus," Quarterly Journal of Geological Society, vol. xii., 1856, p. 204. 3 These remains are described in eight memoirs by the writer, published in the third and fourth volumes of the Transactions of the Zoological Society of London. The description of the first fragment of the bone, indicative of the Dinornis, is in vol. iii., p. 39, pl. 3. Mammalia extinct tridactyle wingless birds have been determined—e.g., Dicynodon ingens, D. struthoides, D. rhodites, D. dromedarius, D. camaricus, D. robustus, D. crassus, D. geraudensis, D. curvis. With these remains have been found bones of a bird the size of a swan, but of an extinct genus (Apteryx); also those of a large coot (Neosilurian Mantellia), which, though ordinarily fossil remains, were afterwards discovered living in the Middle Island of New Zealand. Two species of Apteryx not distinguishable from the existing kinds, were contemporaries with the gigantic Dicynodon, and the writer has received evidence that the D. elephantipes afforded food to the natives at probably no very remote period. Some of the smaller kinds of Dicynodon may yet be found living on the Middle Island.

In Madagascar portions of metatarsal bones, indicating a three-toed bird as large as, but generically distinct from, the Dicynodon giganteus, have been discovered in alluvial banks of streams; and with them entire eggs, measuring from 13 to 14 inches in long diameter. The contents of one of these eggs is computed to equal those of six ostrich eggs, or of one hundred and forty-eight hen eggs.

In the neighbouring island of Mauritius the dodo (Dodo dodo) has been exterminated by man within the period of two centuries; and in the islands of Bourbon and Rodriguez the "solitaire" (Pezophaps) has also become extinct. Both these birds had wings too short for flight.

Class III.—Mammalia.

(Warmed-blooded, Air-breathing, Viviparous Vertebrates.)

Every calcified part of an animal, whether coral, shell, crust, tooth, or bone, can preserve its form and structure when buried in the earth during the changes there gradually operated in it, until every original particle may have been removed and replaced by some other mineral substance previously dissolved in the water percolating the bed containing the fossil. A bone, or other part so altered, is said to be "petrified." Not only are all its outward characters preserved, but even the minutest structure may be, and in most cases is, demonstrable in the fine sections under the microscope.

Fossil bones and teeth have been discovered in every intermediate stage of alteration, from their recent state to that of complete petrification. Recent bones consist of a soft—commonly called animal or organic—bone, hardened by earthy salts, chiefly phosphate of lime. Fishes have the smallest proportion of birds the largest proportion, of the earthy matter in their bones. The soft part is chiefly a gelatinous substance.

Proportions of Hard and Soft Matter in the Bones of the Vertebrate Animals.

| FISHES | Salmon | Carp | Cod | |--------|--------|------|-----| | Soft | 60-62 | 40-40| 34-30| | Hard | 39-38 | 59-60| 65-70|

| REPTILES | Frog | Snake | Lizard | |----------|-----|-------|--------| | Soft | 35-50| 31-04| 46-67 | | Hard | 64-50| 69-96| 53-33 |

| MAMMALS | Porpoise | Lion | Man | |---------|----------|------|-----| | Soft | 35-50 | 31-00| 27-70| | Hard | 64-10 | 69-00| 72-30|

| BIRDS | Goose | Turkey | Hawk | |-------|-------|--------|------| | Soft | 32-91 | 30-49 | 26-72| | Hard | 67-09 | 69-51 | 73-28|

The chemical nature of the hardening particles, and of the soft basis of bone, is exemplified in the subjoined table, including a species of each of the four classes of Vertebrata:

| Ingredients | Hawk | Man | Terrible | Cod | |-------------|------|-----|----------|-----| | Phosphate of lime, with traces of flint of lime | 64-39 | 59-63 | 52-66 | 57-29 | | Carbonate of lime | 7-03 | 7-33 | 12-53 | 4-90 | | Phosphate of magnesia | 0-94 | 1-32 | 0-82 | 2-40 | | Sulphate, carbonate, and chlorate of soda | 0-92 | 0-69 | 0-90 | 1-10 | | Glutin and chondrin | 27-73 | 29-70 | 31-75 | 32-31 | | Oil | 0-99 | 1-33 | 1-34 | 2-00 |

The most common change which bones first undergo is the loss of more or less of their original soft and soluble basis. This effect of long interment is readily tested by applying the specimen to the tongue, when the affinity of the pores of the earthy constituent, after having lost the gelatine, for fluid is so great, that the specimen adheres to the tongue like a piece of dry chalk. Bones and teeth in this state quickly absorb a solution of gelatine, and thus their original tenacity may be restored. Petrified fossils need no such treatment; they are usually harder and more durable than the original bone itself.

The interpretation of such fossil remains requires a comparison of them with the corresponding parts of animals now living, or of previously determined extinct species. In the case of the vertebrate animals, such comparison is limited to the osseous and dental systems. The interpretation of a vertebrate fossil, therefore, presupposes a knowledge of the various modifications of the skeleton and teeth of the existing vertebrate animals; and the more extensive and precise such knowledge may be, the more successful will be the efforts, and the more exact the conclusions, of the interpreter.

The determination of the remains of quadrupeds is best left as Cuvier truly remarks, with such difficulties thrown into the organic fossils. Shells are usually easily exposed, and with all the characters by which they may be compared with their analogues in the museum, or with figures in the illustrated books, of naturalists. Fishes frequently present their skeleton or their scale covering more or less entire, from which may be gathered the general form of their body, and frequently both the generic and specific characters which are derived from such internal or external hard parts. But the entire skeleton of a fossil quadruped is rarely found, and when it occurs, it gives little or no information as to the hair, the fur, or the colour of the species. Portions of the skeleton with the bones dislocated, or scattered pell-mell,—detached bones and teeth, or their fragments merely,—such are the conditions in which the petrified remains of the mammalian class most commonly present themselves in the strata in which they occur.

Prior to the time of Cuvier but little progress had been made in the interpretation of such fragmentary remains. The striking success which attended the application of the great comparative anatomist's science to this previously neglected field of study, was referred by Cuvier to principles in the organization of animal bodies, which he termed the "Correlation of Form and Structure," and the "Subordination of Organs"—principles which his philosophical biographer, M. Florens, in common with most contemporary philosophers, has regarded as the most effective and successful instrument in the restoration of extinct animals. They will be exemplified in the course of the present and concluding section of the article PALEONTOLOGY.

A terminal phalanx—suitable to fit a hoof may give, as Cuvier declared, the modifications of all the bones of the fore limb that relate to the absence of a rotation of the fore leg, and all the modifications of the jaw and skull that relate to the mastication of food by broad-crowned complex molars.

But there are certain associated structures for the coincidence of which the physiological law is unknown. "I doubt," writes Cuvier, "whether I should have ever divined, if observation had not taught it me, that the ruminant hoofed beasts should all have the cloven foot, and be the only beasts with horns on the frontal bone." We know as little why horns should be in one or two pairs on the frontal bone of those Ungulates only which have hoofs in one or two pairs; whilst in the horned Ungulates with three

---

1 That this combination of phosphorus and calcium has ever taken place in nature, save under the influence of a living organism, remains to be proved.

2 The writer's experience of this effect led him to suggest the application of a similar process to the long-buried ivory ornaments from the ruins of Nineveh in the British Museum; it proved successful.

3 Éloge Historique et Analyse Raisonnée des Travaux de G. Cuvier, 12mo, Paris, 1841, p. 42.

4 Ossaments Fossiles, 8vo, ed. 1834, tom. i., p. 184. Mammalia, hoofs, there should be either one horn, or two horns placed one behind the other in the middle line of the skull; or why the Ungulates with one or three hoofs on the hind foot should have three trochanters on the femur, whilst those with two or four hoofs on the hind foot should have only two trochanters.

"However," continues Cuvier, "since these relations are constant, they must have a sufficient cause; but as we are ignorant of it, we must supply the want of the theory by means of observation. This, this observation will serve to establish empirical laws almost as sure in their application as rational ones."

"That there are secret reasons for all these relations, observation may convince us independently of general philosophy." The constancy between such a form of such organ, and such another form of another organ, is not merely specific, but one of class, with a corresponding gradation in the development of the two organs."

"For example, the dentary system of non-ruminant Ungulates is generally more perfect than that of the Bivalves; in each of the former have almost always both incisors and canines in the upper as well as the lower jaw; the structure of their feet is in general less complex, inasmuch as they have more digits, or hoofs less completely enveloping the phalanges, or more bones distinct in the metacarpus and metatarsus, or more numerous tarsal bones; or a more distinct and better developed fibula; or a concomitance of all these modifications. It is impossible to assign a reason for these relations; but, in proof that it is not an affair of chance, we find that whenever a bivalve animal shows in its den-

tition any tendency to approach the non-ruminant Ungulates, it also manifests a similar tendency in the conformation of its feet. Thus the camels, which have canines and two or four incisors in the upper jaw, have an additional bone in the tarsus, resembling from its substructure not being connected with the sub-tibia, and the small bovinae have generally very small phalanges. The musk-deer, which have long upper canines, have the fibula co-extensive with the tibia, whilst the other ruminants have a mere rudiment of fibula articulated to the lower end of the tibia."

"There is then a constant harmony between two organs to all appearance quite strangers to each other, and the gradations of their forms correspond uninterruptedly even in the cases where one can render no reason for such relations." But in thus availing ourselves of the method of observation as a supplementary instrument when theory abandons us, we arrive at astonishing details. The smallest articular surface (facette) of a bone, the smallest process, present a determinate character relating to the use to which it is subjected, and to the species to which it belongs; so that whoever possesses merely the well-preserved extremity of a bone, he may, with application, aided by a little tact (adresse) in discerning analogies, and by sufficient comparison, determine all these things as surely as if he possessed the entire animal."

There have been, of course, instances, and will be, where, for want of the "efficacious comparison," and the "tact in discerning likenesses," such results have not rewarded the endeavours of the palaeontologist; and these shortcomings, and the mistakes sometimes made, even by Cuvier himself, have been cast in the teeth of his disciples, as arguments against the principles by which they believed themselves guided in their endeavours to complete the "last and edifice of which their master had laid the foundation."

The writer has thought it proper, from the well-known "Preliminary Discourse" to Cuvier's great work on Fossil Remains, with a view to neutralize the efforts of statements reiterated in apparent ignorance of the clear and explicit manner in which Cuvier there defines the limits within which the law of correlation of animal structures may be successfully applied, and indicates the instances in which—the physiological condition being unknown, and the coincident structures being understood empirically,—careful observation and rigorous comparison must supply the place of the physiologically-understood law.

Those who deny the existence of design in the construction of any part of an organized body, and who protest against the deduction of the organic plan from the parts, at the point of the eye-ball, repudiate the reasoning which the palaeontologist carries out from the hoof to the grinder, or from the carunculated molar to the retractile claw, through the guidance of the principle of a pre-ordained mutual adaptation of such parts; but such minds are not, nor have been, those who have contributed to the real advance-

ment of physiology or paleontology.

By reference to the "Table of Strata" (fig. 1), it will be seen that the earliest evidence of a vertebrate animal is of the cold-blooded water-breathing class in the upper Silurian period. Next follows that of a cold-blooded, but air-breathing vertebrate, probably from the upper Devonian; but, after the beginning of life, certainly in the carboniferous period. The warm-blooded air-breathing classes are first indicated, as birds, by foot-marks in a sandstone of probably triassic but not older age; and, as Mammals, by fossil teeth from bone-beds of the upper triassic system in Würtemberg, and of the same age near Froome, Somersetshire. Mammalian remains have also been found in a coal-field in North Carolina, which may be earlier, but cannot be later, than the lias formation.

Genus MICRACERTES.—The mammalian teeth from German and English strata indicate a very small insectivorous quadruped, to which the above generic name was given by Professor Plieninger. The German specimen were discovered in 1847 in a bone-breadth near Dielgerbach, about two miles from Stuttgart. The singular relations of which are still待解明, as between this and Keuper sandstone. The teeth of Micracertes from Froome, submitted to the writer by the discoverer, Mr Charles Moore, F.G.S., in 1858, are four in number, two being molars of the upper jaw, each with four fangs; one a molar with a narrower crown and two fangs from the lower jaw; and the fourth a small, pointed, front tooth. The crowns of the molars are short vertically in proportion to their breadth; the distinct enamel contrasts with the cement-covered fangs; the grinding surface shows a wide and shallow depression, surrounded by small, low, obtuse cusps, three of equal size being on one side, a larger cusp near one end, and smaller and less regular cusps on the opposite side. The lower molar shows a similar type, but with the trigonidial cusp less prominent. The crown of the largest of the upper molars does not exceed one line in its longest diameter. Amongst existing Mammals, some of the small molars of the marsupial and insectivorous Myrmecodons of Australia offer the nearest resemblance to these fossil teeth; but a still closer one is presented by the small tubercular molars of the extinct oolitic Mammal called Plagiolagus (fig. 88, ss. 1 and 2).

Genus DROMOTHERIUM.—It would appear that the Mammal from the American triassic or liassic coal-bed (Dromotherium splendens, Emmons) also found its nearest living analogue in Myrmecodon; for each ramus of the lower jaw contained 10 small molars in a continuous series, 1 canine, and 3 conical incisors, the latter being divided by short intervals.

Genus AMPHITHERIUM (Thylacotherium, Val.)—This genus is


found upon a few specimens of lower jaw, one ramus of which (fig. 80) gave the entire dentition of its side,—viz., three small conical incisors (i), one rather larger canine (c), six premolars, unicuspids, with a small point at one or both sides of the base (p, l–6), and six quinque-cuspid molars (m, l–6) not departing very far from the type above-described. The molars, and most of the premolars, are implanted by two roots. The condyle of the jaw is convex, and is a little higher than the level of the teeth; the coronoid process is broad and high; the angle projects backward, with a feeble projection inward. It is, again, to the marsupial Myrmecodon, amongst existing forms, that the present genus is most nearly allied. The remains of Amphitherium are from the lower oolitic slates of Stonefield (fig. 82, stratum 8).

Genus AMPHILESTERUS.—This genus is founded on a ramus of the lower jaw, from the Stonesfield oolitic slate, showing true molars of a compressed form, with a large middle cusp and a smaller, but well-marked, one at the fore and back part of its base; the "cin-

---

1 "Puisque ces rapports sont constants, il faut bien qu'ils aient une cause suffisante; mais comme nous ne la connaissons pas, nous devons supposer un défaut de la théorie par le moyen de l'observation." (Tom. cit., p. 184.)

2 "En effet, quand on forme un tableau de ces rapports, on y remarque non-seulement une constance spécifique, si l'on peut s'exprimer ainsi, entre telle forme de tel organe, et telle autre forme d'un organe différent; mais l'on aperçoit aussi une constance de classe et une gradation correspondante dans le développement de ces deux organes, qui montrent, presque aussi bien qu'un raisonnement effectif, leur influence mutuelle." (Tom. cit., p. 185.)

3 Tom. cit., p. 187.

4 For the full description and demonstration of the mammalian nature of this much-discussed fossil, see Owen, History of Brit. Fossil Mammals, Svo, p. 29.

5 Owen, Hist. Brit. Foss. Mam., p. 58, fig. 19 (Amphitherium Broderipii). Mammalia, palaeo, or basal ridge, peculiar to mammalian teeth, traverses the inner ridge of the crown, where it develops three small cusps, one at the base of the larger outer or principal cusp, and the other two forming the anterior and posterior ends of the crown. This form of tooth is unknown in existing Mammalia, but is well adapted for crushing the cases of coleopterous insects (elytra of which are found fossil in the same oolitic matrix) as are any of the multicuspid molars of small opossums, shrews, and bats. The Amphilestes Broderipii was somewhat larger than Amphilestes Prehistoric.

Genus Phascolotherium.—Although the evidence of the very slight degree of inflection of the angular process of the lower jaw of Amphilestes may favour its affinity to the placental Insectivora, yet the range of variety to which that mandibular character is subject in the different genera of existing Marsupials warrants again laying undue stress upon its value as developed in the extinct genera of the marsupial branch, and incites us to look with redoubled interest at whatever other indications of a marsupial character may be present in the fossil remains of other genera and species of Mammalia that have been detected in the Stonesfield slate.

In the specimen of Phascolotherium (fig. 81) presented to the British Museum by William J. Broderip, Esq., F.R.S., its original describer, which is as perfect in regard to the dentition as the jaw of the Amphilestes above described, the marsupial characters are more strongly manifested in the general form of the jaw, and in the extent and position of the inflected angle, while the agreement with the genus Didelphys in the number of the premolar and molar teeth is complete. The forms of the crowns of those teeth differ from those in Didelphys, and correspond so closely with those in the Amphilestes Broderipii, as to show the closer affinity of the Phascolotherium with the latter oolitic Insectivora; and, accordingly, whatever additional evidence of marsupiality is afforded by the Phascolotherium, seems to justify the conclusion that both Amphilestes and Thylacoleoidea are to be admitted into the marsupial group. The general form and proportions of the coronoid process of the jaw of Phascolotherium resemble those in the zoophagous Marsupials; and especially with that of the Thylacinus in regard to the depth and form of the entering notch between this process and the condyle.

The base of the inwardly-bent angle of the lower jaw progressively increases in Didelphys, Dasyurus, and Thylacinus; and judging from the fractured surface of the corresponding part of the fossil, it most nearly resembles the jaw of Thylacinus. The condyle of the jaw is nearer the plane of the inferior margin of the ramus in the Thylacinus than in the Dasyurus or Opossums; and consequently, when the inflected angle is broken off, the curve of the jaw continues from the condyle along the lower margin of the jaw is least in the Thylacinus. In this particular, again, the Phascolotherium resembles that Australian Carnivore. In the position of the dental foramen, the Phascolotherium, like the Amphilestes, differs from the zoophagous Marsupials and placental Carnivora and Insectivora, and resembles the Hypsognathus, a marsupial Herbivore, that orifice being near the vertical line dropped from the last molar tooth. In the direction of the line of the symphysis, the Phascolotherium resembles the Opossums more than the Dasyurus or Thylacinus. It is probable that the teeth at the fore part of the jaw showed the same correspondence. In the number of the molar series, the Phascolotherium differs from Amphilestes, Amphilestes, and Myosorex, and resembles the Thylacinus and Opossum, but without having the premolars (p, i, 2, 3) distinguished, as in them, from the true molars (m, i, 2, 3, 4), by smaller and more simple crowns. As, however, these two kinds of teeth can only be determined by their order of development and succession, the Phascolotherium may well have had three premolars and four true molars.

The difference between these teeth in the lower jaw of Didelphys is shown by the addition, in the true molars, of a pointed tubercle on the inner side of the middle cone. In Phascolotherium a mere basal ridge or cingulum extends along the inner side of the middle cone. Such a ridge is present in the last molar of Sarcophagus, but not in the other molars; but in these there are two small hind cusps on the same transverse line, whilst that cusp appears to be single in Phascolotherium. The cingulum, moreover, in the second to the penultimate of the molar series of this fossil, extends so far as to form a small talon at the fore end, but partly on the crown, thus making five points, which are very distinct in the third to the penultimate tooth inclusive; and by this character the dentition of Phascolotherium differs materially from any existing Marsupial, and repeats the type of molar which, as yet, would seem to be peculiar to the Insectivora of the oolitic epoch. There is a feeble indication of this structure in the antepenultimate and penultimate molars of Thylacinus, but the hinder division of the crown shows two small cusps on the same transverse line, besides the rudimental hindmost one; and there is no cingulum. Upon the whole, it would seem that, though the affinity may not be close, Phascolotherium most resembles Thylacinus amongst existing Marsupials; but Thylacinus is now confined to Tasmania, and is there fast verging to extinction.

The resemblance shown by the lower jaw and its teeth of the Amphilestes and Phascolotherium to marsupial genera not confined to Australia and Tasmania, leads one to reflect upon the increasing correspondence between other organic remains of the Oxfordshire oolite and other existing forms now confined to the Australian continent and surrounding sea. Here, for example, swims the Cestracion, or Port-Jackson shark, which has given the key to the nature of the "palates" from our oolites, now recognised as the teeth of congenerous larger forms of cartilaginous fishes. Mr Broderip, in his Memoir above quoted, observes, "that it may not be uninteresting to note that a recent species of Triposia has very lately been discovered on the coast of Australia, that land of marsupial animals. Our specimen lies imbedded with a number of fossil shells of that genus." Not only Triposia but Tetraclitidae exist, and the

---

Fig. 81. Lower Jaw and Teeth of the Phascolotherium (nat. size in outline), Lower Oolite.

Fig. 82. (After Fenton.) 1. Rubbly Limestone (coarse). 2. Clay, with Terrestratites. 3. Limestone rock. 4. Blue clay. 5. Sandstone rock. 6. Silty clay. 7. Oolitic rag, or limestone. 8. Sandy bed containing the Stonesfield slate.

---

1 Zoological Journal, vol. iii., p. 408, pl. xi., 1828. Mammalia, latter abundantly, in the Australian seas, yielding food to the Cetaceans, as their extinct analogues doubtless did to the allied Pliocene glossothorax with crushing teeth, called Arctodus, Pannonotherium, etc., Arctoceras and cycadaceous plants, like those found fossil in oolitic strata, flourish on the Australian continent, where marsupial quadrupeds now abound; and thus appear to complete a picture of an ancient condition of the earth's surface, which has been superseded in our hemisphere by other strata and a higher type of mammalian organization.

Fig. 82 represents a section of the strata overlying the slates whence the fossil mammalian jaws, with associated Megalosaurus, Pterodactylus, and other oolitic organisms, have been obtained at Stonesfield in Oxfordshire. The vertical thickness of the strata through which the shaft is sunk to the gallery is 62 feet; on the side opposite the right hand is marked the depth of the horizontal gallery, where the slate is dug which contains the fossils; on the opposite side the strata are numbered in succession.

Genus Stereognathus.—The last evidence of a mammalian animal discovered in the Stonesfield slate is of peculiar interest, because it exhibits a type of grinding teeth quite distinct from any of the previously acquired jaws from that locality, and affords evidence of a small vegetable-feeding or omnivorous quadruped. It consists of a portion of a lower jaw, imbedded in the characteristic matrix (fig. 83), about 9 lines in extent, containing three molar teeth (a, b, c). It is straight; the side exposed is convex vertically; a slight bend downwards, and decrease of vertical diameter towards the end, indicates it to be part of a left ramus. This is unusually shallow, broad or thick below, the side passing by a strong convex curve into the lower part; a very narrow longitudinal ridge, continuing after its disappearance by a few fine lines, forms a trough which divides the lateral from the under surface; elsewhere the bone is smooth, without conspicuous vascular perforations. The depth or vertical diameter of the ramus is not more than 2 lines. Of the three teeth remaining in this portion of jaw, the middle one is the least mutilated. The crown of this tooth (fig. 84, B) is of a quadrate form, 3 milli-metres by 3½ millimetres, of even lineal height, and supports six subequal cusps in three pairs, each pair being more closely connected in the antero-posterior direction of the tooth than transversely.

The outer side of the crown (fig. 83, b), supported by a narrow fang which contracts as it sinks into the socket, shows two principal cones or lobes, and a small accessory basal cusp. The hard and shining enamel which covers these parts of the crown contrasts with the lighter cement that coats the root. The two outer lobes or cones are subcompressed, and placed obliquely on the crown, so that the hinder one (o', fig. 84) is a little overlapped externally by the front one (o), the fore part of the base of the hinder one being prolonged inwards on the inner side of the base of the front cone. The two middle cones (k, l) are subcompressed laterally, with the fore part of their base a little broader than the back part. The two inner cones (p, p') have their inner surface convex, with their summits slightly inclined forwards. The fore part of the base of the hinder cone extends obliquely towards the entire length of the crown, beyond the contiguous end of the base of the front cone; this causes an arrangement like that of the two outer cones (o, o'), the obliquity of the posterior cone of both the outer and the inner pairs being such that they slightly converge as they extend forwards.

This type of tooth differs from that of all other known recent or extinct Mammals. The nearest approach to it is made by the middle true molar of Philotherium simplicipes, a small extinct herbivorous Mammal from the London clay (fig. 91, m, n).

That the fragment in question is the jaw of a Mammal is inferred from the implantation of the tooth by two or more roots. Most Mammals are known to have certain teeth so implanted. Such complex mode of implantation in bone has not been observed in any other class of animals. Why two or more roots of a tooth should be peculiar to viviparous quadrupeds, giving suck, is not precisely known. That a tooth, whether it be designed for grinding hard or cutting soft substances, should do both the more effectively in the ratio of its firmer and more extended implantation, is intelligible. That a more perfect performance of a preliminary act of digestion should be a necessary corollary, to be in harmony with a more complete conversion of the food into chyle and blood,—and that such more efficient type of the whole digestive machinery should be correlated, and necessarily so, with the hot blood, quick-beating heart and quick-breathing lungs, with the higher instincts, and more vigorous and varied acts of a Mammal, as contrasted with a cold-blooded reptile or fish,—is also conceivable. To the extent to which such and the like reasoning may be true, or in the direction of the secret cause of the constant relations of many-rooted teeth discovered by observation,—to that extent will such relations ascend from the empirical to the rational category of laws.

The interest which the above-described fossil from the Stonesfield oolite slate excites is not exclusively due to its antiquity, its uniqueness, or its peculiarity; much is attached to its relations as a test in palaeontology of the actual value of a single tooth in the determination of other parts of the organization of the animal. According to our opinion of these unseen parts, we frame our expression of the nature and affinities, or of the place in the zoological system, of the extinct species. From the resemblance of the lower molars of Stereognathus to those of Philophas, which, though not close, is closer than to the teeth of any other known animal, it is probable that the Stereognathus was hoofed, and consequently herbivorous, or deriving the chief part of its subsistence from the vegetable kingdom. Cuvier has written: "La première chose à faire dans l'étude d'un animal fossile est de rechercher la forme des os dans ses parties, et de voir une part-la s'il est carnivore ou herbivore, et dans ce dernier cas, on peut s'assurer, jusqu'à un certain point de l'ordre d'herbivores auquel il appartient." In the case in question the form of the molar teeth of one jaw is recognisable, but the herbivory of the fossil is not thereby determined. We can only infer it to be more probable that the fossil was a Herbivore than an Insectivore or a mixed-feeding Carnivore.

Admitting the herbivory of the fossil, it is not certain that it was hoofed; there is nothing in the form and structure of the tooth to prove that. Both form and structure are compatible with the hoofless multicuspid type of herbivorous Mammal, as shown by the Manatees. It is a small size of the Stereognathus which renders it less probable that it was a distinctive kind of Manatee, and more probable that it was a diminutive form of Ungulate. But seeing the manifold diversities of the multi-cuspid form of molar teeth in recent and extinct insectivorous ungulate quadrupeds, it is not impossible but that the Stereognathus may have belonged to that order; there is no known physiological law forbidding it.

The form of the cusps, and their regular symmetrical arrangement in the Stereognathus, as compared with the known modifications of multi-cuspid molars in certain small extinct forms of hoofed quadrupeds, constitute the grounds upon which an opinion is formed of its most probably belonging to the same section of Ungulata.

Then, is it not true, it may be asked, that by virtue of certain established laws of correlated structures, an extinct animal may be re-constructed from a single tooth or from a fragment of bone? Is the Cuvierian basis, or what has been so regarded, of palaeontology unsound? Not necessarily from anything that has been said or written on the subject of the Stereognathus. We do not know the comparative anatomy of the family of quadrupeds to which the Stereognathus belonged. What we do know of its teeth suggests that that family may have had modifications of the skeleton so far differing from those of any, the modifications of which are known, as to have constituted a type of its own, a special family; but a type as well marked and as distinct as the type of skeleton of the Felidae independently studied in the feline Canivora (fig. 105), and in the omnivorous Herbivora (fig. 107), and by which preliminary study he was enabled to enunciate that beautiful law of the "correlation of forms and structures" to which allusion has been already made, and which will be illustrated by examples, and its mode of application pointed out, in another part of the present article.

In certain instances of constant coincidences of structure, as demonstrated by comparative anatomy, the sufficiency—i.e. recognisable, Mammalia, intelligible, or physiological—cause of them is not yet known. But, as Cuvier in reference to such instances truly remarks, "Since these relations are constant, there certainly must be a sufficient cause for them." In certain other cases Cuvier believed that he could assign that "sufficient cause," and he selects as such the correlated structures in a feline Carnivore, and in a hoofed Herbivore.

The physiological knowledge displayed by him in his explanation of the condition of those correlations is most exact; its application in the restoration of the Amphoterium and Palaeotherium most exemplary.

In the ratio of the knowledge of the reason of the coincidences of animal structures—in other words, as these coincidences become "correlations"—is our faith in the soundness of the conclusions deduced from the application of such rational law of correlations; and with the certainty of such application is associated a greater facility of its application. A knowledge of the physiological conditions governing the relations of the contents of the cavities of bones to the flight and other modes of locomotion in birds both enabled the writer to infer from one fragment of a skeleton that it belonged to a terrestrial bird deprived of the power of flight, and so predicted that such a bird, but at least of the size of the ostrich, would almost certainly be found in New Zealand.

This principle, however,—these modes of thought—which Cuvier affirmed to have guided him in his interpretation of fossil remains, and which he believed to be a true clue in such researches, were repudiated or contested by some of his contemporaries.

Geoffroy St Hilaire denied the existence of a design in the construction of any part of an organized body; he protested against the deduction of a purpose from the contemplation of such structures as the valves of the veins or the converging lens of the eye.

Beyond the co-existence of such a form of flood-gate with such a course of the fluid, or of such a course of light with such a converging medium, Geoffroy affirmed that thought, at least his mode of thinking, could not satisfy, or ought not to go.

The following is the writer's attempt at a briefest summary of the arguments which have been adduced by teleologists and antiteleologists from Democritus and Plato down to Conste and Whewell. The writer would merely remark, that in the degree in which the reasoning faculty is developed on this planet and is exercised by our species, it appears to be a more healthy and normal condition of such faculty,—certainly one which has been productive of most accession to truths, as exemplified in the mental workings of an Aristotle, a Galen, a Harvey, and a Cuvier,—to admit the instructive impression of a design or purpose in such structures as the valves of the vascular system and the dioptric mechanism of the eye.

In regard to the few intellects,—they have ever been a small and unfruitful minority,—who do not receive this impression and will not admit the possibility of existence of final causes in physiology, the writer has already expressed his belief that such intellects are not the higher and more normal examples, but rather manifest some, perhaps congenital, defect of mind, allied or analogous to "colour-blindness" through defect of the optic nerve, or the insaudibility of notes above a certain pitch through defect of the acoustic nerve.

The truth of a physiological knowledge of the condition of a correlated structure, and of the application of that knowledge to palaeontology, is not affected by instances adduced from that much more extensive series of coincident structures of which the physiological condition is not yet known. Nor is the power of the application of the physiologically interpreted correlation less certain because the only empirically recognised coincidences have failed to recur, with the same certainty and to the same extent, an extinct form of animals.

Certain coincidences of form and structure in animal bodies are determined by observation. By the exercise of a higher faculty the reason, or a reason, of these coincidences is discovered, and they become correlations; in other words, it is known not only that they do exist, but how they are related to each other. In the case of coincidences of the latter kind, or of "correlations" properly so called, the mind infers with greater certainty and confidence, in their application to a fossil, than in the case of coincidences which are held to be constant only because so many instances of them have been observed.

Besides, the application of the latter kind of coincidences is limited by the actual amount of observations at the period of such application, and because mistakes have been made through a miscalculation of the value of such amount, it has been argued that a rational law of the correlation of animal forms is inapplicable to the determination of a whole from a part; and it has not only been asserted that the results of such determination are unsound, but that the philosopher who believed himself guided by such law deceived himself and misconceived his own mental processes! But the true state of the case is that the non-applicability of Cuvier's law in certain cases is not due to its non-existence, but to the limitations to which it is subjected.

The consciousness of that limitation led the enunciator of the law to call the attention of palaeontologists expressly to the extent to which it could then be applied, as, for instance, to the determination of the class, but not the order; or of the order, but not the family or genus, &c.; and to caution them also as to the extent of the cases in which, the coincidences being only known empirically, he consequently enjoins the necessity of further observation, and of caution in their induction. Cuvier expresses, however, his belief that such coincidences must have a sufficient cause, and that cause once discovered, they then become correlations and enter into the category of the higher law. Future comparative anatomy will have that great consummation in view, and its result, doubtlessly, will be the vindication of the full value of the law in the interpretation of fossil remains as defined by the illustrious founder of palaeontology.

**Genus SPALACOTHERIUM, Orw.—**The next stratum overlying the older colites in which mammalian remains have been detected, is a member of the newest colitic series at Purbeck, Dorsetshire, called the "marly" or "dirt-bed." In a series of fossils discovered there by Mr W. H. Brodie, and transmitted for determination in 1854 to the writer, amongst the remains of fishes and small Saurians, constituting the majority of the specimens, were detected three unequivocal evidences of a mammalian species, which were described under the name of *Spalacotherium tricuspis.* The specimen here selected (fig. 85) to exemplify the condition of this interesting Mammal, is a right ramus of the lower jaw. The posterior half contains four teeth, and extends backward beyond the dental series; but instead of showing the compound structure which that part of the jaw exhibits in the lizard tribe, it continues undivided; the convex surface showing a smooth depression for the insertion of the temporal muscle; the concave boundary answering to that going to the condyle and ramus of the jaw, and the upper one to that going to the external process in the ramus of the jaw of the turtle and all shrews. The crowns of the teeth are long, narrow, and tricuspid, the inner part of the crown being produced into a point both before and behind the longer cusp which forms the chief outer division of the crown. Each of these teeth is implanted by a fang divided externally into two roots, in a distinct socket in the substance of the jaw. The multicuspid crown, the divided root of the tooth, its complex implantation, and the undivided or simple structure of the ramus of the jaw, all concurred, therefore, to prove the mammalian nature of this fossil.

The other specimens showed that the *Spalacotherium* had ten molar teeth in each ramus of the lower jaw, preceded by a small canine and incisor. The anterior molars are compressed, increase in height and thickness toward the sixth, and from the seventh decrease in size to the hindermost, which seems to be the last of the series. The sharp multicuspid character of so much of the dental series as is here preserved repeats the general condition of the molar teeth of the small insectivorous Mammals in a striking degree: one sees the same perfect adaptation for piercing and crushing the tough chitinous cases and elytra of insects. The particular modification of the pointed cusps, as to number, proportion, and relative position, resembles in some degree that of the Cape mole (*Chrysochloris capensis*), but both in these respects and in the number of molars, the dentition accords more closely with that of the extinct *Dinotherium.*

The chief interest in the discovery of the *Spalacotherium* is derived from its demonstration of the existence of mammals about midway between the older colites and the oldest tertiary periods.

Both the Oxford colitic slate and the Purbeck marly shell-beds give evidence of insect life; in the latter formation abundantly. The association of these delicate Invertebrata with remains of plants allied to Zannia and Cycas, is indicative of the same close interdependency between the insect class and the vegetable kingdom, of which our power of surveying the phenomena of life on the present

---

1 Discours sur les Révolutions de la Surface du Globe, 4to, 1826, p. 59. 2 Transactions of the Zoological Society, vol. iii., p. 32, pl. 3. 3 De Blainville, Cartographie, 4to, fasc. 1, 1839, p. 34. 4 Prof. Huxley, "Lecture on Natural History," &c., Royal Institution of Great Britain, Feb. 15, 1856. 5 From *κατά*, a mole, *βαρύς*, a beast. Mammalia, surface of the earth enables us to recognize so many beautiful examples. Amongst the numerous enemies of the insect class ordained to maintain its due numerical relations, and organized to pursue and secure its countless and diversified members in the air, in the waters, on the earth and beneath its surface, bats, lizards, shrews, and moles now carry on their petty warfare simultaneously; in warmest latitudes work together, etc., the localities, in their allotted time, have been found to be thereby felt at the discovery that Mammals and Lizards co-operated simultaneously and in the same locality at the same task of restraining the undue increase of insect life during the period of the deposition of the Lower Purbeck beds.

Genus Triconodon, OW.

Sp. Triconodon mordax.—This name is proposed for a small zoophagus Mammal, whose generic distinction is shown by the shape of the crown of the molar teeth of the lower jaw, which consist of three nearly equal cones on the same longitudinal row, the middle one being very little larger than the front and hind ones; and these cones are not complicated by any cingular or accessory basal cusp. The convex condyle is below the level of the alveoli, and there is no angular process projecting beneath it. The coronoid process is broad and high, with its hinder point not extended so far back as the condyle; the depression marking the insertion of the temporal muscle extends nearly to the lower border of the jaw. There are the obscure remains of three broken incisors, and the point of apparently a canine; next come the two stumps or broken roots of a small premolar; then the crown of a second double-rooted premolar, which shows a principal cone and a smaller cone on each of the next teeth wanting; then there is a larger premolar, with the two tangs raised some way out of their socket; the crown of this tooth shows a principal cone, with a small anterior and large posterior talon; it rises, apparently from partial displacement, higher than the succeeding molars; these are three in number, and present the characteristic three-coned structure already described; each cone is smooth, and convex externally. The three cones seem to answer to the three middle or principal cones of the molars of Amphilestes and Phacolotherium, but the front and hind cones are raised to near equality with the middle cone in Triconodon.

The lower jaw of this species, in the relation of the condyle to the lower border, resembles that of Phacolotherium more than that of Amphilestes; but it differs from both; there is not the same gradual curve from the condyle to the symphysial axis as in Phacolotherium; and, besides the lower level of the condyle, it is divided by a less deep notch from the coronoid process. This process is larger in proportion to the entire jaw; approaches more nearly to the quadrato or rhomboid form, the upper border being less curved; it affords a more extensive surface of attachment to the principal biting muscles than in most predatory extinct or recent quadrupeds. This character, with the depth and strength of the jaw, suggested the specific name. From the shape of the exposed part of the ramus, we may conclude that the part answering to the angle is bent inwards, and that Triconodon was a genus of the marsupial order. The specimen was discovered by Mr. Beocles in the "dirt-bed" at Purbeck at the spot at which Phacolotherium was found.

Genus Plagiulax, Fr.—The most remarkable of Mr Beocles' discoveries in the above formation are the mammalian jaws indicative of the genus above named, of which two species have been determined by Dr Falconer.

Sp. Plagiulax Beoclesi, Fr.—Two specimens exemplified the but implanted by a thick root in the fore part of the jaw, like the Mammalia. large lower incisor of a kangaroo or wombat. The three anterior teeth in place have compressed oval-shaped crowns, and rapidly augment in size from the first (1) to the third (4). They are followed by sockets of two much smaller teeth, shown in other genera to have subcylindrical crowns resembling those of Microtus. The large front tooth of Plagiulax is formed to pierce, retain, and kill; the succeeding teeth, like the carnassials of Carnivora, are, like the blades of shears, adapted to cut and divide soft substances, such as flesh. As in Carnivora also, these sectorial teeth are succeeded by a few small tubercular ones. The jaw conforms to this character of the dentition. It is short in proportion to its depth, and consequently robust, sending up a broad and high coronoid process (b), for the adequate grasp of a large temporal muscle; and the condyle (c) is placed below the level of the grinding teeth,—a character unknown in any herbivorous or browsing Mammal; whilst the lever of the coronoid process is made the stronger by the condyle being placed farther behind it than in any known carnivorous or herbivorous animal. The angle of the jaw makes no projection below the condyle, but is slightly bent inwards, according to the marsupial type.

Sp. Plagiulax minor, Fr.—In this species the first premolar (fig. SS, p. i) is preserved; the rest (p, 2, 3, and 4) show nearly the same shape and proportions as in P. Beoclesi. The first molar (m, i) has a broad depression on the grinding surface, surrounded by tubercles, of which three are on the outer border; the external tubercles of the second smaller tooth are smaller and more numerous.

In the general shape and Plagiulax minor (four times nat. size), proportions of the large Purbeck (after Lyell).

Fig. 87.

Plagiulax Beoclesi (twice nat. size), Purbeck.

shape and proportions of the entire jaw of this species (fig. 87). The foremost tooth (i) is a very large one, shaped like a canine,

---

1 An abbreviation for Plagiulaxodon, from ἀπλοῦς, oblique, and ὀδούς, groove; having reference to the diagonal grooving of the premolar teeth.

VOL. XVII. Genus Coryphodon, Ow.—Rarely has the writer felt more misgiving in regard to a conclusion based, in palaeontology, on a single tooth or bone, than that to which he arrived after a study of the unique fragment of jaw with one tooth dredged up off the Essex coast, and on which he founded the genus Coryphodon.

The marked contraction of the part of the jaw near one end of the tooth seemed, at first view, clearly to show it to be the narrower fore part of the ramus; in that case the tooth would have been a premolar, and of comparatively little value in the determination of a genus or species. But a closer inspection showed the line of abrasion of the summits of the two transverse ridges of the tooth to be on one side, and the general law of the relative positions and reciprocal action of the upper and lower grinders in tapiroid Pachyderms determined that those oblique lines abraded must be on the hinder side of the ridge. These smaller characters carried conviction against the showing of the larger and more catching ones. So, in determining the position of the nautilus in its pearly abode, when the animal without its shell was first brought to England in 1831, the reasons afforded by some small and inconspicuous parts in like manner outweighed the first impressions from more obvious appearances, as well as the bias from the general analogies of testaceous Univalves. Some contemporary naturalists asserted, and for a time it was believed, that the nautilus had been put upside down in its shell; just as some contemporary anatomists surmised that the writer had mistaken the fore for the back part of the jaw of his Coryphodon, which, in that case, might only be the known Lophiodon. In both instances the writer's conclusions have been confirmed, and proved to be correct. And the writer would maintain that, in the course of his experience, he has often found that the prominent appearances which first catch the eye and indicate a conformable conclusion are deceptive, and that the less obtrusive phenomena which require searching out, more frequently, when their full significance is reasoned up to, guide to the right comprehension of the whole. It is as if truth were whispered rather than outspoken by Nature.

Truth, it is sometimes said, lies at the bottom of a well. The first additional glimpse that the writer obtained of the veritable nature of one of our most ancient tertiary Mammals was derived from the inspection of a fossil tooth brought up from a depth of 160 feet, out of the "plastic clay," during the operations of sinking a well in the neighbourhood of Cambridge, near London. It was a canine tooth, belonging from its size (near 3 inches in length) to a large quadruped, and, from the thickness and shortness of its crown, crown not to a carnivorous but to a hoofed Mammal, most resembling in shape, though not identical with, that of the crown of the canine tooth of some large extinct tapiroid Mammals, which Cuvier had referred to his genus Lophiodon, but which has proved to belong to Coryphodon.

The last lower molar of Lophiodon has three lobes; the molar determined to be the ultimate one, in the fragment of lower jaw above referred to, resembles that of the tapir in the absence of a third or posterior lobe, but the posterior ridge or part of the cingulum is less developed than in the tapir. It presents the divisions in the form of transverse ridges or eminences, the front ridge being the largest, and with its ridge more entire. From the outer end of each division the ridge is inclined obliquely forward, inward and upward; the anterior one extends to the antero-internal angle of the base of the crown; the posterior one terminates at the middle of the interspace between the two chief divisions of the crown. The trenchant summit of the anterior ridge is slightly concave toward the fore part of the tooth, as in that of Lophiodon; but its outer and inner ends rise higher, and appear as more distinct cones or points; whence the generic name of Coryphodon. The posterior division is lower than the anterior one, and is bicuspid; the trenchant margins connecting the outer and inner points does not extend across the crown parallel with the anterior ridge, as in Tapirus and Lophiodon, but bends back so as to form an angle, the apex of which rises into a third point.

Some lophiodontoid fossils from the lignites of Solsona and Lecos, and from the phillie clay of Mondain in France, including the upper molar tooth figured by Cuvier in the chapter of the Osseux Fossiles entitled "Animaux voisins de Tapirus," pl. vii., fig. 6, belong to the genus Coryphodon. Cuvier compares this tooth with one from Bastberg, which he figures in pl. vi., fig. 4, and which is certainly the last upper molar of a true Lophiodon, and points out truly that the Solsona tooth differs in the external border passing into the posterior one, so that, instead of being quadrangular, its crown is triangular; but he explains this difference on the hypothesis that the Bastberg tooth was a penultimate molar. The reduction of the second or posterior ridge to a semi-Mammalia circular one, developed at its middle and hindmost part into a prominent cone, so far agrees with the modification of the same part of the last molar of the lower jaw of the Coryphodon as to render it very probable that the last upper molar from Solsona, figured by Cuvier in pl. vii., fig. 6, above quoted, also belongs to the genus Coryphodon. Cuvier states that the entire skeleton was found, indicative of an animal as long and almost as large as a bull; but that the workmen employed in the sandpit (submarine) preserved only that one tooth. Both the lower molar from Harwich, and the upper one from Solsona, indicate an animal of at least double the size of the American tapir.

Professor Hebert has recently described a very instructive series of teeth and bones from the oldest eocene deposits in France, which he refers to the genus Coryphodon; the last molar is identical in form with the tooth from the plastic clay of Essex, on which the genus was originally founded.

Genus Philocephalus, Ow.—The most complete and instructive example of a Mammal from the next overlying division of the eocene teritories, viz., the "London clay," is that which the writer has described under the name of Philocephalus vulpisceps. It is a hoofed Herbivore, but presents a dentition not exhibited by any later or existing species of Mammal.

The length of the skull (fig. 89) is 4 inches, its extreme breadth 2 inches 2 lines, the height of the cranium opposite the first premolar tooth 9 lines. Its shape and characteristics determine the hoofed nature of this species, and its affinities to the Perissodactyla, or the order of Ungulata with toes in odd number. The extent and well-defined boundary of the temporal fossae by the occipital (3), parietal (7), and post-frontal ridges, and their free communication with the orbits, give almost a carnivorous character to this part of the cranium. The brain-case has been placed by Hyrax and Palaeotherium, the greatest cerebral expansion is at the middle and toward the rear of the fossae, with a contraction toward the occiput; the brain-case not continuing to enlarge backward to beyond the origin of the zygomatics, as in the fox. The zygomatic arches have a less outward span than in the Carnivora. In this part of the cranial structure Philocephalus resembles Palaeotherium more than it does any existing Mammal; but the post-frontal processes are longer and more inclined backward. The incompleteness of the orbit occurs in both Anoplotherium and Palaeotherium, as in Rhinoceros, Tapirus, and the hog tribe; but in the extent of the deficient rim, Philocephalus is intermediate between Palaeotherium and Tapirus. The orbit is not so low placed as in Palaeotherium, Tapirus, and Rhinoceros; nor so high as in Hyrax. The straight upper margin of the skull (V to 15) is like that in the hog tribe and Hyrax, and differs from the convex contour of the same part in the Anoplotherium and Palaeotherium. The size of the antorbital foramen (a) indicates no unusual development of the muzzle or upper lip. In the conformation of the nasal aperture by four bones (two maxals, 15, and two premaxillaries, 22), Philocephalus resembles the horse, Hyrax, hog tribe, and Anoplotherium, which have the maxillaries, as well as the nasals and premaxillaries, entering into the formation of the external bony nostril.

The ungulate and herbivorous character of Philocephalus is most distinctly marked by the modifications of the lower jaw, especially by the relative dimensions of the parts of the ascending ramus. Mammalia, which give the extent of attachment of the biting (temporal) and grinding (masseteric and pterygoid) muscles respectively. In the shape of the mandible Philotherium most resembles Tupirus among existing Mammals, and the Palaeotherium among the extinct ones in which that shape is known. As in almost every species of eocene quadruped yet discovered, the Philotherium presents the type-dentition of the placental diphyodont series, viz.:

\[ \begin{array}{cccc} 1 & 3 & 2 & 1 \\ 3 & 1 & 4 & 3 \\ \end{array} \]

The incisors are preserved in the lower jaw with marks of attrition on their crowns demonstrating corresponding teeth of the same number (6), and of similar size, in the upper jaw, from which the alveolar part of the premaxillaries had been broken away.

The canines are small in both jaws; they are separated by a vacant space from the outer incisors, and by a longer interval from the first premolars. These form a continuous series with the remaining teeth in the upper jaw, but are separated by a space of about half their breadth from the second premolar in the lower jaw. The succeeding teeth (1, 2, 3, 4) increase in size to the penultimate molar in the upper, and to the last molar in the lower jaw (p. 4, in figs. 90 and 91), which tooth has a third lobe.



In the last premolar upper jaw (fig. 90, p. 4) the cingulum is uninterrupted along the outer side from its anterior well-developed talon (c) to the back part. The two outer cones resemble those of the true molars; but there is only one inner cone, and the crown of p. 4 differs accordingly from that of m. 1, being triangular rather than square. A ridge is sustained from the interspace between the anterior talon (c) and the outer anterior lobe obliquely inward and backward to the inner lobe, swelling into a small tubercle at the middle of its course.

The first molar (m. 1) presents four low thick cones, two internal and two external: each external cone is connected with its opposite internal one by a low ridge, swelling into a tubercle at the middle of its oblique course. The cingulum (c) seems to be continued uninterruptedly round the crown of this tooth, thickest at the fore and back part, and at the interspace of the inner lobes; and developing the small accessory antero-external tubercle. The second molar (m. 2) is similar to, but rather larger than, the first; the tubercle on the oblique ridge connecting the two front lobes is less developed. The cingulum is obliterated on the inner side of the posterior lobe.

The last molar is rather narrower behind than m. 2; the tubercle on the anterior of the oblique connecting ridges is smaller: that on the posterior ridge is almost obsolete.

In the last lower premolar (fig. 91, p. 4) the division and development of the anterior lobe gives rise to a pair of cones, one external (a), the other internal (b), connected anteriorly by a basal ridge, in front of which is the fore part of the cingulum. The low posterior lobe (e) shows the rudiment of a second internal cone (d).

The first molar (fig. 91, m. 1) has a pair of front lobes and a pair of hind lobes, with an oblique ridge continued from postero-internal lobe to the interspace between the front pair.

The second molar (m. 2) shows an increase of size; but its chief and most interesting modification is the development of a tubercle (e) between the two anterior lobes, making three cones on the same transverse line, and thus repeating the character of the molar tooth of Stereognathus (fig. 92, e). The oblique ridge from the outer and hinder lobe (e) abuts against the intermediate tubercle (e). The nearest approach to the above dentition is made by the extinct Hyracotherium, also a fossil from the London clay.

The third trochanter on the femur of Philotherium, and the association of three metatarsals in one portion of the matrix, as if belonging to the same limb root, confirm the essentially perissodactyl affinities of that genus as shown by the skull and teeth. Philotherium and Hyracotherium form a well-marked section in the lophiodont family, which seems to have preceded the palaeotherian family in the order of appearance, and to have retained more of the general ungulate type than that family. This is shown by the graduation of the tapiroid modification of the molar teeth into one more nearly resembling that of the Anthracotheria and Choropodonta; by the absence of the postero-internal cone on the ultimate premolar, by which all the premolars are, as in artiodactyls, less complex than the true molars, by the form and position of the nasal bones and the structure of the external nostril.

**Genus Lophiodon**, Cuv.—In the year 1800 Cuvier first announced the discovery of the fossil remains of a quadruped allied to that of the tapir, in the tertiary deposits of the Montagne Noire, near Isel, department of Aude in Languedoc. The outer incisor of the lower jaw was shortened to give room to the longer corresponding incisor above, as in the tapir; the canines offered the same proportional development, but the three first molars (premolars) of the lower jaw presented a more simple structure, having the crown compressed, and forming two cones, the front one being the largest—in short, a structure, the type of which is presented only by the first of the three premolars (p. 2) in the genus Tupirus.

Years elapsed ere Cuvier obtained clear evidence of the structure of the upper molars of this new fossil Mammal. Such detached teeth as had been obtained from the fresh-water formations near Isel were referred, owing to the way in which they departed from the type of the true molar teeth of the Tupirus to the genus Rhinoceros. This fact is indicative of the uncertain affinities of the Lophiodon in the perissodactyl series. Besides the character of form, the upper molar series of Lophiodon differs, like the lower cone, from that in Tupirus, in the greater simplicity of the last two premolars; these teeth have a single cone on the inner side in Lophiodon; they have there two cones in Tupirus, forming the inner terminations of two transverse ridges, as in the true molars. These teeth in the Lophiodon differ from those in the Tupirus in the greater fore-and-aft expanse of the outer terminations of the transverse ridges, and the less depth of the cleft between them—a more complete coalescence of those parts causing a more entire outer wall of the crown, completing the transition to the Rhinoceros type, toward which the Palaeotherium is the next stage.

**Genus Palaeotherium**, Cuv.—This extinct genus of quadruped was restored (fig. 93) by Cuvier through a series of admirably instructive steps, ultimately verified by a complete series of fossils, obtained chiefly from the upper eocene gypseous formation at Montmartre and other parts of France. The molar teeth of Palaeotherium (fig. 94) approach nearer to those of the rhinoceros; but in the number, kind, and general arrangement the entire dentition resembles that of Philotherium. The skull affords indications that the Palaeotherium possessed a short proboscis. It had three toes on each foot, each terminated by a hoof;

---

1 See art. ODONTOLOGY, vol. xvi., p. 448, for Cuvier's "Homologies of the Teeth," and explanation of their symbols. 2 Bulletin des Sciences, Paris, Nivose, an. viii., No. 34. 3 Art. ODONTOLOGY, p. 471, fig. 136, p. 2. 4 Ibid., p. 470. Mammalia, the middle one being the largest. The femur had a third trochanter, and the dorsal-lumbar vertebrae were 21 in number. Several species of Palaeotherium have been described, ranging from the size of a sheep (P. sericea) to that of a horse (P. major). Fig. 94 gives the grinding surface of an upper molar of this species from the upper oolite of the Bembridge beds, Isle of Wight. The crown is divided into an anterior (a) and posterior (b) part by an oblique fissure (c), continued from near the middle of the inner surface of the crown obliquely across two-thirds of the tooth. Each division is subdivided partially into an outer (ab) and an inner (cd) lobe; the former is marked by the terminal expansion (i) of the fissure (c), the posterior one by the fissure (g). The lobes (c and d) are bordered near their base by a ridge. This is the type of grinding surface, in which are superinduced the modifications of that surface in the upper molars of the rhinoceros and horse. The dental formula of Palaeotherium is $3\cdot3\cdot1\cdot1\cdot4\cdot4$, $m=44$. The canines exceed in length the other teeth and there are consequently vacancies in the dental series for the lodgement of the crowns of the canines when the mouth is shut.

Genus Anoplotherium, Cuv.—With the same dental formula as in Palaeotherium. The present genus, like Dichodon, had no interval in the series of teeth; neither the canine nor any other tooth rising above the general level.

The grinding surface of the molar teeth somewhat resembles and presages the ruminant type; in the upper jaw the crown (fig. 95) is divided into a front (f) and a back (b) lobe by a valley (a) extending two-thirds across. A second valley (g) crosses its termination at right angles, forming a curved depression in each division, which it thus subdivides into two lobes, concave towards the outer side of the tooth. There is a large tubercle (m) at the wide entry of the valley (e). The Anoplotherium (fig. 96) was of a lighter and more elegant form than the Palaeotherium; its limbs terminated each in two digits, with the metapodial bones distinct, and the last phalanx hoofed. Some transitory characters of the embryo ruminant were retained throughout life by the Anoplotherium. The species restored in fig. 96 was about the size of a fallow-deer.

Fig. 95.

Upper Molar, Anoplotherium commune (Eocene Gypse).

Fig. 96.

Restoration of the Anoplotherium (Eocene Gypse).

had a long and strong tail, and was probably of aquatic habits. Smaller and less developed species of Anoplotherium from upper oolite strata have been referred to distinct genera by later palaeontologists. The researches of Barron Cuvier, which resulted in the restoration of the Palaeotherium and Anoplotherium, are the most instructive which the palaeontologist can study. They form the third volume of the 4th edition of the Osteums Faunae, 1822-5.

Genus Dichodon, Ow.—The upper oolite beds of Hampshire have yielded evidence of an extinct form of even-toed (artiodactyle) hoofed quadruped, most interesting as a transitional form between the Anoplotheroids and the true Ruminants. Like the Anoplotherium the dental series is continuous, without break—a character which is Mammalia only manifested by mankind among existing Mammals; the crowns of the teeth, in Dichodon, being all of nearly equal height, as in the teeth of man. On each side of both upper and lower jaws there are in the Dichodon (art. Odontology, fig. 118) three incisors (i, i, i, i), one canine (c), four premolars (p, p, p, p), and three true molars (m, m, m)—in all forty-four teeth, constituting the typical Diphyodont dentition which so many mammalian genera, on their first appearance in the oolite strata, exhibit. It is formalized as follows: $i\cdot3\cdot1\cdot1\cdot4\cdot4$, $m=44$. From the first incisor to the third premolar the teeth have all a trenchant, and, after the canine, a somewhat trenchant character. The back of the third premolar (p), and all the fourth premolar, show the crushing form of crown, which in the true molars, after the wearing down of the first sharp cusp, produces the double crescentic lines of enamel which are now peculiar to the Ruminants amongst hoofed quadrupeds. The extinct species showing the above characters, and on which the genus was founded, was nearly the size of a fallow-deer; it is called Dichodon cupulidens, in reference to the number of sharp points on the unworn molars. The dentition indicates that its food may have been of a peculiar character, perhaps not exclusively of a vegetable nature.

In the same upper oolite formation of Hampshire have been found instructive examples of some smaller members of the extinct anoplotheroid family.

Genus Xiphodon, Owen—This genus Xiphodon (from xiphos, a sword, colis), was proposed by Cuvier, in the second edition of his Osteums Faunae, 4to, tom. iii., 1822, p. 64, for the Anoplotherium minus of the original Memoir in the Annales de Muséum, tom. iii., 1803, and for the A. leptorhynchus of the 4th edition, 1822, tom. i., pl. 2, fig. 3; and tom. iii., pp. 70 and 251. It is closely allied to the anoplotheroid genus Xiphodon; the dental formula is the same, only there is a slight interval between the canine and the first premolar in both jaws; the first three premolars are sub-compressed, sub-trenchant, but less elongated from behind forwards than in Xiphodon. Besides the two normally-developed and functional digits on each foot, there be one, sometimes two, supplementary digits.

The best illustration of the structure of the upper true molars is afforded by the figure of one of these teeth in the Proceedings of the Geological Society, May 29, 1846, published in the Quarterly Journal, vol. ii., p. 420. "The anoplotheroid character of the tooth is shown by the large size of the lobe (p, a, fig. 1), and the subgenomic peculiarity by the continuation of its dentinal base with that of the inner and anterior lobe (mf), at the early stage of attrition presented by the crown of the tooth in question. In the large and typical Anoplotherium, the lobe (p) preserves its insular form and uninterrupted contour of enamel until the crown is much more worn down than in the present tooth (fig. 1). In this respect, as in the modifications of the lower molar teeth, the genus Dichodon shows its closer affinity to the true Ruminants; but the little fold of enamel dividing the lobe (mf) from p distinguishes the upper molar tooth in question from that of any Ruminant." (T. 421.)

A more interesting species of this genus, called D. cupulidens, has been founded on an almost entire lower jaw with the permanent dental series, wanting only the four middle incisors, which now forms part of the palaeontological collection in the British Museum. The dental formula, as shown by the mandibular teeth, and by the evidence on their crowns of the presence of the teeth of the upper jaw, is the typical one in diphyodont Mammalia, viz., $i\cdot3\cdot1\cdot1\cdot4\cdot4$, $m=44$. The canine, with a crown like that of the first premolar, and not longer, is separated from it by an interval of half the breadth of the crown, and by a narrower interval from the outer incisor. The first premolar is divided by an interval of scarce a line's breadth from the second. The rest of the molar series are in contact. The total length of the lower jaw is 5 inches 11 lines (0m=148); that of the molar series is 2 inches 11 lines (0m=75); that of the three true molars is 1 inch 4½ lines (0m=35). The near equality in height of the crowns of all the teeth, and their general character, show that the animal belonged to the anoplotheroid family. The dentition of the present species differs from that of Dichodon in the absence of the accessory cusp on the inner side of the base of the true molars; and both from Dichodon cupulidens and Xiphodon gracilis in the minor antero-posterior extent of the premolars; it corresponds with Dichodon (as represented by the D. leptorhynchus, Cuvier) in the proportions of the premolars and in the separation of the canine from the adjoining teeth; to this genus, therefore, the fossil is referable, provisionally, in the absence of knowledge of the molars of the upper jaw, which are the most characteristic; and the writer has proposed to call the

---

1 See art. Odontology, p. 439. 2 Quarterly Journal of the Geological Society, tom. iv., 1847, p. 36, pl. 4. Genus XIPHOPODON.—The genus Xiphoodon was indicated, and its name proposed by Cuvier, for a small and delicate, long and slender-limbed anoplotheroid animal, which, in his first Memoir (Annales de Muséum, tom. iii., p. 55, 1803), he had called Anoplotherium macrurus; but he altered the name, in the second edition of the Osteum Fossile (tom. iii., pp. 69 and 231, 1822), to that of Anoplotherium gracile.

The distinction indicated by Cuvier is now accepted by palaeontologists as a generic one, and a second species (Neohyodon goleensis) has been added by M. Gervais (Paléontographie Française, 4to, 1845, p. 90) to the type-species Xiphoodon gracile, of which he figures an instructive portion of the dental series of both jaws, obtained from the ligature of Dibrége near Apt. The dental formula of Xipho- don is the typical one, viz.: $I^1 \cdot I^1 \cdot C^1 \cdot P^4 \cdot M^3 = 44$.

The teeth are arranged in a continuous series in both jaws. The canines and first premolars have the crowns more extended antero-posteriorly, lower, thinner transversely, and more trenchant, than in the type Anoplotherium (whence the name Xipho- don, or sword-tooth). The feet are didactyle, with metatarsals and meta- tarsals distinct. The tail is short. The lower true molars have two pairs of crescentic lobes with the convexity turned outwards.

Genus MICROCHERUS.—Entire crania of Microcherus, from the lacustrine calcareous marls of the Pay-de-Dôme, are in the British Museum, and these show that the hinder division of the upper true molars was complicated by the additional (third) cusp.

With regard to Microtherium, the unusually perfect fossil skulls of that small herbivorous animal did not exceed in size the delicate chevrotains of Java and other Indo-Australian islands, e.g., Tragulus kanchil.—One of importance in regard to the question of their alleged affinity to the Ruminants, on account of the demonstration they give of the persistent and functional upper incisor teeth. The little eocone even-toed Herbivores, like the larger Anoplotheroids, thus departed from the characters of the true Ruminants of the present day, in the same degree in which they adhered to the more general type of the artiodactyls. Had M. de Blainville, who be- lieved them to be Ruminants, possessed no other evidence of the Microtherium than of the Dichobune mariae and Dichobune obesa, Cuv., he would have had the same grounds for referring the Micro- theria, to the Dichobunes, to the genus Tragulus or Moschus (les Chevrotains); but the entire dentition of the upper jaw of the spe- cies of Microtherium is unlike that of any other species by Cuvier to his genus Dichobune, must be known before the existence of Ruminants in the upper eocone gypsum of Paris can be inferred.

No doubt the affinity of these small Anoplotheroids to the Chevrotains was very close. Let the formative force be transferred from the small upper incisors to the contiguous canines, and the transition would be effected. We know that the Ruminant stomach of the species of Tragulus is simplified by the suppression of the palearium or third bag. The stomach of the small Anoplotheroids, whilst preserving a certain degree of complexity, might have been somewhat more simplified. The certain information which the gradations of dentition display by the above-cited extinct species import, however, the artificial character of the order Ruminants of the modern systems, and to the natural character of that wild group of even-toed hoofed animals for which has been proposed the term ARTIODACTYLA.

Genus HYENODON, Laiz.—With the delicate and beautiful Her- bivora of the upper eocone and lower miocene periods, there co- existed carnivorous quadrupeds, which, to judge by the character of their flesh-cutting teeth (carnassials), were more fell and deadly in their destructive task than modern wolves or tigers. Of these old extinct Carnivora a species of the remarkable genus Hyenodon, about the size of a leopard, had left its remains in the upper eocone of Hordwell, Hampshire. Fig. 113, art. ONTOLOGY (vol. xvi., p. 464), shows the dentition of the upper jaw of another species of the same genus, found in the beds at Débrége and Alais, France. The carnassial teeth ($m_1$, $m_2$, $m_3$), instead of being one in number in each ramus of the jaw, as in modern Felins, were three in number, equally adapted by their trenchant shape, to work like scissors- blades on the teeth of the upper jaw, in the act of cutting flesh. After the small incisors came a pair of large piercing and prehensile canines ($c$), followed by four compressed pointed and trenchant premolars ($p_1$, $p_2$, $p_3$, $p_4$) in each side of the jaw; the whole of this carnivorous dentition conforming to the diphyodont type—

$$I^1 \cdot I^1 \cdot C^1 \cdot P^4 \cdot M^3 = 44.$$

Genus AMPHICRON.—With the foregoing predecessor of the digitigrade Carnivora was associated a forerunner of the planti- grade family, viz., a large extinct species having the molars tuber- culated, after the pattern of those of the bears; but retaining, like Mammalia, Hyenodon, the perfect type of diphyodont dentition. Fig. 114, art. ONTOLOGY (vol. xvi., p. 464), shows the teeth of one side of the upper jaw of the Amphicron giganteus. The first and second molars ($m_1$ and $m_2$) have each two tubercles on the outer side and one on the inner side; the last tubercular molar ($m_3$) is of very small size. Fossil remains of Amphicron have been found principally in the miocene deposits at Sansans, south of France. Those of a smaller species from the miocene at Eppelsheim, have been referred to the wolverine genus as Gulo diaphorus, Kaup.

The proofs of the abundant mammalian inhabitants of the eocone continued to be obtained by Cuvier from the fossilized remains in the deposits that fill the great Parisian excavation of the chalk. But the forms which that great antiquary restored were all new and strange, specifically, and for the most part peculiarly distinct from all known existing quadrupeds. By these restorations the naturalist was first made acquainted with the aquatic cloven- hoofed Anoplotheria, and with its light and graceful congeners, the Dichobunes and Xiphocon, with the great Palaeotheres, which may be likened to hornless rhinoceroses, with the more tapiroid Lopho- don, with the large peccary-like Pachyderm called Choropotaenus, and with about a score of other genera and species of placental Mammalia.

Among the sole exception to the generic distinction of these eocone forms from modern ones was yielded by the opossum of Montmartre (Didelphis Gypsom, fig. 97), and what made this discovery the more remarkable was the fact that all the known existing species of that marsupial genus are now confined to America, and the greater part to the southern division of that continent. An opossum appears to have been associated with the peccary-like Hyacrothe- rium in the eocone sand of Suffolk; where likewise, a porcine beast with tusks like ordinary canines (Choropotaenus), and some remains of a marsupial (Erythrops) have been found. With respect to the Didelphis Gypsomus, its generic relations were deduced from characters of the lower jaw and teeth; but these were asso- ciated with other parts of the skeleton in the same block of stone. When Cuvier expressed his convictions from the teeth and other parts first examined, his scientific associates were incredulous. He invited them, therefore, to witness a crucial test. The outline of the back part of the pelvis was exposed, the fore part buried in the matrix. By his delicate use of the graving-tool, Cuvier brought to light the fore-part of the pelvis with the two marsupial bones (fig. 97, a, a) in their natural position. He thus demonstrated that there had been buried in the soft fresh-water deposits, hardened in ages ago into the building- stone of Paris, an animal whose genus at the present day is pecu- liar to America; it is not without striking remark that the Peccary, the nearest existing ally to the old Choropotaenus, is, like the opossum, now peculiar to America; and that two species of tapir, the nearest living allies to the Lophodon and Palaeotheres, exist in South America.

The marine deposits of the miocene epoch show the remains of extinct genera of dolphins (Zyphius and Dioploodon) and of whales (Balæodon). Petrified cetaceous teeth and ear- bones, called "cetotolites" (fig. 98) have been washed out of previous strata into the red crag of Suffolk. These fossils belong to species distinct from any known existing Cetacea, and which, proba- bly, like some contem-

---

1 Quarterly Journal of the Geological Society, vol. iv., 1847. Mammalia, perary quadrupeds, retained fully-developed characters which are embryonic and transitory in existing cognate Mammals. The teeth of these Cetaceae were determined in 1840, the ear-bones in 1843.

The vast numbers of these fossils, and the proportion of phosphate of lime in them, led Professor Henslow¹ to call the attention of agricultural chemists to the red clay as a deposit of valuable manure. Since that period it has been in large supply, with many thousands of pounds annually, of the superphosphate. The red clay is found in patches from Walton-on-Naze, Essex, to Aldbro', Suffolk, extending from the shore to 5 or 15 miles and more inland. It averages in thickness 10 feet, but in some places 40 feet. Broken-up septarian nodules form a rude flooring to the clay, left by the washing off of the London clay, and called "rough stone." The phosphatic fossils, or "cops," as they are now locally termed, occur in greatest abundance immediately above the "rough stone." Thousands of cubic acres of earlier strata must have been broken up to furnish the cetacean nodules of the "red clay." This is a striking instance of the profitable results of a seemingly most unpromising discovery in pure science,—the determination of what in 1840 was regarded as a rare, unique, and most problematical British fossil.²

Our knowledge of the prehistory of mammalian life during the miocene period is derived chiefly from continental fossils. These teach us that one or two of the generic forms most frequent in the older tertiary strata still lingered on the earth, but that the rest of the eozone Mammalia had been superseded by new forms, some of which present characters intermediate between those of eozone and those of pliocene genera. The Dinotherium and narrow-toothed Mastodon, for example, diminish the interval between the Lophiodon and the elephant; the Anthracotherium and Hippohippus, that between Choropithecus and Hippopotamus; the Acheratherium was a link connecting Palaeotherium with Rhinoceros.

One of the most extraordinary of the extinct forms of the cetaceous order has been restored from fossil remains discovered in formations of the miocene age in Arkansas, North America. The tooth of this carnivorous whale, for which the generic name Zeuglodon seems likely to be generally accepted, was first described and figured by the medieval palaeontologist Scilla, in his treatise entitled De Corporibus Marinis (470, 1747, tab. xii., fig. 1), and have since given rise to various interpretations. The originals were obtained from the miocene strata at Malta, and are now preserved in the Woodwardian museum at Cambridge.

The remains of a gigantic species of the same genus, discovered by Dr Harlan in miocene formations of Arkansas, Mississippi, were described and figured by him as those of a reptile, under the name of Basiliscusaurus.³ Teeth of a smaller species discovered by M. Grateloup, in miocene beds of the Gironde and Herault, were ascribed by him also to a reptile, under the name of Synododon.⁴ In 1829 Dr Harlan brought over his specimen of Basiliscusaurus to London, and submitted them to the writer's inspection, by whom they were determined to be mammalian and cetaceans. The entire skeleton has since been obtained from miocene deposits in Alabama, revealing a length of body of about 70 feet. The skull is very long and narrow; the nostril single, with an upward aspect, above and near the orbits. The jaws are armed with teeth of two kinds, set wide apart; the anterior teeth have sub-compressed, conical, slightly-recurred, sharp-pointed crowns, and are implanted by a single root; the posterior teeth are larger, with more compressed and longitudinally extended crowns (see fig. 65, art. ODONTOLOGY), conical, but with a more obtuse point, and with both front and hind borders strongly notched or serrated. The crown is contracted from side to side in the middle of its base so as to give its transverse section an oval or egg form (see fig. 66, ODONTOLOGY), and the opposite wide longitudinal grooves which produce this form become deeper as the crown approaches the socket, until they meet and divide the root into two fangs. The name Zeuglodon (yoke-tooth) refers to this structure. The mode of succession of the teeth in this genus conforms to the general mammalian type more than does that of any of the existing carnivorous Cetaceans. In the figure given by Dr Carus⁵ of a portion of the jaw of Zeuglodon coticoides, a deciduous molar (fig. 65, a, ODONTOLOGY) is about to be displaced and succeeded, vertically, by a second larger molar. This mode of succession is not known in the Pliocene or Inio, which among existing true Cetacea present teeth most like those of Zeuglodon; but it is a mode of succession and displacement affecting certain teeth in the herbivorous Cetacea or Sirenia; and we thus seem to have in the Zeuglodon another of those numerous instances of a more generalized character of organization in older tertiary Mammalia. In systematic characters, Zeuglodon typifies a distinct family or group, intermediate between Cetacea Mammalia proper and Sirenia.

Of the latter family or order, however, represented at the present day by the Dugong, Manatees, and Stellerians or Arctic Manatee (if the supposed sirenians), there were abundant and more widely distributed representatives during the miocene period, having, upon the whole, the nearest affinity with the existing African Manatee (Manatus Sireniae), but with associated characters of the Dugong (Halicore). There were, e.g., two incisive tusks in the upper jaw, and four or five small incisors along the deflected part of each ramus of the lower jaw. The upper molars, with three roots, were thickly enamelled, like those of the Manatees, but with a pattern of grinding surface which led Cuvier to attribute detached specimens to a small species of Hippospotamus. The lower molars had two roots. All the bones have the dense or solid structure of those of the Sirenia. On the remains of this remarkable amphibious Mammal, discovered by Kaup in 1838, in the miocene beds at Eppelsheim, belong to the genus Halicore. Other remains have been discovered in Pleistocene, Asté, and many parts of France, under the name "calcaire grösier" of the Gironde, containing Lophiodontidae fossils, up to the pliocene near Montpellier; at which period the Halicoreans seem to have become extinct.

Genus MACROCHERIUM, Lartet.—The edentate order, which is so abundantly and variously represented in South America, which has its Orycteropterus and Pangolins in Africa, and its Manatees in tropical Asia, has no living representative in Europe. Perhaps the most unexpected form of Mammal to be revealed by fossil remains from European tertiary deposits, after a Marsupilai, was a member of the edentate order. Cuvier, by whom the evidence of this extinct animal was first made known, prefaced his description of the single mutilated phalangeal bone, on which this evidence was founded, by the remark, that nothing proves better the importance of the laws of comparative osteology than all the consequences which one may legitimately draw from a single fragment.¹ One willingly admits the proof so afforded of the former existence of animals now unknown; but one may demur to the conclusion that their extinction was due to some sudden catastrophe.

The single mutilated ungual phalanx on which Cuvier based his conclusions in regard to the species in question was discovered, associated with remains of Rhinoceros, Manatees, Dinotherium, and Tapir, in a formation near Eppelsheim, Hesse-Darmstadt, which is now determined to belong to the miocene division of the tertiary series. This phalanx shows two distinctive characters of the edentate order:—1st, its posterior surface for articulation with the antepenultimate phalanx is a double pulley, bolted out on each side, with a salient cleft between, terminating the finest kind of ginglymoid joint peculiar to certain Edentata.² 2nd, The concave arch formed by this pulley carries furthest backward at its upper end, which would prevent the claw being retracted upward, as in the cat tribe, and constrain the flexion downward—"ainsi c'est nécessairement un ongual d'edentat."³ To the foregoing characters are joined two others which Cuvier believed to determine "as necessarily" the genus. The species of Myrmecophaga have on the upper part of the pointed end of the claw-phantas a groove, indicative of a disposition to bifurcate; in the species of Manis the bifurcation is complete, the cleft extending as far as the middle of the claw-bone. The Pangolins (Manis) have not those bony sheaths, which, in the sloths, some ant-eaters and armadillos rise from the base and cover the root of the claw; there was also absence of any claw-sheath in the fossil. The single fossil claw-bone has no trace of such a sheath. In the case of these of the Manis; and, according to all the laws of co-existence, it is impossible to doubt that the most marked relations of the animal that bore it should have been with that genus of quadrupeds.⁴ But what must have been its size? The phalanx was not one of the largest on the foot—for it had not those slight raised borders which one sees in the large claw-bones of the Pangolins. This question, which Cuvier answered by the proportions of the short-tailed Manis, at 24 French feet, has had a more reasonable reply given to it by certain other bones of the skeleton subsequently discovered in the miocene terraces of France. These discoveries have likewise rectified and moderated the absolute application of the correlative law, to the necessary determination of the genus, as well as of the order. The relations of the double-pulleyed and claw-phalanx to the Edentata is beautifully confirmed; but the additional fossils, and especially the evidences of teeth, have shown that it belonged to a peculiar and now extinct genus intermediate between the Manis and the Orycteropterus. And these relations are deeply interesting on account of the geographical position of both those edentate genera, on

¹ Proceedings, and Quarterly Jour. Geol. Soc., 1843. ² Medial and Physical Researches, p. 333. ³ Nova Acta Car. Leop. Carol., vol. xxiii., tab. xxxix. B., fig. 2, p. 340. ⁴ Ibid., p. 194. ⁵ Hist. de Brit. Fossil Mammals, 8vo, p. 536. ⁶ Att. Soc. Léon. de Bordeaux, 1840, p. 201. ⁷ Ossuements Fossiles, 4to, t. v., pl. I., p. 193. Mammalia, tracts of land, viz., which are now most contiguous to the continent containing the remains of the extinct oculant genus.

The locality in France is near the village of Saman, near Auch, department of Gers, Haute-Pyrénées. The formation is a lacustrine stratum of the miocene period.

Portions of two molar teeth have been found, 1 inch 3 lines in greatest transverse diameter; the tooth preserving the same size and shape through the whole length of the portion—viz., 1½ inch. The structure of the crown of the Orycteropus is less tubular and have not the same tubular tissue. Their microscopical texture appears not to have been analysed; it would be important to determine whether it resembled that of the teeth of the sloths or armadillos. The humerus differs from that of the ant-eaters and armadillos by its greater length in proportion to its breadth, and by the peculiar flattening from before backwards of its lower half, and especially at the condyles, above which it is expanded transversely by both external and internal supra-condyloid ridges. It is not perforated above the inner condyle, as the same bone is in both the Manis and Orycteropus. In the degree to which it departs from the type of the ant-eaters it approaches that of the Megatheroids and sloths—viz., in its relative length, flattening at the distal end, and the imperforate character of the condyles. The radius also presents a sloth-like character in its greater proportionate length, which exceeds that of the humerus; and in the compression of its lower slightly-expanded end. In both the Pangolin and Orycteropus the radius is shorter than the humerus. The ulna differs likewise from both that of the Pangolin by the much smaller development of the olecranon, whereby, again, it more resembles that of the sloths. The femur is relatively longer and more slender than that of the terrestrial and fossorial Edentata; it has not the third trochanter which characterizes it in the Orycteropus, nor so marked a development of the great and small trochanters as in the Pangolin. In the flattened form of the shaft of the femur, and the absence of any rounded projection at the distal end, it resembles the femur of the Megatherium and Mylodon. It is shorter than the humerus; whereas, in both the Pangolin and Orycteropus the femur is longer; in this respect the femur of the Macrotheres resembles that of the sloths. The great width of the popliteal space dividing the condyles is an endentate and more especially a megatheroid character. The internal condyle is much broader than the external one, as it likewise is in the Megatheroids; it is certainly with the femur of the latter family of the Edentata, rather than with that of the Proboscideans or Pachyderms, that one should compare the femur of the Macrotheres: it is not so strong or so slender relatively as in the sloths. The tibia is much shorter than the femur, owing to an expansion of the proximal end and its relative length to the femur. It resembles that of the Megatheroids more than that in the Pangolin or Orycteropus; it was not anchylosed to the tibia as in the Armadillos, Glyptodon, and Megatherium, but a distinct bone, as in the Mylodon and sloths.

Genus MEGATHERIUM, Gerv.—In the same miocene deposits of the south of France as those which contained the Macrotherium, fossil remains of two kinds of Quadrumania, resembling a small and large species of Hylobates, have been discovered. The smaller of these extinct apes, called Platycebus antiquus by Gervais, is based upon the lower jaw and dentition. The teeth occupy an extent of 1½ inch; the two incisors are narrower, the canine less, and the last molar is larger than in the siamang (H. syndactylus). As in this species the second premolar is more developed than the first, and the second is more produced than in the chimpanzee and gorilla, and to the degree in which the fore-and-aft diameter of the tooth exceeds the transverse one, it departs farther from the human type; in the degree of the development of the talon or third lobe of the last lower molar, the Platycebus resembles the tailed monkeys (Semnopithecus and Imaus).

Genus DRYOPITHECUS, Lart.—In the larger miocene ape (Dryopithecus Fontani, Lart.) the canine is relatively larger than in the Hylobates, and the incisors, to judge by their alveoli, are relatively narrower than in the chimpanzee and human subject. The first premolar has the outer cusp pointed, and raised to double the height of that of the second premolar, and the anterior lobe is more rudimental than in the chimpanzee, and departs progressively from the human type. The posterior lobe or talon of the second premolar is more developed, and the fore-and-aft extent of the tooth greater, than in the chimpanzee, thereby more resembling the second premolar of the siamang, and less resembling that of the human subject. The last (third) molar is undeveloped in the fossil jaw of the Dryopithecus, and its amount of departure from the human type, and approach to that of Imaus, cannot be determined. The canine is more vertical in position than in Troglodytes or Pithecus, but this Mammalia character is offered by some of the small South American apes, and cannot be cited as a mark of real affinity. From the portion of humerus associated with the jaw of Dryopithecus, the arm would seem to have been proportionally longer and more slender than in the chimpanzee and gorilla, more like that in the long-armed apes (Hylobates), and less like the arm of the human subject.

The characters of the nasal bones, orbits, mastoid processes, relative length of skull, relative length of forearm, relative length of arm to forearm, relative length and size of thumb, relative length of lower limb; and, above all, the size of the hallux and shape of the astragalus and calcaneum, must be known before any opinion can be trusted as to the proximity of Dryopithecus to the human subject.

Genus MESOPTHECUS, Wagn.—In tertiary formations of Greece, at the base of Pentelicon, remains of a Quadrumania have been found, which Professor Wagner regards as transitional between Hylobates and Semnopithecus; the third lobe of the last molar is however, as well developed as in the latter genus.

In the pliocene deposits of Montpellier, remains of a monkey occur, referred by Christol to a Cercopithecus; and in pliocene brick-earth in Essex, the writer has determined part of the fossil jaw and teeth of a Megatherium.

Genus DINOTHERIUM, Cuv. and Kp.—This name was given by Kaup to the huge bilophodont Mammal, first made known by Cuvier under the name of "Tapir giganteus," after the discovery of the singular shape and armature of the lower jaw. The length of the skull, from f to d, in fig. 99, is 3 feet 8 inches. The teeth

Skull of Dinotherium giganteum (Miocene, Eppelsheim).

in this skull, in addition to the two large deflected tusks of the lower jaw, are five in number on each side of both jaws. A study of the changes of dentition in fossils of young Dinotheres show that the first two teeth answer to the third and fourth premolars, as signified by the symbols p, 3 and 4; and that the rest are true molars (m, 1, 2, 3). Of these, the first tooth (p, 3), is rather trenchant than triturant; the third tooth (t) has three transverse ridges; the other grinders have two transverse ridges. This "bilophodont" or two-ridged type is shown by the molars of the Tapir, Lophodon, Macrotherium, Dryopithecus, Remington, and Momotus. In the general shape of the skull and aspect of the nostrils Dinotherium most resembles Momotus. Bone of limbs have not yet been found so associated with teeth as to determine the ordinal affinities of Dinotherium. Yet cranial and dental evidences of the genus have been discovered in miocene deposits of Germany, France, Switzerland, and Perina Island, Gulf of Cambay.

Genus MASTODON, Cuv.—The earliest appearance of this genus of proboscidian or elephantoid Mammal is in tertiary strata of miocene age, and by a species in which the fore part of the lower Mammalia, jaw was produced into a pair of deep sockets containing tusks; but these are only slightly deflected from the line of the grinding teeth (fig. 100, C). This species of Mastodon, discovered in the miocene of Eppelsheim, was called longirostris by Kaup; but he afterwards recognised it as the same with a species which had been previously called Mastodon arvernaeis (Cuvier and Jobert). Both belong to that section of Mastodon in which the first and second true molars have each four transverse ridges, and for which Dr Falconer proposes the name Tetralophodon. In the newer tertiary deposits of North America remains of a Mastodon (M. Ohioensis) have been discovered, in which the transverse ridges of the grinders are in shape more like those of the Dinotherium than in any other Mastodon; the first and second, moreover, are bilophodont, the third trilophodont; but this is followed by two three-ridged molars and a last larger molar with four or five ridges.

An almost entire skeleton of a Mastodon (M. tetriceps) has been discovered in the pleistocene deposits of Anté, Périgord, and has been described and figured by Professor Stanhope of France whose beautiful Messalina, though taken after the total length, from the tail to the end of the tusks, is 17 feet. The teeth have the same narrow shape and multi-mammillate structure as in M. arvernaeis, but in the numerical character of transverse divisions of the crown this species agrees with M. Ohioensis. The Mastodons were elephants with the grinding teeth less complex in structure, and adapted for

Fig. 100.

Mastodon tetriceps (Pliocene): A, B.—M. Ohioensis; C.—M. longirostris.

bruising coarser vegetable substances. The genus was represented by species ranging, in time, from the miocene to the upper pleistocene deposits, and in range, cosmopolitan with tropical and temperate latitudes. The transition from the mastodontal to the elephantine type of dentition is very gradual.

Genus Elephas, L.—The latest form of true elephant which obtained its sustenance in temperate latitudes is that which Blumenbach called primigenius, the "Mammoth" of the Siberian collectors of its tusks (fig. 101). It remains occur chiefly, if not exclusively, in pleistocene deposits, and have even been found in turbaries near Holyhead. Its grinders are broader, and have narrower and more numerous close-set transverse plates and ridges, than in other elephants. The mammoth is more completely known than most other extinct animals by reason of the discovery of an entire specimen preserved in the frozen soil of a cliff at the mouth of the river Lena in Siberia. The skin was clothed with reddish wool, and with long black hair. It had never proceeded at St Petersburg, together with the skeleton (fig. 101). This measures across the base part of the skull to the end of the mutilated tail, 16 feet 6 inches; the height, to the top of the dorsal spine, is 9 feet 4 inches; the length of the tusks, along the curve, is 9 feet 6 inches. Parts of the skin of the head, the eye-ball, and of the strong ligament of the nape which helped to sustain the heavy head and teeth, together with the hoofs, remain attached to the skeleton. These huge elephants, adapted by their clothing to endure a cold climate, subsisted on the branches and foliage of the northern pines, birches, willows, &c.; and during the short summer they probably migrated northward, like their contemporary the monk-buffalo, which still lingers on, to the 70th degree of N. latitude, retreating during the winter to more temperate quarters. The mammoth was preceded in Europe by other species of elephant,—e.g., Elephas meridionalis (Nodi), which, during the plioce period, seem not to have gone northward beyond temperate latitudes. An elephant, hardly distinguishable from the African, also roamed at that period in Europe.

Genus Hippopotamus, L.—The discovery, in lacustrine and fluviatile deposits of Europe, of the remains of an amphibious genus of Mammal now restricted to African rivers, gives scope for speculating on the nature of the land which, uniting England with the Continent, was excavated by lakes and intersected by rivers, with a somewhat warmer temperature than at present, to judge by a few S. American shells which occur in the fossil forest strata, e.g., at Gray's Essex, where remains of the last existing Hippopotamus have been found. The smallest of known jaws (fig. 102) was discovered in similar deposits in Norfolk coast. Other localities are specified in the writer's History of British Fossil Mammals. The hippopotamus is first met with in plioce strata. The remains of H. major have hitherto been found only in Europe; they are common along the Mediterranean shore, and do not occur north of the temperate zone. In Asia this form of Pachyderm was represented, perhaps at an earlier period, by the genus Hemiprotodon,—essentially a hippopotamus, with six incisor teeth, instead of four, in each jaw.

Genus Rhinoceros, L.—The rhinoceros, like the elephant, was represented in plioce and pleistocene times, in temperate and northern latitudes of Asia and Europe, by extinct species. One

---

1 Beiträge zur Näheren Kenntniss der Urweltlichen Säugethiere, 4to, 1837, p. 19. The name augustidentis was first applied by Cuvier to teeth of this type or species. 2 First demonstrated by Kaup, Ossaeum Fossile de Darmstadt, 4to, 1835. 3 Odontography, p. 617, pl. 144. 4 Ibid., p. 618. 5 See art. ODONTOLOGY, p. 475, fig. 142. 6 Falconer "On the Species of Elephant and Mastodon occurring in a fossil state in England," Proc. Geol. Soc., June 1857. Mammalia. (Rhinoceros leptocephalus) associated with the Hippopotamus major in fresh-water pliocene deposits; another (R. tichorhinus) with the mammoth in pleistocene beds and drift. The discovery of the case of the tichorhinus rhinoceros in frozen soil, recorded by Pallas Mammalia in his Voyages dans l'Asie Septentrionale, showed the same adaptation of this, at present tropical, form of quadruped to a cold cli-

mate, by a twofold covering of wool and hair, as was subsequently demonstrated to be the case with the mammoth. Both the above-named fossil rhinoceroses were two-horned; but they were preceded, in the pliocene and miocene periods, by species devoid of horns, yet a rhinoceros in all other essentials (Acrotherium, Kaup). Not fewer than twenty species of extinct rhinoceroses are entered in Palaeontological catalogues.

**ORDER RUMINANTIA.**

Of other forms of beasts subsisting on the vegetable productions of the earth, and more akin to actual European Herbivora, there co-existed with the foregoing now extinct genera a vast assemblage of species, nearly all of which have passed away. The quadrupeds called "Ruminants," from the characteristic second mastication of the partly-digested food by the act called "ruminant" or "chewing the cud," constituted at this period a circular division of the group of Mammalia, which Cuvier believes to be the most natural and best-defined order of the class. He characterized it by having incisive teeth only in the lower jaw (fig. 107, c), which were replaced in the upper jaw by a callosus gum. Between the incisors and molars is a diastema, in which, in certain genera only, may be found one or two canines. The molars (fig. 107, b), almost always six on each side of both jaws, have their crown marked by two double crescents, with the convexity turned inwards in the upper set, outwards in the lower. The four legs are terminated by two toes and two hoofs, flattened at the contiguous sides, so as to look like a single hoof cloven; whence the name "cloven-footed," also given to these animals. The perfect circumscription and definition of this order, so desirable by the systematic zoologist, is indeed invaded, in the actual Ruminantia, by certain peculiarities of the camel tribe.

In entering upon the evidences of the first appearance in this place of the order of animals, which now are the most valuable to man, it may be well to call to mind the characters of the Acrotherium. The upper true molars have two double crescents, convex inwards, one of the inner ones being encroached on by a large tubercle, the reduced homologue of which may be seen in the internal interspace of the crescents in the ox and some other Ruminants. The lower true molars also, at one stage of attrition, form crescentic islands of enamel, with the convexity turned outwards, as in Ruminants, the last molar having the accessory crescent behind. The functional hoofs were two in number on each foot, but must have resembled those of the camel tribe in shape; the scaphoid and cuboid of the tarsus were distinct also, as in the Camelidae; and the metacarpal and metatarsal bones were divided into anterior and posterior divisions, as in the modern ruminants. The molar dentition of the extinct Dichodon made a still nearer approach to that of the Ruminants. The chief distinction of this and other extinct Herbivores with double crescentic molars is the completion of the upper series of teeth by well-developed incisors. But the premaxillaries in the new-born camel contain each three incisors, one of which becomes fully developed. The Camelidae are hornless, like the Acrotherioids and Dichodonts; and with one exception—the giraffe—all Ruminants are born without horns.

Thus the Acrotherioides, in several important characters, resembled the embryo Ruminant, retaining throughout life those marks of adhesion to a more generalized mammalian type. The more modified or specialized form of booted animal, with cloven feet and ruminating stomach, appears at a later period.

**FAMILY I.—BOVIDE.**

Fossil molars of the ruminant type and bovine character have hitherto been found, with unequivocal evidence, to the writer's knowledge, only in beds or breccias of pliocene and pleistocene age. At those periods in Britain there existed a very large species of bison (Bison priscus), and a larger species of ox (Bos antiquus), from pleistocene fresh-water beds; while a somewhat smaller but still stupendous white ox (B. priscus) has left its remains in pleistocene deposits of England and Scotland. With this was associated an aboriginal British ox of much smaller stature and with short horns (B. longicornis), which continued to exist until the historical period, and was probably the source of the domesticated cattle of the Celtic races before the Roman invasion. A buffalo, not distinguishable from the musk kind (Bubalus moschatus), now confined to the northern latitudes of North America, roamed over similar latitudes of Europe and Asia in company with the hair-clad elephants and rhinoceroses.

**Family II.—Cervidae.**

Cuvier first made known the fact of teeth with the character of ruminant molars, and of portions of antlers, being associated with more or less mutilated antlers, which had been shed, to a species he called C. diceroceros. The beam rises from 1 to 2 inches without sending out any branch or brow-snag; it then sends off a branch larger and oblique that the beam seems here to bifurcate; the anterior prong is, however, the smallest and shortest. The writer has received similar shed and mutilated antlers from the red clay of Sussex, which seems to contain a melange of broken-up beds of eocene, miocene, and pliocene age.

The cervine Ruminants have been divided into groups according to the forms of the antlers. Of the group with antlers expanded and flattened at top, of which the fallow-deer (Dama) is the type, no fossil examples have been found in Britain. Cuvier has described and figured antlers of great size from the pliocene deposits of the valley of the Somme, near Abbeville, which, from the relative position and direction of the brow-snag and mid-ring, and from the terminology used, he regards as a large extinct species of fallow-deer, the name Cerus somnus has since been attached to this species. But there once existed a group (Megaloceros, fig. 103) characterized by a form of antler at present unknown amongst existing species of deer. With a beam (b) expanding and flattening towards the summit, and a brow-snag (p), as in the Dama tribe, this antler shows a back-snag (s). Moreover, in antlers which, from their size and form, seem to have been developed by the deer at its prime, the brow-snag expands and sometimes bifurcates—a variety never seen in the fallow-deer, but which becomes exaggerated in the reindeer group. The representative of the present Megaloceros is one extinct species (M. Intermedius, fig. 103), remarkable for its great size, and especially for the great relative magnitude and noble form of its antlers. It is the species commonly but erroneously called the "Irish elk," because it is a true deer, intermediate between the fallow- and rein-deer; and because, though most abundant in Ireland, it is not peculiar to Ireland. In that country it occurs in the shell-marl underlying the extensive turberies. In England its remains have been found in lacustrine beds, brick-earth, red clay, and siliceous caves.

The reindeer (Cerus tarandus) has peculiar antlers (fig. 104), and proportionally the largest of any of existing species. The beam is somewhat flattened throughout, but expands only and suddenly at its extremity, a similar expansion characterizing the brow-snag (br) and mid-ring (m). From this point the points are developed from all these expansions in fully-developed antlers. The brow-snag is remarkable for its length. There is also frequently a short back-snag. It is plain, therefore, from the presence of this snag, from the great relative size of the antlers, from the complex brow-snag, and the terminal expansion of the beam, that we have in the reindeer the nearest of kin to the extinct Megaloceros.

The existing species (Tarandus) is restricted to northern latitudes, ranging to extreme ones in Europe, and in America from the Arctic Circle southward to the latitude of Newfoundland, where the large variety called "Caribou" still exists. Reindeer of similar size ranged over continental Europe appear to have been seen by Caesar in Germany, and have left good evidence of their existence in the parts of England. The specimen figured (fig. 104) was found in pliocene "tilt" at Bilney Moor, East Dereham.

A large deer, with compressed ramified antlers, slightly expanding at the base of the terminal divisions, but differing from the reindeer in the absence of the brow-snag, has left its remains in the pliocene sands of Rillé, near Pézenas, France. It is the Cerus martialis of Gervais; and seems to have been an intermediate form between the reindeer (Tarandus) and the elk (Alces).

There is no existing representative of this interesting extinct form of deer.

In formations of corresponding age in France, called "alluvions volcaniques" by Gervais, fossil antlers of two other extinct species of deer have been discovered, in which, as in Alces, the brow-anterl is absent, but in which the beam does not expand into a palm.

In North America remains of a large deer (Cerus americanus fossile, Harlan), much resembling the Wapiti (Cerus canadensis) have been found in pliocene deposits on the banks of the Ohio. In South America Dr Lund discovered fossil antlers of two species in bose-caves in Brazil: they were associated with remains of an antelope (Antilope maguarum, Lund) of which genus no living representative is now known in South America.

---

1 *Osteums Fossiles*, 4to, tom. iv., p. 104, pl. viii., figs. 5 and 6. 2 *Owen, Hist. of Brit. Fossil Mammals*, p. 444. 3 Quarterly Jour. of the Geol. Soc., vol. xlii., 1856, p. 217, figs. 14–16. 4 Zoologie et Palaeontologie Francaise, 4to, p. 82. Mammalia. Of deer with antlers of the type of the existing red-deer (C. elaphus), a species is indicated in pleistocene beds and bone caves.

Genus Felis, L.—As it is by this form of perfect Carnivore that Cuvier chiefly illustrated his principle of the correlation of animal structures, it will be exemplified more particularly in this place, and by the aid of the subjoined cut (fig. 106). The founder of paleontology thus enunciates the law which he believed to be so operative in his labours of re-constructing extinct species:

"Every organized being forms a whole, a single circumcumbered system, the parts of which mutually correspond and concur to the same definitive action by a reciprocal reaction. None of these parts can change without the others also changing, and consequently each part, taken separately, indicates and gives all the others."

Cuvier did not predicate that law by an a priori method, by any of those supposed short cuts to knowledge, the fallacy of which Racine well expressed; he arrived at the law indirectly, and after many dissections had been made to the same facts of the law of the Carnivore being strong by virtue of certain proportions; of its having a peculiarly shaped and articulated condyle, with a plate of bone of breadth and height adequate for the implantation of muscles, with power to inflict a deadly bite—a process grasped by muscles of such magnitude as necessitated a certain extent of surface for their origin from the cranium, with concomitant strength and curvature of the zygomatic arch; the facts of the modified occiput and dorsal spines in relation to vigorous uplifting and retraction of the head when the prey had been gripped; the size and shape of the piercing, lacerating, and trenchant teeth; the mechanism of the retractile claws, and of the joints of the limb that wielded them—these facts were not until after Cuvier had recognised these facts, and studied them, and their correlations in a certain number of typical Carnivora, that he felt justified in asserting that "the form of the tooth gives that of the condyle, of the blade-bone (a), and of the claws, just as the equation of a curve evolves all its properties, and exactly as, 'in taking each property by itself as the base of a particular equation, one discovers both the ordinary equation and all its properties, so the claw, the blade-bone, the condyle, the femur, and all the other bones individually, give the teeth, or are given thereby reciprocally; and in consequence by any of them, whoever possesses rational laws of the organic economy will be able to re-construct the entire animal.'"

The principle is so evident that the non-anatomical reader will have little difficulty in satisfactorily comprehending it by the aid of the subjoined diagram.

Fig. 105. Antler of Red-deer, from alluvium, Ireland.

ORDER CARNIVORA.

The quadrupeds which subsist by preying upon others co-existed under corresponding varieties of form, and in adequate numbers, with the numerous and various Herbivora of the newer tertiary periods. A brief description has already been made of some of the singular forms, the genera of which are extinct, that lived in eocene and miocene times.

Genus Galecyrus, Ow.—In 1829 the fossil skeleton of a Car-

nivore, of the size of a fox, was discovered by Sir Roderick I. Mur-

chison in the pliocene schist of Gliningen. On a close comparison of this specimen, the writer finds that the first premolar is smaller, and the third and fourth larger than in the fox, and all the teeth are more close-set and occupy a smaller space than in the genus Canis; the bones of the feet are more robust; and these, with other characters, indicate an extinct genus intermediate between Canis and Viverrina. The unique specimen is now in the British Museum.

Fig. 104. Skull and Antlers of Cervus Turanensis.

which rivalled the Megaceros in bulk (Strongyloceros speleus); and with this are found, in similar places of deposit, remains of a red-deer with antlers equaling or surpassing the finest that have been observed within the historical period.

Fig. 105 represents one of a pair of antlers from the bed of the Boyne at Drogheda, near Dublin, in the museum of Sir Philip Egerton, Bart., which measures 30 inches in length, and sends off not fewer than fifteen branches or "snags," a is the "brow-snag," which rises immediately above the "burr;" b the second, c the third, and d the "crown" or terminal branch, the only one which gave to the deer developing them at the period of his full perfection the title of "crowned hart."

The roebuck, like the red-deer, appears from its fossil remains to have continued to exist from the prehistorical pleistocene times to the present period.

Fig. 106. Palaeontological characters of a Feline Carnivore.

---

1 See Quarterly Journal of the Geological Society, vol. iii., 1847, p. 55. 2 Geology of Fossils, 4to, tom. i. (1812), p. 58. Mammalia; the movement of the lower upon the upper jaw. The lower jaw (a) is short and strong; it articulates to the skull by a transversely extended convexity or condyle (d), received into a corresponding concavity (e), forming a close-fitting joint which gives a firm attachment to the jaw, and almost restricts its movements to one plane as in opening and closing the mouth. The plate of bone, called coronoid process (r), which gives the surface of attachment to the chief biting muscle (coronohyoid or temporal) is broad and high; the surface on the side of the skull (temporal fossa, r) from which that muscle arises is correspondingly large and deep, and is augmented by the extension of ridges of bone from its upper and hinder periphery.

The bar or bridge of bone (zygomatic arch) which spans across the muscle, bends strongly outwards to augment the space for its passage; and as it gives origin to another powerful biting muscle (masseter), the arch is also bent outwards to form the strongest point of resistance during the gripes of that muscle. From almost every periphery of the back surface of the skull there is a strong pitted ridge, affording extensive attachment to powerful muscles which raise the head, together with the animal's body, which the lion may have seized with his jaws; this beast of prey being able to drag along the carcass of a buffalo, and with ease to raise and bear off the body of a man. If we next examine the framework of the fore limb, which is associated with the above-defined structure of the skull, we find that the fore paw consists of five digits (1-5); the innermost and shortest (1) answering to our thumb, and having two bones; the other four digits having each three bones or "phalanges." All those digits enjoy a certain freedom of motion and power of reciprocal approximation for grasping; but their chief feature is the reduction of the terminal phalanx, which is enlarged, compressed, subtriangular, or nearly so; with a plate of bone as it were, reduced forwards from that base, from which the pointed termination of the phalanx projects like a peg from a sheath. A powerful, compressed, incurved, sharp-pointed, hard, horny claw is fixed upon that peg, its base being firmly wedged into the interspace between the peg and the sheath. The toe-joint so armed is retractile. This complex, prehensile, and destructive paw is articulated to the two bones of the fore leg (radius, s, and ulna, u); they are both strong, are both distinct, are firmly articulated to the arm-bone (b) by a joint, which, although well built, allows great extent and freedom of motion in bending and extension; and, besides this, the two bones are reciprocally joined as to rotate on each other, or rather the radius upon the ulna, can be rotated with it, by the greater expansion of its lower end, the whole paw, which can thus be turned "prose" or "cupping," whereby its application as an instrument for seizing and tearing is greatly advantaged.

The humerus or arm-bone (A) is remarkable for the extension of strong ridges from the outer and inner sides, just above the elbow-joint. These ridges indicate the size and force of the supinator, pronator, flexor, and extensor muscles of the paw. To defend the main artery of the fore leg from compression during the action of these muscles, a bridge of bone (a) spans across it as it passes near their origin. The upper end of the arm-bone is equally well marked by powerful ridges for muscular implantation, especially for the deltoid; but these ridges do not project beyond the round "head" of the bone, so as to impede its movement in the socket.

The blade-bones (scapulae, s) is of great breadth, with well-developed processes (spine, acromion, coracoid) for muscular attachments; the size and shape of this bone relate closely to the volume of the muscles which act upon the arm-bone and fore limb. A small clavicular bone (b) is interposed between a muscle of the head and arm of the arm, giving additional force and determination of action reciprocally to both muscles.

Such are some of the modifications of the teeth and framework of a beast of prey, which concur, and were deemed by Cuvier to be correlated, in the organization of such animals.

Let us compare them with those of the corresponding parts in an ox (fig. 107). The teeth answering to the great laniaries in the lion are absent; at most, one recognises the homologues of the lower canines, reduced in size and altered in shape, so as to form the outer teeth (c) of a beat row of incisors terminating in the lower jaw. The back teeth (h), instead of being trenchant, have broad and flat crowns, roughened with hard edges, opposing each other with a grinding action like stones. The lower jaw is long and slender; it articulates to the skull by a flat condyle (d), admitting of rotational movements upon a flattened articular surface on the skull, and limiting the extent of opening and shutting the mouth. The coronoid process (r) is very slender, and the fossa which marks the size of the temporal muscle (i) is correspondingly small. The zygomatic arch (e) is short and feeble, and its span is narrow; it is almost straight, or with a slight bend downwards. The part of the skull (pterygoid plates), which afford attachment to the rotating muscles of the jaw, and the (angular) part (f) of the jaw into which they are inserted, are of great extent. The ox masticales Mammalia; grown with great efficiency; it inflicts no injury to other animals with its teeth. The horns are its weapons, and they are chiefly defensive.

The fore foot of the ox is reduced to two principal toes, with two rudimentary ones dangling behind. Each of these has its extremity enveloped by a thick horny case, or hoof; this modification is accomplished by a junction or coalescence of the radius and ulna (u), preventing reciprocal rotation or movement of those bones on each other,—by a joint restricting the movement of the fore arm (antebrachium) upon the arm (brachium or humerus, h) to one plane,—by a long and narrow blade-bone (s), with a stunted coracoid and no clavicle; in short, by modifications adapting the limbs to perform the movements required for locomotion, and almost restricting it to such. This type of fore limb is always associated with broad grinding teeth, and with the modifications of jaw and skull above defined. The due amount of observation assured Cuvier that these several modifications, like the contrasted ones in the Carnivora, were correlated, and he examined the physiological grounds of that correlation.

Those grounds may be traced to a certain degree in the secondary modifications of the feline order. If the retractility of the claw be suppressed, the carnassialness of the teeth is reciprocally modified. If the ungulculate foot is reduced from the digitigrade to the plantigrade type, the dentition is still more altered, and made more subservient to a mixed diet.

By the application of the correlative principle to the fossil mammalian remains of pliocene and later deposits, the Herbivora have been distinguished from the Carnivora, and most of the latter have been re-constructed; outlines visible of the felinae viverrinae, urinae, and other families of the order. In England and continental Europe a peculiarly destructive feline quadruped existed, with the upper canines much elongated, trenchant, sharp-pointed, sabre-shaped; whence the name Machairodus proposed for this feline sub-genus. It was represented by species as large as a lion (M. cultridens and M. latidens); and by others of the size of a leopard (M. poliedrus and M. megacephalus). This form is first found in the miocene of Auvergne and of Eppelsheim; next in the pliocene of the Val d'Arno; and finally in cave breccia in Devonshire.

The finest examples of the large pleistocene lion (Felis spelaea) have been discovered in bone-caves,—e.g., in those of Banwell, Somersetshire, and of Halipium. The production of the size of the marsupials of the former period is seen in the size of the lion; proves this species to have been a lion, not a tiger. It roamed over pliocene and pleistocene Europe, and has left its remains in many stratified deposits of the former period in Britain. Under similar circumstances have been found, more abundantly in Germany, the remains of the gigantic bear (Ursus spelaea), and more abundantly in England those of the great hyena (Hyaena spelaea), probably a spotted one, like the fierce "Crocuta" of the Cape. Wolves, foxes, badgers, otters, wolverines, and martin-cats, four-marts and weasels, have left their remains in the newer tertiary deposits and bone-caves. Bats, moles and shrews were then, as now, the forms that preyed upon the insect world in Europe. The majority of these Carnivora, like the hares, rabbits, voles, and other Rodents, are not distinguishable from the species which still exist. These smaller ungulculate Mammals, like the smaller pleistocene Mammalia, Ruminants, seem to have survived these changes during which the larger species perished. It is probable that the horse and ass are descendants of a species of pleistocene antiquity. At the pliocene period there existed a species similar in size to a zebra. There is no certain character by which the present wild boar can be distinguished specifically from the Sus, which was contemporary with the mammoth.

ORDER RODENTIA.

The small size of the great majority of the species of this order leads to the neglect or the oversight of their fossil remains by the labourers in quarries and other deposits of bone, to whom the palaeontologist is usually indebted for his first acquaintance with characteristic fossils of such formations. No evidence has yet been obtained of any unequivocal remains of a rodent animal in strata more ancient than the eocene tertiary deposits. Cuvier detected remains of Rodents allied to the dormouse (Myosorex) and squirrel (Sciurus) in the building-stone of the Montmartre quarries near Paris. The lacustrine marls of the middle (miocene) tertiary period have yielded evidences of not fewer than eleven genera of Rodentia distinct from any now known to exist. The deposits at Ephesus, in Asia Minor, of the same miocene age have given evidence of Rodents akin to the dormouse and the beaver. The more recent tertiary formations and the bowersaves in England have furnished fossil remains not distinguishable from the existing beaver, hare, and rabbit, water-vole and field-vole, as well as remains of a Pica, or tailless hare, belonging to the genus Lagomys, now confined, as an existing species, to Asia; and of a very large Rodent, akin to the beaver, called Trogontherium. Similar fossil remains have been abundantly found in the pliocene and pliocecene formations of continental Europe; whilst the coeval deposits of America have yielded fossil remains of extinct species belonging to genera—e.g., Lagostomus, Echimys, Cratogeomys, Ctenomys, and other Cavius—now restricted to South America. In North America, several genera of Rodent of comparatively gigantic size have recently been discovered in the parts of the skeleton, and more especially the dentition of these rodent orders are highly characteristic,—the form of the articulate surface of the lower jaw, which is a longitudinal groove, the molars, especially of the phyllodonte kinds, crossed by enamel plates more or less transverse, or, these, with the long, curved, chisel-shaped incisors, two in each jaw, suffice to determine the ordinal relations of the fossil. The incisors alone would not be always so safe a guide, for the rodent modification of these teeth is repeated in the marsupial wombat and the lemurs' eye-eye.

With regard to the Rodentia, the great beaver (Trogontherium) seems to have become extinct in England and the Europo-Asiatic continent before the historical period, whilst the smaller pliocene beaver continued to exist with us like the wolf, until hunted down by man. It still survives in a few of the great continental rivers. Of the little Lagomys of our native saves, no living example remains in either England or Europe. The species, indeed, may be extinct: its genus is now limited to Central and Southern Asia.

GEOGRAPHICAL DISTRIBUTION OF PLEISTOCENE MAMMALS.

A most interesting generalization has been deduced from the mass of facts relating to the fossil Mammals of the later tertiarys,—viz., their close correspondence between the fauna of those and present periods in the Europo-Asiatic expanse of dry land. For here species continue to exist of nearly all those genera which are represented by pliocene and post-pliocene mammalian fossils of the same natural continent and of the immediately adjacent island of Great Britain.

The bear still abounds in both Europe and Asia; the beaver of the Rhone and Danube resembles the great Trogontherium; the Lagomys and the tiger exist on both sides of the Himalayan mountain chain; the hyena ranges through Syria and Hindustan; the bactrian camel typifies the huge Merychippus of the Siberian drift; the elephant and rhinoceros are still represented in Asia, though now confined to the south of the Himalayas. The true macaques are peculiar to Asia, and though most abundant in the southern parts of the continent and the Indian Archipelago, also exist in Japan; a closely-allied sub-genus (Simus) is naturalized on the rock of Gibraltar at the present day. A fossil species of Macaca was associated with the elephant and rhinoceros in England during the period of the deposition of the newer pliocene freshwater beds. The two extraordinary extinct forms of Mammalia, called Elephasmirus and Sinatherium, have their nearest existing pachydermal and ruminant analogues in the elephants and to which those fossils are peculiar. Cuvier places the Elasmotherium between the horse and rhinoceros; the existing four-horned antelopes, like their gigantic extinct analogues, the Elasmotherium and Bramatherium, are peculiar to India. It may be regarded as part of Mammalia, the same general concordance of geographical distribution, that the genus Hippopotamus, extinct in England, in Europe, and in Asia, should continue to be represented in Africa, and in none of the remoter continents of the earth.—Africa also having its hyenas, its elephant, its rhinoceroses, and its great feline Carnivores. The discovery of extinct species of Cassidoparadise in both Europe and Asia, of which genus the sole existing representative is now, like the hippopotamus, confined to Africa, adds to the propriety of regarding the three continuous continental divisions of the Old World as forming, in respect to the geographical distribution of pliocene, post-pliocene, and recent mammalian genera, one great natural division. The only large extinct animal (Pinguinus giganteus, Cuvier; Macrauchenia patachonica) hitherto found in the tertiary deposits of Europe, manifests its nearest affinities to the genus Manis, which is exclusively Asiatic and African.

Extending the comparison between the existing and the latest of the extinct series of Mammalia to the continent of South America, it may be first remarked that, with the exception of some carnivorous and canine species, no representatives of the above-cited mammalian genera of the Old World of the geographer have yet been found in South America. Buffon long since enunciated a similar generalization with regard to the existing species and genera of Mammalia; it is almost equally true in respect of the fossil. Not a relic of an elephant, a rhinoceros, a hippopotamus, a tapir, or a sloth has yet been detected in the caves or the more recent tertiary deposits of South America. On the contrary, most of the fossil Mammalia from those formations are as distinct from the Europo-Asiatic forms as they are closely allied to the peculiarly South American existing genera of Mammalia.

The genera Equus, Tapirus, and the still more ubiquitous Mastodon, form the chief, if not sole exceptions. The representation of Equus during the pliocene period by distinct species in Asia (E. primigenius) and in South America (E. curvifrons), is analogous to the geographical distribution of the species of Tapirus at the present day. South America alone is now inhabited by species of sloth, of armadillo, of cavy, aguti, ctenomyss, and platyrhine monkey; but no fossil remains of a quadruped referable to any of these genera have yet been discovered in Europe, Asia, or Africa. The types of Bovidae and Dasyproctidae are however, richly represented by diversified and gigantic specific forms in South America during the geological periods immediately preceding the present. The skeleton of one of these forms of the sloth tribe is represented in fig. 108;

---

Fig. 108.

Extinct Terrestrial Sloth, Mylodon robustus (Pleistocene, S. America).

It measures, from the fore part of the skull to the end of the tail, 11 feet. It was discovered buried 12 feet deep in the fluviatile deposits seven leagues north of the city of Buenos Ayres in the year 1841. It forms the subject of a work entitled Description of the Skeleton of an Extinct Gigantic Sloth (Mylodon robustus), in which are set forth in detail the grounds for regarding it as a member of the same natural family as the present small arboreal sloth, and as being modified to obtain its leafy food by uprooting and prostrating trees. Numerous species of terrestrial sloth co-existed with the Mylodon in South America. Its skeleton, now complete in the British Museum, measures 18 feet; its dentition agrees as to number and kind of teeth with that of the sloths (Bradypus). But the Mammalia, molars (fig. 109) are longer, more deeply implanted, of more complex structure, and with grinding surfaces of the biloiodont type. The elephants, which subsist on similar food to that of the Megatherium, had their grinding machinery maintained by a numerous succession of teeth. The same end was attained in the Megatherium, by a constant growth and renovation of the same teeth. The formative pulps were lodged in the deep basal cavities, exposed in the section figured (fig. 109). The molar teeth were five in number on each side of the upper jaw, and four in number on each side of the lower jaw (fig. 110). In this bone the fore part is much prolonged, and grooved above, to support a long, cylindrical, posteriorly directed tooth, by which, like the giraffe, the Megatherium stripped off the small branches of the trees its colossal strength enabled it to prostrate. The dentition of Mylodon differed only from that of Megatherium in the shape of the teeth. The same may be said of the allied genera called Mylodonys and Scelidotherium. They were contemporary and geographically associated genera of the same, now quite extinct, family of great terrestrial sloths.

In like manner, the small horioted and banded quadrupeds of South America called armadillos were represented in pleistocene times in South America by a well-defended species, rivalling the Megatherioids in bulk. The specimen of the almost entire skeleton and bony armour (fig. 111) is one of the smaller species of these great extinct non-banded armadillos; yet it measures from the snout to the end of the tail, following the curve of the back, 9 feet; Lower Jaw and Teeth of Megatherium (Pleistocene, South America).

These large extinct species differ from the modern armadillos, in having bands of plates in their coat of mail, for the purpose of contracting to a ball. They also differ in the fluted form of the teeth (fig. 112); whence the generic name (Glyptodon) assigned to them. The species are distinguished, like their puny representatives (Dasypus), by peculiar patterns of the outer surface of the constituent ossicles of the tesselated mail. In the species figured (G. clariops), a large raised central circular plate is surrounded by smaller portions. The species named G. reticulatus, G. rubroceps, G. oemus, &c., have their names from other modifications of the sculptured surface of their armour. Above the principal figure in cut 111 are shown the front and back margins Mammalia of the body-armour; below it, opposite the left hand, are upper and under views of the cranium, which was defended by a tesselated bony casque. The tail also had its independent osseous sheath, supported by the vertebrae within, as shown in the figure opposite the left hand.

Other evidences of extinct South American Mammals, matched only by species now peculiar to that continent, might be adduced if space permitted.

Australia in a like manner yields evidence of an analogous correspondence between its last extinct and its present aboriginal mammalian fauna, which is the more interesting on account of the very peculiar organization of most of the native quadrupeds of that division of the globe. That the Marsupialia form one great natural group, is now generally admitted by zoologists; the representatives in that group of many of the orders of the more extensive placental sub-class of the Mammalia of the larger continents have also been recognized in the existing genera and species: the dasyuruses, for example, play a part of the greatest importance in Australia (Phascoga) of the Insectivora, the phalangers of the Quadrumenidae, the wombat of the Rodentia, and the kangaroos, in a remoter degree, that of the Ruminantia. The first collection of mammalian fossils from the ciferaceous caves of Australia brought to light the former existence on that continent of larger species of the same peculiar marsupial genera—some, as the Thylacine, and the dasyurine sub-genus represented by the D. urinus, are now extinct on the Australian continent, but one species of marsupial still exists on the adjacent island of Tasmania; a few recent extinct marsupials, phalangers, potoroes, and kangaroos—some of the latter (Macropus Atlas, M. Tenuis) being of great stature. A single tooth, in the same collection of fossils, gave the first indication of the former existence of a type of the marsupial group, which represented the Pachyderms of the larger continents, and which seems now to have disappeared from the face of the Australian earth. Of the great quadruped so indicated under the name Diprotodon in 1838, successive subsequent acquisitions have established the true marsupial character and the near affinities of the genus to the kangaroo (Macropus), but with an occasional relationship with the herbivorous wombat (Vombatus) of the same island. Australia (fig. 113) has lately been acquired by the British Museum, showing in situ the tooth (i) on which the genus was founded. This skull measures 3 feet in length; that of a man

---

Fig. 109. Section of Upper Molar Teeth, Megatherium (one-third nat. size), Pleistocene, South America.

Fig. 110. Lower Jaw and Teeth of Megatherium (Pleistocene, South America).

Fig. 111. Extinct gigantic Armadillo.

Fig. 112. Teeth of great extinct Armadillo (Glyptodon clariops), Pleistocene, South America.

---

1 Mitchell's (Sir Thos.) Three Expeditions into the Interior of Australia, Svo, 1838, vol. ii., p. 359. Mammalia, is inserted in the cut to exemplify the huge dimensions of the primitive kangaroo. Like the contemporary gigantic sloth in South America, the Diprotodon of Australia, while retaining the dental formula of its living homologue, shows great and remarkable modifications of its limbs. The hind pair were much shortened and strengthened, compared with those of the kangaroo; the fore pair were lengthened as well as strengthened; yet, as in the case of

Fig. 113.

Skull, gigantic Pachydermoid Kangaroo (Diprotodon Australis) Pleistocene, Australia.

the Megatherium, the ulna and radius were maintained free, and so articulated as to give the fore paw the rotatory actions. These, in Diprotodon, would be needed, as in the herbivorous kangaroo, by the economy of the marsupial pouch. The dental formula of Diprotodon was \( \frac{4}{3} \cdot \frac{2}{3} \cdot \frac{4}{3} \cdot \frac{3}{3} = 28 \), and, as in Macropus major, the first of the grinding series (p) was soon shed; but the other four two-ridged teeth were longer retained, and the front upper molar (i) was very large and scaliform, as in the wombat. The zygomatic arch sent down a process for augmenting the origin of the masseter muscles, as in the kangaroo. The foregoing skull, with parts of the skeleton of the animal to which it belonged, were discovered in a lacustrine deposit, probably pleistocene, intersected by creeks, in the plains of Darling Downs, Australia.

The same formation has yielded evidence of a somewhat smaller extinct herbivorous genus (Nototherium), combining, with essential affinities to Macropus, some of the characters of the Koala (Phascolarctos). The writer has recently communicated descriptions and figures of the entire skull of the Nototherium Mitchelli to the Geological Society of London. The genus Phascolomys was at the same period represented by a wombat (Pogona) of the dimensions of a tapir, one of the grinding teeth of which is represented, of the natural size, in fig. 114.

The pleistocene marsupial Carnivora presented the usual relations of size and power to the Herbivora, whose undue increase they had to check. Fig. 115 represents an almost entire skull, with part of the lower jaw of an animal rivalling the lion in size, the marsupiality of which is demonstrated by the position of the lacrymal foramen (i) in front of the orbit; by the palatal vacuity (e), by the loose tympanic bone, the development of the tympanic bulla in the sphenoid, by the very small relative size of the brain, and other characters detailed in a Memoir lately communicated by the writer to the Royal Society of London. The canine tooth is 2 inches 3 lines in longitudinal extent, or nearly double the size of that of the lion. The upper tubercular tooth (m, i) resembles in its smallness and position that in the placental Felidae. But in the lower jaw the carnassial (p) is succeeded by two very small tubercular teeth (m, i and 2), as in Pliognathus (fig. 88); and there is a socket close to the symphysial of the lower jaw of Thylacoleo which indicates that the canine may have terminated the dental series there, and have afforded an additional feature of resemblance to the Pliognathus.

The foregoing are some of the more interesting illustrations of the law, that "with extinct as with existing Mammalia, particular forms were assigned to particular provinces, and that the same forms were restricted to the same provinces at a former geological period as they are at the present day." That period, however, was the more recent tertiary one.

In carrying back the retrospective comparison of existing and extinct Mammals to those of the eocene and oolitic strata, in relation to their local distribution, we obtain indications of extensive changes in the relative position of sea and land during those epochs, in the degree of incongruity between the generic forms of the Mammalia which then existed in Europe and any that actually exist on the great natural continent of which Europe now forms part. It would seem, indeed, that the further we penetrate into time for the recovery of extinct Mammalia, the further we must go into space to find their existing analogues. To match the eocene Palaeotheres and Lophiodons, we must bring Tapirs from Sumatra or South America, and we must travel to the antipodes for Myrmecobians, the nearest living analogue to the Amphitheres of our oolitic strata.

On the problem of the extinction of species little can be said; and of the more mysterious subject of their coming into being, nothing profitable or to the purpose at present. As a cause of extinction in times anterior to man, it is most reasonable to assign the chief weight to those gradual changes in the conditions affecting a due supply of sustenance to animals in a state of nature which must have accompanied the slow alternations of land and sea brought about in the vicissitudes of geological time. Yet this reasoning... is applicable only to land-animals; for it is scarcely conceivable that such operations can have affected sea-fishes. There are characters in land-animals rendering them more obnoxious to extirpating influences, which may explain why so many of the larger species of particular groups have become extinct, whilst smaller species of equal antiquity have survived. In proportion to its bulk is the difficulty of the contest which, as a living organism, the individual of such species has to maintain against the surrounding agencies that are ever tending to dissolve the vital bond, and subjugate the living matter to the ordinary chemical and physical forces. Any changes, therefore, in such external agencies as a species may have been originally adapted to exist in, will militate against that existence in a degree proportionate to the bulk of the species. If a dry season be gradually prolonged, the large Mammal will suffer from the drought sooner than the small one; if such alteration of climate affect the quantity of vegetable food, the bulky Herbivore will first feel the effects of stinted nourishment; if new enemies be introduced, the large and conspicuous animal will fall a prey, while the smaller kinds conceal themselves and escape. Small quadrupeds are more prolific than large ones. Those of the bulk of the Mastodons, Megatheria, Glyptodonts, and Diprotodon are uniparous. The actual presence, therefore, of small species of animals in countries where larger species of the same natural families formerly existed, is not the consequence of degeneration,—of any gradual diminution of the size of such species,—but is the result of circumstances which may be illustrated by the fable of the "oak and the reed;" the smaller and feebler animals have bent and accommodated themselves to changes to which the larger species have succumbed.

That species should become extinct, appears, from the abundant evidence of the fact of extinction, to be a law of their existence. Whether, however, it be inherent in their own nature, or be relative and dependent on inevitable changes in the conditions and theatre of their existence, is the main subject for consideration. But admitting extinction as a natural law, which has operated from the beginning of life, it might be expected that some evidence of it should occur in our own time, or within the historical period. Reference has been made to several instances of the extirpation of species, certainly, probably, or possibly, due to the direct agency of man. But this cause avails not in the question of the extinction of species at periods prior to any evidence of human existence: it does not help us in Palafax-y-Melzl the explanation of the majority of extinctions, as of the races of aquatic Invertebrata and Vertebrata which have successively passed away. Within the last century, Academicians of Petersburg and good Naturalists described and gave figures of the body and the perishable parts, including the alimentary canal, of a large and peculiar carnivorous Sireniian,—an amphibious animal like the Manatee, which Cuvier classified with his herbivorous Cetacea, and called Stelleria, after its discoverer. It inhabited the Siberian shores and the mouths of the great rivers there disemboguing. It is now believed to be extinct, and this extinction has not been due to any special quest and persecution by man. We may discern in this fact the operation of changes in physical geography, which have at length so affected the conditions of existence of the Stelleria as to have caused its extinction. Such changes had operated, at an earlier period, to the extinction of the Siberian elephant and rhinoceros: a future generation of zoologists may have to record the final disappearance of the Arctic buffalo (Ovibos moschatus). Remains of Orobos and Stelleria show that they were contemporaries of Elephas primigenius and Rhinoceros tichorhinus. The great auk (Auka impennis, L.) existed in the last century: no specimen has been obtained within the present. Scandinavian naturalists believe it to be extinct. It has not been specially hunted down, like the dodo and dinornis; but by degrees has become more and more scarce. Physical changes, slowly operating, seem to have affected its sources of food and other circumstances favourable to its well-being. The numbers of its bones on the shores of Iceland, Greenland, and Norway, attest the abundance of the bird in former times. A consideration of such instances of modern extinctions may best throw light, and suggest the truest notions, of the causes of ancient extinctions.

As to the successions, or coming to be, of new species, one might speculate on the possibility of a variety of auk being occasionally hatched with a somewhat longer winglet and a dwarfish stature,—of such a variety better adapting itself to the changed climatal conditions than the old type,—of such an origin, for example, of Auka torda—but to what purpose? Past experience of the chance aims of human fancy, unchecked and unguided by observed facts, shows how widely they have ever glanced away from the gold centre of truth.

(P.O.)