Bible Encyclopedias
Reptiles

1911 Encyclopedia Britannica

(Lat. Reptilia, creeping things, from reptilis; refere, to creep; Gr. g pirav, whence the term " herpetology," for the science dealing with them). In the days before Linnaeus, writers comprised the animals which popularly are known as tortoises and turtles, crocodiles, lizards and snakes, frogs and toads, newts and salamanders, under the name of oviparous quadrupeds or four-limbed animals which lay eggs. Linnaeus, desirous of giving expression to the extraordinary fact that many of these animals pass part of their life in the water and part on land,' substituted the name of Amphibia for the ancient term. Subsequent French naturalists (Lyonnet 2 and Brisson 3) considered that the creeping mode of locomotion was a more general characteristic of the class than their amphibious habits, and consequently proposed the scarcely more appropriate name of Reptiles. As naturalists gradually comprehended the wide gap existing between frogs, toads, &c., on the one hand, and the other oviparous quadrupeds on the other, they either adopted the name of Batrachia for the former and that of Amphibia for the latter, or they restricted the term Amphibia to Batrachians, calling the remainder of these creatures reptiles. Thus the term Amphibia, as used by various authors, may apply (1) to all the various animals mentioned, or (2) to Batrachians only (see Batrachia). The term Reptiles (Reptilia) is used (i) by some for all the animals mentioned above, and (2) by others, as in the present article, for the same assemblage of animals after the exclusion of Batrachians.

Equally varying are the limits of the term Saurians, which occurs so frequently in every scientific treatise on this subject. At first it comprised living crocodiles and lizards only, with which a number of fossil forms were gradually associated. As the characters and affinities of the latter became better known, some of them were withdrawn from the Saurians, and at present it is best to abandon the term altogether.

I. History Of Herpetology Certain kinds of reptiles are mentioned in the earliest written records or have found a place among the fragments of the oldest relics of human art. Such evidences, however, form no part of a succinct review of the literature of the subject such as it is proposed to give here. We distinguish in it six periods: (I) the Aristotelian; (2) the Linnaean (formation of a class Amphibia, in which reptiles and Batrachians are mixed); (3) the period of the elimination of Batrachians as one of the reptilian orders (Brongniart); (4) that of the separation of reptiles and Batrachians as distinct subclasses; (5) that of the recognition of a class Reptilia as part of the Sauropsida (Huxley); (6) that of the discovery of fossil skeletons sufficiently well preserved to reveal, in its general outlines, the past history of the class.

1. The Aristotelian Period. - Aristotle was the first to deal with the reptiles known to him as members of a distinct portion Aristotle. of the animal kingdom, and to point out the characteristics by which they resemble each other and differ from other vertebrate and invertebrate animals. The plan of his ' " Polymorpha in his amphibiis natura duplicem vitam plerisque concessit." 2 Theologie des insectes de Lesser (Paris, 1 745), i. 91, note 5. Regne animal divise en neuf classes (Paris, 1756).

work, however, was rather that of a comparative treatise of the anatomical and physiological characters of animals than their systematic arrangement and definition, and his ideas about the various groups of reptiles are not distinctly expressed, but must be gleaned from the terms which he employs. Moreover, he paid less attention to the study of reptiles than to that of other classes. This is probably due to the limited number of kinds he could be acquainted with, to which only very few extraEuropean forms, like the crocodile, were added from other sources. But while we find in some respects a most remarkable accuracy of knowledge, there is sufficient evidence that he neglected everyday opportunities of information. Thus, he has not a single word about the metamorphoses of Batrachians, which he treats of in connexion with reptiles.

Aristotle makes a clear distinction between the scute or scale of a reptile, which he describes as 40Xis, and that of a fish, which he designates as AE7ris. He mentions reptiles (I) as oviparous quadrupeds with scutes, viz. Saurians and Cheloniai s; (2) as oviparous apodals, viz. Snakes; (3) as oviparous quadrupeds without scutes, viz. Batrachians. He considered the first and second of these three groups as much more nearly related to each other than to the third. Accurate statements and descriptions are sadly mixed with errors and stories of, to our eyes, the most absurd and fabulous kind. The most complete accounts are those of the crocodile (chiefly borrowed from Herodotus) and of the chameleon, which Aristotle evidently knew from personal observation, and had dissected himself. The other lizards mentioned by him are the common lizards (aavpa), the common seps (X aXKis or "ctyvis) and the gecko (aaKaXa(3c. Trts or Kop8bXos). Of snakes (of which he generally speaks as 6 s) he knew the vipers ( g xis or g XLBva), the common snake (155pos), and the blindworm (roc/Aim ac/as), which he regards as a snake; he further mentions the Egyptian cobra and dragons (5peuc v) - North-African serpents of fabulous size. Of Chelonians he describes in a perfectly recognizable manner land tortoises (XEAc ? vrt), freshwater turtles (Eµos) and marine turtles (XEAc.vn OaXarria). Passing over eighteen centuries, we find the knowledge of reptiles to have remained as stationary as other branches of natural history, perhaps even more so. The reptile fauna of Europe was not extensive enough to attract the energy of a Belon or Rondelet; popular prejudice and the difficulty of preserving these animals deterred from their study; nor was man sufficiently educated not to give implicit credence to the fabulous tales of reptiles in the i 5th and 16th centuries. The art of healing, however, was developing into a science based upon rational principles, and consequently not only those reptiles which formed part of the materia medica but also the venomous snakes became objects of study to the physician, though the majority of the writers were ignorant of the structure of the venom-apparatus, and of the distinction between non-venomous and venomous snakes.

Nothing can show more clearly the small advance made by herpetology in this long post-Aristotelian period than a glance at the celebrated work, De Differentiis Animalium Libri decem (Paris, 1552), by Edward Wotton (1492 '555). Wotton treats of the reptiles which he designates as Quadrupedes oviparae et Serpentes in the sixth book of his work. They form the second division of the Quadrupedes quae sanguinem habent, and am subdivided in the following " genera ": Crocodilus et scincus (cap. cv.); Testudinum genera (cvi.); Ranarum genera (cvii.); Lacertae (cviii.); Salamandra et seps quadrupes (cix.); Stellio (cx.); Chamaeleo (cxi.); Serpentes (cxii.), a general account, the following being different kinds of serpents: Hydrus et alii quidam serpentes aquatiles (cxiii.); Serpentes terrestres et primo aspidum genera (cxiv.); Vipera, dipsas, cerastes, et hammodytes (cxv.); Haemorrhus, sepedon, seps, cenchris, et cenchrites (cxvi.); Basiliscus et alii quidam serpentes quorum venenum remedio caret (cxvii.); Draco, amphisbaena, et alii quidam serpentes quorum morsus minus affert periculi (cxviii.).

Wotton's work might with propriety be termed " Aristoteles redivivus." The plan is the same, and the observations of the Greek naturalist are faithfully, sometimes literally, reproduced.

It is surprising that even the reptiles ^f his native country were most imperfectly known to the author.

With the enlargement of geographical knowledge that of reptiles was also advanced, as is sufficiently apparent from the. large encyclopaedic works of Gesner, Aldrovandi and Johnston. The last-named author especially, who published the various portions of his Natural History in the middle of the 17th century, was able to embody in his compilations notices of numerous reptiles observed by Francisco Hernandez in Mexico and by Marcgrave and Piso in Brazil. As the author had no definite idea of the Ray-Linnaean term " species," it is not possible to give the exact number of reptiles mentioned in his work. But it may be estimated at about fifty, not including some marine fishes and fabulous creatures. He figures (or rather reproduces the figures of) about forty - some species being represented by several figures.

2. Linnaean Period: Formation of a Class Amphibia.- Within the century which succeeded these compilatory works (1650-1750) fall the labours which prepared the way sors of for and exerted the greatest influence on Ray and Linnaeus. Although original researches in the field of herpetology were limited in extent and in number, the authors had freed themselves from the purely literary or scholastic tendency. Men were no longer satisfied with reproducing and commenting on the writings of their predecessors; the pen was superseded by the eye, the microscope and the knife, and statements were tested by experiment. This spirit of the age manifested itself, so far as the reptiles are concerned, in Chara's and Redi's admirable observations on the viper, in Major's and Vallisnieri's detailed accounts of the anatomy of the chameleon, in the researches of Jacobaeus into the metamorphoses of the Batrachians and the structure of lizards, in Dufay's history of the development of the salamander (for Batrachians are invariably associated with reptiles proper); in Tyson's description of the anatomy of the rattlesnake, &c. The natural history collections formed by institutions and wealthy individuals now contained not merely skins of crocodiles or serpents stuffed and transformed into a shape to correspond with the fabulous descriptions of the ancient dragons, but, with the discovery of alcohol as a means of preserving animals, reptiles entire or dissected were exhibited for study; and no opportunity was lost of obtaining them from travellers or residents in foreign countries. Fossils also were now acknowledged to be remains of animals which had lived before the Flood, and some of them were recognized as those of reptiles.

The contributions to a positive knowledge of the animal kingdom became so numerous as to render the need of a methodical arrangement of the abundance of new facts more and more pressing. Of the two principal systematic attempts made in this period the first ranks as one of the most remarkable steps of the progress of natural history, whilst the second can only be designated as a signal failure, which ought to have been a warning to all those who in after years classified animals in what is called an, " artificial system." As the latter attempt, originating with Klein (1685-1759), did not exercise any further influence on herpetology, it will be sufficient to have merely Ray. mentioned it. John Ray (1628-1705) had recognized R the necessity of introducing exact definitions for the several categories into which the animals had to be divided, and he maintained that these categories ought to be characterized by the structure of animals, and that all zoological knowledge had to start from the " species " as its basis. His definition of reptiles as " animalia sanguinea pulmone respirantia cor unico tantum ventriculo instructum habentia ovipara " fixed the class in a manner which was adopted by the naturalists of the succeeding hundred and fifty years. Nevertheless, Ray was not a herpetologist; his knowledge of reptiles is chiefly derived from the researches of others, from whose accounts, however, everything not based upon reliable demonstration is critically excluded. He begins with a chapter treating of frogs (Rana, with two species), toads (Bufo, with one species) and tortoises' (Testudo, with fourteen species). The second group comprises the Lacertae, twenty-five in number, and includes the salamander and newts; and the third the Serpentes, nine species, among which the limbless lizards are enumerated.

Except in so far as he made known and briefly characterized a number of reptiles, our knowledge of this class was not advanced by Linnaeus. That he associated in the 12th edition cartilaginous and other fishes with the reptiles under the name of Amphibia Nantes was the result of some misunderstanding of an observation by Garden, and is not to be taken as a premonitory token of the recent discoveries of the relation between Batrachians and fishes. Linnaeus places reptiles, which he calls Amphibia, as the third class of the animal kingdom; he divides the genera thus: Order I. Reptiles. - Testudo (15 species); Rana (17 sp.); Draco (2 sp.); Lacerta (48 sp., including 6 Batrachians).

Order 2. Serpentes. - Crota1US (5 species); Boa (10 sp.); Coluber (96 sp.); Anguis (15 sp.); Amphisbaena (2 sp.); Caecilia (2 sp.).

None of the naturalists who under the direction or influence of Linnaeus visited foreign countries possessed any special knowledge of or predilection for the study of reptiles; all, however, contributed to our acquaintance with tropical forms, or transmitted well-preserved specimens to the collections at home, so that Gmelin, in the 13th edition of the Systema Naturae, was able to enumerate three hundred and seventy-one species. The man who, with the advantage of the Linnaean method, - first treated of reptiles monographically, was Laurenti. In a small book 2 he proposed a new division of these animals, of which some ideas and terms have survived into our times, characterizing the orders, genera and species in a much more precise manner than Linnaeus, giving, for his time, excellent descriptions and figures of the species of his native country. Laurenti might have become for herpetology what Artedi was for ichthyology, but his resources were extremely limited.

The circumstance that Chelonians are entirely omitted from his Synopsis seems due rather to the main object with which he engaged in the study of herpetology, viz. that of examining and distinguishing reptiles reputed to be poisonous, and to want of material, than to his conviction that tortoises should be relegated to another class. He divides the class into three orders: I. Salientia, with the genera Pipa, Bufo, Rana, Hyla, and one species of " Proteus," viz. the larva of Pseudis paradoxa. 2. Gradientia, the three first genera of which are Tailed Batrachians, viz. two species of Proteus (one being the P. anguinus), Triton and Salamandra; followed by true Saurians- Caudiverbera, Gecko, Chamaeleo, Iguana, Basiliscus, Draco, Cordylus, Crocodilus, Scincus, Stellio, Seps. 3. Serpentia, among which he continues to keep Amphisbaena, Caecilia and Anguis, but the large Linnaean genus Coluber is divided into twelve, chiefly from the scutellation of the head and form of the body.

The work concludes with an account of the experiments made by Laurenti to prove the poisonous or innocuous nature of those reptiles of which he could obtain living specimens.

The next general work on reptiles is by Lacepede. It appeared in the years 1788 and 1790 under the title Histoire naturelle des quadrupedes ovipares et des serpens (Paris, Lacepede. 2 vols. 4to). Although as regards treatment of details and amount of information this work far surpasses the modest attempt of Laurenti, it shows no advance towards a more natural division and arrangement of the genera. The author depends entirely on conspicuous external characters, and classifies the reptiles into (1) oviparous quadrupeds with a tail, (2) oviparous quadrupeds without a tail, (3) oviparous 1 In associating tortoises with toads, Ray could not disengage himself from the general popular view as to the nature of these animals, which found expression in the German Schildkrote (" Shieldtoad ").

Specimen medicum exhibens Synopsin Reptilium emendatam cum experimentis circa venena et antidota Reptilium Austriacorum (Vienna. 1768, 8vo, pp. 214, with 5 plates).

XXIII. 5 'a ' 0v 4 - Der , I 3, 1 31 "4°, lac,i.ti,ng,1 3 cj, +004d, 14 0,14 t. ??&????41????`"s'? V/6-ct 4'?,1a1, 1???1?? ^l; n,i?' 1s t ?ll y ?S i f,4,1 merely regarded as Amphibia because they closely resemble the genera which are proved to have been gill-breathers when immature. All these genera, however, so far as known, agree with the existing Amphibia in the production of their large parasphenoid bone as far forwards as the vomers to form a rigid and complete basicranial axis (fig. i, A). Those genera which less resemble the typical Labyrinthodonts are characterized by the reduction of the parasphenoid bone so that it no longer reaches the vomers; in these animals the weakened skull exhibits a secondary basicranial axis formed by the approximation of the pterygoids to the median line (fig. i, B). The latter condition is universal in existing reptiles, and may therefore perhaps be regarded as a diagnostic feature. If so, the oldest known undoubted reptile is Palaeohatteria, from the Lower Permian of Saxony.

In the structure of the skull Palaeohatteria is much like the existing Sphenodon, the cheek-plates which cover the temporal and masseter muscles on each side being pierced by two great vacuities, one superior-temporal, the other lateral-temporal. The majority of the earliest reptiles, however, either resemble the Labyrinthodonts in having the biting muscles completely covered with a roof of bony plates, or exhibit a slight shrinkage of this investment so that a superior-temporal vacuity appears. As the various groups or orders become differentiated, this shrinkage or reduction continues, while the shape of the ossifying ear-capsule changes, and the squamosal bone, which covers the organ of hearing in the fishes, and presumably also in the Palaeozoic Batrachia, is gradually thrust outwards from all connexion with this capsule except at its hinder angle. The resultant modifications are diagrammatically represented in fig 2. In one series of orders, comprising the Anomodontia, Chelonia, Sauropterygia and Ichthyopterygia (fig. 2, B, C), the superior-temporal vacuity (s) first appears, and the cheekplates in the broad temporal arch thus formed may be variously fused together, sometimes even irregularly perforated - showing at first, indeed, the usual inconstancy of a new and not completely established feature. From the earliest members of this series of reptiles, palaeontology seems to demonstrate that the Mammalia (with one robust temporal arcade or zygomatic arch) arose. In a second series, comprising the orders Rhynchocephalia, Dinosauria, Crocodilia and Ornithosauria (fig. 2, D), the broad arch of cheek-plates is regularly pierced by a lateraltemporal vacuity, which leaves a narrow bar above, another narrow bar below, and uncovers the middle part of the quadrate bone. By the constant loss of the lower, and the frequent loss of the upper, bar, some members of this series eventually pass into the order Squamata (Lacertilia+Ophidia), in which the quadrate bone is completely exposed and loosely attached to the skull (fig. 2, E); other reptiles exhibiting a similar modification may readily have acquired the typical Avian skull (fig. 2, F) by the loss of the upper and the retention of the lower temporal bar in question.

In view of these and other palaeontological considerations, the Reptilia may be classified into orders as follows: - Orders Of Class Reptilia I. Anomodontia. - Bones of postero-lateral region of skull forming a complete roof over the temporal and masseter muscles, or contracted into a single broad zygomatic arch, leaving a superior-temporal vacuity. Pineal foramen present. Ribs completely or imperfectly doubleheaded. No abdominal ribs. A large separately ossified epicoracoid. Limbs for support as well as progression; third and fourth digits with not more than three phalanges. Dermal armour feeble or absent. Range. - Permian and Triassic.

2. Chelonia. - Postero-lateral region of skull as in Anomodontia, except bones of ear-capsule more modified. No pineal foramen. Ribs single-headed. No sternum. Pectoral and pelvic arches unique in being situated completely inside the ribs. No epicoracoid. Abdominal ribs replaced by three or four pairs of large plates, which, with the clavicles and interclavicle, form a plastron. Limbs only for progression; third and fourth digits with not more than three phalanges. A regular dorsal carapace of bony plates intimately connected with the neural spines, and ribs of seven to nine dorsal vertebrae. Range. - Upper Triassic to Recent.

3. Sauropterygia. - Bones of postero-lateral region of skull contracted into a single broad zygomatic arch, leaving a superior-temporal vacuity. Pineal foramen present. No fused sacral vertebrae. All dorsal ribs single-headed, articulating with transverse processes of the neural arches. Abdominal ribs forming dense plastron. Apparently no sternum. Coracoid, pubis and ischium in form of much-expanded plates. Limbs modified as paddles, with not more than five digits, of which the third and fourth always have more than three phalanges; all digits usually consisting of numerous phalanges. No dermal armour. Range. - Upper Triassic to Cretaceous.

4. Ichthyopterygia. - Bones of postero-lateral region of skull contracted into a single broad zygomatic arch, leaving a superiortemporal vacuity. Pineal foramen present. Vertebral centra short and deeply biconcave, with feeble neural arches which are almost or completely destitute of zygapophyses. No fused sacral vertebrae. Cervical and dorsal ribs double-headed, articulating with tubercles on the vertebral centra. Abdominal ribs forming dense plastron. Apparently no sternum. Coracoid an expanded plate, probably with cartilaginous epicoracoid. Pelvis very small, not connected with vertebrae. Limbs modified as paddles, with digits of very numerous short phalanges, which are closely pressed together, sometimes with supplementary rows of similar ossicles. No dermal armour. A vertical triangular caudal fin, not supported by skeletal rays. Range. - Triassic to Cretaceous.

5. Rhynchocephalia. - Bones of postero-lateral region of skull contracted into two slender zygomatic bars, leaving a superiortemporal and a lateral-temporal vacuity, and partly exposing the quadrate bone from the side. Pineal foramen present or absent. Ribs single-headed. Abdominal ribs present. Sternum present. Epicoracoid cartilaginous. Limbs only for progression; third and fourth digits with four or five phalanges. Dermal armour feeble or absent. Range. - Lower Permian to Recent.

6. Dinosauria. - Postero-lateral region of skull as in Rhynchocephalia. No pineal foramen. Cervical and dorsal ribs doubleheaded. Rarely abdominal ribs. Sternum present, but apparently no clavicular arch. Limbs for support as well as progression; third and fourth digits with four and five phalanges respectively, Dermal armour variable. Range. - Triassic to Cretaceous.

7. Crocodilia. - Postero-lateral region of skull as in Rhynchocephalia. No pineal foramen. Cervical and dorsal ribs doubleheaded. Abdominal ribs present. Sternum present; also interclavicle, but no clavicles. Limbs only for progression on land or swimming; third and fourth digits with four or five phalanges. Dermal armour variable. Range. - Lower Jurassic to Recent.

Ornithosauria. - All bones extremely dense, light and hollow, the organism being adapted for flight. Postero-lateral region of skull as in Rhynchocephalia. No pineal foramen. Cervical and dorsal ribs double-headed. Abdominal ribs present. Sternum present, and keeled for attachment of pectoral muscles; no clavicular arch. Fifth digit of hand much elongated to support a wingmembrane, but with only four phalanges. Hind limb feeble. No dermal armour. Range. - Lower Jurassic to Cretaceous.

0000 0 N l A After Credner. After C. W. Andrews.

FIG. I. - A, Palate of Palaeozoic Amphibian (Archegosaurus decheni). B, Palate of Mesozoic Reptile (Plesiosaurus macrocephalus). b.occ, basioccipital; bs, basisphenoid; eept, ectopterygoid; i. pt, interpterygoid vacuity; j, j ugal; mx, maxilla; pas, parasphenoid; pl, palatine; pmx, premaxilla; pt, pterygoid; pt. nar, posterior p ares; qu, quadrate; s.o, suborbital vacuity; v, vomer.

0000000 (1u 9. Squamata. - Bones of postero-lateral region of skull much reduced and partly absent, never forming more than a slender superior-temporal bar, thus completely exposing the quadrate, which is only loosely attached to the cranium at its upper end. Pineal foramen present. Ribs single-headed. No abdominal ribs. Sternum present when there are limbs. Limbs, when present, only for progression; third and fourth digits at least with more than three phalanges. Dermal armour feeble or absent. Range. - Cretaceous to Recent.

Order r. Anomodontia. - The Anomodonts are so named in allusion to the peculiar and unique dentition of the first-discovered genera. They are precisely intermediate between the and India, but they are best represented in the Karoo formation (Permian and Triassic) of South Africa. The Pariasauria most closely resemble the Labyrinthodont Amphibia, but have a single occipital condyle. Pariasauria itself is a massive herbivorous reptile, with a short tail, and the limbs adapted for excavating in the ground. It is known by several nearly complete skeletons, about 3 metres in length, from South Africa and northern Russia. Elginia, found in the Elgin sandstones of Morayshire, Scotland, is provided with horn-like bony bosses on the skull. Another apparently allied genus (Otocoelus) has a carapace suggesting that it may be an ancestral Chelonian. The Therio ?z 'D. ' .?+ ??.

-qj. ' 'w ' qu. From A. S. Woodward, Outlines of Vertebrate Palaeontology. Fin. 2. - Diagram of the Cranial Roof in a Labyrinthodont Amphibian, various types of Reptiles, and a Bird. A, Labyrinthodont Amphibian (Mastodonsaurus giganteus). B, Generalized Anomodont or Sauropterygian, passing with slight modification into the Chelonian (sutures dotted to denote inconstancy in fusion of elements). C, Ichthyosaurus. D, Generalized Rhynchocephalian, Dinosaurian, Crocodilian, or Ornithosaurian. E, Generalized Lacertilian, often losing even the arcade here indicated. F, Generalized Bird.

fr, frontal; j, jugal; 1, lateral temporal vacuity; la, lachrymal; mx, maxilla; n, narial opening; na, nasal; o, orbit; pa, parietal; pmx, premaxilla; prf, prefrontal; ptf, postfrontal; pto, postorbital; q.j, quadrato-jugal; qu, quadrate; s, supratemporal vacuity; s.t, supratemporals and prosquamosal; sq, squamosal. Vacuities shaded with vertical lines, cartilage bones dotted.

Labyrinthodont B atrachia and the lowest or Monotreme Mammalia. They flourished at the period when the former are known to have reached their culmination, and when the latter almost certainly began to appear. Many of them would, indeed, be regarded as primitive Mammalia, if they did not retain a pineal foramen, a free quadrate bone, and a complex mandible. The term Theromorpha or Theromora is thus sometimes applied to the order they represent. So far as known, they are all land-reptiles, with limbs adapted for habitual support of the body, and their feet are essentially identical with those of primitive mammals. Most of them are small, and none attain a gigantic size. They first appear in the Permian of Europe and North America, and also occur in the Triassic both of Europe dontia exhibit the marginal teeth differentiated (in shape) into incisors, canines and molars (fig. 3). They have two occipital condyles, as in mammals. They seem to have been all carnivorous, or at least insect,:. 3rous, but the malariform teeth vary much in shape in the different genera. Cynognathus (fig. 3) and Lycosaurus have cutting teeth, while Tritylodon and Gomphognathus possess powerful grinders. The Dicynodontia have one pair of upper tusks or are toothless: their occipital condyle is trefoil-shaped, as in Chelonia. Dicynodon itself occurs in the Karoo formation of S. Africa, while other genera are represented in India, N. Russia and Scotland.

Order 2. CHELoNIA. - This order occurs first in the Upper Triassic of Wurttemberg, where a complete " shell" has been into five " sense " series, and each series into three orders, one comprising forms of superior, the second of medium and the third of inferior development. In the generic arrangement of the species, to which Fitzinger devoted himself especially in this work, he equally failed to advance science.

We have now arrived at a period distinguished by the appearance of a work which superseded all its predecessors, which formed the basis for the labours of many succeeding years, and which will always remain one of the classical monuments of descriptive zoology - the Erpetologie generale on histoire naturelle complete des reptiles of A. M. C. Dumeril and G. Bibron (Paris, 8vo). The first volume appeared in 18 34, and the ninth and last in 18J4. No naturalist of that time could have been better qualified for the tremendous undertaking than C. Dumeril, who almost from the first year of half a century's connexion with the then largest collection of Reptilia had chiefly devoted himself to their study. The task would have been too great for the energy of a single man; it was, therefore, fortunate for Dumeril that he found a most devoted fellow-labourer in one of his assistants, G. Bibron, whose abilities equalled those of the master, but who, to the great loss of science, died (in 1848) before the completion of the work. Dumeril had the full benefit of Bibron's knowledge for the volumes containing the Snakes, but the last volume, which treats of the Tailed Batrachians, had to be prepared by Dumeril alone.

The work is the first which gives a comprehensive scientific account of reptiles generally, their structure, physiology and literature, and again each of the four orders admitted by the authors is introduced by a similar general account. In the body of the work 121 Chelonians, 468 Saurians, 586 Ophidians and 218 Batrachians are described in detail and with the greatest precision. Singularly enough, the authors revert to Brongniart's arrangement, in which the Batrachians are co-ordinate with the other three orders of reptiles. This must appear all the more strange as Von Baer' in 1828, and J. Miiller 2 in 1831, had urged, besides other essential differences, the important fact that no Batrachian embryo possesses either an amnion or an allantois, like a reptile.

4. Period of the Separation of Reptiles and Batrachians as Distinct Classes or Subclasses. - In the chronological order which we have adopted for these historical notes, we had to refer in their proper places to two herpetologists, Blainville and Latreille, who advocated a deeper than merely ordinal separation of Reptiles from Batrachians, and who were followed by F. S. Leuckart. But this view only now began to find more general acceptance. J. Miiller and Stannius were guided in their classification entirely by anatomical characters, and consequently recognized the wide gap which separates the Batrachians from the Reptiles; yet they considered them merely as subclasses of the class Amphibia. The former directed his attention particularly to those forms which seemed to occupy an intermediate position between Lacertilians and Ophidians, and definitely relegated Anguis, Pseudopus, Acontias to the former, and Typhlops, Rhinophis, Tortrix, but also the Amphisbaenoids to the latter. Stannius interpreted the characteristics of the Amphisbaenoids differently, as will be seen from the following abstract of his classification: Subclassis: Amphibia Monopnoa (Leuckart).

[[Sect. I. Streptostylica]] (Stann.). Quadrate bone articulated to the skull; copulatory organs paired, placed outside the cloacal cavity.

Ordo I. Ophidia.

Subordo I. Eurystomata or Macrostomata (Mull.).

The facial bones are',loosely connected to admit of great extension of the wide mouth.

Subordo 2. Angiostomata or Microstomata (Mull.).

Mouth narrow, not extensile; quadrate bone attached to the skull and not to a mastoid.

' Entwicklungsgeschichte der Thiere, p. 262.

s Tiedemann's Zeitschrift fur Physiologie, vol. iv. p. 200.

Siebold and Stannius, Handbuch der Zootomie - Zootomie der Amphibien (2nd ed., Berlin, 1856,1856, 8vo).

Ordo 2. Sauria.

Subordo I. Amphisbaenoidea.

Subordo 2. Kionocrania (Stann.) =Lizards.

Subordo 3. Chamaeleonidea.

[[Sect. 2. Monimostylica]] (Stann.). Quadrate bone suturally united with the skull; copulatory organ simple, placed within the cloaca.

Ordo I. Chelonia.

Ordo 2. Crocodilia.

This classification received the addition of a fifth Reptilian order which with many Lacertilian characters combined important Crocodilian affinities, and in certain other respects differed from both, viz. the New Zealand Hatteria, which by its first describers had been placed to the Agamoid Lizards. A. Gunther, 4 who pointed out the characteristics of this reptile, considered it to be co-ordinate with the other four orders of reptiles, and characterizes it thus: Rhynchocephalia. - Quadrate bone suturally and immovably united with the skull and pterygoid; columella present. Rami of the mandible united as in Lacertilians. Temporal region with two horizontal bars. Vertebrae amphicoelian. Copulatory organs, none.

5. Period of the Recognition of a Class of Reptilia as Part of the Sauropsida

Although so far the discovery of every new morphological and developmental fact had prepared naturalists for a class separation of Reptiles and Batrachians, it was left to T. H. Huxley to demonstrate, not merely that the weight of facts demanded such a class separation, but that the reptiles hold the same relation to birds as the fishes to Batrachians. In his Hunterian Lectures (1863) he divided the vertebrates into Mammals, Sauroids and Ichthyoids, subsequently substituting for the last two the terms Sauropsida and Ichthyopsida. 5 The Sauropsida contain the two classes of birds and reptiles, the Ichthyopsida those of Batrachians and fishes.

6. Period of the Consideration of Skeletons of Extinct Reptiles

SIR R. Owen, while fully appreciating the value of the osteological characters on which Huxley based his division, yet admitted into his consideration those taken from the organs of circulation and respiration, and reverted to Latreille's division of warmand cold-blooded (haematothermal and haematocryal) vertebrates, thus approximating the Batrachians to reptiles, and separating them from birds. 6 The reptiles (or Monopnoa, Leuck.) thus form the highest of the five subclasses into which, after several previous c l assifications, Owen' finally divided the Haematocrya. His division of this subclass, however, into nine orders, makes a considerable step in the progress of herpetology, since it takes into consideration for the first time the many extinct groups whose skeletons are found fossil. He shows that the number of living reptilian types bears but a small proportion to that of extinct forms, and therefore that a systematic arrangement of the entire class must be based chiefly upon osteological characters. His nine orders are the following: a. Ichthyopterygia (extinct) - Ichthyosaurus. b. Sauropterygia (extinct) - Plesiosaurus, Pliosaurus, Nothosaurus, Placodus. c. Anomodontia (extinct) - Dicynodon, Rhynchosaurus, Oudenodon. d. Chelonia.

e. Lacertilia (with the extinct Mosasaurus). f. Ophidia.

g. Crocodilia (with the extinct Teleosaurus and Streptospondylus). h. Dinosauria (extinct) - Iguanodon, Scelidosaurus and Megalosaurus. i. Pterosauria (extinct) - Dimorphodon, Rhamphorhynchus and Pterodactylus. Owen was followed by Huxley and E. D. Cope, who, however, restricted still more the selection of classificatory characters by relying for the purposes of arrangement on a few parts of the 4 " Contribution to the Anatomy of Hatteria (Rhynchocephalus. Owen)," in Phil. Trans. (1867), part ii.

5 An Introduction to the Classification of Animals (London, 1869, 8vo), pp. 104 seq.

s Anatomy of Vertebrates (London, 1866, 8vo), vol. i. p. 6. Op. cit. p. 16.

skeleton only. They attempted a further grouping of the orders which in Owen's system were merely serially enumerated as cosubordinate groups. Huxley used for this purpose almost exclusively the position and character of the rib-articulations to the vertebral centra, the orders themselves being the same as in Owen's system: A. PLE Urospond Ylia. Dorsal vertebrae devoid of transverse processes and not movable upon one another, nor are the ribs movable upon the vertebrae. A plastron. Order I, Chelonia.

B. The dorsal vertebrae (which have either complete or rudimentary transverse processes) are movable upon one another, and the ribs upon them. No plastron.

a. The dorsal vertebrae have transverse processes which are either entire or very imperfectly divided into terminal facets (Erpetospondylia). a. Transverse processes long; limbs well developed, pad dles; sternum and sternal ribs absent or rudiment ary. Order 2, Plesiosauria (= Sauropterygia, Ow.). a. Transverse processes short.

aa. A pectoral arch and urinary bladder. Order 3, Lacertilia.

bb. No pectoral arch and no urinary bladder. Order 4, Ophidia.

b. The dorsal vertebrae have double tubercles in place of transverse processes (Perospondylia). Limbs paddle-shaped. Order 5, Ichthyosauria Ichthyopterygia, Ow.). c. The anterior dorsal vertebrae have elongated and divided transverse processes, the tubercular being longer than the capitular division (Suchospondylia). a. Only two vertebrae in the sacrum. Order 6, Croco Dilia.

0. More than two vertebrae in the sacrum.

aa. Manus without a prolonged ulnar digit.

aa. Hind limb Saurian. Order 7, Dicynodontia (= Anomodontia, Ow.).

f30. Hind limb Ornithic. Order 8, Ornitho Scelida (= Dinosauria, OW.).

bb. Manus with an extremely long ulnar digit. Order 9, Pterosauria.

Cope,' by combining the modifications of the quadrate and supporting bones with the characters used by Huxley, further developed Owen's classification, separating the ' Proc. Amer. Assoc. for the Advancement of Science, 10th meeting (Cambridge, 1871, 8vo), pp. 230 sq.; Amer. Naturalist (1889), vol. xxiii. p. 863.

Syllabus of Lectures on the Vertebrata (Philadelphia, 1898, 8vo), P. 54.

Ribs two-headed; interclavicle not distinct; external digits greatly elongated to support a patagium for flight.

Order 8, Ornithosauria.

Ribs two-headed; no interclavicle; acetabulum open; ambulatory. Order 9, Dinosauria.

Ribs two-headed; an interclavicle; acetabulum closed; ambulatory. Order to, Loricata.

Ribs one-headed; an interclavicle; acetabulum closed, a large obturator foramen; ambulatory. Order II, Rhynciiocephalia.

II. The quadrate bone loosely articulated to the cranium and at the proximal end only (Streptostylica). No distinct supramastoid, nor opisthotic; one or no postorbital bar; scapular arch, when present, external to ribs; ribs one-headed. Order 12, Squamata.

While this classification was being considered and prepared, both Cope and G. Baur made a special study of the bones which surround the quadrate and arch over the biting muscles in the various groups of reptiles. This led to a series of discussions which ended in the idea, that the class could be most naturally divided into two great subclasses, the one culminating in tortoises and mammals, the other in crocodiles, lizards, snakes and birds. Professor H. F. Osborn in 1903 3 therefore proposed the following classification: - Subclass Synapsida. Primarily with single or undivided temporal arches. Giving rise to the mammals through some unknown member of the Anomodontia. Orders Cotylosauria, Anomodontia, Testudinata and Sauropterygia. Subclass Diapsida. Primarily with double or divided temporal arches. Giving rise to the birds through some unknown type transitional between Protorosauria and Dinosauria. Orders Diaptosauria (=Protorosauria, Pelycosauria and Rhynchocephalia), Phytosauria (=Belodon, &c.), Ichthyosauria, Crocodilia, Dinosauria, Squamata and Pterosauria. The most exhaustive and modern general work on reptiles is by Dr C. K. Hoffmann in Bronn's Klassen and Ordnungen des Thierreichs (1879-90). A most useful and less technical treatise is the volume on Amphibia and Reptiles contri buted by Dr H. Gadow to the Cambridge Natural History (London, 1902). (A. C. G.; A. S. Wo.) II. General Characters Of The Class Reptilia Reptiles, as known in the existing world, are the modified, and in many respects degenerate, representatives of a group of lung-breathing vertebrate animals which attained its maximum development in the Mesozoic period. So far as can be judged from the skeleton, some of the members of this group then living might have become mammals by very slight change, while others might as readily have evolved into birds. It is therefore probable that the class Reptilia, as now understood, comprises the direct ancestors both of the Mammalia and A y es. Assuming that its extinct members, which are known only by skeletons, were organized essentially like its existing representatives, the class ranks higher than that of the lowest five-toed vertebrates (class Batrachia) in the investment of the foetus by two membranous envelopes (the amnion and allantois), and in the total absence of gills even in the earliest embryos. It ranks below both the Mammalia and A y es in the partial mixture of the arterial blood with the venous blood as it leaves the heart, thus causing the organism to be cold-blooded; it also differs both from Mammalia and A y es in retaining a pair of aortic arches, of which only the left remains in the former, while the right one is retained in the latter. No feature in the endoskeleton is absolutely distinctive, except possibly the degeneration of the parasphenoid bone, which separates the Reptilia from the Amphibia. In the exoskeleton, however, the epidermis forms horny scales, such as never occur in Amphibia, while there are no traces of any structures resembling either hairs or feathers, which respectively characterize Mammalia and Ayes.

There is little doubt that true reptiles date back to the latter part of the Palaeozoic period, but at that epoch the Amphibia approached them so closely in the characters of the skeleton that it is difficult to distinguish the members of the two classes among the fossils. Some of the Palaeozoic Amphibia - a few of the so-called Labyrinthodonts - are proved to have had welldeveloped gill-arches in their immature state, while there are conspicuous marks of slime-canals on their skulls. Others are Mem. American illus. Nat. Hist. (November 1903), vol. i. art. viii.

Pythonomorpha and Rhynchocephalia as distinct orders from the Lacertilia. He eventually' elaborated the following classification, based entirely on osteological characters: I. The quadrate bone immovably fixed to the adjacent elements by suture.

A. Scapular arch external to ribs; temporal region with a complex bony roof; no longitudinal postorbital bars.

A tabular and supramastoid bones and a presternum; limbs ambulatory; vertebrae amphicoelous. Order I, Cotylosauria.

AA. Scapular arch internal to ribs; temporal region with complex roof and no longitudinal bars.

A presternum; limbs ambulatory. Order 2, Chelydo Sauria.

AAA. Scapular arch internal to ribs; sternum extending below coracoids and pelvis; one postorbital bar.

No supramastoid; a paroccipital; clavicle not articulating with scapula. Order 3, Testudinata.

Aaaa. Scapular arch external to ribs; one longitudinal postorbital bar (Synaptosauria). A supramastoid and paroccipital bones; ribs two-headed on centrum; carpals and tarsals not distinct in form from metapodials; vertebrae amphicoelous. Order 4, Ichthyopterygia.

A supramastoid; paroccipital not distinct; a postorbitosquamosal arch; ribs two-headed; a clavicle; obturator foramen small or none; vertebrae amphicoelous. Order 5, Theromora.

No supramastoid; paroccipital not distinct; a quadratoj ugal arch; scapula triradiate; no clavicle; ribs oneheaded. Order 6, Plesiosauria.

Aaaaa. Scapular arch external to ribs; two longitudinal postorbital bars (paroccipital arch distinct) (Archosauria). a. A supramastoid bone.

Ribs two-headed; a clavicle and interclavicle; acetabulum closed; no obturator foramen; ambulatory; vertebrae amphicoelous. Order 7, Pelycosauria.

aa. No supramastoid.

bipeds (Chirotes and Pseudopus), (4) serpents, - an arrangement in which the old confusion of Batrachians and reptiles and the imperfect definition of lizards and snakes are continued, and which it is worthy of remark we find also adopted in Cuvier's Tableau elementaire de l'histoire naturelle des animaux (1798), and nearly so by Latreille in his Histoire naturelle des reptiles (Paris, 1801, 4 vols. 12 mo). Lacepede's monograph, however, remained for many years deservedly the standard work on reptiles. The numerous plates with which the work is illustrated, are, for the time, well drawn, and the majority readily recognizable.

3. The Period of Elimination of Batrachians as one of the Reptilian Orders. - A new period for herpetology commences with Alex. Brongniart,' who in 1799 first recognized Wart. the characters by which Batrachians differ from the other reptiles, and by which they form a natural passage to the class of fishes. Caecilia (as also Langaha and Acrochordus) is left by Brongniart with hesitation in the order of snakes, but newts and salamanders henceforth are no more classed with lizards. He leaves the Batrachians, however, in the class of reptiles, as the fourth order. The first order comprises the Chelonians, the second the Saurians (including crocodiles and lizards), the third the Ophidians - terms which have been adopted by all succeeding naturalists. Here, however, Brongniart's merit on the classification of reptiles ends, the definition and disposition of the genera remaining much the same as in the works of his predecessors.

The activity in France in the field of natural science was at this period, in spite of the political disturbances, so great that, only a few years after Lacepede's work another, almost i dentical in scope and of the same extent, appeared, viz. the Histoire naturelle generale et particuliere des reptiles of F. M. Daudin (Paris, 1802-3, 8 vols. 8vo). Written and illustrated with less care than that by Lacepede, it is of greater importance to the herpetologists of the present day, as it contains a considerable number of generic and specific forms described for the first time. Indeed, at the end of the work, the author states that he has examined more than eleven hundred specimens, belonging to five hundred and seventeen species, all of which he has described from nature. The system adopted is that of Brongniart, the genera are well defined, but ill arranged; it is, however, noteworthy that Caecilia takes now its place at the end of the Ophidians, and nearest to the succeeding order of Batrachians.

The next step in the development of the herpetological system was the natural arrangement of the genera. This involved a stupendous amount of labour. Although many isolated contributions were made by various workers, this task could be successfully undertaken and completed in the Paris Museum only, in which, besides Seba's and Lacepede's collections, many other herpetological treasures from other museums had been deposited by the victorious generals of the empire, and to which, through Cuvier's reputation, objects from every part of the world were attracted in a voluntary manner. The men who devoted themselves to this task were A. M. C. Dumeril, Oppel and Cuvier himself. Oppel was a German who, during his visit to Paris (1807-1808), attended the lectures of Dumeril and Cuvier, and at the same time studied the materials to which access was given to him by the latter in the most liberal manner. Dumeril 2 maintains that Oppel's ideas and information were entirely derived from his lectures, and that Oppel himself avows this to be the case. The passage, 3 however, to which he refers is somewhat ambiguous, 1 Bull. Acad. Sci. (1800), Nos. 35, 36.

2 Erpet. gener., i. p. 259.

" Wire es nicht die Ermunterung ... dieser Freunde gewesen, so wiirde ich iiberzeugt von den Mangeln, denen eine solche Arbeit bei aller mOglichen Vorsicht doch unterworfen ist, es nie gewagt haben, meine Eintheilung bekannt zu machen, obwohl selbe Herr Dumeril in seinen Lectionen vom Jahre 1809 schon vorgetragen, and die Thiere im Cabinet darnach bezeichnet hat " (preface, p. viii). A few lines further on he emphatically declares that the classification is based upon his own researches.

and it is certain that there is the greatest possible difference between the arrangement published by Dumeril in 1806 (Zoologie Analytique, Paris, 8vo) and that proposed by Oppel in his Ordnungen, Familien, and Gattungen der Reptilien (Munich, 1811, 4to). There is no doubt that Oppel profited largely by the teaching of Dumeril; but, on the other hand, there is sufficient internal evidence in the works of both authors, not only that Oppel worked independently, but also that Dumeril and Cuvier owed much to their younger fellow-labourer, as Cuvier himself indeed acknowledges more than once.

Oppel's classification may be shortly indicated thus: - Order I. Testudinata Or Cheloniens. I. Chelonii (gen. Mydas, Coriacea). 2. Amydae (gen. Trionyx, Chelys, Testudo, Emys). Order 2. Squamata.

Sect. A. Saurii.

I. Crocodilini (gen. Crocodilus, Gavialis, Alligator). 2. GECxoIDES (gen. Gecko, Stellio, Agama). 3. Iguanoides (gen. Camaeleo, Draco, Iguana, Basiliscus, Lophyrus, Anolis). 4. Lacertini (gen. Tupinambis, Dracaena, Lacerta, Tachydromus). 5. Se1NCOIDES (gen. Scincus, Seps, Scheltopusik, Anguis). 6. Chalcidici (gen. Chalcides, Bimanus, Bipes, Ophisaurus). Sect. B. Ophidii.

I. Anguiformes (gen. Tortrix, Amphisbaena, Typhlops). 2. Constrictores (gen. Boa, Eryx) Hydri (gen. Platurus, Hydrophis). Pseudoviperae (gen. Acrochordus, Erpeton). 5. Crotalini (gen. Crotalus, Trigonocephalus). 6. Viperini (gen. Vipera, Pseudoboa). 7. Colubrini (gen. Coluber, Bungarus). Order 3. Nuda Or Batracii.

In this classification we notice three points, which indicate a decided progress towards a natural system. (I) The four orders proposed by Brongniart are no more considered cosubordinate in the class, but the Saurians and Ophidians are associated as sections of the same order, a view held by Aristotle but abandoned by all following naturalists. The distinction between lizards and snakes is carried out in so precise a manner that one genus only, Amphisbaena, is wrongly placed. (2) The true reptiles have now been entirely divested of all heterogeneous elements by relegating positively Caecilia to the Batrachians, a view for which Oppel had been fully prepared by Dumeril, who pointed out in 1807 that " les cecilies se rapprochent considerablement des batraciens auxquels elles semblent her l'ordre entier des serpens." 4 (3) An attempt is made at arranging the genera into families, some of which are still retained at the present day.

In thus giving a well-merited prominence to Oppel's labours we are far from wishing to detract from the influence exercised by the master spirit of this period, Cuvier. Without his guidance Oppel probably never would have found a place among the promoters of herpetological science. But Cuvier's principal researches on reptiles were incidental or formed part of some more general plan; Oppel concentrated his on this class only. Cuvier adopts the four orders of reptiles proposed by Brongniart as equivalent elements of the class, and restores the blindworms and allied lizards and, what is worse, also the Caecilias, to the Ophidians. The chameleons and geckos are placed in separate groups, and the mode of dividing the latter has been retained to the present day. Also a natural division of the snakes, although the foreign elements mentioned are admitted into the order, is sufficiently indicated by his arrangement of the " vrais serpens proprement dits " as (1) non-venomous snakes, (2) venomous snakes with several maxillary teeth, and (3) venomous snakes with isolated poison-fangs. He distinguishes the species of reptiles with a precision not attained in any previous work.

Cuvier's researches into the osteology of reptiles had also the object of discovering the means of understanding the fossil remains which now claimed the attention of French, English and German naturalists. Extinct Chelonian and Crocodilian Memoires de zoologie et d'anatomie comparee (Paris, 1807, 8vo), P. 45.

Fam. Fam.

Fam. Fa m. Fam.

Fam.

Fam. Fam.

Fam. Fam. Fam. Fam. Fam. Fam. Fam.

remains, Pterodactylus, Mosasaurus, Iguanodon, Ichthyosaurus, Teleosaurus, became the subjects of Cuvier's classical treatises, which form the contents of the 5th volume (part 2) of his Recherches sur les ossemens fossiles, oit l'on retablit les caracteres des plusieurs animaux dont les revolutions du globe oat detruit les especes (new ed., Paris, 1824, 4to).

All the succeeding herpetologists adopted either Oppel's or Cuvier's view as to the number of orders of reptiles, or as to the position Batrachians ought to take in their relation to reptiles proper, with the single exception of D. DE Blainville. He divided the " oviparous subtype " of Vertebrates into four classes, Birds, Reptiles, Amphibians and Fishes,' a modification of the system which is all the more significant as he designates the reptiles " Squammiferes Ornithoides, ecailleux," and the amphibians " Nudipelliferes, Ichthyoides nus." In these terms we perceive clear indications of the relations which exist to the class of birds on the one hand, and to that of fishes on the other; but, unfortunately, Blainville himself did not follow up the ideas thus expressed, and abandoned even the terms in a later edition of his systematic tables.

The direct or indirect influence of the work of French anatomists manifested itself in the systems of the other herpetologists of this period. The Crocodiles, especially, which hitherto (strange to say, even in Cuvier's classification) had been placed as one of the families of Saurians, now commence to be separated. from them. Merrem (Versuch eines Systems der Merrem Amphibien, Marburg, 1820, 8vo) distinguishes two classes of " Amphibians," Pholidota and Batrachia.

The Pholidota (or Reptiles) are divided into three orders, distinguished chiefly by osteological and splanchnological characters: - I. Testudinata.

2. Loricata (=Crocodiles).

3. Squamata (=Oppel's Squamata, excluding Crocodiles).

Merrem's subdivision of the Squamata into (I) Gradientia (=limbed Lacertilia), (2) Repentia (=limbless Lacertilia), (3) Serpentia (= Snakes and Amphisbaena), (4) Incedentia (= Chirotes), and (5) Predentia (= Chamaeleons) was based chiefly on the modifications of the limbs, and not adopted by his successors. The greater part of his work is occupied with a synopsis of all the species of Reptiles known, each being shortly characterized by a diagnosis; but, as only a small proportion (about one hundred and seventy) were known to him from autopsy, this synopsis has all the faults of a compilation.

Latreille, who commenced the study of reptiles as early as 1801, had kept pace with the progress of science when he published, in 1825, his Families naturelles du regne Latreille. animal (Paris, 1825, 8vo). He separated the Batra chians as a class from the Reptiles, and the latter he divides into two sections only, Cataphracta and Squamosa - in the former Crocodiles being associated with the Chelonians. He bases this view on the development of a carapace in both, on the structure of the feet, on the fixed quadrate bone, on the single organ of copulation. None of the succeeding herpetologists adopted a combination founded on such important characters Gray. except J. E. Gray, who, however, destroyed Latreille's idea of Cataphracta by adding the Amphisbaenians 2 as a third order.

A mass of new materials now began to accumulate from all parts of the world in European museums. Among others, Spix had brought from Brazil a rich spoil to the Munich Museum,. and the Bavarian Academy charged JoH. Wagler wagler to prepare a general system of reptiles and batra chians. His work, 3 the result of ten years' labour, is a simple but lasting monument to a young naturalist, 4 who, endowed with an ardent imagination, only too frequently misinterpreted the evidence of facts, or forced it into the service of preconceived ideas. Cuvier had drawn attention to certain resemblances in 1 Bull. Sci. Soc. Philomat., July 1816.

2 Catalogue of the Tortoises, Crocodiles and Amphisbaenians in the Collection of the British Museum (London, 1844, 16mo), p. 2.

3 Natiirliches System der Amphibien mit vorangehender Classification der Seiugethiere and Vogel Beitrag zur vergleichenden Zoologie (Munich, 1830, 8vo).

4 Wagler was accidentally killed three years after the publication of his System. some parts of the osseous structure of Ichthyosaurus and Pterodactylus to dolphins, birds, crocodiles, &c. Wagler, seizing upon such analogical resemblances, separated those extinct Saurians from the class of Reptiles, and formed of them and the Monotremes a distinct class of Vertebrates, intermediate between mammals and birds, which he called Gryphi. We must admit that he made free use of his imagination by defining his class of Gryphi as " vertebrates with lungs lying free in the pectoral cavity; oviparous development of the embryo (within or) without the parent; the young fed (or suckled?) by the parents." By the last character this Waglerian class is distinguished from the reptiles.

Reptiles (in which Wagler includes Batrachians) are divided into eight orders: Testudines, Crocodili, Lacertae, Serpentes, Angues, Caeciliae, Ranae and Ichthyodi. He has great merit in having employed, for the subdivision of the families of lizards, the structure of the tongue and the mode of insertion of the teeth in the jaws. On the other hand, Wagler entirely failed in arranging snakes in natural families, venomous and non-venomous types being mixed in the majority of his groups.

L. Fitzinger was Wagler's contemporary; his first work 5 preceded Wagler's system by four years. As he says in the preface, his object was to arrange the reptiles in Fitz- " a natural system." Unfortunately, in order to lager. attain this object, Fitzinger paid regard to the most superficial points of resemblance; and in the tabula affinitatum generum which he constructed to demonstrate " the progress of nature " he has been much more successful in placing closely allied generic forms in contiguity than in tracing the relationships of the higher groups. That table is prepared in the form of a genealogical tree, but Fitzinger wished to express thereby merely the amount of morphological resemblance, and there is no evidence whatever in the text that he had a clear idea of genetic affinity. The Batrachians are placed at the bottom of the scheme, leading through Hyla to the Geckos (clearly on account of the digital dilatations) and through Caecilia to Amphisbaena. At the top Draco leads through Pterodactylus to the Bats (Pteropus), Ichthyosaurus to the Cetaceans (Delphinus), Emys to the Monotremes, Testudo to Manis, and the Marine Turtles to the Divers and Penguins.

In Fitzinger's system the higher groups are, in fact, identical with those proposed by Merrem, while greater originality is shown in the subdivision of the orders. He differed also widely from Wagler in his views as to the relations of the extinct forms. The order of Loricata consists of two families, the Ichthyosauroidea and Crocodiloidea, the former comprising Iguanodon, Plesiosaurus, Saurocephalus and Ichthyosaurus. In the order Squamata Lacertilians and Ophidians are combined and divided into twenty-two families, almost all based on the most conspicuous external characters: the first two, viz. the Geckos and Chameleons, are natural enough, but in the three following Iguanoids and Agamoids are sadly mixed, Pterodactyles and Draco forming one family; Megalosaurus, Mosasaurus, Varanus, Tejus, &c., are associated in another named Ameivoidea; the Amphisbaenidae are correctly defined; the Colubroidea are a heterogeneous assemblage of thirty genera; but with his family of Bungaroidea Fitzinger makes an attempt to separate at least a part of the venomous Colubrine Snakes from the Viperines, which again are differentiated from the last family, that of Crotaloidea.

If this little work had been his only performance in the field of herpetology his name would have been honourably mentioned among his fellow-workers. But the promise of his early labours was not justified by his later work, and if we take notice of the latter here it is only because his name has become attached to many a reptile through the pedantic rules of zoological nomenclature. The labours of Wiegmann, Muller, Dumeril and Bibron exercised no influence on him, and when he commenced to publish a new system of reptiles in 1843, 6 of which fortunately one fasciculus only appeared, he exhibited a classification in which morphological facts are entirely superseded by fanciful ideas of the vaguest kind of physiosophy, each class of vertebrates being divided 5 Neue Classification der Reptilien nach ihren natiirlichen Verwandtschaften (Vienna, 1826, 4to).

6 Systema Reptilium (Vienna, 1843, 8vo).

¦ Blainvule. found (Proganochelys). Its members are proved to have been toothless since the Jurassic period, and have only changed very From A. S. Woodward, Outlines of Vertebrate Palaeontology. FIG. 3. - Skull of an Anomodont (Theriodont) Reptile (Cynognathus crateronotus), one-fifth natural size. - Karoo formation (Permian or Triassic), South Africa.

d, dentary; j, jugal; l.t.f, incipient lateral temporal vacuity; la, lachrymal; mx, maxilla; na, nasal; orb, orbit; pa, parietal; pmx, premaxilla; prf, prefrontal; pto., postorbital; ptf, postfrontal; s.t, supratemporal (prosquamosal); sq, squamosal.

slightly since their first appearance. The marine turtles seem to have first acquired elongated paddles and vacuities in the shell during the Cretaceous period, and the Trionychia, destitute of epidermal shields, apparently arose at the same time.

Order 3. Sauropterygia. - TheSe are amphibious or aquatic reptiles (fig. 4). The head is comparatively small in most effective paddles with elongated digits, and as the genera are traced upwards in the geological formations it is possible to observe how the arches supporting the limbs become more rigid until the maximum of strength is reached. A few genera, such as Pliosaurus from the Jurassic and Polyptychodon from the Cretaceous of Europe, are distinguished by their relatively large head and stout neck. Some of the largest Upper Jurassic and Cretaceous species must have been ro metres in length. They were cosmopolitan in their distribution, but became extinct before the dawn of the Tertiary period.

Order 4. Ichthyopterygia. - The Ichthyosaurians are all fish-shaped, with a relatively large head and very short neck. Both pairs of paddles are retained, but the hinder pair is usually very small, and locomotion seems to have been chiefly effected by a large caudal fin. This fin, as shown in impression by certain fossils from Wurttemberg and Bavaria, is a vertical, triangular, dermal expansion, without any skeletal support except the hindermost part of the attenuated vertebral column, which extends along the border of its lower lobe (fig. 5). Another triangular fin, without skeletal support, is known to occur on the back, at least in one species (fig. 5). Some of the genera are proved to have been viviparous. Like the Sauropterygia, the Ichthyopterygia appear to have originated from terrestrial ancestors, for their earliest Triassic representatives (Mixosaurus) have the teeth less uniform and the limbs slightly less paddleshaped than the latter genera. In this connexion it is noteworthy that their hollow conical teeth exhibit curious infoldings of the wall, like those observed in many Labyrinthodonts, while their vertebrae almost exactly resemble those of the Labyrinthodont Mastodonsaurus and its allies. As the Ichthyosaurs are traced upwards in geological time, some genera become almost, or quite, toothless, while the paddles grow wider, and are rendered more flexible by the persistence of cartilage round their constituent bones (Ophthalmosaurus). They were cosmopolitan in distribution, but disappeared from all seas at the close of the Cretaceous period. The largest forms, with a skull 2 metres in length, occur in the Lower Lias.

Order 5. Rhynchocephalia. - TheSe are small lizard-shaped reptiles, which have scarcely changed since the Triassic period. Though now represented only by Sphenodon or Hatteria, which survives in certain islands off New Zealand, in the Mesozoic epoch they ranged at least over Europe, Asia and North America. They comprise the earliest known reptile, Palaeohatteria, from the Lower Permian of Saxony, which differs from the Triassic and later genera in having an imperfectly ossified pubis and ischium, more numerous abdominal ribs, and the fifth metatarsal.

FIG. 4. Plesiosaurus rostratus: restoration of skeleton by W. G. Ridewood.- Lower Lias, Dorsetshire.

genera, and the neck is usually elongated though not flexible. The tail is insignificant, generally short, and both pairs of paddles seem to have been concerned in progression. The order appears to have arisen from a group of land-reptiles, for its earliest members, from the Triassic of Europe (Lariosaurus) and from the Permo-Carboniferous of S. Africa (Mesosaurus) and Brazil (Stereosternum), are all amphibious animals. They are comparatively small, and their limbs are only just becoming paddle-like. The skull suggests affinities with the terrestrial FIG. 5. - Ichthyosaurus quadriscissus: outline of specimen showing dorsal and caudal fins, about one-sixth natural size. - Upper Lias, Wurttemberg. (After E. Fraas.) The irregularities behind the triangular dorsal fin are torn pieces of skin.

Anomodontia, and the shape of the scapula seems to show some connexion with the Chelonia. The truly aquatic Sauropterygians of the Jurassic (fig. 4) and Cretaceous periods possess most bone normal. They are also represented in the Permian, chiefly of North America, by the so-called Pelycosauria, which have sharp teeth in sockets, and are remarkable for the extreme short, biconcave elongation of the spines of their cervical and dorsal vertebrae (Dimetrodon, fig. 6). They seem to include various Triassic From Prof. E. C. Case's Revision of the Pelycosauria of North America, by permission of the Carnegie Institution of Washington.

FIG. 6. - Dimetrodon incisivus: restoration of skeleton by E. C. Case, about one-eighteenth natural size.

genera (e.g. Aetosaurus, Belodon), which may perhaps belong to the ancestral stock of the Dinosauria and Crocodilia. Other Triassic genera (Hyperodapedon, Rhynchosaurus) scarcely differ from Sphenodon, except in the dentition and in the absence of the pineal foramen in the skull. In the late Cretaceous and early Eocene periods one genus (Champsosaurus) was truly aquatic, with gavial-shaped head.

Order 6. DIN0sAURIA. - The dinosaurs are land reptiles which flourished on all the continents during the Jurassic and Cretaceous periods, in the interval between the decline of the Anomodontia and the dominance of the Mammalia. They first appeared as carnivorous reptiles in the Triassic period in Europe, India, S. Africa, and N. America, but afterwards comprised numerous massive herbivores in nearly all parts of the world except the Australian and New Zealand regions. The skeleton in the carnivorous dinosaurs, or Theropoda, is of very light construction, the vertebrae and limb bones being hollow, with thin, dense walls and often perfectly fitting joints. The fore limbs are small, and the hind limbs are adapted for running, jumping or hopping on the toes. The sabre-shaped cutting teeth are fixed in sockets, and all the claws are sharp. Anchisaurus and Hallo pus, from the Trias of N. America, and Scleromochlus from the Elgin sandstones of Scotland, are comparatively small animals. Ceratosaurus and Megalosaurus, from the Jurassic of North America and western Europe respectively, must have attained a length of from 5 to 6 metres. Tyrannosaurus, from the Cretaceous of Montana, U.S.A., has a skull more than a metre in length. The herbivorous Dinosaurs of the suborder Ornithopoda resemble the Theropoda in general shape, but are heavier in build, with a pelvis constructed more nearly on the plan of that of a running bird. It has, indeed, been suggested that certain arboreal Dinosaurs of bipedal gait may have been the ancestors of the class A y es. The bestknown Ornithopod is Iguanodon (fig. 7), from the Wealden of W. Europe, with species from 5 to Io metres in length. Claosaurus, from the Cretaceous of N. America, is I nearly similar, and is represented by at least one complete skeleton in the Yale University Museum. There are also members of the same group with a heavy armour of bony plates and spines, sometimes termed Stegosauria. Stegosaurus itself occurs in the Upper Jurassic of Colorado, and Omosaurus, from the Kimmeridge and Oxford clays of England, is a nearly similar reptile. Polacanthus, from the Wealden of the Isle of Wight, has the hip-region armoured with a continuous bony shield. Triceratops (fig. 8) and its allies, from the Upper Cretaceous (Laramie) of western N. America, are the latest members of the group, with a bony frill over the neck, a pair of bony horncores above the eyes, and a median bony horn-core on the nose. The skull with the bony frill sometimes measures nearly two metres in length. Another suborder of herbivorous Dinosaurs, that of Sauropoda, comprises the largest known land animals of any age, some measuring from 17 to 25 metres in total length. They have a small head, long neck, and long tail, and must have been quadrupedal in gait. Their teeth are adapted for feeding on succulent water weeds, perhaps with an admixture of small animals living among these; and their vertebrae are of very light construction, while the ribs are raised high on the neural arches to increase the size of the body cavity, perhaps for unusually large lungs or air sacs. Their massive limbs have five toes, of which the three inner alone bear outwardly curved claws. Diplodocus and Brontosaurus, from the Jurassic of Wyoming and Colorado, U.S.A., are the best-known genera. Atlantosaurus, from the same formation, is usually noteworthy for size. Cetiosaurus, from the Jurassic of England, is also known by large parts of the skeleton in the British Museum and the Oxford Museum, indicating species nearly 20 metres in length.

[[General Characters] Fig]]. 7.-Iguanodon bernissartensis: restoration of skeleton by O. C. Marsh, one-eightieth natural size. - Wealden, Bernissart, Belgium.

8

Triceratops prorsus: restoration of skeleton by O. C. Marsh, one-eightieth natural size. - Cretaceous, Wyoming.

Order 7. Crocodilia. - Typical crocodiles can be traced downwards to the Lower Lias at the base of the Jurassic FIG.

Lacertilia. The zygomatic arch of the Mammalia is formed (cf. also Agamidae) out of the supratemporal arch of Sphenodon, pm  ? ?

FIG. 17. - Ventral Aspect of Skull of Chelys matamata. bo, basioccipital; bs, basisphenoid; mdl, mandible; oh, opisthotic; pl, palatine; pm, premaxilla; po, prootic; pb, pterygoid; q, quadrate; s, squamosal; v, vomer.

m FIG. 18.--Lateral Aspect of Skull of Chelys matamata. an, angular; ar, articular; bo, basioccipital; d, dentary; op, opisthotic; m, maxilla; pa, parietal; pm, premaxilla; pr, prefrontal; ps, postfrontal; pt, pterygoid; q, quadrate; s, squamosal; sg, supra-angular.

after the loss of the postorbital element and of the quadratojugal, the squamosal gaining connexion with the upper, not posterior and ventral, branch of the jugal or malar bone.

The mandibular halves form a complete osseous symphysis, the only instance in reptiles; all the other elements retain their sutures. The articular portion of the articular bone forms several shallow cups and a slight anterior knob, best developed in Chelone. The angular bone does not help to form the posterior upper angle. The coronoid, or complementary element, is often small; the supra-angular and the splenial or opercular are always present, mostly also a pre-splenial wanting in Testudinidae (cf. G. Baur).

The hyoid apparatus is well developed, and sometimes assumes large dimensions, especially in Chelys. The two pairs of " horns " are the first and second branchial arches, whilst the hyoid arches are reduced to a pair of small, frequently only cartilaginous nodules, attached near the anterior corners of the basis linguae, which generally fuses with the os entoglossum in the tip of the tongue. In Chelydidae the long median basal or copular piece forms a semi-canal for the reception of the trachea.

In the skull of the Lacertilia the arcades over the temporal region vary much in composition and numbers. There are at most two arcades and two windows. First the posttemporal arcade, enclosing the posttemporal fenestra, which is framed mainly by the large paroccipital process below and the long parietal process above, both meeting distally, and the quadrate is carried by the paroccipital process. In the corner, in front, where the three bones meet, lies the squamosal, connecting parietal and quadrate. This squamosal, when not too much reduced, has an upper parietal and an anterior horizontal arm; the latter is essential for the formation of the second horizontal arcade, which makes the lower border of the supra-temporal window. The infra-temporal arcade, namely a quadrato-jugal +jugal arch, is absent in all Lacertilians owing to the complete absence of the quadrato-jugal element.

In Heloderma and Geckos the posttemporal is the only arcade. In the Amphisbaenids and in Aniella, practically also in Anelytropsis, all the arcades are lost. All the other families FIG. 19. - Skull of Chlamydosaurus kingii (old male), showing much differentiated teeth. I, ventral aspect; 2, posterior; 3, profile, showing the enormous process at the hinder end of the lower jaw.

of lizards and the chameleons have two arcades. We begin the description of the horizontal arcade with those families in which it is most complete, and most like that of Sphenodon. In Varanus it is formed by four bones. The postfrontal is short; to it is attached the postorbital, which sends a long horizontal process to join the squamosal 1 splint, and this connects with the 1 There is a much-debated question of the homologies of the one or two elements, both apparently membrane bones, which connect the upper end of the quadrate with the parietal and with the supratemporal arch. The question becomes acute in the snakes, whether the single element connecting skull and quadrate has to be called squamosal or supratemporal. Space forbids here to expound the matter, which has been very ably reviewed by S. W. Williston (" Temporal Arches in the Reptilia," Biolog. Bulletin, vii. No. 4, 1904, pp. 1 751 9 2; cf. also F. W. Thyng, Tufts College Studies, II. 2, 1906). About ten different names have been applied to these two elements, and two, namely, squamosal and supratemporal, are being used quite promiscuously. When only one element is present, the present writer uses the term squamosal, and there are reasons making it probable that this element is the squamosum of mammals. When both elements are present, the more ventral or lateral of the two is termed squamosal, that which always helps to form the upper anterior end of the quadrate; between the quadrate, the squamosal and the long parietal process lies the likewise splint-like supratemporal, attached by most of its length to the parietal process. The jugal has only one arm, and this connects the maxilla with the postorbital, completing the posterior orbital border. There is a wide gap between jugal and quadrate. In Tejidae the arcade is the same, but the squamosal reaches the jugal, both meeting the postorbital. In Lacerta the arcade is essentially the same, but the window is completely filled up by the postfrontal, which extends so far back as to reach the supratemporal. In the Agamidae the arcade is strong and simplified. Postfrontal and postorbital are represented by one forked piece. This squamosal and the postfrontal mass are connected by the upper, much up-curved end of the jugal, which is thrust between them. This arrangement is further emphasized in Iguana, the upper end of the jugal being much enlarged so as to form the greater portion of the arcade, and keeping ^*w the postfrontal mass and the simple squamosal widely asunder. In Heloderma postand prefrontals are in contact with each other, separating the frontal bone from the orbit; the jugal joins only the prefrontal, and there is no further arcade whatever. A vestige of a supratemporal (?) lies on the outside of the base of the squamosal, between s and q in fig. 20.

The chameleons are peculiar. The posttemporal arcade, spanning a wide space, is formed by a long process of the supra temporal - squamosal, which is directed up and backwards to join the parietal, which ex tends back by a long me unpaired process. The P horizontal arch is broad and short, squamosal and postfrontal, forming a broad suture; below they are joined by the jugal; above the suture lies, in chameleon, a tiny piece, perhaps a vestige of the dislodged post - orbital.

The jugal bones, to continue the description of the appendi cular parts of the skull, are firmly joined to lateral processes of the pterygoids by the ectopterygoids; further forwards they are extensively connected with the maxillaries. These rest against strong transverse palatine processes. The palatines form a medium symphysis; posteriorly they diverge together with the pterygoids, which articulate with the quadsupratemporal bridge, generally with the postorbital, sometimes also with the jugal. The more dorsal element is mentioned as supratemporal; it is always smaller, and mostly restricted to the corner between the squamosal and the parietal process against which it rests. Either of these two elements articulate with the quadrate. Both elements are present in Labyrinthodonts and in most of the extinct groups of reptiles; among recent forms in Lacertidae, Varanidae, Tejidae; one three-armed piece in Sphenodon, chameleons and crocodiles, without, in Sphenodon at least, any trace of a compound nature; one piece, forked, in Agamidae; one simple piece in most of the other Lacertilia, and in snakes.

rates and with the basisphenoid by a pair of strong basipterygoid processes. A slender vertical rod of bone, the columella cranii, arises from the dorsal surface of each pterygoid and, passing at a distance from the cranial capsule, is sutured to a short lateroventral process of the parietals Such a pair of columellm exists in nearly all Lacertilia (distinguished by many systematists as Kionocrania) with the exception of the chameleons and the Amphisbaenidae. In many lizards, however, this columella, or epipterygoid, does not quite reach the parietal, leaning instead against the proOtic; possibly it has been evolved out of the alisphenoid, and Chelonians seem to support this view. The premaxillary bone is single, except in the Skinks and in some Geckos; ventrally it touches the vomers which vary much in size; they are always paired although suturally connected; posteriorly they pass into, and fuse with, the palatines before these send off their maxillary processes. Between the vomer and its maxillary is a longitudinal hole. Often, e.g. in Lacerta, the vomers enclose a median hole near their anterior end, for Jacobson's organ. Dorsally the premaxilla sends a median process backwards to the nasals. These are paired, and fuse together only in Uroplates and in Varanus. The external nasal fossae are sometimes very large, and their anterior half appears blocked by the ossified turbinals, e.g. in Varanus and Tejus. Prefrontals are always present, often fused with the lacrymals; in Heloderma, in Aniella and in chameleons the prefrontals extend so far back as to meet the postfrontals, excluding thereby the frontals from the orbital rim. The frontals are either paired, as in Varanus, Lacertidae, Heloderma, Anguidae, Scincidae,Anelytropsidae, Aniella, Amphisbaenidae, and in some Geckoninae; or they are fused into one bone, as in the Eublepharinae, chameleons, Tejidae, Iguanidae, Agamidae, Xenosaurus. The parietals are double in the Geckos, in Uroplates and Xantusia; in all the others they form one coossified mass, generally with a pineal foramen, except in Eublepharinae, Amphisbaenidae, Tejidae, in Aniella and other degraded forms. In the majority the pineal foramen lies in the middle of the parietal, but in the Iguanidae it is near the frontal, and actually in the frontal in chameleons.

As regards the brain-case, there is a cartilaginous interorbital septum, connected posteriorly with the slender, bony presphenoid; ventrally on to this is fused a vestige of the parasphenoid, a narrow and thin splint which sometimes can be dislodged. The whole of the anterior wall of the brain-case is membranous, excepting a pair of separate ossifications, which do but rarely touch any of the cranial bones, as frontal, parietal or prodtics. The ossifications are irregular in shape, each sending out a downward process which curves inwards almost to meet its fellow; between these issue the olfactory lobes. W. K. Parker recognized them as the alisphenoids; E. D. Cope named them postoptics, and remarked that in Sphenodon they coexist with an orbitosphenoid bone. The probtic has a notch in its anterior lateral margin for the passage of the trigeminal nerve. The opisthotic portion of the petrosal mass is intimately fused with the lateral occipital bones and their paroccipital process, and sometimes, e.g. Tejus, encloses with them many intricate recesses of the middle ear-chamber, which extend also into hollow and swollen thick downward processes of the basioccipital. These cavities of both sides communicate with each other through the cancellous substance of the basioccipital and basisphenoid. There are no Eustachian tubes opening into the mouth through the base of the skull.

The occipital condyle is tripartite, the lateral occipitals partaking of the articulation; very rarely, e.g. in Amphisbaenidae (see fig. 22), the basioccipital portion is so much reduced that the skull articulates by two very broad condyles.

The halves of the under jaw are but loosely united, either by ligament only or by an at least very movable suture. The jaw is compound and the numerous constituent bones mostly retain their sutures. Besides the dentary and articular, angular and supra-angular on the lateral side, and the opercular or splenial on the inner side, there lies on the dorsal side the coronoid, six pairs in all. The posterior angle of the jaw r FIG. 20. - Dorsal aspect of skull ofHelodermahorridum. f, frontal ;j, j ugal; 1, lachrymal; m, maxilla; n, nasal; pa, parietal, pm, premaxilla; pr, pref rontal; ps, postfrontal; pi, pterygoid; q, quadrate; s, squamosal; so, supraoccipital.

ps FIG. 21. - Skull of Chamaeleon vulgaris. ag, angular; ar, articular; bs, basisphenoid; d, dentary; j, jugal; m, maxilla; me, median ethmoid; p l and p 2 , parietals; pi, palatine; pr, prefrontal; pt, pterygoid; q, quadrate; sg, supra-angular; so, supraoccipital; sq, squamosal.

and as such is crossed by the auditory columellar chain. The infra-temporal bridge or jugal arch is formed by the jugal (qj in fig. 12), which joins the descending process of the squamosal, and the quadrato-jugal, which is very small and partly fused with the lateral side of the quadrate. Now, between the quadrate on the one side and the squamoso+quadrato-jugal+jugal on the other, is enclosed a gap, met with only in Sphenodon of recent reptiles. This fourth, or quadrato-squamosal foramen, with its squamoso-quadrato-jugal bridge, is, as a rule, not mentioned, being too small to be obvious. The quadrate is very firmly fixed. On the ventral side of the cranium we notice the broad and long bony palate, the large vomers, and the pterygoids meeting in the middle line; aside of the vomers are the long posterior p ares; posteriorly the pterygoids diverge to rest upon short basi-sphenoid processes, and they articulate by short flanges with the quadrates.

The occipital condyle is kidney-shaped, triple, composed of the basi and the lateral occipitals. The dorsal median roof of the cranium is formed by the paired parietals, near their anterior symphysis with the large pineal foramen, the paired frontals, nasals and premaxillaries. The outer nares are surrounded by the premaxillaries, maxillaries and nasals. Prefrontals and postfrontals exist. There is a complete cartilaginous, interorbital septum, and a cranial columella, a pair of upright buttresses arising in the alisphenoidal walls, connecting the parietals with the pterygoids. The hyoid apparatus consists of a narrow base, with three pairs of arches; of these the first or hyoid arch is variously connected with the cranium near the paroccipital process, or with the extracolumella (see Middle Ear, below); the others are a long and stout pair of first and a smaller pair of second branchial arches.

Crocodiles

The temporal region is still bridged over by three arches, dividing the whole fossa into three, very much as in Sphenodon. The supratemporal foramen is bordered by the parietal, postfrontal (postorbital absent) and squamosal. The posttemporal foramen is very much reduced, sometimes to a narrow passage between the parietal, occipitals and squamosal, because the latter bone forms an extensive suture with the paroccipital process. The infratemporal or lateral fossa is wide and rather shallow, bordered above by the postfrontal and squamosal, in front by the postfrontal and jugal, below by the jugal and quadrato-jugal, behind by the latter, the quadrate, tip of the paroccipital and the squamosal. The quadrato-jugal being long and in an almost horizontal position, being wedged in between the jugal and nearly the whole length of the lateral edge of the quadrate, and there being no squamoso-quadratojugal bridge, the fourth foramen of Sphenodon is absent. The middle-ear cavity is reduced to a complicated system of narrow passages; one for the passage of the extra-columellar-mandibular string of the auditory chain (see Ear, below), between the quadrate, paroccipital and lateral occipital bones; another passage (Eustachian) opens in the roof of the mouth, between basioccipital and basisphenoid; a third joins that of the other side and forms with it a median opening between the same bones, just behind the posterior pterygoid border of the choanae. These nares, being in the recent crocodiles shifted as far back as possible, communicate with the outer nostrils by very long passages, formed by the whole length of the pterygoids, palatines, maxillaries, vomers and pre-maxillaries, all of which form a long median suture. But this long bony palatal roof is interrupted by a pair of large palatal foramina, bordered usually by palatine, pterygoid, ectopterygoid, or transverse bone and maxillary. On the dorsal side of the cranium we notice the parietals fused into an unpaired bone, without a pineal hole and the likewise unpaired frontal. There are a pair of postfrontals, prefrontals and lacrymals perforated by the naso-lacrymal duct. The nasals vary much in length, mostly in conformity with that of the maxillaries; as a rule they reach the short premaxillaries, but not always the nasal groove. (For taxonomic detail see under Crocodile.) The occipital condyle is formed mainly by the basioccipital, which always borders part of the foramen magnum, but the lateral occipitals each send a flange to it, which in immature specimens still partakes of the articulation with the atlas. The opisthotic and epiotic bones fuse early with the lateral and with supraoccipital bones; only the proOtic remains longer as a separate element, anteriorly with a large hole for the exit of the third branch of the trigeminal nerve. The basisphenoid is scarcely visible, being overlaid by the pterygoids. The presphenoid is larger, continued forwards and upwards into the inter-orbital septum, which remains mostly cartilaginous. Near the anterior and upper margin of the pre-sphenoid is a large notch on either side for the passage of the optic nerve, the three eyemuscle nerves and the first branch of the trigeminal. The place of the orbitosphenoids is taken by membrane or cartilaginous continuations of the interorbital septum, but the alisphenoids are large and abut upwards against the frontals and with a lateral flange against the postfrontals. These send down a conspicuous process which forms sutures with an upward process of the jugal and another of the ectopterygoid; it is this compound pillar which, partly divides the orbit from the infratemporal or lateral fossa. The size of these and the upper temporal fossae stand in an inverse ratio to each other. The upper fossae are still comparatively large in the long-snouted Gavialis and Toznistoma, whilst these holes almost completely disappear in the alligators, namely, in the broadand short-snouted members of the order, which chew their prey. In extinct Crocodilians the upper fossae were the larger. The temporo-mandibular muscle which lifts or shuts the lower jaw arises from the walls of the upper fossa, passes beneath the jugal-arch and is inserted upon the supra-angular portion of the lower jaw. In the more recent crocodiles this muscle is more and more superseded by the pterygo-mandibular muscle, which, arising chiefly from the dorsal surface of the much-broadened pterygoid, fills the widened space between the latter and the quadrate, and is inserted into the outer surface of the angular bone. The arrangement of this muscle secures a more advantageous leverage of the jaw, and is capable of more powerful development than the other, which is consequently on the wane - a nice illustration of onward, orthogenetic evolution. The dentary bones of the under jaw form a suture, later a symphysis; this is very long in the long-snouted genera, in which the splenials likewise form a long symphysis; in the others the mandibular symphysis is much shorter and the splenials remain widely separated. The articular bone is short, forms a transverse cup for the quadrate, or a saddle-shaped cup, and is perforated by the Siphonium (see below under Ear). The angle is upturned, formed by the articular, angular and, laterally, by the supra-angular bone; the opercular or counterpart of the splenial lies on the outer side, forming part of the anterior border of the oval foramen in the jaw.

The Chelonian skull agrees in many important features with that of Sphenodon and of the crocodiles, but it is composed of fewer bones, the ectopterygoids, lacrymals and postorbitals being absent, often also the nasals, unless they are fused with the prefrontals. The vomer is unpaired and forms a septum between the nasal passages, which, except in Sphargis, are ventrally roofed over to a variable extent by wings sent out by the palatines, joining the sides of the voiner. Most of the configurations of the other cranial bones are well represented in the accompanying figures. The palatines form a continuous broad floor with the pterygoids, which are extensively and firmly joined to the quadrates and to the basisphenoid. There are no Eustachian tubes. The occipital condyle is distinctly triple and the basioccipital is frequently excluded from the foramen magnum. The lateral occipitals early send out a pair of stout wings, the ventral of which joins a stout ventrilateral process of the basioccipital, both forming a thick knob especially in Chelone, and a dorsolateral wing, which broadly joins the large opisthotic bone. This connects the lateral occipital and the supraoccipital with the upper portion of the quadrate. On the top of the quadrate and upon the lateral dorsal portion of this compound transverse process (which of course corresponds to the paroccipital process of crocodiles, &c.) lies the squamosal, about which more presently. The two wings of the lateral occipital, part of the opisthotic, the quadrate, and part of the I pterygoids, form the bony borders of the middle ear-cavity, which is open behind; through it extends horizontally the columellar rod, received with its outer portion by a notch on the posterior side of the quadrate. This is of very complicated shape. Its outer margins form most of the tympanic frame; the posterior margins being curved backwards leave a wide notch behind in the Cryptodira and in Sphargis, but in the Pleurodira this part of the quadrate is transformed into a trumpet, the rim of which, forming a complete ring, carries the tympanic membrane. The tympanic cavity thus formed often leads into a deep recess which extends into the hollowed-out squamosal (e.g. in Testudo) towards the opisthotic and bears some resemblance to the intricate tympanic recesses which pervade that region of the crocodile's skull. With its upper anterior and FIG. 15. - Side view of skull of Testudo tabulata (from nature).

an, angular; ar, articular; d, dentary; f, frontal; j, jugal; m, mandible; n, naso-prefrontal; pa, parietal; pl, palatine; ps, postfrontal; q, quadrate; qj, quadrato-jugal.

inner portion the quadrate joins the large prodtic bone which is usually completely fused with the rest of the opisthotic, but in Sphargis it remains separate, and in this turtle the sutures between the otic bones and the supraoccipital also persist. In front of the prodtics the bony lateral walls of the brain-case end in Sphargis, but in most of the other Chelonians bony alisphenoids are represented by a pair of epipterygoids which rest upon short upward processes of the pterygoids and are joined by much longer, rather thin, but broad descending lamellae from the parietals. They represent of course the columellae cranii or pterygoidal columellae; if they are of alisphenoidal origin the term epipterygoids is a misnomer; the same applies to these structures in other reptiles. Through the space enclosed by the pterygoid, basioccipital, opisthotic and quadrate, enters the cranial carotid artery, sometimes piercing the posterior rim of the pterygoid; then the canal runs along the dorsal side of this bone and opens near the cranial columella. The arcades over the temporal region are most variable. Potentially Chelonians possess all the three arcades of the crocodiles, but it so happens that never more than one fenestra is present. The false roof over the temporal region is most complete in Sphargis and in the Chelonidae. Excepting Sphargis the supraoccipital extends far beyond the back of the cranium in shape of a long unpaired crest, which never diverges, or sends out lateral processes, but it is joined, and partly overlaid for a great part of its length, by the parietals in Chelonidae and Sphargis. In these genera the much-enlarged parietal, the equally large postfrontal, with the squamosal behind, the jugal below, and a large quadrato-jugal, form one continuous bony roof over the whole temporal fossa, which is widely open behind, the space being bordered by supraoccipital, opisthotic, squamosal and parietal. All other Chelonians show a great reduction of this roof. The parietal does not send out dorsolateral expansions; and the postfrontal likewise forms no expansions. It joins the rather short malar, forming the posteriororbital bridge, which posteriorly is connected by the quadrato-jugal with the upper portion of the quadrate and with the squamosal. The latter rests upon the quadrate and is in no connexion with the parietal. Consequently the whole temporal fossa is quite open. The horizontal bridge or arcade is to a certain extent homologous with the infra-temporal arcade. All the bones which border the temporal fossa vary much in extent. The greatest reduction has taken place in Cistudo and in Geoemyda, the latter an Indian genus of Testudinidae, in which the quadrato-jugal is lost, leaving a wide gap in the horizontal arcade. - The Chelonians form an instructive parallel to mammalian conditions by the broad contact of the squamosal with the malar, e.g. in Chelone, whilst the quad as FIG. 16. - Dorsal Aspect of Skull of Chelys matamata. bo, basioccipital; eo, exoccipital; f, frontal; j, jugal; m, maxilla; pm, premaxilla; pa, parietal; pr, prefrontal; ps, postfrontal; pt, pterygoid; q, quadrate; s, squamosal; so, supraoccipital.

rato-jugal, having in all Chelonians lost its original ventral connexion with the jugal, may actually get lost as in all the so, -pa tvs ,f formations, but all the Jurassic and some of the Cretaceous genera have the secondary bony plate less extended backwards than that in the Tertiary and existing genera, while their vertebrae have flattened or concave ends, instead of exhibiting a balland-socket articulation. Some of the Upper Jurassic crocodiles (Metriorhynchus) were more truly aquatic than any now living, with the fore limbs degenerate, the hind limbs much enlarged for swimming, and the dermal armour lacking. The end of the vertebral column is bent downwards, as in Ichthyosaurus, so they doubtless possessed a similar triangular tail-fin. Typical crocodiles and alligators date back to the close of the Cretaceous period, and they did not become extinct in Europe until the beginning of the Miocene period. Remains of an extinct alligator (Diplocynodon) are common in the Upper Eocene sands of the Hordwell cliffs, Hampshire.

Order 8. Ornithosauria. - The flying reptiles or Pterodactyls (fig. 9) are completely evolved at their earliest known FIG. 9. - Pterodactylus spectabilis, natural size, from the Lithographic Stone. h, humerus; ru, radius and ulna; mc, metacarpals; pt, pteroid bone; 2, 3, 4, digits with claws; 5, elongated digit for support of wing-membrane; st, sternum, crest not shown; is, ischium; pp, prepubis. The teeth are not shown. (After H. von Meyer.) appearance in the Lower Lias (Dimorphodon), and exhibit little essential change as they are traced upwards through the Mesozoic formations. The latest Cretaceous genera, however, comprise the largest species, which have been found in Europe, N. America and Brazil. Some of these (Pteranodon) are toothless, and their wings are so large that for adequate support the pectoral arch is fixed to the vertebrae like a pelvis. The wings occasionally have a span of from 5 to 6 metres. The wingmembranes are only known in the European Jurassic genus, Rhamphorhynchus (fig. Io), found well preserved in the finegrained lithographic stone of Bavaria. In this genus there is also a rhomboidal flap of membrane at the end of the tail.

Order 9. Squamata. - The ancestors of the lizards and snakes can only be traced back definitely to the latter part of the Cretaceous period. They were then represented by two suborders of aquatic reptiles, the Dolichosauria and Pythonomorpha(or Mosasauria),which are in many respects intermediate between the existing Lacertilia and Ophidia. The Dolichosauria, from the Upper Cretaceous of Europe, are small and snake-like in shape, but with completely formed limbs. The Pythonomorpha are known from Europe, N. and S. America and New Zealand, and sometimes attained a very large size, the typical Mosasaurus camperi from Maastricht being about 15 metres in length. Their limbs are powerful paddles. Their trunk and FIG. Io. - Rhamphorhynchus phyllurus, from the Solenhofen Lithographic Stone, one-fourth natural size, with the greater part of the wing-membranes preserved. x, caudal membrane; st, sternum; h, humerus; sc, scapula' and coracoid; wm, wingmembrane. (After O. C. Marsh.) tail are often much elongated, so that their shape is snake-like, as shown by Clidastes (fig. 11), from the Chalk of Kansas, U.S.A. The Lacertilia and Ophidia, so far as known, are exclusively Tertiary and Recent reptiles. Marine snakes (Palaeophis) occur in the Eocene of the London and Hampshire basins.

AuTHORITIES

General Works on Extinct Reptiles. - K. A. v. Zittel, Handbuch der Palaeontologie, vol. iii. (Munich, 1887 - 1889). - H. A. Nicholson and R. Lydekker, Manual of Palaeontology, vol. ii. (Edinburgh, 1889). - R. Lydekker, Catalogue of the Fossil Reptilia and Amphibia in the British Museum, vols. i.--iv. (London, 1888-90). - A. S. Woodward, Outlines of Vertebrate Palaeontology (Cambridge, 1898). - K. A. v. Zittel, Text-book of Palaeontology, ed. C. R. Eastman, vol. ii. (London, 1902). Anomodontia: R. Owen. Catalogue of the Fossil Reptilia of South Africa in the Collection of the British Museum (London, 1876). - E. D. Cope, " The Reptilian Order Cotylosauria," Proc. Amer. Phil. Soc. vol. xxxiv. (1896), p. 43 6, and vol. xxxv. (1896), p. 122. - E. T. Newton, " Some New Reptiles from the Elgin Sandstones," Phil. Trans., vol. 184E (1893), P. 431. - Various papers by R. Owen in Quart. Journ. Geol. Soc., 1876-1884, by H. G. Seeley in Phil. Trans. (1889-1895), and by R. Broom in Proc. Zool. Soc., Ann. S. African Museum and Trans. S. African Phil. Soc. (from 1900 onwards). Chelonia: G. Baur, " Bemerkungen fiber die Phylogenie der Schildkroten," Anat. Anzeiger, vol. xii. (1896), p. 561. - Technical papers by F. A. Quenstedt in Wiirtt. Jahresh. vol. xlv. (1889), p. 120 (Proganochelys). - G. R. Wieland in Amer. Journ. Sci. ser. 4, vol. ii. (1896), p. 399 (gigantic Cretaceous leathery turtle), and E. C. Case, Journ. Morphol. vol. xiv. (1897), p. 21 (ditto). Sauropterygia: G. A. Boulenger, " On a Nothosaurian Reptile from the Trias of Lombardy, apparently referable to Lariosaurus," Trans. Zool. Soc. vol. xiv. (1896), p. i. - H. G. Seeley, " The Nature of the Shoulder Girdle and Clavicular Arch in Sauropterygia," Proc. Roy. Soc. vol. li. (1892), p. 119, the supra-, infra-, and post-temporal, which subdivide the whole temporal fossa into four foramina. The supratemporal bridge is formed by the squamosal and post-orbital, the latter (j in fig. 12) being continued forwards and fused with the post-frontal. These three bones, with the parietal, enclose the supratemporal foramen. The postorbital joins an ascending branch of the jugal, both together forming the hinder border of the orbit, and this is bordered below chiefly by the maxillary. The posteriortemporal bridge is formed. by the parietal and squamosal, extends laterally over the quadrate and encloses a wide space between itself and the buttress-like transverse expansion of the lateral occipital.

FIG. I I. - Skeleton of Clidastes. (After Cope.) and vol. liv. (1893), p. 160. Ichthyopterygia: `E. Fraas, Die Ichthyosaurier der siiddeutschen Triasand Jura-Ablagerungen (Tubingen, 1891). - J. C. Merriam, " Triassic Ichthyosauria," Mem. Univ. California, vol. i. No. I (1908). - Also technical papers by E. Fraas on fins in Wiirtt. Jahresh. (1894), p. 493, and Foldtani Kozlony, vol. xxviii. (Budapest, 1898), p. 169. Rhynchocephalia: G. A. Boulenger, " On British Remains of Homoeosaurus, with Remarks on the Classification of the Rhynchocephalia," Proc. Zool. Soc. (1891), p. 167. - J. H. McGregor, " The Phytosauria," Mem. Amer. Mus. Nat. Hist. vol. ix. pt. ii. (1906) - E. C. Case, Revision of the Pelycosauria of North America (Carnegie Institution, Washington, 1907). - Technical papers by H. Credner in Zeitschr. deutsch. geol. Ges. vol. xl. (1888), p. 488 (Palaeohatteria), T. H. Huxley in Quart. Journ. Geol. Soc. vol. xliii. (1887), p. 675 (Hyperodapedon), and L. Dollo in Bull. Soc. Belg. Geol. vol. v. (1891), Mem. p. 151 (Champsosaurus). Dinosauria: O. C. Marsh, " The Dinosaurs of North America," Sixteenth Ann. Rep. U.S. Geol. Survey (1896). - Technical papers by L. Dollo in Bull. mus. roy. d'hist. nat. Belg. vols. i. - iii. (1882-84) (Iguanodon), O. C. Marsh in Amer. Journ. Sci. ser. 3, vol. 1. (1895), pl. viii. (restorations), J. B. Hatcher in Mem. Carnegie Museum, vol. i. No. 1 (1901), and W. J. Holland in Mem. Carnegie Museum, vol. ii. No. 6 (1906). Crocodilia: T. H. Huxley, " On Stagonolepis robertsoni, and on the Evolution of the Crocodilia," Quart. Journ. Geol. Soc. vol. xxxi. (1875), p. 423. - E. Koken, " Thoracosaurus macrorhynchus, BI., aus der Tuffkreide von Maastricht," Zeitschr. deutsch. geol. Ges. (1888), P. 754. - E. Fraas, " Thattosuchia," Palaeontogr. vol. xlix. (1902), p. I. - L. Dollo, " Premiere note sur les crocodiliens de Bernissart," Bull. mus. roy. d'hist. nat. Belg. vol. ii. (1883), p. 309. - G. A. Boulenger, Catalogue of the Chelonians, Rhynchocephalians and Crocodiles in the British Museum (London, 1889). Ornithosauria: K. A. von Zittel, " Ueber Flugsaurier aus dem lithographischen Schiefer," Palaeontogr. vol. xxix. (1882), P. 49. - E. T. Newton, " On the Skull, Brain and Auditory Organ of a New Species of Pterosaurian," Phil. Trans. vol. 179E (1888), p. 503 - H. G. Seeley, Dragons of the Air (London, 1901). - Technical papers by O. C. Marsh in Amer. Journ. Sci. ser. 3, vol. xxiii. (1882), p. 251 (wing-membranes), S. W. Williston in Kansas Univ. Quarterly, vol. vi. (1897), p. 35 (restoration of Pteranodon), and G. F. Eaton in Amer. Journ. Sci. ser. 4, vols. xvi., xvii. (1903-4). Squamata: R. Owen, " On the Rank and Affinities of the Reptilian Class of the Mosasauridae, Gervais," Quart. Journ. Geol. Soc. vol. xxxiii. (1877), p. 682, and vol. xxxiv. (1878), P. 748. - G. A. Boulenger, Catalogue of the Lizards in the British Museum, vols. i. - iii. (London, 1885-87); Catalogue of the Snakes in the British Museum, vols. i., ii. (London, 18 93-94). - Technical papers by A. Kornhuber in Abh. k. k. geol. Reichsanst. Wien. vol. v. (1873), No. 4, and vol. xvii. (1893), No. 3 (Dolichosauria), F. Noposa in Beitr. Paleiont. Oesterr.- Ungarns, vol. xxi.(1908), and S. W. Williston in Kansas Univ. Quarterly, vols. i., ii., vi. (1892-1897) (Mosasauria). (A. S. Wo.) III. Anatomy Of Reptiles The Skull. Sphenodon has the most primitive and still most complex skull, the salient features of which it is easy to derive from Stegocephalian and early, generalized reptilian conditions; whilst in other directions, mostly by reduction, the skull of this " living fossil " affords the key to that of all the other groups of at least recent reptiles. The main features are the following. There are, in the temporal region, three complete bony arches, After Gunther. FIG. 12. - Skull of Sphenodon. 1, Ventral aspect; 2, lateral aspect; 3, lateral aspect of mandible. ar. articular; bo, basioccipital; bs, basisphenoid; c, coronoid; ca, columella auris; d, dentary; j, postorbital; m, maxilla; n, nasal; pa, parietal; p1, palatine; pm, premaxilla; pr, prefrontal; ps, postfrontal; pt, pterygoid; q, quadrate in the upper figure, quadrato-jugal in the middle figure; qj, jugal; s, squamosal; sp, splenial; v, vomer.

bone (these "parotic processes " are made up of the lat. occipital, parotic and opisthotic bones); this is the posttemporal foramen. The space enclosed between this occipital buttress, the quadrate and the pterygoidal support of the latter represents the wide and large cavity of the middle ear, 23 not by the angular the middle of the jaw; it is fused with the articular in Geckos, some Tejidae, Amphisbaenidae, and some other burrowing kinds. The splenial is absent in chameleons; near the vanishing point in some of the Agamidae. The coronoid is P ° always present, for the insertion of masseter muscles. In the pleurodont lizards the outer wall of the dentary forms a ledge, against the inner side of which are fixed the teeth with cementum.

The snakes' skull shows many peculiarities, and most of the bones cranial capsule fuse together without sutures. The occipital condyle is triple, the lateral occipitals and the basioccipital taking equal share in its composition; the basioccipital is excluded from the foramen magnum; frequently one common epiphysial pad covers this tripartite condyle. The supraoccipital is likewise excluded from the margin of the foramen magnum by the lateral occipitals. The basisphenoid is prolonged forwards into a long presphenoidal rostrum, on the upper surface of which the trabeculae cranii, which persist as cartilages, extend forwards to blend with the median ethmoidal cartilage. There are no aliand no orbitosphenoids, their places being taken by downward extensions of the frontal bones, which descend to this sphenoidal rostrum and then turn inwards to meet together on the floor of the cranial cavity. There is consequently no interorbital septum. The parietals also descend laterally, but unite with the basisphenoid by suture. On .pr m FIG. 23. - Skull of Python sebae. ar, articular; ca, columella auris; d, dentary; f, frontal; m, maxilla; p, parietal; pm, premaxilla; po, prootic; pr, prefrontal; ps, postfrontal; pt, pterygoid; q, quadrate; s, squamosal; t, transversum; tb, turbinal.

the base of the skull we note various processes for the insertion of ventral cervicooccipital muscles, much used during the act of vigorous striking. Boidae have a long sphenoidal ridge and thick basipterygoid processes; others have one or more median knobs or crests, and the Viperidae have a very prominent and large ridge. The parietals fuse together into an unpaired mass whence arises mostly a strong median crest which projects a little beyond the occiput; there is no parietal or pineal foramen. There are paired frontals, postfrontals, prefrontals and nasals; the latter are said to coossify in Charina only. The position of the prefrontals is variable. In the boas, for instance, they meet, separating the nasals from the frontals; they are in contact with the nasals in the boas, burrowing snakes and in Xenopeltis, but more or less widely separated from them, and often from each other, in the Colubridae and Viperidae. The premaxillary is single; and only in Glauconiidae connected with the maxillaries; in the others it is but loosely connected with the ethmoidal end of the skull, for instance, with the turbinals, which are osseous and well developed in pythons.

The whole appendicular apparatus is most loosely attached to the skull, at least in the typical snakes, and since they do not chew their prey but only hook it in, so to speak, during the act of swallowing, the whole apparatus is as movable as possible.

The whole palatal apparatus shows many modifications, but the maxillaries, palatines and pterygoids always remain widely asunder, and from the mid-line. Some of the modifications, so far as they are used for taxonomic purposes, are mentioned in the article Snakes: Classification. In the majority of snakes the maxillaries form the borders of the mouth, and they are but loosely attached to the other bones, to their palatine processes, to the palatines, and with their posterior ends, by the ectopterygoids to the pterygoids. In the Viperidae the maxillaries are much shortened and articulate extensively with the prefrontals; they can be erected, or rather pushed forwards, by the ectopterygoids (see Snakes); they are not connected with the palatines. The pterygoids diverge posteriorly and articulate loosely with the quadrates; in the original condition the articulation is near the distal end of the quadrate, e.g. in Boidae, and the pterygoids may form an additional attachment with the mandibles; in the Viperidae the pterygoids are somewhat shortened and are attached to about the middle of the quadrate shafts; in the Amblycephalidae they are still shorter and do not reach these bones. The ectopterygoids are lost by the burrowing Typhlopidae and Glauconiidae. The quadrate is always extremely movable; besides being in a most curious way connected with the outer end of the columellar rod (see below, Ear), it is suspended from the skull by the squamosal. The squamoso-quadrate connexion is very loose; that of the squamosal with the skull varies much. In the majority of snakes it slides quite freely upon the parietal; it is much longer than the quadrate in the boas, much shorter than the elongated and slender quadrate in most of the poisonous snakes. Lastly, in most of the ancient burrowing snakes, e.g. Typhlops, Glauconia, Ilysia and Uropeltis, the squamosal has worked its way into the cranial wall so that the quadrate, itself also much shortened, rests directly upon the cranium.

The Vertebral Column. The vertebrae of all reptiles are gastrocentrous, that is to say, the centra or bodies of the vertebrae are formed by the originally paired, interventral cartilages, while the basiventrals are reduced, persisting either as so-called intercentra or wedge-bones, or as intervertebral pads, or disappearing altogether; the basidorsal elements form the neural arch. At the earlier stages of development the gastrocentrous vertebrae behave in the same way as in the Urodela, except that the interdorsal pair of elements is suppressed from the beginning (the very elements which in is always formed by the articular bone, which lies on the ventral side, about FIG. 22. - Skull of Monopeltis sphenorhynchus. I, dorsal aspect; 2, ventral aspect; 3, lateral aspect; 4, posterior aspect. ar articular; bs, basisphenoid; d, dentary; f, frontal; in, maxilla; n, nasal; oc, oc, occipital condyles; of, occipital foramen; pal, palatine; pa, parietal; pm, premaxilla; ptg, pterygoid; q, quadrate; so, supraoccipital; sq, squamosal; v, vomer.

of the d FIG. 24. - Skull of Vipera nasicornis. ar, articular; ca, columella auris; d, dentary; f, frontal; m, maxilla; pf, poison fang; pm, premaxilla; pr, prefrontal; ps, postfrontal; pt, pterygoid; q, quadrate; s, squamosal; transversum or ectopterygoid.

Stegocephali and most Anura form the centre), therefore the typical batrachian vertebrae are notocentrous. If the remaining three pairs of constituent elements of each vertebra (the neural arch, the centrum and the intercentra) remain separate, the vertebrae are called temnospondylous (Attvco, I cut, virovluXos, a vertebra). If the neural arches and the centra are suturally united, or are fused with each other, the vertebrae are called stereospondylous (6rcp€6c, solid). In many fossil reptiles most or many of the vertebrae are temnospondylous; in most of the recent. Amniota l they are consolidated, but the atlas or first vertebra remains usually in a relatively primitive condition, and is temnospondylous but for the usual modification that its centrum becomes attached to that of the second vertebra and forms its odontoid process. The composition of gastrocentrous vertebrae is best illustrated by the first and second cervical vertebrae of crocodiles, whence by reduction and fusion the structure of every other vertebra can be explained. We have only to add that the ribs are genetically derived from lateral outgrowths of the basiventral elements, whilst the chevron bones are mere ventral outgrowths from the same basal cartilages. The most primitive vertebral column is that of the Geckos. The I.D. ' 't .V.

7 / 8'/ 9 10 FIG. 25. - Composition of Vertebrae of Reptiles. In all the figures the right side looks towards the head.

1. Diagram showing the relative position of the four pairs of arcualia which constitute a complete quadripartite vertebra. B.D., Basidorsal; B.V., basiventral; I.D., interdorsal; I.V., interventral, shaded vertically in all figures; N., position of axil of the spinal nerve, i.e. behind the neural arch of its vertebra. 2, 3. Side views of the constituent cartilaginous blocks of a caudal vertebra (2) and a trunk vertebra (3) of Archegosaurus, as typical examples of temnospondylous quadripartite and tripartite vertebrae. For comparison with Reptilian vertebrae. 4. Temnospondylous tripartite vertebra of the trunk of Eryops, a Permian reptile. 5. Composition of the second vertebra of a crocodile. 6. A vertebra of which the vasiventrals are reduced to an " interventrum." 7. Side view of the first and second: cervical vertebra of a crocodile. 8. The same analysed. NI, N2 and N3, position of the first, second and third spinal nerves; S.D., occasionally called Proatlas, the detached spinous process, or supradorsal, of the atlas or first vertebra. 9. The first three vertebrae of Sphenodon. Io. The complete atlas vertebra of an adult Trionyx, still typically temnospondylous.

vertebra consists chiefly of a large neural arch which rests broadly upon the centrum; this is a tube, more or less calcified and ossified, with a narrow waist in the middle, widening headand tailwards. The tube is hollow, the chorda dorsalis passing through the whole column, and there are no proper joints between the centra, which are amphicoelous. Between the centra lies a separate element, the so-called intercentrum, which is ring-shaped and acts as an interarticular pad instead of a joint. The first of these rings forms the ventral half of the atlas ring; the second is attached to the cranial surface of the second centrum, and produces, like some of the next following ones, a vertical median blade of bone, a true hypapophysis. Such intercentra exist throughout the length of the vertebral column; in the tail they are enlarged and carry a pair of chevrons, which are cartilaginous and have the tendency of fusing by superficial 1 There remained a flaw in the correctness of the view that the bodies of the amniotic vertebrae are formed by the paired interventral pieces, since the bodies were known always to appear from the first as unpaired, cartilaginous masses, until G. B. Howes found them to consist of a right and left pair in the embryos of Sphenodon. ossification on to the caudal ends of the centrum next in front, to which they do not belong genetically. Exactly in the middle of each vertebra the thin shell of the centrum forms a cartilaginous septum, of what is often wrongly called chordal cartilage. When this septum is complete, and this seems to be the normal condition in the tail, the chorda is here rent asunder, otherwise it is only constricted. This septum is but slightly invaded by ossification, and consists of large cells which retain the appearance of young or embryonic cartilage. It coincides exactly with the line of transverse division of most of the caudal vertebrae into an anterior and a posterior half, the division gradually extending right through the bone of the neural arch. The same kind of division, and from the same causes, exists in Sphenodon and in many lizards, in fact in all those reptiles which can reproduce their broken-off tail. It is from the septal cartilage that the regeneration starts 2 (fig. 26).

Sphenodon also has biconcave vertebrae owing to the persistence of the chorda dorsalis in the intervertebral region; otherwise the vertebrae are solid. Intercentra occur from the atlas regularly into the tail, where they carry chevron bones. The atlas-ring (fig. 25, 9) is composed of the first intercentrum and a pair of neural arches which remain quite separate and carry on the dorsal side a pair of ossicles, the disconnected supradorsal elements of the atlas, erroneously supposed to be the remnants of the " proatlas." Crocodiles. - Remnants of the chorda persist in the middle of the centra, which, in recent species, are mostly procoelous, and with a convex knob behind, but the first caudal is strongly biconvex. Cartilaginous intercentral rings, pads or menisci, occur throughout the column; in the tail they carry chevrons. For the instructive detail of the composition of the first and second cervical vertebrae see fig. 25, 7 and 8. Some of the posterior neck and anterior thoracic vertebrae have an unpaired hypapophysis arising from the centrum. The vertebrae have the usual processes, viz. spinous process, a pair of anterior and posterior zygapophyses arising from the neural arch, diapophyses likewise from this arch for the articulation with the tubercular portion of the rib; short parapophyses from the centra for the capitular ends of the ribs; the transverse processes of the 12th vertebra, and following, carry the whole rib, and are like the processes of the lumbar vertebrae diapapophyses; the so-called transverse processes of the tail are mainly the anchylosed or fused ribs themselves.

Chelonians

The vertebrae are sometimes in the various regions of the same column opistho-pro-or amphicoelous, or even biconvex. Intercentra occur regularly on the first two or three cervicals, and on the tail as paired or unpaired nodules, or as chevrons, which articulate mostly with the previous centra and occasionally fuse with them. Intercentral, fibrocartilaginous disks occur regularly, mostly in the shape of rings; the first is the transverse ligament of the atlas-ring. In the Trionychidae (fig. 25, io), but also in some other tortoises, the various pieces of the atlas do not anchylose, and the first centrum remains also movably attached to the second, although it sometimes carries, 2 Regeneration of the tail can take place in Sphenodon, all Geckos, Anguidae, Gerrhosauridae, Lacertidae, most Scincidae, and in many Tejidae and Iguanidae; certainly not in chameleons, Varanus, Agamidae, snakes, crocodiles and tortoises. Often the tail is so brittle and the muscular cones are so loosely connected that part can be thrown off by the muscular exertion of the creature itself. The reproduced tail is, however, only a sham tail, since neither centra nor arches, but only a non-segmented rod or tube of fibrocartilage is produced. It is, however, invested with new muscles and with skin, but the scales often differ considerably from those of the normal organ, sometimes showing reversion to an ancestral form. For further detail see G. A. Boulenger, P.Z.S. (1888), p. 351, and (1891), p. 466.

S.D. Chorda.

4 FIG. 26. - Vertical section of four (7th to loth) caudal vertebrae of Sphenodon. a, line passing through the middle of centrum and through part of the neural arch, where the vertebrae break off. (After Gunther.) -Jd'en% form a bundle opposed to the rest; the fibulare and tibiale are fused into one bone; the fused fifth and fourth distal tarsals form a very large half-globular piece for the three outer toes, whilst the second toe is carried by the third distal tarsal, besides which there are three more small cartilages, one of which may be the displaced second tarsal or the still independent central. The tarsus of Sphenodon is like that of typical lizards, but none of its distal tarsals are fused on to metatarsals. The Crocodilian foot marks an advance. The astragalus is large, articulating well with tibia and fibula, and against the fibulare, which forms a typical, heel-shaped calcaneum. The fifth and fourth distal tarsals carry the fourth toe and the hook-shaped fifth metatarsal to which the fifth toe is reduced. The third, second and first distal tarsalia scarcely contain osseous nodules; they form together a wedge-shaped cartilaginous pad between the astragalus and the first and second toes. This attachment of the distal tarsals to the metatarsals reminds us of the Lacertilian condition, the result in either case being a still more marked intertarsal joint in addition to the cruro-tarsal.

Most well-footed reptiles retain all the five toes; only the crocodiles and a few tortoises have lost all the phalanges of the fifth toe. The phalangeal numbers are in the Lacertilia 2, 3, 4, 5 and 3 in the fifth toe; in chameleons 2, 3, 4, 4, 3; in most tortoises 2, 3, 3, 3, 2; but in Homopus, Pyxis and Cinixys 2, 2, 2, 2, o; in the crocodiles 2, 3, 4, 4, o. The embryos of crocodiles are said to be hyperphalangeal; i.e. as many as 7 phalanges on the fourth; 5 or 6 on the fifth finger; 6 on the fourth toe, and there are traces of the fifth toe. In the adult the fourth toe remains without a claw. Burrowing and living in sand, or humus, is in many lizards correlated with reduction of the limbs and their girdles. The vestiges of the hind limbs come to lie as near the vent as possible. The reduction occurs in various families, independently.

In most cases the fore 3 limbs disappear first, but ` p 2 in the Amphisbaenidae, cf. Chirotes, and in the Tejidae, the reverse takes place. Whilst degeneracy of the shoulder-girdle is delayed long after the loss of the anterior limbs, that of the pelvic arch precedes the loss of the hind limbs. Cope has drawn up a tabular statistic of p the loss of digits, limbs and, 1 their girdles on pp. 202-3 of his work, Crocodiles, Lizards and Snakes of North America (Washington, 190o). The 2 i L peculiar hind limbs of the Dibamidae are described in the article Lizard.

The majority of snakes have lost all traces of the limbs and their girdles, except the so-called Peropoda (see Snakes: Classification). The vestiges of a Boa and of a Glauconia are shown in fig. 35.

Tegulnentary System. The skin of reptiles is characterized by the strong development of its horny stratum; on the outside of it exists a thin cuticular or epitrichial layer. An important feature in most lizards and in the snakes is the existence of a " subepirdemoidal " or transitional layer which is produced by the migration of ectodermal cells into the cutis. The immigration takes place during' the embryonic development, observed first by Kerschner, who, however, misinterpreted the process. Pigment cells, black chromatophores also, make their first appearance in the epiderm and then migrate into the transitional stratum, as has been first correctly stated by F. Maurer. The horny stratum is shed periodically, several times during the year, and as one entire piece in snakes and a few lizards, e.g. Anguidae; in most lizards, chameleons, geckos and in Sphenodon the thin, transparent colourless layer comes off in flakes. In crocodiles it is not shed except for the usual wear and tear, nor in tortoises, although in some e.g. Chrysemys, a periodical peeling of the large shields has been observed.

In all reptiles the cutis is raised into papillae, or folds. When the papillae are small the skin appears granular; when they are large, flat, mostly imbricating, they form scales; when they are very broad-based and still larger, they are called scutes or shields. The overlying epidermal covering partakes of these elevations, often e.g. in many snakes, with a very fine system of ridges of its own. Such a scale, cutis and horny sheath, may form spikes, or crests. They all have only basal growth. Thus, for instance, a shield of a tortoise-shell is a much flattened scale, or cone, with the apex more or less in the centre, surrounded by marginal ridges which indicate the continuous additional growth at the base. The central " areola " represents in fact the size of the shield at the time of hatching.

Of very common occurrence is the development of bone in the cutaneous portion of the scales; such osteoderms occur in many lizards, very strongly developed in the scutes of the crocodiles, especially on the back; they also occur in the skin of tortoises especially on their legs and on the tail, and they probably constitute the peculiar shell of Sphargis, the leathery turtle (see Tortoise). Sphenodon and chameleons are devoid of such osteoderms, in geckos they are likewise absent, but calcifications occur in their tubercular skin. A similar process seems to have produced the egg-tooth of crocodiles and tortoises (see under Teeth below). Calcareous deposits, or at least deposits of guanine and more commonly of carbonate of lime, play a considerable role in the skin of lizards and snakes. These waste products of the metabolism are always deposited within cells, and a favourite place is the subepidermal layer. In combination with su p erimposed yellow or red pigment, and with the black chromatophores as a foil, partial or complete screen to the light, as the case may be, these mineral deposists are to a great extent answerable for the colours and their often marvellous changes in the skin (see Chameleon).

Peculiar pits in the scales of snakes and crocodiles are described under Sense-Organs below.

The skin of reptiles is very poor in glands, but the few which exist are well developed. Crocodiles possess a pair of glandular musk bags which open by rather large slits on the under jaw, against the inner side of the jaw. Another pair of musk glands are the anal glands. During great excitement all these glands can be everted by the crocodiles. Sphenodon and snakes have only the anal pair. Water tortoises have inguinal glands, which secrete a strongly scented fluid, opening near the posterior rim of the bridge. Trionyx has additional glands opening near the anterior part of the plastron. Peculiar glandular structures are the femoral pores of many lizards. They lie in a line from the inner side of the knee to the anterior margin of the anal region, to which they are restricted in the limbless Amphisbaenidae. Each pore leads into a subcutaneous pocket, sometimes with slightly acinous side chambers, the walls of which produce a smeary, yellowish matter consisting chiefly of the debris of disintegrated cells which dries or hardens on the surface in the shape of a little projecting rod. They occur in both sexes, but are most active in males during the pairing season. Their use is unknown. It would be far-fetched to liken them to forerunners of the sebaceous portions of milk glands, although not so imaginary as to see in them and in the sensory pits of snake scales the forerunners of the mammalian hairs!

Claws, scarcely indicated in Batrachia, are fully developed in all limbed reptiles. The base is sunk into the skin like our own finger nails; the dorsal and ventral halves are differentiated into a harder, more curved dorsal sheath-like portion, and into the beginning of a sole, especially in crocodiles and in blunt-toed tortoises. The first claw to be reduced is that of FIG. 34. - Vestiges of pelvic limb - I, of Lialis bartonii; 2, of Anguis fragilis; 3, of Amphisbaena fuliginosa. f, femur; il, ilium; ip, iliopectineum; p, pubis; t, tibia.

P. FIG. 35. - I, Vestigial pelvis and limb of Glauconia macrolepis. 2, The same parts of Boa (after Fiirbringer). f, femur; il, ilium; ip, bone called " iliopectineum " by Fiirbringer; p, pubis; t, tibia.

the fifth digit. The claws of many geckos are " retractile," like those of cats; the adhesive lamellae on the under side of their digits have already been described (see Gecko).

Nervous System. The hemispheres are still much longer than broad, and pass, especially in lizards, gradually into the olfactory lobes, into which continue the ventricles of the hemispheres. The dorsal walls of these are thin, especially in crocodiles, although they possess already a considerable amount of grey matter. The basal masses of the fore-brain bulge into the roomy ventricles like cushions. Fibres referable to a corpus callosum are scarcely separated from those of the still much stronger anterior commissure. The epiphysis comes to the surface between the hinder parts of the hemispheres. The pineal eye is described below under Sense Organs. The hypophysis has but a shallow infundibulum. The mid-brain shows a pair of dorsal globular swellings, each with a cavity; they separate the hemispheres from the cerebellum. Of the hindbrain, the middle portion is by far the largest; although the dorsal wall of this cerebellum is thick, and rich in grey matter, its surface is still quite smooth and it shows no trace of an arbor vitae. It covers but a small portion of the wide fourth ventricle.

The spinal cord shows a brachial and a lumbar longitudinal swelling, especially marked in tortoises, but without a rhomboidal sinus. The cord is continued into the end of the tail.

The cranial nerves of the reptiles agree in their arrangement and distribution more with those of birds and mammals than with those of the Batrachia. The facial nerve sends a palatine branch to the palate and to the superior maxillary of the trigeminus, and a strong mandibular branch joins the third of the trigeminal, and further ramifications supply the sphincter muscle of the neck. The vagus and glossopharyngeus leave the cranium separately. The vagus then goes towards the heart, which in the cb opt P ch ° Sauropsida is far re moved from the head, and there possesses another ganglion, vari- P. ously called ganglion o J t trunci vagi or g. nodosum. It is connected by a nerve with the large ganglion supremum of the sympathetic. From the cardiac ganglion, and from the continuation of the vagus, are sent off several branches in succession, which, having to pass below or tailwards from the transverse carotic, aortic and Botallian vessels, have to take again a headward course to the larynx and pharynx; a side branch enters the heart by its truncus. The main mass of the vagus then supplies lungs, stomach and further viscera. The accessory or II th cranial nerve arises with about half a dozen roots which extend often beyond the second cranial nerve; they collect into a thin stem which leaves the cranium together with the vagus, with which it is often fused; it supplies the cucullaris s. trarepius muscle.

The hypoglossus arises by two ventral roots, leaving the skull by two holes through the lateral occipital bone, near the condyle. The united stem is invariably joined by strong branches from cervical nerves, always from the first, mostly also from the second, sometimes also from the third. The details vary much; occasionally there are three cranial roots and foramina, and then only the first cervical joins the hypoglossus; this often fuses with the glossopharyngeal or with the vagus. In the broad and well-muscularized tongue of the crocodiles the right and left hypoglossal branches form a complete ansa, an arrangement in which A. Schneider saw the infraoesophageal nerve ring of Invertebrata!

The spinal nerves each issue behind, or through, the neural arch of the vertebra to which they belong genetically. The first spinal, or suboccipital, nerve has no dorsal roots, and, having lost its vertebra, an apparently anomalous arrangement has come to pass, in this way, that there are x cervical vertebrae, but x cervical nerves, a condition prevailing in, and characteristic of, all Amniota. The hypoglossal-cervical plexus is separated from the brachial plexus by several metameres, according to the length of the neck. The brachial plexus is composed of about 5 nerves; the variations have been studied chiefly by M. Fiirbringer. It is interesting to note that the brachial plexus still persists in snakes, although they have completely lost the anterior girdle and the limbs (Albertina Carlsson). A disturbance in the pelvic region likewise indicates in snakes the former existence of a pelvic or lumbo-sacral plexus, which in limbed reptiles is composed of about 5 nerves, the last of which is weak and in many cases (by no means the rule) issues between the two sacral vertebrae, sending one branch to the ischiadic, another to the public plexus which supplies the cloacal region. (For details of these plexuses see the papers by Mivart, Jhering and Gadow.) The sympathetic system shows considerable modifications in the various orders and even families of the reptiles. In the neck region, in Sphenodon and most lizards it is, on the right and left side, composed of two portions. One, more lateral and placed deeply, runs along the side of the vertebral column, starting from the first and second spinal nerves, with which it is connected by so-called rami communicantes; it is not connected with the other spinal nerves until it reaches, in the thorax, the first stem of the brachial plexus, and hereabout lies the so-called second thoracic ganglion. The other, superficial and more ventral, portion arises from the petrosal ganglion of the glossopharyngeal, and from the vagus ganglion, and then forms a long loop which joins the second thoracic ganglion. In its long course it sometimes, e.g. in Varanus, forms one common stem with the vagus before it splits off. At a variable distance, but not far above the heart, the vagus possesses a big swelling, the ganglion trunci vagi, and the sympathetic stem, in the same level, or farther down, has likewise a large ganglion, the g. supremum vagi, or first thoracic ganglion. The vagus ganglion receives several nerve strands from this big sympathetic ganglion, and then divides as described above.

In the crocodiles the deep portion of the sympathetic begins at the vagus and extends in rope-ladder fashion into the thorax, there being, as in birds, regular transverse communicating branches with the spinal nerves, and the longitudinal strands run through the transverse foramina between the capitular and tubercular portions of the cervical ribs. The other, ventral, portion starts by a right and a left branch from the vagus ganglia, but both branches unite at once into one unpaired stem, which is deeply embedded in the middle line between the ventral muscles of the cervical vertebrae. Very thin branches connect this unpaired stem with the right and left sympathetic portions; small ganglia are embedded in the unpaired nerve.

The so-called second thoracic ganglion is in reality a compound of all the sympathetic ganglia of the four or five metameres of the brachial plexus. It forms the point of juncture of the deep and the superficial cervical sympathetic portions. From the posterior region of the thorax backwards the right and left strands run along their side of the vertebral column, with a communicating branch and a ganglion for each metamere; sometimes one or more successive ganglia are combined, for instance near the cloaca. After having supplied the latter, the sympathetic system appears exhausted and is continued into the tail by but a very thin strand, which runs between the caudal vein and artery. The best illustrations of the sympathetic system are those by Vogt (neck of crocodile), J. G. Fischer (many) N Cjav et. t -6 ch py FIG. 36. - Brain of Lacerta agilis. (After Leydig.) I, Dorsal aspect; 2, vertical longitudinal section. cb, cerebellum; ch, cerebral hemisphere; m, medulla oblongata; olf, olfactory lobes; on, optic nerve; opt, optic lobes; p, pineal body or epiphysis; py, base of pituitary body.

vulgaris, 4 or 5 in Anolis, to 1-3 in some other iguanids, skinks and geckos. Uroplates fimbriatus has 14, and the last four pairs are separated from the dorsal portions of their ribs; similar discontinuity occurs in geckos, the median portions bearing a striking. although not fundamental, resemblance to parasternal ribs.

In the lizards with much reduced fore limbs, the sternum loses its connexion with the ribs from behind forwards; two sternal ribs existing in the Tejid Ophiodes and in the Scincoid Acontias, one only in Pygopus, none in Ophisaurus s. Pseudopus and Anguis (in the latter one rib is still connected in the embryo). The sternum is likewise quite free in Chirotes in spite of its functional limbs; the sternum is still a large plate, with a window, and ending in two long, xiphoid processes.

Lastly, the sternum has vanished without a trace, as in the snakes, in some species of Acontias, in the Anelytropidae, Dibamus and Aniella (Furbringer). In the limbless genera of Amphisbaenidae the sternum is very much reduced; in Trogonophis alone it is still represented by a narrow transverse bar connecting the ossicular vestiges of the shoulder-girdle; in the other genera the sternum has shrunk to a pair of nodules or to a single nodule.

The pectoral or shoulder-girdle in its completest condition consists of a right and left scapula, coracoid, precoracoid and clavicles, and an unpaired interclavicle or episternum. The dorsal portion of the scapula remains cartilaginous, with or without calcification, and is usually distinguished as suprascapula. The ventral portion of the precoracoidal and coracoidal mass remains likewise more or less cartilaginous, rather unnecessarily distinguished as epicoracoid. Ossification begins near the glenoid cavity and thence spreads, eventually with the formation of a dorsal and a ventral centre. The resulting suture separates the dorsal or scapular from the ventral or coraco-precoracoidal mass. A kind of landmark, not always reliable, between coracoid and precoracoid is the exit of the supra-coracoidal nerve. The ventral margins of the coracoids articulate in tenon and mortice fashion with the antero-lateral margins of the sternum. The interclavicle, usually T-shaped, is a dermal bone and rests upon the ventral side of the girdle. The paired clavicles, sometimes fused together, rest upon the anterior end of the interclavicle and extend transversely to the acromial process of the scapula; the detail of the attachments varies much.

The girdle is most complete in Sphenodon and in Lacertilia. In Sphenodon the coracoid forms one continuous mass with the precoracoid, without further differentiation; the clavicles are fused with the interclavicle into one T-shaped mass, the cross-arms of which are attached to the acromia by ligaments. In the lizards (except Heloderma) the much-broadened central and anterior halves of the girdle are fenestrated; the windows, always closed by membranes, are bordered by bony processes, distally by unossified cartilage. The first window to appear, or the most constant, lies between the coracoid and its precoracoid; in Anguis it is the only window, in this case not a primary feature. In other lizards, e.g. Uromastix, a second window occurs between precoracoid and scapula, and even a third window can appear in the scapula itself, causing in many Iguanidae, e.g. Amblyrhynchus (see fig. 33, ms.), the socalled mesoscapula; an analogous window within the coracoid produces the mesocoracoid; unnecessary distinctions of little morphological value considering the great variability of these fenestrations in closely allied genera.

The chameleons have lost the clavicles and the interclavicle, and the scapula, which is very slender and long, is devoid of an acromial process. The coracoid forms one mass with the precoracoid, through the middle of which `passes the supracoracoidal nerve; the coracoids articulate by their whole bases with the sternum.

Geckos possess a complete shoulder-girdle; the ventral portion shows, e.g. Hemidactylus, three pairs of windows; only FIG. 33. - Sternum and Shoulder-Girdle of Amblyrhynchus subcristalus (after Steindachner). cl, clavicle; co, coracoid; h, humerus; ic, interclavicle; mc, mesocoracoid; ms, mesoscapula; pc, precoracoid; s, scapula; st, sternum.

one in Uroplates. In the latter the interclavicle is much reduced; the clavicles meet each other and are slender rods. In the Geckoninae and Eublepharinae the ventral halves of the clavicles are dilated and possess each a foramen; the interclavicle is cross-shaped.

In the more or less limbless genera of lizards the shouldergirdle is much reduced. In Chirotes, which still has functional fore limbs, the clavicles and the interclavicle are absent, the coracoids are not divided from the precoracoids; in the limbless Amphisbaenidae the girdle is reduced to a pair of cylindrical ossicles in Amphisbaena, Blanus and Trogonophis; no vestiges exist in Rhineura, Lepidosternon and Anops. Foramina in the broadened clavicles occur also in various Lacertae, for instance in the Iguanid Laemanctus, in the Scincoid Trachysaurus, in Plestiodon, Zonosaurus and in Lacerta simonyi, but not in L. agilis. In Mabuia the median portions are especially broad and show each two foramina. Their presence can be of but very doubtful taxonomic value.

The girdle of the Crocodiles is considerably simplified. Scapula and coracoidae, movably united, at least in younger specimens. The precoracoid is slightly indicated by a process of the coracoid, which is perforated by the supra-coracoidal nerve near the glenoid cavity. Clavicles are absent. The interclavicle is reduced to a long, flat splint-bone, which is firmly fused on to the sternal cartilage. The Chelonian shouldergirdle shows several very remarkable modifications. Instead of lying outside the trunk, it has been transferred into the cavity of the trunk, the carapace with the 'ribs covering it from the outside. An explanation of the changes implied in this transposition is still extant. Chelonians are, moreover, the only reptiles besides Pterosauria in which the scapula is attached to the skeleton of the trunk. The scapulae stand in a more or less vertical position, and their dorsal end rests against the inside of the nuchal plate, where this is sutured to the first neural and the first costal plate, a little in front of and sidewards from the first short rib. From near its ventral end the scapula sends off a long process, which converges transversely with its fellow. This process, the clavicle(!) or the precoracoid of many authors, is the acromial process, the Plesiosauri giving the clue as to how an acromion can assume such an abnormal position. The coracoid, with a suture between it and the scapula, is very long and extends horizontally backwards, not meeting that of the other side. The sternum being i 2 FIG. 32. - Rudiments of pectoral arch - I, of Acontias meleagris; 2, of Typhlosaurus aurantiacus (after Furbringer).

absent, and clavicles and interclavicles forming the epi-and endo-plastral elements of the plastron, the shoulder-girdle is nowhere in contact with the skeleton except at its dorsal end.

The Fore Limbs

The humerus has near its upper end a median process, and at a variable distance a lateral process, near which is the biceps-fossa. Above the radial or outer condyle exists a foramen for the passage of the radial nerve in Sphenodon, in the Lacertilia, and in many Chelonians, e.g. Cholone and Sphargis; such an ectepicondylar foramen is absent in crocodiles. Above the ulnar condyle exists, but only in Sphenodon, the entepicondylar foramen, for the passage of the nervus medianus and brachial vessels. Thus Sphenodon alone possesses both foramina, the crocodiles neither.

Ulna and radius always remain distinct; the former is generally the stouter although not always the larger bone. The carpus may contain as many as 12 separate elements: ulnare, intermedium, radiale, 2 centralia, a pisiform on the ulnar and a small nodule in a corresponding position on the medial side, and 5 distal carpals. In Sphenodon the centralia are sometimes fused into one, and the radial nodule is absent; the numbers of phalanges are, 2, 3, 4, 4 and 3 proceeding from the first to the fifth finger. The carpus of the Chelonia is likewise primitive, with various unimportant reductions; Chelydra possesses one or two centralia, whilst pisiform and extra radial are absent; both these bones are present in Emys, but the centrale fuses with the radial carpal, and the fourth and fifth distal carpal are fused together. In Testudo the pisiform is small; intermedium, centrale and radiale are represented by one bone only, and the first, second and third distal carpals are fused, whilst the two remaining are free. In the marine turtles the fore limbs are transformed into paddles; the ulna is considerably shorter than the radius; all the normal nine carpal elements remain distinct; the pisiform is much enlarged, helping to increase the paddling surface, and it has moved from the ulnar carpal to the side of the fifth distal carpal. The three middle fingers and toes have mostly 3 phalanges; the pollex and hallux have always 2; the number of phalanges of the fifth finger varies from 3 to I, of the fifth toe from 2 to o. The greatest reduction occurs in Testudo and its allied genera of typical land-tortoises, Homo pus, Pyxis and Cinixys, the formula for the fingers being 2, 2, 2, 2, 2 or r, and 2, 2, 2, 2, o for the toes. In Pelomedusa all the fingers possess 2 free phalanges only, owing to fusion of the first and second phalanges with each other.

Considerable advance is marked by the Crocodiles. The intermedium and centrale are lost, the pisiform is small, ulnar and radiale are considerably elongated and enlarged. Of the distal carpals the two last are fused into one bone, and the three first, together with the central, are transformed into a pad-like cartilaginous and ligamentous piece between the large radial and the first and second finger, to which the pad is firmly attached. The other fingers articulate with the " humatum." The result of the whole arrangement is the formation of two main joints, one between fore arm and carpus, the other intercarpal. The number of phalanges is 2, 3, 4, 4, 3.

The conditions prevailing in Lacertilia are connected with those of Sphenodon. The intermedium is lost, the other normal carpalia are present, also the pisiform; the first distal carpal is much reduced and the correspondingly enlarged radial carpal comes into articulating contact with the first metacarpal. The numbers of phalanges are 2, 3, 4, 4, and 2 or 3 for the fifth finger. The hand of the chameleons is most modified; the first three fingers form an inner bundle opposed to the outer or fourth and fifth fingers; in correlation herewith the third and fourth distal carpals are fused into one rather large mass; the other elements remain free, and A. Stecker has found a small intermedium present in the young, in a position which indicates that its subsequent absence is due to loss, not fusion with neighbouring elements.

The Pelvic Girdle

The ilium is attached to the vertebral column by means of the two sacral_ribs.' The ischia and the 1 In all reptiles, except a few fossil groups, the ilio-sacral connexion is post-acetabular, i.e. it lies in a transverse plane tailwards from pubic bones join the ilium at the acetabulum, which is not perforated, except in crocodiles. The ischia and pubes invariably form symphyses at their ventral ends, except the so-called pubes of the crocodiles, and these two symphyses are further continuous with each other, dividing the pubo-ischiadic space into a right and left foramen obturatum of very variable size. They are small and round in Testudo, divided by a broad, bony bridge, larger in Chelone, separated by a chiefly ligamentous, partly cartilaginous string; largest they are in Sphenodon and in the Lacertilia. Frequently the symphysial portion at the anterior end of the pubic symphysis remains cartilaginous, unpaired, e.g. in most Chelonians and Lacertilians, comparable with the epipubis of Urodela. A corresponding cartilage, the os cloacae or hypoischium, is continued backwards, from the ischiadic symphysis towards the vent, serving for the attachment of sphincter muscles; it occurs in many lizards and tortoises. In the Chelonians the pubic bones are generally much stronger than the ischia, and they send out each a strong lateral pubic process, directed forwards and outwards; the obturator nerve passes through the wide obturator foramen. In the pleurodirous tortoises the ends of the ilia and those of the lateral processes of the pubes are much broadened and firmly anchylosed with the posterior costal plates and with the xiphiplastron respectively. The whole pelvis, like the shoulder-girdle, lies inside the body. The pelvis of Sphenodon is essentially like that of the Lacertilia. The pubes are slender; they send out a pair of lateral processes, near the base of which the obturator nerve pierces the shaft of its pubis. This lateral process is the homologue of the long, slender pubis of birds. The chameleons' pelvis is peculiar. The pubes are devoid of lateral processes, but from their anterior end arises a pair of small cartilages, in a transverse direction; their ends are connected by ligament with the median anterior portion of the ischiadic symphysis. The crocodilian pelvis is very aberrant. The ilium is broad and sends two processes to the acetabulum, which retains a foramen; the posterior process articulates movably with the ischium; the preacetabular process fuses in very young specimens with a separate, ossifying, cartilaginous piece, which then forms a rough joint with the anterior portion or process of the ischium, which closes the acetabulum on its ventral side. To this anterior ischiadic process is attached the freely-movable, clubshaped bone, generally called pubis. The homologies of these club-shaped bones and of the small bone mentioned above are not clear. The club-shaped bones remain asunder; the ischia form a long and firm symphysis. The obturator nerve passes out of the pelvis between the ischium and the club-shaped bone, close to the posterior margin of the latter.

The posterior limbs show essentially the same composition as the fore limbs, but the modifications in the various reptilian orders are much greater. The femur has generally a wellmarked neck. Fibula and tibia remain distinct; the former usually shows a reduction in thickness. In the tarsus we observe never more than two proximal tarsal elements, a reduction due either to the suppression of the intermedium or to its enlargement and concomitant loss of the tibial element. The least-modified foot-skeleton is that of the Chelydridae, the lowest Chelonians. The proximal row is composed of a fibulare, and a much larger piece articulates with both tibia and fibula, the " astragalus"; the centrale is present; the first three distal tarsals remain separate, each carrying a toe. The fused fourth and fifth tarsals carry the fourth toe, and, laterally attached, the hook-shaped fifth metatarsal. Chelone shows the same arrangement, except that the centrale is fused with the astragalus; in Testudo, Emys, the fibulare, astragalus and centrale are fused into one broad mass, with the result of forming a crurotarsal and an intertarsal joint. The same arrangement reached by the Testudinidae is universal in the Lacertae, with the further modification that the three first distal tarsals fuse on to the proximal ends of their respective metatarsals. Most aberrant is the tarsus of Chameleons, in which the first and second toe one passing through the acetabulum. In birds it is likewise postin mammals pre-acetabular.

LNA10.4eAt-.(Tif. and fuses with, the second intercentral piece. The entire atlas remains in a primitive, typically temnospondylous condition. On the other hand, in some Pleurodira, e.g. Platemys and Chelys, all the constituent parts of the atlas cobssify and form a complete, solid vertebra, which articulates by a concave-convex joint with the true centrum of the second vertebra. The normal number of cervical vertebrae is eight in all Chelonians. The last cervical has sometimes, e.g. Chelydra, a very peculiar shape with strangely modified articular facets, in correlation with the retractile neck. The neural spines of the trunk vertebrae broaden out and fuse with the neural plates of the carapace. A tertiary modification takes place in many Pleurodira with the reduction of the neurals by the costal plates, which then meet in the dorsal line and cover the neural spinal processes. The caudal vertebrae are often much reduced in size, although not always in numbers, when the tail is very short, as in the marine turtles. In various species of Testudo about half a dozen of the last caudal vertebrae fuse together into a veritable urostyle, which is covered with a clawor nail-shaped sheath of horn. In some of the gigantic tortoises of Mauritius this caudal vertebral complex is fully 3 in. long and 2 in. broad, of an extraordinary appearance.

The vertebrae of the Lacertae, or Lizards proper, are a direct further development of those of Sphenodon. The chorda disappears; the vertebrae are procoelous, with an articulating knob behind. Intercentrals, in the shape of osseous, unpaired nodules or wedges, persist on most of the cervical vertebrae; they are absent in the trunk and reappear in the tail, either as wedges or with chevrons. The first intercentral forms the central half of the atlas, with the neural half of which it is connected by suture. The second fuses mostly with the cranial end of the second centre and with the caudal and ventral surface of the odontoid, forming a downward-directed hook. Frequently the fusion remains incomplete, or the wedges may completely merge into the epistropheal mass without leaving any outward traces. Boulenger has made the important observation that the intercentra of the tail are sometimes paired, e.g. in Heloderma. When the caudal vertebrae are strongly procoelous, the knob is very long and the chevrons are attached to its neck, having shifted on to the vertebra in front, while their basal intercentral piece, or pieces, remain in the original position. In Ophisaurus the chevrons are absolutely fused with the caudal ends of the centra and thus assume a superficial resemblance to the vertebrae of Urodela. The splitting of the tail-vertebrae and regeneration have been described on a previous page. The trunk-vertebrae of the Tejidae and the larger Iguanidae possess additional articulating processes and facets, besides the usual processes. The Zygosphene is a wedge-shaped process with two articular facets, which projects forward from the anterior side of each neural arch. The Zygantrum forms a corresponding excavation with a pair of articular surfaces on the hinder side of the arch. The crests on the tail and trunk of many lizards, e.g. Iguanidae, are entirely tegumentary structures and not supported by the axial skeleton, except in some chameleons, e.g. Ch. cristatus, and in the peculiar genus Brookesia; in these the accessory much-complicated processes are enormously elongated and support the high cutaneous crest which arises from the back, especially in B. ebenaud. The vertebrae of the snakes are procoelous (figs. 27, 28, 29). Besides the zygapophyses, they have zygosphenes on the neural arches; the ribs articulate with the parapophyses. Long, unpaired hypapophyses arise from the centre of the anterior neck and trunk vertebrae to a variable extent. In Dasypeltis and Rhachiodon a considerable number of these processes perforate the oesophagus and act as crushers of the shell of the eggs which these snakes swallow. The oftenrepeated statement that these processes are capped with enamel is erroneous. The caudal vertebrae are devoid of chevron bones, but they carry paired hypapophyses, and they have transverse processes which also are generally bent downwards.

Lastly, the numbers of vertebrae composing the whole column and its various regions. In the snakes we can distinguish only between atlas and epistropheus, trunk and tail. The numbers vary exceedingly, in the trunk up to several hundred.

FIG. 27. - Lateral aspect of two trunk vertebrae of Python. a, articular processes of the zygapophyses; na, neural arches; ns, neural spines; t, parapophyses. zs, zygosphene.

The tail may contain only a few, e.g. in the burrowing Typhlops, Glauconia, Uropeltis; or it may be very long, as for instance in Boa. There is no obvious reciprocal correlation between the length of the trunk and the tail. In the other orders of reptiles the neck is well marked, except in the snake-shaped lizards. If we define as first thoracic vertebra that which is the first connected with the sternum, all those anterior being cervical, the neck vertebrae number in chameleons, 7 in Sphenodon, 8 in the Chelonians and in the lizards, with the exception of the majority of Varanus, which have 9 like the Crocodilia. THE Number Of Vertebrae Of Some Specimens In The Museum Of Zoology, Cambridge, England The ribs, having arisen as lateral, separated off processes from the basiventral elements, show many modifications in their proximal attachments. These can be best studied on the skeleton of a young crocodile (fig. 25, 7 and 8). The first pair of ribs is very long and broad, attached to the unpaired ventral piece of the atlas-ring; the tubercular portion is indicated by a very small rugosity. The second pair of ribs is still larger; the capitulum attached to the second intercentral piece which fuses. with the odontoid process; the tubercular process is weak or represented only by a ligamentous connexion with a small knob of the odontoid process; consequently the tuberculum has. shifted its attachment away from the second vertebra. The other cervical, and the anterior thoracic, ribs have complete FIG. 28. - Posterior aspect of a trunk vertebra of Python (from nature). a, zygapophyses; h, ball on the surface of the centrum; t, parapophysis; zg, zygantrum.

ns FIG. 29. - Anterior aspect of a trunk vertebra of Python (from nature). a, zygapophyses; c, cup on the surface of the centrum.

capitular and tubercular processes, which, articulating with the bodies and with dorsolateral processes of the neural T,r arches of their vertebrae, enclose typical transverse canals. In the posterior thoracic region L the ribs are attached entirely to transverse processes of the neural arches, both capitular and tubercular portions having left the bodies or centra; the same arrangement prevails in the tail, but the ribs are very short and soon fuse with the processes. The two sacral ribs are very thick, articulating with the centra and the bases of their neural arch, and they even form part of the intervertebral joint ! In Sphenodon the first three ribs are represented by bands of con nective tissue only, with similar attachments as in crocodiles. The other cervical ribs are osseous; their short capitula retain their partly intercentral attachment, while the tubercula are carried by low processes of the centra. In the thorax both capitulum and tuberculum merge into one facet, which is gradually shifting farther tailwards and upwards until the attachment reaches them, and then lies upon the neuro-central suture. The first caudal vertebrae also possess ribs, very short and soon fusing with the diapophyses of the neural arches. In the cervical region of the Chelonia the ribs seem to be absent. In the thorax they retain their primitive intercentral position throughout life, assuming (except the first pair, which remains short and least modified) an absolutely intervertebral position. From the lumbar or presacral region backwards the capitula are gradually shifting upon short processes of the centra, until in the tail the vestigial ribs are carried by the diapophyses of the neural arches. In Sphargis (fig. 31) all the ribs are free; in the other Chelonians the ribs, generally in the recent species, flatten and become surrounded by the growing membrane bone of the dorsal plates, and the cartilage of the ribs (except the capitular and neck portion of the rib, which cannot be got at by the dermal bones) undergoes a pro cess of calcification.

Ultimately this is resorbed and its place is taken by the dermal bone, which forms, so to speak, a cast of the rib. Several of the short presacral ribs, and of course the postsacrals, are not drawn into these enormous changes, although the carapace covers, and indirectly affects, them.

Certain changes initiated in Sphenodon are more marked in the ribs of the Lacertilia; cervical ribs are often long in the lower neck. In the trunk the capitular portions are often much reduced, and in these cases the ribs are suspended mainly by their tubercular portions, usually from the diapophyses of the neural arches near the anterior end.

In the snakes all the vertebrae, from the second cervical to the tail, carry ribs. These are very movable, articulating with a rather large, more or less vertically placed facet, which is borne by the parapophysis or transverse process; sometimes the rib retains traces of the original division into a capitular and tubercular portion. The ribs of the snakes, although long, consist only of their dorsal portions. In snake-shaped lizards, e.g. Pseudopus, rather long ribs begin with the fourth vertebra.

Uncinate processes are developed only in Sphenodon and in the Crocodilia. They are not homologous structures, arising in the former from the posterior margin of the middle of the dorsal portions of the ribs, overlapping the shaft of the next following rib; in the crocodiles they arise out of the middle portion of the ribs, remaining cartilaginous, whilst the middle portion coossifies with the dorsal. Only in Sphenodon and Crocodiles the thoracic ribs consist of three successive pieces; in the Lacertilia they consist only of the dorsal and the ventral or costosternal. The latter remain cartilaginous, or they calcify, but they never ossify.

The sternum and further modifications of the ribs of the trunk.

The sternum of most reptiles consists (I) of an anterior portion (presternum, Parker; prosternum, Furbringer; mesosternum of Gegenbaur), which is generally broad, more or less rhomboid and carries the shoulder-girdle, and on its posterior sides several pairs of ribs; (2) of a posterior portion (mesosternum and xiphisternum of Parker; xiphisternum of Furbringer; metasternum of Gegenbaur), which is narrow, sometimes metameric, carries several pairs of ribs, and generally divides into a right and left xiphoidal half, each of which is continued into one or more ribs. These ribs tend to lose their connexion, and in these cases the sternum ends in two typical xiphoid processes. The distinction between preand metasternum is arbitrary. In Sphenodon the broad sternal plate carries only three pairs of ribs, the 8th to loth, and there is no xiphisternum. The other ribs of the trunk are long and compound, but they remain free and do not approach the mid-line. From the posterior edge of the sternum to the pelvis extends the complicated parasternum, embedded in the abdominal wall; it is composed of about two dozen sets of abdominal ribs, each set containing a Tight and a left and a median chevron-shaped piece. In the Crocodilia the presternum carries only two or one pair of ribs, always that of the 10th vertebra. The narrow, more or less metameric metasternum carries seven or eight ribs, the last one to three being xiphoidal. The post-thoracic ribs gradually decrease in length; about three presacral vertebrae have no ribs, and so are typically lumbar. The sacral ribs are generally the 25th and 26th in Crocodilus and Alligator; sometimes the 24th and 25th in Gavialis. The parasternum consists of only seven or eight transverse sets, each composed of two right and two left narrow splint-bones. All these parasternal elements belong to the category of dermal bones, together with those of the plastron of tortoises, inherited from Stegocephalian conditions.

The Lacertilia present an almost endless variety. The presternum is rhomboid and broad; it carries from three to six pairs of ribs, mostly four or five; the first thoracic rib is that of the 9th vertebra, the only exceptions being the chameleons with only five cervical vertebrae, and Varanus, which has usually nine cervicals like the crocodiles. The last cervical rib in these long-necked lizards is very long and has all the appearance of having but recently severed its connexion with the sternum. The presternum of Lacertilia sometimes has a window, e.g. some species of Lacerta, Phrynosoma, Iguana, or a pair of windows, e.g. Agama, Liolepis, Goniocephalus. The xiphisternum carries a variable number of ribs; it is either scarcely distinguished from the anterior plate, or it is long, and in these cases either double, e.g. Iguana, Gerrhonotus, Varanus, Zonurus, Agama, Cyclodus, Lacerta; or single, e.g. Zonosaurus. The poststernal ribs shorten gradually in the majority of the Lacertae, and there is sometimes a ribless lumbar vertebra, e.g. in Iguana; in many Lacertilia, however, the ventral cartilaginous halves of the ribs are connected with those of the other side, either by ligaments, or they join together, forming complete hoops of thin cartilages. Such ribs occur in all Geckones and Chameleons, but also in many Iguanidae, Scincidae, and even in the Anelytropidae; their numbers vary much, from 27 in the Scincoid Acontias meleagris, 7 - Io in Polychrus, 8 in Chamaeleo  ?


FIG. 30. - Lateral aspect of Three Thoracic Vertebrae of Crocodilus vulgaris (after Mivart). c, cup on the anterior surface of centrum; cp, capitula of ribs; ns, neural spines; t, tubercula of ribs; u, uncinate processes; vr, dorsal or vertebral portions of the ribs; rc, ventral or sternal cartilaginous portions of ribs.

FIG. 31. - Three Vertebrae of Sphargis coriacea. c, vertebral centra; n, neural arches; r, ribs.

lizards), H. Gadow (cloaca of crocodile), J. F. v. Bemmelen (Sphenodon and others), W. H. Gaskell and H. Gadow (heart of tortoise) .

Sense Organs. i. Tegumentary Organs of some Tactile or other Sense. - Reptiles possess apparently no traces of those tegumentary sense organs which, belonging to the domains of the trigeminal and vagus nerves, have spread far over the body in fishes and batrachia. They were developed by those classes in correlation with their essentially aquatic life. This does not apply to the reptiles which, as a class, are of absolutely terrestrial origin. Nevertheless all recent reptiles possess numerous low sense-organs, " tactile bodies," in most parts of the skin, connected with the regional, spinal nerves. They are most obvious in snakes, appearing as one or more little colourless spots near the apex of each scale on the back. The spot is formed by a little cluster of epidermal cells, connected with a sensory nerve. Their lowest stage they show in Sphenodon and in lizards, whilst in crocodiles they have reached a higher stage, at the bottom of the pit, since the tactile bodies, mostly several together, have sunk into the cutis, below the epiderm, forming a little pit, mostly near to the anterior margin of the flat scutes. They are most obvious on the belly of crocodiles, whilst in the American alligator such pits are scarcer, not because the organs are absent, but because these have sunk still farther into the skin. The last stage is that met with in tortoises, which possess such tactile bodies in considerable numbers in the softer subepidermal layers, beneath the large horny shields which themselves show no traces of them.

2. Taste

The respective organs do not seem to have been investigated. That they exist is amply proved by the careful predilection for certain kinds of food which is shown especially by vegetarian tortoises and lizards, independent of smell. Many lizards are, for instance, very fond of sugar.

3. Nose

The sense of smell is well developed in all reptiles. In none is the olfactory organ degraded; that the nasal passages, the nose itself, are never degraded is explained by the fact that all reptiles invariably breathe through the nose, except snakes during the act of swallowing their prey. The nostrils, always paired, are frequently provided with valves, to shut out the water, or sand. In some water tortoises, e.g. Trionyx, Chelys, the nostrils are prolonged into a soft, unpaired proboscis. Double tubes exist in the snake Herpeton (see Snakes, Opisthoglypha). The nostril leads into an antrum or vestibulum, this again into the nasal cavity proper, at the dorsal farther end enters the olfactory nerve, whilst ventrally it leads into the nasolaryngeal duct, with its posterior narial opening, or choana. The ducts are short in snakes and lizards, the choanae lying in the front part of the palate, but in tortoises and crocodiles they are placed far backwards, as has been described under Skull above. Into the nasal cavity projects, from the septum, a concha, least developed in tortoises, most in lizards and snakes. Crocodiles show a beginning of separation into several conchae as in birds and mammals. A large nasal gland lies against the lateral, or ventral, side of the outer wall of the nasal cavity, into which also opens the naso-lacrymal duct. Jacobson's organ, of uncertain function, is present in most reptiles. It is paired. In tortoises it is still placed within its nasal cavity, against the median wall, and is still nothing but a recess of the same and its mucous lining. In lizards and snakes the organ has become completely separated from the nasal cavity, lying below it and opening, each by a separate passage, into the palate mouth, close to or still within the choanae. In snakes it is mushroom-shaped, with a very short stalk. It lies immediately below the floor of the nasal capsule, and the membranous wall of the cavity on which it lies is covered and protected by a bone, commonly called the turbinal, which extends out from the median nasal system to the maxilla. In crocodiles these organs are vestigial and soon disappear.

Ear.

In crocodiles the outer ear lies in a recess, dorsally overhung by the lateral edge of the bony squamoso-frontal bridge; it carries a flap of skin, provided with muscles, to close the ear tightly. In lizards the outer ear is quite unprotected, and when the meatus is very short and wide, the drum is quite exposed. No reptiles possess cartilages comparable to the mammalian outer ear. Sphenodon, chameleons, snakes have no outer ear, the skin passing over the region. So also in tortoises, but in some of the aquatic kinds its position is well indicated by softer and thinner skin; in others, for instance marine turtles, a thick leathery plug, or a bigger scale marks the former position. In various lizards, chiefly burrowing in sand, the ear passage is very narrow, or closed. The middle ear or tympanic cavity is quite obliterated in snakes, Amphisbaenas and some other snake-shaped lizards. In Anguis may exist individual traces. The cavity communicates with the mouth. In lizards the communication is a wide recess, lined with black pigment, so that in these creatures the whole auditory chain can easily be inspected from the opened mouth. In tortoises the recesses are contracted into the Eustachian tubes, each of which opens by a separate aperture into the roof of the mouth. In the crocodiles part of the cavities is transformed into an intricate system of canals and passages. The two Eustachian tubes open together in the mid-lines protected by a valve, between the basioccipital and basisphenoid; thence arises a median passage which with lateral arms and loops extends upward through the occiput into the cranial roof, communicating with the tympanic cavity, and further continued through the quadrates and beyond into the mandibles, by the siphonium.

In spite of the obliterated tympanic cavity of snakes, and the closed up outer ear passage and absence of a tympanic membrane in snakes and tortoises, these creatures can hear very well. The same applies to Sphenodon, but it seems doubtful whether chameleons can hear.

Through the whole middle ear, from the fenestra ovalis to the drum-membrane, stretches the chain of auditory ossicles or cartilages, partly attached to the posterior wall by the common lining membrane. The arrangement appears simplest in snakes, in chameleons and in tortoises, not because it is primitive but because it is so much reduced, partly in correlation with the abolition of the outer ear. In these creatures the columella goes as a bony, slender rod straight to the middle of the quadrate, against which it leans, or with which it articulates by a short piece of cartilage, the extra-columella. Here the whole chain ends. It looks like a proof that columella = stapes, extra columella =incus, and quadrate = malleus; or, with the usual ignoring of the little extra-columellar piece, that quadrate = incus, Gegenbaur's favourite impossibility. In those lizards which have a tympanic membrane conditions are far less reduced. The extra-columellar piece sends out three distal processes; one leans on to the middle of the tympanic membrane, the second usually is fastened to the bony dorsal rim of the meatus, the third is directed downwards and is continued as a thin ligament towards the inner angle of the articular of the mandible, but before reaching this it comes to grief, being squeezed in between the quadrate and the posterior end of the pterygoid. The hyoid proper is of no account in snakes and tortoises, since it is reduced to very short distal pieces attached to the base of the tongue; but in lizards it remains in its original length, or it even lengthens, and shows many vagaries in its position and attachments. In embryos of Sphenodon and lizards it arises from near the junction of the columella with the extra-columella. It becomes very long, too long for the available space (perhaps correlated with lingual functions), and it forms a high loop, thereby causing the peculiar loop of the chorda tympani; the upward bend of the hyoid becomes connected with the parotic process of the cranium. Next aborts the portion between this connexion and the original proximal end of the hyoid, near the columellar mass. The upper end of the hyoid either remains attached to the parotic process (various lizards and Sphenodon) whence the lingual apparatus remains suspended, or the hyoid, having broken loose, leaves a little cartilage, Versluy's cartilage, behind, at the end of the parotic process, and the hyoid horn remains free, in the majority of lizards. In Sphenodon, whilst passing the distal portion of the extra-columella, part of the hyoid fuses with it, often forming thereby a little hole, the remnant of imperfect fusion.

In the crocodiles the arrangement is at first complete and diagrammatically clear, not obscured by vagaries of the hyoid, which is free and much reduced. In the embryo the large extra-columellar cartilage, abutting against the tympanic membrane, and with another process against the quadrate, sends its third, downward, process as a thick rod of cartilage to the posterior inner angle of the mandible with which it is directly in cartilaginous continuity. It was W. K. Parker's mistake to call this cartilage the cerato-hyal. In young embryos it looks like an upward continuation of Meckel's cartilage, much resembling mammalian conditions. But in nearly ripe embryos this cartilage is already reduced to a string of connective tissue, cartilage remaining only at the upper end, and where this string enters the mandible lies the siphonium, the tube which connects the air cavities of the mandible with the Eustachian passages, the long connecting channel becoming - side by side with the extracolumellar-mandibular ligament - embedded into a canal of the quadrate, so that in older stages, and above all in the adult, the proper display of the whole arrangement requires a 5 FIG. 37. - Diagram showing Evolution cf the Ossicular Chain of the Ear. 1. Hyostylic Elasmobranch. H, hyoid; Hm, hyomandible; M, mandible; P Q, palatoquadrate. 2. Lacertilian. Co, columella or stapes; and E, extra-columella with supra-, extraand infra- "stapedial " processes. 3. Hypothetic stage between 2 and 4, Sphenodon. Par = parotic bone. 5. Lacertilian. Parotic process with a piece of cartilage at its end, remnant of piece of the hyoid; connexion of infra-stapedial process with mandible vanishing. 6. Embryo of Crocodile. Continuous cartilaginous connexion of extra-columella with Meckel's cartilage. 7. Embryonic Mammal; for comparison. Cd, the new condyle, articulating with Sq, squamosal; Cor, coronoid process; quadrate transforming into tympanic ring.

little anatomical skill. The whole string, whether cartilaginous or ligamentous, which connects the downward extracolumellar process with the articulare, is of course homologous with the continuation of Meckel's cartilage into the malleus of foetal and young mammals; and the chain of bones and cartilages between the auditory capsule, fenestra ovalis, and the proximal part of the mandible is also homologous wherever such a chain occurs; lastly, fenestra ovalis and membrana tympani are fixed points. Consequently columella =stapes, extracolumella of Sauropsida= lentiform+incus+malleus of Mammalia.

The inner ear has been studied minutely and well by C. Hasse, E. Clason and G. Retzius. It is enclosed by the periotic bones. The fenestra rotunda is surmounted by the opisthotic, the fenestra ovalis by the same and by the pro-otic, and this protects also the anterior vertical semicircular canal. The posterior canal is opisthotic, the horizontal is proand opisthotic. The anterior canal is the largest of the three, a feature characteristic of the Sauropsida. The lagena, with its own acoustic papilla, begins to show a basilar membrane with papilla, at the expense of that in the sacculus. In Sphenodon and lizards a slight curving of the lagena indicates the beginning of a cochlea, and a scala is developed in crocodiles, but neither cochlea nor scala is specially twisted. The endo-lymphatic ducts end as closed sacs, in lizards and snakes, in the roof of the skull, between the occipital and parietal bones. They reach an enormous development in many geckos, where they form large twisted sacs beneath the skin, covering the sides of the neck, which then assumes a much swollen appearance. They contain white otolithic masses, with lymph. It is remarkable that the extent of these sacs varies not only in allied species, but even individually, independent of sex and age, although they are naturally liable to increase with age.

5. Eyes are present in all reptiles, although in many of the burrowing snakes and lizards they may be so completely covered by the skin as to have lost their function. Most reptiles have upper and lower lids, moved by palpebral muscles, and a third lid, the nictitating membrane, which can be drawn over the front of the cornea from the inner angle obliquely up and backwards. Its mechanism is simplest in lizards. A muscle, a split from the retractor muscle of the eyeball, arises from the posterior part of the orbit, is attached to the posterior wall of the eyeball, and there forms a pulley for the long tendon which arises from the median side of the orbit and passes over the back of the ball forwards into the nictitating jmembrane. Contraction of this muscle draws the membrane backwards and over the eye. In crocodiles and tortoises the tendon of the nictitating membrane broadens out into a muscle (M. pyramidalis), which arises from the median side of the posterior portion of the ball; above the optic nerve it crosses over the broad insertion of the retractor of the ball, without being much guided by it, although this muscle by its contraction slightly prevents the nictitating tendon and muscle from touching the optic nerve.

It is easy to recognize the mechanism of birds as a combination of the two types just described; their ,muse. quadrates s. bursalis is of course the single muscle of the lizards, but now restricted to, and broadened out upon, the eyeball.

Special Modifications of the Lids

In the snakes the upper and lower lids are reduced to the rim, and the nictitating membrane has become the permanent cover, which protects the eye like a watch-glass, leaving between itself and the cornea a space, drained by the naso-lacrymal duct, and behind this space the eyeball moves as freely as in other animals. A similar arrangement exists in the true geckos, not in the Eublepharidae, which still possess the outer lids. In some lizards, especially such as live in deserts, the middle of the lower lid has a transparent disk, and it is always the lower lid which is drawn over the eye, the upper in nearly all Sauropsida being much smaller and less movable; for instance, some specimens of the Lacertine genus Eremias in Africa and India. In the Indian genus Cabrita, and in Ophiops of Africa and India, the lower lid is permanently fused with the rim of the shrunken upper lid and forms a transparent window superficially looking like that of the snakes. Exactly the same arrangement has been developed by Ablepharus, one of the Scincidae.

The eyeball is provided with the usual rectus and obliquus muscles, in addition to a retractor oculi. Apparently all reptiles possess a pair of Harderian or nictitating glands, which open in front, in the nasal, inner corner, and lacrymal glands which open likewise into the conjunctival sac, but near the outer or temporal corner. The secretion of both is drained off through the lacrymal canals, which in lizards open below in the outer wall of the posterior p ares; in snakes they open into the mouth by a narrow aperture on the inner side of the palatine bone.

The walls of 'the anterior half of the sclerotic of lizards, tortoises and Sphenodon contain numerous cartilaginous or osseous plates, which imbricate in ring shape; they are absent in snakes and crocodiles. Internally the eye of most reptiles possesses at least traces of a pecten; very small indeed in tortoises, or in crocodiles where it is represented by only a few mosslike, pigmented vessels. In many lizards these vessels, arising from near the optic nerve, form a network which extends right up to the posterior side of the lens; in others, especially in Iguanidae, is developed a typical, large pecten, deeply pigmented with black, fan-shaped or umbrella-shaped, sometimes folded. In chameleons it is a short cone; apparently XXIII. 6 quite absent in Sphenodon. A falciform process and other remnants of a campanula are absent. In most of those reptiles which have but a rudimentary pecten, the retina is supplied by hyaloid vessels which spread over the surface of the vitreous body; such superficial vessels disappear with a greater development of the pecten, and the retina receives a choroid supply; special retinal arteries from the a. centralis retinae, and veins, exist in snakes.

Ciliary processes of the choroid are usually small, a proper ciliary body being least developed in crocodiles; all reptiles have a ciliary muscle. The shape of the contracted pupil varies from round to a vertical slit; the latter is most marked in Sphenodon. The retina shows usually a fovea centralis, sometimes but slightly indicated by a shallow depression; it is well marked in chameleons. The retina contains only cones, rods being absent; fat-drops on the apex of the cones are common; their usual colours are green and blue.

6. The pineal, median or parietal eye is the terminal organ of the epiphysis of the brain, with which it is connected by a nerve-containing string. Among recent reptiles it exists in Sphenodon and in the Lacertilia, with vestiges in snakes. It is embedded in the median parietal foramen. Externally its presence is generally marked by the scales being arranged in a rosette, with a transparent central scale. The organ itself is distinctly a dioptric apparatus, with all the essential features of an eye; a pigmented retina of the arthropodous simple type surrounds an inner chamber which is nearly filled by a cellular globular mass which projects into it from above; this is the so-called lens, in reality much more like the corpus vitreum in its still cellular condition, while the real lens has to be looked for in the superimposed tissue. The whole organ is best developed in Sphenodon, even in the adult; but whether it is still functional, and what its function is, remain unknown. The throwing of a beam of light upon this eye, by means of a lens, produces no effect. Whilst in Sphenodon the " lens " is rather dull and the efferent nerve is still present, in various lizards the " lens " is more perfect, but the nerve is degenerated. We conclude that the whole organ is now without the least visual function, whilst in various extinct groups of reptiles and Stegocephali it was fully developed. It has been well investigated by de Graaff, W. B. Spencer and A. Dendy.

The Muscular System. A useful account of the differentiation of the muscles in the main reptilian groups, with their almost endless modifications in correlation with walking, climbing, swimming, gliding and burrowing, with limbs complete or absent, would fill several pages of this article and would necessitate many illustrations. The literature is great; it comprises many good detailed descriptions of various kinds of reptiles, and several monographs. M. Fiirbringer has devoted a whole series to the muscles of the neck, shoulder-girdle and fore limbs. Hand in hand with these investigations went that of the innervation, without which myology would lack scientific value. The present writer has devoted much time to the muscles and nerves of the pelvis and hind limbs, and has, in tabular form, compared them with those of other vertebrates. The results of all these labours are rather disappointing, except for the study of myology as such, which raises many interesting questions. Broadly speaking, the muscles of typical reptiles, crocodiles and lizards are more highly differentiated (by no means always more numerous, but more individualized by origin and insertion, the behaviour of the tendons), more effectively disposed according to mechanical principles, than in Batrachia, and less than in birds and mammals. This can easily be proved, whether we take for comparison the muscles of the neck, of the larynx or hyoid, or limbs. Lowest in general stands Sphenodon, next to it the lizards, highest the crocodiles, while tortoises and snakes show the greatest reduction and specialization. In the tortoises it is the non-yielding box of carapace and plastron which has caused great changes within the region of the trunk proper. First, all the epiaxial muscles have vanished; the same applies to the costal muscles; but traces of dorso-lateral muscles occur on the inside of the posterior half of the carapace, extending as a longitudinal system from one transverse process to the next in many of the lower aquatic tortoises, as perfectly useless vestiges; or more striking, these muscles exist in the young, and disappear with age, for instance in Testudo. Secondly, it is rather surprising that the rigid shell has offered so little or no inducement to the muscles of the girdles, neck and tail to transfer their origins upon it. Thirdly, the retractile neck of the typical cryptodirous tortoises is correlated with a pair of long retractor muscles, which in the shape of a pair of broad, vertical ribbons (between which is received the S-kinked neck) extend far back along the vertebral column, almost to the level of the pelvis.

In snakes, owing to the loss of limbs and girdles, only the spinal and costal muscles remain, besides of course those of the abdomen and the visceral arches. The vestigial muscles of the limbless lizards and of the peropodous snakes have been monographed by Fiirbringer in much detail without great results.

Respiratory Organs. All reptiles breathe by lungs, and they possess no vestiges of gills, not even during their embryonic stages, although gill clefts are invariably present in the embryo. Nor does any part of the outer skin assist respiration, as is so commonly the case in Batrachia; yet, strictly speaking, the lungs are not the only organs of respiration in the class of reptiles, since various tortoises possess additional breathing apparatus in the anal sacs and in certain recesses of the throat, to be mentioned farther on.

The Larynx, instead of lying at the bottom and; far back in the throat, as in the Batrachia, is considerably moved forwards so as to rest upon the hyoid and to project into the pharyngeal cavity. A pair of arytenoid cartilages, enclosing the glottis, rest upon several more or less fused tracheal cartilages, which thus represent the cricoid, but there is no thyroid cartilage. A small process from the anterior median edge of the cricoid is .the beginning of an epiglottis. Vocal chords are indicated by lateral projecting folds of the inner membranous lining of the larynx, and are in a few cases effective in producing a voice. Crocodiles and alligators have a powerful, loud, bellowing voice; many tortoises utter weak, piping sounds, especially during the pairing season; and also various lizards can emit a feeble squeak, for instance, Psammodromus hispanicus, and the geckos. Sphenodon, at least the males, can grunt. Snakes have no voice; they can only hiss like all other reptiles, but a curious modification exists in the larynx of the North American Coluber s. Pityophis, e.g. C. melanoleucus: the epiglottis is more enlarged, and laterally compressed so that the hissing sound is much strengthened by the vibration of the epiglottis. The larynx possesses a constrictor and a dilator muscle, which arise from the arytenoids and from the cricoid respectively, and are attached to the hyoid. Chameleons have bladder-shaped sacs which can be filled with air from a slit immediately below the larynx. For further modifications see G. Tornier.

The Trachea is furnished with cartilaginous rings and semirings, which extend to the lungs. As a rule the trachea is straight; in Crocodilus americanus it forms a loop; and similar curvings occur in various tortoises in correlation with the retractile neck. The two bronchi are shortest in Sphenodon, very long in most tortoises, where they begin frequently already half down the neck. In Sphargis most of the trachea is divided by a longitudinal partition. It is an advance upon amphibian conditions that the bronchus enters its lung no longer at its apex, since an anterior, pre-bronchial lung-portion has come into existence. This is still very short in Sphenodon, while in crocodiles, tortoises and in the highly developed Varanidae the bronchus enters near the middle of its lung, so that the anterior portion is nearly as long as the posterior. The shape of the trunk influences that of the lungs. In the snake-shaped forms, both snakes and lizards alike, the lungs have become very asymmetrical, one of them being much larger than the other, which is often quite aborted.

The simplest form of lungs is that of Sphenodon; the pre, bronchial part is still small. Each lung is still a sac with one large lumen, the walls being honeycombed. In the lizards the walls are more spongy, and several septa begin to extend more or less far from the walls into the lumen, towards each bronchus. Some of these septa begin to cut the lung into lobes, especially in Varanus and in chameleons. In the latter exists a further specialization, a side-departure, in the shape of several long, hollow processes which are sent out from the posterior portions of the lungs and extend far into the bodycavity and between the viscera. By means of them these creatures can " blow " themselves out. They are of morphological interest since they are first stages of air-sacs so marvellously developed in birds, and possibly also in various Dinosaurs. In the Amphisbaenids the left lung alone remains.

The lungs of crocodiles have reached a considerably higher stage. They alone in reptiles are, on the ventral side, completely shut off from the viscera by a pleural, partly muscularized, membrane. From each bronchus extend a number of broad septa towards the periphery, dividing the originally single lumen into many chambers, perhaps a dozen, from the walls of which wide secondary or parabronchial canals extend into the alveolar meshwork, in very regular arrangement, in series like organ-pipes.

The lungs of the tortoises are, in adaptation to the peculiar shape of the body, stowed away along the back, as far as the pelvis, and only their ventral surface is covered by a strong peritoneal membrane which receives muscular, diaphragmatic fibres. The inner division of the lungs into chambers has progressed so much that a sort of mesobronchus has become discernible; the arrangement of the side-bronchi is far less regular than in crocodiles; the whole lung is much more honeycombed, meshy and spongy.

The mechanism of breathing of tortoises is not such a puzzle as it is sometimes stated to be. Of course the rigid box of the trunk excludes any costal, or abdominal breathing, but by protruding the limbs or the neck, piston-like, an effective vacuum is produced in the box. Moreover, the throat is distended and worked considerably by the unusually large and very movable hyoid apparatus, by which air is pumped into the lungs.

The lungs of the snakes are very thin-walled, with a very wide lumen, and only for about the first half from the heart backwards the walls are alveolar enough for actual respiratory function, while towards the blind end the sacs are so thin and sparsely vascularized that they act mainly as reservoirs of a large amount of air. Frequently their posterior portions receive blood vessels not from the pulmonary arteries but directly from those of the trunk. In correlation with the long, cylindrical body, the lungs are much elongated and they are not equally developed. The asymmetry shows great differences in the various groups, consequently the asymmetry has been developed independently in those groups. It is usually stated that the left lung is much smaller than the right. This is but rarely the case. The most recent observations are those of E. D. Cope (Proc. Am. Phil. Soc. (1894), xxxiii. 217). In Boidae both lungs are large, although unequal: the left or more dorsally placed one being the larger. In Ilysia the right is functional, the left is ventral and vestigial. In Rhinophis the right is very small, the left larger. In Glauconia and Typhlops the right lung alone is developed: the left is quite aborted. In Colubridae the left lung alone is functional, while the right is vestigial. There is no trace of the right in Elapinae and Hydrophinae and most Viperidae. In the Colubridae the right, or ventral, lung is, when present at all, reduced to a length of from 2-5 mm., and it then communicates with the anterior portion of the left lung by a foramen, in level of the heart, whilst the right bronchus is aborted.

A further complication is the so-called tracheal lung, which is present in Typhlopidae, Ungalia of the Boidae, in Chersydrus of the Acrochordinae, in the Hydrophinae and Viperidae. This peculiar organ is a continuation of the anterior portion of the functional lung, extending far headwards, along the trachea, with the lumen of which it communicates by numerous openings. In Chersydrus this mysterious organ is " composed of coarse cells and without lumen, extends from the heart to the head, and is discontinuous with the true lung; the trachea communicates with it by a series of symmetrical pores on each side." In Typhlops it extends likewise from the heart to the throat, as a cellular body but without lumen or connexion with either trachea or lung.

Thyroid and Thymus. The Thyroid of the reptiles is a single, unpaired organ, placed ventrally upon the trachea and one or other of the arterial trunks, more or less distant from the heart. In snakes it lies on the mid-line near the heart; a little farther up in Sphenodon; still farther in lizards, and chameleons near the root of their gular sac. In tortoises it is globular, at the division of the carotic trunk. In crocodiles it is bilobed.

The Thymus is paired. It is largest in crocodiles, extending on either side of nearly the whole neck, along the carotids and jugulars. In the tortoises they are much shorter; in Sphenodon and lizards are two pairs, more or less elongated; in the snakes are sometimes as many as three pairs, elongated but small, attached to the carotis near the heart. As usual the thymus bodies become much reduced with age.

The Spleen. The Spleen varies much in shape and position. In lizards it is mostly roundish, elongated in Sphenodon, and placed near the stomach; in crocodiles it lies in the duodenal loop behind the pancreas; similarly situated in snakes, but in the tortoises it is much concentrated, large and attached to the hind-gut.

The Body Cavity. The body cavity of the reptiles is subdivided into several sacs or cavities by serous membranes of peritoneal origin. The number of these subcavities differs much in the various groups. The pericardial sac is always complete. In tortoises the lungs are retro-peritoneal, a dense serous membrane spreading over their ventral surface from the walls of the carapace forwards to the liver and shutting off a saccus hepato-pulmonalis from the rest of the peritoneal cavity. Snakes possess, besides the modifications mentioned above, separate chambers for the stomach, right and left liver, and for the gut, whilst the pleural cavities as such have been destroyed. In lizards a " post-hepatic septum " divides liver, lungs and heart from the rest of the intestines. This transverse vertical septum is best developed, almost complete, in some of the Tejidae, in others it seems to be more imperfect, and it is probably a further development of the suspensorial ligament of the liver, which is ultimately inserted Upon the ventral wall of the body.

The subdivisions have reached their highest development in the crocodiles, there being, besides the pericardial and the two pleural cavities and the usual peritoneal room, a right and left hepato-pericardiac, an hepato-gastric, and an hepato-pulmonal sac. The caudal and ventral edges of these liver-sacs are fused on to the ventral body-wall, thus producing a complete transverse partition, headwards of which lie the lungs, liver and heart. This partition, morphologically not homologous with the mammalian diaphragm, more resembling the imperfect structure in birds, acts, however, as a perfect diaphragm, since it is well furnished with muscular fibres. These are attached to its whole periphery, with centripetal direction, especially on the ventral half. These fibres are transgressors upon this septum from a broad sheet of muscles, which, inserted together with the septum upon the body-wall, arise from the iliac bones, the pubes, and the greater portion of the last pair of abdominal ribs. This broad muscular sheet, covering the intestines, is the so-called abdominal diaphragm or peritoneal muscle. Its continuation upon the transverse septum is the crocodilian musc. diaphragmaticus, and in functional effect very similar to that of the Mammalia, whilst the abdominal diaphragm undoubtedly causes abdominal respiration. We have seen that these crocodilian conditions do not stand quite alone, but are connected with simpler features in the other reptiles. Two recent, very lengthy papers have been written on this subject by I. Bromann (1904) and by F. Hochstetter (1906), besides two in 1902 by G. Butler.

The Heart. The Heart of all reptiles is removed from the head and is placed well in the thorax, in the Varanidae even a little beyond it. Only in snakes the heart lies headwards from the hilus of the lungs, not caudalwards, generally at about the end of the first fifth of the body. The batrachian conus arteriosus is reduced, one set of semilunar valves guarding the entrances into the truncus arteriosus which now issues directly from the heart. A sinus venosus exists still in Sphenodon and Chelonians, in which it may even receive separate hepatic veins, but in crocodiles, lizards and snakes the sinus as such exists no longer, forming part of the right atrium. All the hepatic veins enter the stem of the posterior vena cava, which henceforth enters the heart as inferior vena cava. This, the largest, and the right and left anterior vena cavae, are the only three veins which enter the right atrium. Into the left open the two pulmonary veins. Right and left atrium have in all reptiles a complete septum between them. The ventricular portion shows considerable steps towards the differentiation into a right and a left ventricle, but the partition is very incomplete in tortoises, lizards and snakes, quite complete only in the crocodiles. The most important character of the reptilian heart, absolutely diagnostic of it, is the fact that the systemic vessel which leaves the right ventricle turns to the left to form the left aorta, while the stem which comes from the left ventricular half arches over to the right as the right aorta. It is not at all necessary to conclude that this fact excludes the reptiles from the mammalian ancestry and to hark back to conditions as indifferent as are those of the batrachia. The Foramen Panizzae shows the way to a solution, how ultimately all the arterial blood from the left ventricle may pass, first through the root of the right arch, then through this hole into the left, whilst the rest of the right arch, and the root of the left, obliterate. The difficulty is not much greater than that of deriving the birds' condition from the reptilian. The Foramen Panizzae, which exists only in the Crocodilia, lies exactly where the right crosses dorsally over the left aorta. The whole is not the last remnant of the originally undivided truncus, as is taught generally, but it is a new foramen, a hole dug by the left arterial blood into the venous right aorta. According to the recent observations made by F. Hochstetter the foramen comes into existence in a very late embryonic stage.

Whilst the batrachian single ventricle possesses only one ostium ventriculare or outlet into the truncus, in the reptiles the inter-atrial septum extends considerably downwards into the base of the ventricle, so as to produce a right and a left niche, and correspondingly two ostia instead of one. The atrioventricular valves are still membranous, even in crocodiles; attached to them are muscles, trabeculae carneae, from the very trabecular walls of the ventricle; they are especially spongy in tortoises. By means'of the arrangement of some of these trabeculae, perhaps still more through the confluence of their basal portions, an imperfect ventricular septum is initiated. Certainly even in tortoises, which represent the lowest stage, the venous blood is received into and sent out by the same right side of the ventricle, while the arterial blood is correspondingly managed and dodged by the left side. That there is not very much mixture of the two kinds of blood, in spite of the wide communication in the ventricle, is further due to the peristaltic systole and diastole of the various divisions of the heart. - The heart of Chelonians is broader than long. In correlation with the very much flattened body of Trionyx and its allied genera, the whole heart is dislodged from the middle line, far over to the right side; the vessels of the left side are correspondingly much elongated and have to cross the neck, trachea and oesophagus. - The apex of the heart is attached to the pericardium by a special ligament in the Crocodilia and in many Chelonia, e.g. Testudo, but it is absent in Clemmys. Sometimes this little ligament sends a tiny blood vessel into the liver.

Arterial System. Crocodiles. - The left aorta crosses obliquely beneath the right and gives off only the coeliac, just before joining the right aorta in the level of the eighth thoracic vertebra. The aorta descendens sends off, besides intercostals and other segmentals into the body-wall, the mesenteric, right and left iliac, a pair of renal and ischiadics, a cloacal and the caudal artery. The right aorta forms the main root of the a. descendens. Close to the heart it sends off two coronaries and a short carotis primaria which divides at once into two anonymae, the left of which is the stronger. The right anonyma divides into the subclavia and collateralis colli, the left into subclavia and carotis subvertebralis. Each subclavia sends off an a. vertebralis communis, which runs headwards and, with another longer branch, downwards, giving off intercostals, and then joins the descending aorta.

Tortoises

The left aorta is rather more separated from the truncus, which it crosses ventrally in an oblique forward direction; it sends off a left cardiac to stomach and oesophagus, a coeliac and mesenteric, and then a communicating branch to the right aorta. The a. descendens gives off paired suprarenals, spermatics, very large iliacs, then a pair of renals, hypogastrics and the caudal. Each iliac artery divides into a recurrent intercostal anastomosing with the axillaries, an epigastric (sending off the crural and anastomosing with thoracics and humerals), and other arteries to abdominal muscles and to the shell. The hypogastrics supply the cloacal region and then continue as the ischiadics. But there are many anastomoses which cause great variation in the different tortoises. The right aorta sends off a right cardiac, the coronary, and the right and left anonymae which are quite symmetrical, each dividing into subclavia and carotis; in the angle lies the thymus.

Lizards

Two common carotids arise either side by side, or by one carotis primaria, from the right aortic root. In the majority each common carotis ascends the neck and then divides into the vessels for the head and another branch which turns back and goes into the descending part of the aortic arch. In chameleons two carotid stems ascend the neck and there is no recurrent vessel. In the Varanidae the two common carotids start from a long carotis primaria; there is no recurrent vessel. The vertebral arteries come from the origin of the subclavians and run to the head in a very lateral position. The subclavian arteries (which occur also in limbless lizards) arise far away from the carotids out of the descending arch of the right aorta, in a level often far behind the heart. " Anonymous " arteries are consequently absent in lizards.

Snakes

The left aorta is stronger than the right, both combining soon to form the descending aorta. Owing to the absence of fore limbs and shoulder-girdle the conditions are much simplified. In most snakes the right aorta sends off but one strong carotic vessel which represents the left carotis communis whilst the right is much reduced or even quite absent; further, there is only one vertebral artery, which either runs along the right side of the vertebral column or it divides soon into a right and a left vessel along the neck. In conformity with the reduction of one lung there is usually but one pulmonary vessel.

Venous System. Crocodiles. - Each, right and left, anterior vena cava is composed of a subclavian (axillary and external jugular), an internal jugular, common vertebral and an internal mammary vein. The posterior vena cava is composed of the two revehent renals, veins from the genital glands and ducts, revehent veins of the suprarenals (which, like birds, still have a portal system), and the big vein from the fat body. Thus the vena cava posterior perforates the right liver, receiving from it many hepatic revehent veins and also the big revehent vessel from the left lobe; next it receives the coronary vein and then enters the heart as inferior vena cava. The portal vein arises out of the coccygo-mesenteric (which comes out of the bifurcation of the caudal), collecting the blood from most abdominal viscera and from the thorax and breaks up in the right liver. The rest of the venous system is rather complicated. The big caudal vessel divides near the vent, receives an unpaired cloacal and a rectal vessel, and goes off to the right and left, each of which trunks receives an ischiadic and an inter-sacral vein and then divides into the v. renalis advehens which breaks up in the kidney, and the abdominal vein. The latter are interesting; they run in the abdominal wall, receive the obturator and other pelvic veins, intervertebrals and intercostals, the crurals, and the epigastrics out of the body-wall. Then these two abdominals (Rathke's internal epigastrics) go to the liver, which they enter to either side of the gall bladder, collecting also blood from the stomach and from the vertebral column. Both break up in the liver. Consequently all the blood from " below the heart " passes through some portal system - renal or hepatic - except that which comes from the genital glands and ducts and from the fat body.

Tortoises.

The venous system much resembles that of the crocodiles, but many and wide anastomoses, especially on the inside of the carapace and plastron, exist between often distant vessels, so that one lucky injection may fill the whole system. There are three advehent renal veins which collect on the back of their kidney into one stem; they dissolve completely into a portal system, and leave the kidney on its ventral surface as one v. renalis revehens. The right and left then form the v. c. posterior which perforates the posterior margin of the right liver, then headwards of the liver takes up the hepatic and enters the heart. The three pairs of afferent renal veins are composed as follows. The externa collects from the shell and the abdominal muscles; the posterior collects along the rectum from the genital glands, the bladder, and from parts of other pelvic viscera; the anterior comes from the anterior part of the shell and runs backwards to the kidney, with frequent anastomoses with the other advehent renal veins. The abdominals arise, as in the crocodiles, with the external advehent renal from the lateral continuation of the bifurcated caudal, which takes up vessels from the pelvis, the shell and the crural. The abdominal itself takes up a femoral vein, vessels from the abdominal and pelvic muscles, and from the plastron, and then dives into the body-cavity, receives veins from the fore limbs, and enters the right lobe of the liver, there to break up. The hepatic portal collects from the intestinal tract, spleen and pancreas. Consequently in tortoises all the blood from below the heart passes through some portal system.

The most important peculiarity of the Lizards is the condition of the abdominal veins; they combine into a single stem (after having collected the blood from the fat body and from the ventral body-wall of the pelvic region) which dives into the body-cavity to join, embedded in the ventral hepatic ligament, the left branch of the portal vein. The chief characteristic of the abdominal is that it does not communicate directly with the caudal, and that it forms an unpaired stem. The renal portal system receives its blood from the tail, the hind limbs, the abdominal wall and the urino-genital organs, all the blood passing into a right and a left advehent vein. The suprarenal portal system drains from the abdominal wall and the suprarenal bodies, and issues into the revehent renal s. These, with some intervertebrals and with hepatics, constitute the inferior vena cava.

Lymphatic System. The lymphatic vessels frequently accompany the big arteries of the trunk, either surrounding them with a meshwork or ensheathing them completely, especially in tortoises. The lymphatics from the head and neck combine with stems which accompany the veins of the fore limbs; they join the thoracic ducts and these open into the brachio-cephalic veins, as they do in birds. The lymph from the tail flows into the ischiadic veins or into the advehent renal veins. Reptiles possess only a posterior pair of lymph-hearts; they are placed near the root of the tail against the ends of one of the transverse processes. In snakes they lie in a space protected by the ribs and transverse processes of the original sacral vertebrae. Lymph glands proper are not developed in reptiles, except in the ,shape of the so-called mesenteric' gland of crocodiles.

Blood. The red corpuscles are invariably oval, and, since they still possess a nucleus, biconvex. Numerous measurements have been made by G. Gulliver (P.Z.S., 18 45, pp. 93 -102), their long and short axes range between 0.015-0.023 and 0009-021 mm. respectively.... That means to say they are very much larger than those of mammals, considerably larger than those of most birds, and in turn much smaller than those of amphibia.

Digestive System. Teeth. - All the groups of recent reptiles have teeth, except the tortoises, which have lost even embryonic traces of them. In the under jaw they are restricted to the dentary bones. In the upper they are almost universal in the maxilla and premaxilla, although the latter has lost them in most of the snakes. The pterygoids are toothed in most snakes and in a few lizards, e.g. Lacerta and Iguana. The palatines are toothed in Sphenodon and in some lizards.

Only the young of Sphenodon and the chameleons have a few small teeth on the vomer. The teeth themselves consist of dentine with a cap of enamel and with cementum around their base. In the crocodiles they are planted into separate alveoles in the maxilla, premaxilla and under jaw. In lizards they are either pleurodont, i.e. they stand in a series upon a longitudinal ridge which projects from the lingual side of the supporting bone, or they stand upon the upper rim of the bone, acrodont. In either case they are, when full grown, cemented on to the bone. Acrodont are amongst lizards only the Agamidae; the Tejidae are intermediate, almost acrodont. All the snakes and Sphenodon are acrodont. The latter is in so far peculiar as its broadbased, somewhat triangular teeth are much worn down in old specimens; originally there are several in the premaxilla, but the adults bite with the somewhat curved-down portions of the premaxillaries themselves, or with what remains of the anchylosed bases of the original teeth, which then, together with the bone, look like a pair of large chisel-shaped incisors. The lateral edges of the palatines of Sphenodon likewise carry teeth, those of the mandibles fit into a long slit-like space between the palatine and the maxillary teeth. This is a unique arrangement. Further, it is surprising that in this old, Rhynchocephalian type the supply of teeth has become exhausted, whilst in the other recent reptiles the supply is continuous and apparently inexhaustible. The new teeth lie on the lingual side of the old set, and long before the new tooth is finished part of the base of its older neighbour is absorbed, so that the pulp-cavity which persists in nearly all reptilian teeth becomes free. Ultimately the old tooth is pushed off and the new is cemented into its place. In the crocodiles it has come to pass that several sets of teeth are lodged more or less into one another's bases. Where crocodiles and alligators collect habitually the ground is sometimes found strewn with thousands of teeth, large and small, every creature shedding about seventy teeth many times during its long life.

Some or all teeth of various families of lizards and snakes have a more or less pronounced groove or furrow along their anterior convex curve. The usefulness of this furrow in facilitating the entering of saliva into the bitten wound is merely incidental, but this preformed feature has in many snakes been improved into a fearful weapon. In the Opisthoglypha a few of the most posterior teeth in the maxilla are enlarged, have deeper furrows, and lie in the vicinity of the poison ducts. In the Proteroglypha one or two of the most anterior maxillary teeth are enlarged and furnished with a deep groove for the reception of poison. In the Solenoglypha or Viperidae the 2 enlarged teeth of the Opisthoglypha have moved to the front, owing to reduction of the anterior portion of the maxilla. The latter, much shortened, moves with the firmly anchylosed poison fang upon the prefrontal as its pivot, being pushed forward, or " erected," by the ectopterygoid bone, which connects it with the pterygoid, and this in turn can be moved forwards and backwards, together with the quad rate. (See fig. 24, skull of Vipera nasicornis and the diagram of the mechanism in article Snakes.) In the still unfinished fang the furrow is open, later the edges close together and the end of the duct of the gland itself is surrounded by the substance of the growing basal portion of the tooth, so that the furrow is converted into a canal continuous with that of the gland. The poison is now sure to be projected into the very deepest part of the wound with the precision of a surgical instrument. The Proteroglypha, with their long, non-erectile maxillae, bite, or, like Elaps, deliberately chew their victim; the Viperidae rather strike with the mouth widely open. The teeth of snakes and lizards are often of irregular size; but it is rare that a kind of differentiation into incisors, canines and molars occurs. In many lizards, especially in Iguanidae, some teeth are multicuspid, trilobed, or somewhat serrated; in Tiliqua, universally known as Cyclodus, most of the hinder teeth are roundish crushers.

Lizards and snakes are born with an " egg-tooth " which is lost a day or two after hatching. Its function is the filing through of the eggshell. This tooth, always unpaired, is in Tropidonotus natrix one millimetre long and half a millimetre broad at its base, which rests upon a middle depression of the premaxillary bone; it stands forward above the mouth and is curved upwards. In crocodiles and tortoises the same effect is produced by another organ, which, as in birds, lies well outside the mouth on the top of the end of the snout and consists of a little cone of calcified epidermis.

Tongue

The tongue of the crocodiles is very broad and flat, and with nearly its whole broad base attached to the floor of the mouth; however, in its whole circumference its edge is well marked, and it arises on its hinder border as a transverse fold which meets a similar fold descending from the palate in front of the posterior nares. By these folds the mouth can be completely shut off from the nasal passages into the trachea. The upper surface of the tongue contains several dozen large flat papillae, each with a central pit-like opening; it is not known whether they are gustatory organs. Besides scarce mucous glands on the tongue, there is an absence of salivary glands in the mouth. The tongue of tortoises is likewise short, broad, and not protractile, and there appears to be only a sublingual gland; the surface of the tongue is covered with velvety papillae in the terrestrial, with larger folds in the marine Chelonians. In the Lacertilia the tongue presents a number of variations which have been referred to as diagnostic characters of the various families of Lizards. The chief modifications are the following: Either flat and broad, not protractile, e.g. Agamidae; or the body of the tongue is somewhat cylindrical, elongated, and the whole organ can be protruded; lastly, the anterior half of the tongue, which can be protruded, is retractile or telescoped into the posterior portion, e.g. Anguidae. In nearly all cases the posterior dorsal end of the body of the tongue is well marked off by a margin raised above the root, a character which does not occur in any snake. The upper surface is either smooth or curved with velvety, flat, or scaly, always soft, papillae. In the majority the tip of the tongue is bifid, either slightly niched or deeply bifid. The tips contain tactile corpuscles, although sometimes covered with a horny epithelium. The most specialized is the tongue of the chameleon. The body of this tongue is very thick, clubshaped, fleshy and full of large mucous glands which cover it with a sticky secretion. The base or root is very narrow, composed of extremely elastic fibres and supported by a much elongated copular piece of the hyoid. This elastic part is, so to speak, telescoped over the style-shaped copula, and the whole apparatus is kept in a contracted state like a spring in a tube. A pair of wide blood vessels and elastic bands extend from the base into the thick end, which in an ordinary chameleon can be shot out to a distance of about 8 in.

The tongue of the snakes is invariably slender, smooth and almost entirely retractile into its posterior sheath-like portion. It is always bifid and contains many tactile and other sensory corpuscles by which these creatures seem to investigate. The tongue is always protruded during excitement. How this is done is not very obvious, since the hyoid apparatus itself is much reduced. There is a niche in the middle of the rostral shield to permit protrusion of the tongue whilst the mouth is shut, and probably herewith is correlated the almost universal absence of teeth in the premaxilla. The tongue and the larynx are placed very far forwards in the mouth and, during the act of swallowing, the larynx approaches the chin, or it may even protrude out of the mouth to secure breathing during the often painfully protracted act.

Of Glands, sublingual glands are of general occurrence in reptiles; they open near the root or in the sheath of the tongue. Labial glands seem to be absent in crocodiles and tortoises, but upper and lower labial glands exist in lizards and snakes, generally in considerable numbers. Heloderma is the only lizard in which some of these glands - those along the lower jaw - produce a poisonous secretion, each small gland conducting its secretion towards the base of one of the somewhat furrowed teeth. In the snakes, upper and lower labial glands are well developed for salivation. It is the upper series which attracts our interest by its eventual modification into the deadly poison glands. Probably the saliva of most snakes, like their serum, possesses toxic properties. In most of the harmless Colubrine snakes the glands extend in a continuous series from behind the premaxilla along the whole of the upper jaw, with numerous openings. In the Opisthoglypha a gradual differentiation takes place into an anterior, middle and posterior portion; the middle, extending from below and behind the eye backwards, is the thickest and yellowish in colour; behind it follows a small portion, reddish grey like the anterior portion, with which it is more or less continuous below the middle complex. Thus, still rather indifferent, is Dryophis. In Dipsas, e.g. D. fusca, the middle portion has become predominant; some of its enlarged ducts lead to the pair of posterior, enlarged and well-grooved, maxillary teeth. It is this middle portion which becomes the characteristic poison gland with one long duct. The gland itself retains its position; all the other upper labials, except the anterior series, abort. In the Viperidae the poison duct opens near the base of the perforated fangs, which, owing to the shortening of the anterior portion of the maxilla with its teeth, have come to be the only teeth in the upper jaw. In the Elapine, still more in the Hydrophine snakes, the position of the gland and its duct is the same, but the duct has been carried past the smaller harmless teeth which stand in the maxilla and open at the base of the anterior maxillary teeth. The effect is the same, although the poison fangs are not homologous, in the one case the most posterior, in the other the most anterior, of the maxillary series. In Doliophis, one of the Malay genera of Elapine snakes, each poison gland sends an enormously elongated recess far into the body-cavity. (For Some other details see Snakes; Viper; and Rattlesnake.

The best account of the buccal glands and teeth of poisonous snakes is that by G. S. West, P.Z.S., 1895, pp. 812-826.) Stomach, &c. - In lizards and in Sphenodon the wide pharynx and oesophagus passes gradually into the stomach, which is FIG. 38. - Two Aspects of a Tooth of Heloderma horridum (after Bocourt). I, antero - internal aspect of the tooth, showing a very deep longitudinal groove; 2, postero - external aspect of the same tooth, showing a very faint longitudinal groove.

more or less spindle-shaped, never transversely placed. The walls of the stomach are thrown into longitudinal folds which contain the specific gastric glands, whilst glands are absent in the oesophagus, excepting scattered and very simple slime glands. The circular muscular fibres of the stomach are much stronger than the longitudinal fibres. The end of the stomach is generally marked by a pyloric valve. The walls of the mid gut are said to be devoid of glands. The end gut, marked by a circular valve, is considerably wider and there is a caecum, mostly left-sided, largest in leaf-eating lizards, rarely absent, as, for instance, in Anguis. The absorbent portion of the rectum is always strongly marked off from the cloaca by a circular fold or sphincter, which projects into the widened coprodaeum of the cloaca. In those lizards which, like Varanus, have no urinary bladder, there are two successive sphincters, marking off two chambers, one, the upper or innermost, for the reception of the faeces, the lower for that of the urine. In adult crocodiles the stomach is transformed into a gizzard; it is more or less oval, with a wide fundus and with two opposite apo-neurotic or tendinous disks whence radiate the muscular fibres. The muscular walls remain, however, comparatively thin, like those of birds of prey. There is a distinct pyloric stomach and then follows the pylorus. The inner lining of the stomach is velvetlike with numerous gastric glands which form groups with netlike interstices. There is a distinct duodenal loop which contains the pancreas. The more convoluted mid gut is lined with net-like meshes which farther back assume a longitudinal zigzag arrangement; towards the end gut the walls become quite smooth, but in the end gut the walls again show a very narrow-meshed structure. None of these folds of the mid and hind gut is said to contain digestive glands; they seem to be entirely absorbent. The oesophagus of most tortoises shows longitudinal folds with very numerous mucous glands. In the Chelonidae the pharynx and adjoining part of the gullet are covered with little tubercles upon each of which opens a small gland. Farther down they give way to large, more or less conical papillae, which assume a considerable size, point backwards, and are covered with a somewhat horny epithelium. Similar conical, horny papillae exist also in Sphargis, in which the oesophagus, moreover, makes a long loop half round the stomach before passing into it, an absolutely unique feature. The transition into the stomach is quite gradual. The latter is strongly muscular, partly transversely placed, and possesses often a very distinct pyloric stomach. In Chelone conical papillae extend into the cardiac portion. In the majority of tortoises the inner lining shows longitudinal folds with numerous small glands, mucous and gastric, but their distribution differs much in the various families and even genera. The lining of the mid gut shows either longitudinal folds or a network, without glands, except in some cases, Lieberkiihn crypts, e.g. in Trionyx, not in Testudo and Chelone. The hind gut begins suddenly, but there is no caecum; its inner walls contain numerous glands in Testudo, Emys, not in Chelys, Trionyx, Cinosternum. In the snakes the oesophagus is very thin-walled and passes imperceptibly into the stomach, which continues in a longitudinal direction, scarcely wider in the middle. Its muscular coating is surprisingly weak. There is a small pyloric portion. Mucous and especially long-bodied gastric glands are numerous. The wall of the mid gut carries numerous papillae variably arranged, velvet-like, or densely crowded little blades supported by longitudinal or by meshy folds. The hind gut is short, often constricted into several successive chambers, mostly smooth inside; there is a short, rather wide caecum which seems best developed in Viperidae; sometimes absent. The total length of the snakes' gut is always short, there being only short folds possible or necessary in the body cavity, which itself is of extraordinary length. Yet, while in Typhlops the gut is almost straight, it forms numerous convolutions in Tortrix. Whilst in all other reptiles the gut, at least stomach, liver and mid gut, are suspended by the mesentery from the vertebral column and hang free into the body cavity, in some snakes, especially often described in Boa and Python, the body cavity is cut up into numerous spaces, by peritoneal folds which connect neighbouring twists of the canal into bundles and attach them to the ventral surface of the body-wall. Probably the gut is thereby secured against dislocations in adaptation to the peculiar twisting contortions of the body, especially in the act of climbing. The mesentery of reptiles is remarkable for the possession of smooth, non-striated, muscular fibres. In most lizards, not in other orders, the peritoneum so far as it covers the abdominal cavity shows a deep black pigmentation; this pigment is situated in the connective tissue, not in the epithelial layer; it stops suddenly towards the thorax. In some lizards, e.g. in Anguis, the black pigment extends, more or less scattered, upon the mesentery and thence upon the intestines. The same pigment colours the pharynx with its recesses entirely black in many lizards. There is no compensating correlation between this internal pigment and that in the outer skin.

The Liver of lizards is more or less bilobed; more so in crocodiles; while in tortoises the broad right and left lobes are connected by a narrow isthmus. In the snakes it is much elongated and extends from the heart backwards along the right side of the oesophagus, closely connected in its long course with numerous short branches into, or from, the inferior vena cava and the portal vein. A gall bladder is always present. The ducts into and from the cyst sometimes form a complicated network, for instance in Varanus (F. E. Beddard); the bile is carried by one or more ducts into the duodenal portion of the mid gut. The microscopic structure of the reptilian liver has been compared with that of monotremes by M. Fiirbringer.

The Pancreas is a compact body attached to the duodenal region, which surrounds it by a loop i i the crocodiles, as is the case in birds and mammals.

The Cloaca of the reptiles shows a great advance upon the simple batrachian arrangement. It is no longer one common chamber, but consists of three successive chambers with the further tendency of separating the temporary retention and the passage of the faecal, urinary and genital products from each other. The arrangement is simplest and most typical in the lizards. There is first the proctodaeum or vestibulum of the cloaca, epiblastic in origin. Its outer boundary is formed by the cloacal lips, covered so far by the usual scaly integument. Just within this chamber arise the paired copulatory organs, and, when they are present, as in Sphenodon and snakes, the two anal glands. Secondly, the urodaeum, middle or urinogenital chamber, hypoblastic in origin. It is separated from the proctodaeum by a more or less circular fold which is provided with sphincter muscles, which form the true vent, and this is always round; whilst the outermost opening in lizards `and snakes is a transverse slit. Farther inwards, headwards, the urodaeum is shut off by another circular fold, generally very well marked, especially in its dorsal half, which is higher and thicker. Into the dorsal, and innermost, recess of this urodaeum open the genital and urinary ducts; on the ventral side arises the urinary bladder. The whole chamber is always empty, being only a passage room, and in the female the copulatory chamber. The urine is of course collected in the bladder; when this is absent the fluid is pressed into the third chamber, the coprodaeum, which is often subdivided into two, or even three, successive rooms by circular folds. This coprodaeum serves for the temporary storage of the faeces, eventually mixed with the urine. Micturition and defaecation are in most lizards two successive separate acts.

The snake's arrangement is a side-departure of that prevailing in lizards. The urodaeum is transformed into a dorsal recess into which open above the oviducts, while the ureters open below, in the caudal corner. A horizontal fold imperfectly shuts off the wide urino-genital chamber or recess from the ventral half of the original urodaeum. The coprodaeum is marked above and below by strong sphincters. There is no urinary bladder.

In crocodiles the protodaeum is rather shallow, but long; from its ventral wall arises the unpaired copulatory organ, the basal investing membranes of which continue into the ventral half of the uro-proctodaeal fold, near which open the male ducts. Very young crocodiles possess a typical middle chamber or urodaeum, into the dorso-lateral corners of which open the ureters, but soon the strong circular fold between urodaeum and coprodaeum disappears completely, so that both chambers now form one large oval room, which is used solely for the storage of the urine, there being no bladder. The faeces are kept in the not specially dilated rectum.

The cloacal arrangement of the Chelonia is a further development of early crocodilian conditions, but it has become rather complicated and shows a surprising resemblance to that which still prevails in the Monotremes. The proctodaeum is deep and very long, especially in the males. From its innermost and ventral walls arises the large copulatory organ. From the urodaeum is separated off a deep ventral recess into which open the ureters and the genital ducts, and it is continued by a long neck into the large bladder. Between the dorsal wall of this recess and the ventral wall of the main portion of the urodaeum arises a horizontal fold which, diverging, is continued on to the investing skin of the penis, helping to form the edges of the deep longitudinal furrow on its morphologically dorsal surface. If the lips of this furrow were closed, urine and all the genital products would pass through this urethral canal, but in reality only the semen is conducted through it (the furrow during the state of turgescence being transformed into a closed tube), whilst urine and eggs escape through the wide slit near its inner end. This is an arrangement almost the same as that of Ornithorhynchus. The urodaeum is separated from the rectum by a strong sphincter, and there is, as in the crocodiles and mammals, no special coprodaeum. The Chelonian urodaeum is further complicated by the occurrence of a pair of large anal sacs, thin-walled diverticula on the dorsal side. Such sacs, not to be confounded with the anal glands of other reptiles, exist in many water tortoises, especially in the Chelydidae, also in various aquatic Testudinidae, e.g. Emys, in Platysternum, and sometimes in Trionyx; they are absent in the Chelonidae and in the typically terrestrial tortoises. These sacs have highly vascularized walls and a considerable layer of circular and longitudinal non-striped muscular fibres; their inside is sometimes villous, never glandular. They are incessantly filled and emptied with water through the vent, and act as additional respiratory organs, like a kind of water lungs. When such a tortoise is suddenly taken out of the water it squirts out a stream of water, which is not, as is usually supposed, the urine from the bladder.

In connexion with the cloaca may be mentioned the frequent occurrence of peritoneal canals. In the tortoises their abdominal openings are situated in a recess of the peritoneal cavity close to either side of the neck of the bladder; in the females they extend as funnels, generally blind, into the cloaca on or near the base of the clitoris. In the males they extend, without having communication with the cavities of the corpora cavernosa, and without ramifications, as canals along the dorsum penis and either terminate blindly in the glans (Testudo, Chelone), or they open, each by a small orifice, in the groove at the base of the glans. In crocodiles these canals are short and open near the base of the copulatory organ, protected by a small papilla. They are present in both sexes, but are still closed in newly hatched and very immature specimens. In an adult Nile crocodile they are wide enough to pass an ordinary lead pencil. The function of these outlets from the body cavity is obscure. In Sphenodon the writer has found them as closed funnels which project as soft papillae into the proctodaeum a little to the right and left and caudalwards from the urino-genital papillae.

Urinary Organs. The kidneys of the reptiles show, like those of the birds and mammals, a considerable advance upon those of the Batrachia. They are, in the adult, represented entirely by the metanephros; the segmental tubes have no longer any nephrostomes opening into the body cavity, not even during any time of their development, and it has come to a complete separation of the efferent genital ducts from the kidneys and from their ureters. Yet these differences are but of degree, there being a continuous bridge from Batrachian to Lacertilian conditions. In Lacerta, for instance, in which these features have been studied most thoroughly, the mesonephros continues as the only functional excretory organ during the first year of the young creature until and during its first hibernation, when the formation of the metanephros takes place, and with it the complete separation of the vasa deferentia from the kidneys. Until then the segmental canals remain in the male as common carriers of semen and urine, at least morphologically, not physiologically, since in the immature there is no occasion for the conduction of semen. The kidneys of these young lizards show precisely the same arrangement as that of the Batrachia, excluding the Discoglossidae.

Clearly the metanephros is developed from, and is part of, the posterior portion of the mesonephros, the glomeruli of which no longer open into the segmental duct, but become connected with a new canal, the future ureter, which sprouts from the distal portion of the segmental duct and grows headwards. Or let us put these important changes in another way. Since there are originally several segmental ducts (permanent in the male newt) which tailwards more and more lose their connexion with the testes, until - in the posterior portion of the mesonephrosthey become entirely urinary ducts, the hindmost of these sprouts (in lizards postembryonic, much earlier in birds and mammals) independently, but at the same time as the neighbouring mass of the mesonephros, the growing glomeruli of which then connect with the sprouting processes of the ureter. Phylogenetically and ontogenetically it is evident enough that the kidneys are essentially one organ, the anterior portion of which is the oldest and decays, whilst farther backwards new and more differentiated portions continue to grow. Pro-, mesoand metanephros and successive wave-like stages of the same organ with morphological and functional continuity, until the next, improved portion is ready. It is important that in the Discoglossidae, especially in the male Alytes, an arrangement has come to pass which much resembles that of the Amniota. The mesonephros has, by a simple contrivance, become a metanephros, provided we define the former as a kidney which is still connected with true segmental ducts.

The supra-renal bodies, adrenals, head-kidneys or Nebennieren, are yellowish bodies which lie more in connexion with the generative glands than with the kidneys, always closely attached to the vena cava posterior just above the kidneys. They are very elongated in the snakes, in a 10-foot python they measure about one inch in length; they are flattened in tortoises, roundish in crocodiles.

In all reptiles the kidneys are retroperitoneal, and they do not project into the body cavity. Their position is different in the various groups, and their general shape is much affected by the shape of the body. In the Ophidia they are much elongated, and of course far in front of the pelvic region, which has been moved to the cloaca. They are placed asymmetrically, the right extending farthest forwards. They consist of many transverse lobes, sometimes in such a way as to appear spirally twisted. Each terminates considerably in front of the cloaca. Each ureter begins at the anterior end of the kidney, and thence proceeds on its inner and dorsal border, receiving ducts from the interspaces of the numerous lobes. In the male each ureter opens upon a papilla, together with the vas deferens; in the female the ureter is joined by a blind canal, the vestige of the male duct. No snake has a urinary bladder. The urinary excretion is white, chalky, consisting mainly of uric acid in crystals, with very little fluid.

In the Lacertilia the kidneys are more posteriorly placed than in snakes. They lie between the pelvis and the cloaca and are generally close together, sometimes partly fused with each other. Only in the Amphisbaenids the right kidney extends more forwards. They are usually transversely furrowed. The ureters open dorso-laterally into the urodaeum upon papillae as in the snakes. In the females the remnants of the segmental ducts, or vestigial representatives of the vasa efferentia, are often of considerable length, persistent in chameleon and Uromastix, much reduced in geckos, or disappearing with age as in Lacerta. The urine of most lizards contains much solid uric acid, which is retained in the urodaeum and voided as a rather solid, white mass, not united with the faeces. Those which have a greater amount of fluid urine have a bladder which receives the fluid portion. The opening of this bladder is on the ventral side of the cloaca, not in direct connexion with the ureters. The bladder is very rarely absent, e.g. in Varanidae and Amphisbaenidae.

The Crocodilia have the kidneys placed below the pelvis; their surface shows meandering convolutions separated by furrows. The ureters are for the greater part of their length deeply sunk into the substance of the kidneys, which they leave near the hinder ends, to run freely for a short distance along the dorsal sides of the cloaca, and they open, each separately, and away from the vasa deferentia, into the dorsal side of the urodaeum, which, together with the coprodaeum, forms a large oval chamber, and this being filled with the very fluid urine, functionizes instead of the absent bladder.

In Chelonia the kidneys lie in the pelvis, short and thick, more or less trihedral; the surface is marked with many shallow meandering grooves and fewer deeper furrows. Each ureter, composed of several large successive canals, leaves its kidney near the inner hinder end, and then runs free for a short space, crossing the gut to open into the neck of the urinary bladder, which arises ventrally out of the urodaeum, which itself has become a recess of the cloaca. The bladder is large, often more or less two-horned, attached to the pelvic wall by a peritoneal fold, and it contains very fluid urine.

The kidneys of Sphenodon are very small and far removed from the generative organs. The ureters open, each close to the vas deferens of its side, beneath a little papilla, on the dorsal side, rather near the midline of the urodaeum, whence arises a long-necked bladder.

Reproductive System. The Ovaries are always in pairs, placed headwards at a distance from the kidneys in Sphenodon, lizards and snakes; in the latter the right ovary lies farther forward. In tortoises, and especially in the crocodiles, where they are very long and much twisted or lobated, they are situated close to the kidneys and even accompany them. The ovaries of lizards and snakes contain many and large lymph spaces; those of the other reptiles are much denser in structure. The ripening eggs always cause them to assume the shape of a bunch of grapes. The oviducts are each held by a peritoneal fold which arises from near the dorsal midline. The abdominal ostia are long slits and are turned towards the side, away from the ovaries. The walls of the ducts gradually become thicker, glandular and much folded. Whilst the ripe eggs, often in considerable numbers, receive their shell, each egg lies in a separate chamber; in the geckos, which lay only one pair of eggs, the two respective chambers have become permanent features. In Sphenodon each oviduct opens together with the ureter of its side near the dorsomedian line of the urodaeum. In most lizards the two oviducts and the two ureters have four separate openings in the dorsal wall of the rather deep dorsal recess of the urodaeum. But in Lophura both oviducts unite (like the ureters) and have only one opening, which is placed a little nearer towards the pelvis than the urinary opening, but they are divided by a longitudinal septum which extends almost to their common orifice. In the snakes the oviducts likewise open into the dorsal recess, sometimes by a common ostium, which is provided with a strong sphincter. The whole recess acts like a vagina for the reception of one of the copulatory organs. The oviducts of the crocodiles open in a decidedly ventral position, on either side close to the base of the clitoris, a considerable distance from the openings of the ureters. In the tortoises the oviducts open separately into a wide ventral urino-genital sinus, at the base of the neck of the bladder.

The Testes correspond in position with the ovaries; in snakes and Amphisbaenids the right is placed farther head wards than the left. The usual shape is elongated, sometimes pointed forwards. The Epididymis is sometimes of the same size as the testis and then consists of many meandering convolutions of the vas deferens which is composed of several canals from the testis. The convolutions are held together by a peritoneal lamella. Towards the cloaca they become much smaller and shorter, and the vas deferens passes along the median side of the ureter. In Sphenodon these open separately, each near and below the same papilla near which opens the ureter of the same side. In most lizards the vas deferens unites with its ureter into one short canal which opens beneath or upon a small papilla in the upper corner of the urodaeal recess, far away from the penis. In snakes vas deferens and ureter of each side are likewise commonly united. In the crocodiles each vas deferens passes from the dorsal side of the cloaca to the ventral side, not accompanied by the ureter, and opens into the blind sac which forms the basal continuation of the deep groove on the dorsal side of the penis. In the tortoises the epididymis is very large and the vas deferens is also much convoluted; each opens separately near the neck of the large urinary bladder close to the backward continuation of the deep longitudinal groove of the copulatory organ.

Remnants of the Miillerian ducts run parallel with the vasa deferentia, and similar remnants of the Wolffian ducts accompany the oviducts in crocodiles and tortoises, least degenerated of course in young specimens. Such reciprocal vestiges occur most likely also in lizards, and in female snakes a vestige of the male duct joins its ureter. In a nearly adult male Sphenodon the present writer missed the female remnants.

The copulatory organs show very important modifications. Sphenodon is the only recent reptile which is devoid of such an organ; its imperfect substitute is an unpaired, thin, but high membranous fold which arises from the dorsal middle of the circular fold between urodaeum and coprodaeum. During copulation this part of the cloaca is probably everted to secure conception, a striking resemblance to the arrangement found in the Caecilia. The organs of all lizards and snakes are paired, in their quiescent state withdrawn into deep pockets which open on the right and left posterior corners of the proctodaeum or outer chamber of the cloaca, which for this reason has assumed the shape of a transverse slit in all lizards and snakes. Hence these have sometimes been called Plagiotremata. Each organ can be everted and tucked in like the finger of a glove, a muscle being attached to the inside of the apex; when everted, the muscle extends through the length of the organ; each muscle arises from the ventral side of several transverse processes of the tail vertebrae, at a considerable distance from the cloaca. In the embryo each FIG. 39. - Male copulatory organs of Lacerta agilis (after Leydig).p l, p2, organs of right and left sides - between them is the anal aperture; pp, preanal plate.

organ arises as a conical protuberance, or papilla, which projects out of the vent. Later it becomes inverted. Probably this ontogenetic feature recapitulates the phylogeny of these organs, which have to be looked upon as swelling flaps or portions of the walls of the cloaca which were protruded during copulation, and which in time borrowed, and specialized, muscular fibres from the ventral tail muscles. On the outer everted side of each organ is a furrow for the reception of the semen. The apex is either single or more or less deeply bifurcated, each arm being followed by the likewise divided furrow. The outer investing membrane of these very muscular erectile bodies is epidermal; often, especially in snakes, provided with numerous papillae, folds or other excrescences. In XXIII. 6 a many snakes these are spiny and hard, but according to Leydig this hardness is not due to a horny substance but to the deposition of calcifying matter. E. D. Cope has investigated the almost endless minor modifications of these penial features and uses them for taxonomic purposes in the snakes. Vestiges of these organs occur in females of snakes and lizards. Close to these organs of the snakes lies a pair of anal glands of some size, which pour their very offensive secretion through an opening close to the base of each penis. The same glands occur in the same position in Sphenodon, which has no copulatory organs, and in crocodiles they appear as evertible musk glands. Hence J. E. V. Boas, not knowing of their existence in both sexes of snakes, tried to homologize them with the paired penes of reptiles, an error which has been repeated in C. Gegenbaur's Lehrbuch, vol. ii. p. 533.

The crocodiles and tortoises possess a single, median copulatory organ; it lies on the ventral or anterior end of the cloaca, the outer opening of which is therefore a longitudinal slit, hence the term ucthotremata. In the crocodiles the organ is attached to the caudal corner of the ischiadic symphysis by a strong and roundish fibrous band, which arises single from the ventral sides and forms partly the continuation of the two fibrous halves of the organ; the bulk of the crura, comparable to corpora cavernosa, is not attached to the pelvis, as generally stated, but projects backwards towards and into the pelvic cavity. This portion is especially rich in venous cavernosities. The outer coating of the glans possesses various papillary projections, which are furnished with sensory, hedonic corpuscles. On the morphologically dorsal side of the organ, not on the dorsum penis, is a deep groove which ends towards the crura in a blind sac, into the farther corner of which open the vasa deferentia. In a full-grown Nile crocodile the whole organ is about to in. long. In young females up to a total length of 3 or 4 ft. the clitoris is nearly of the same size as the male organ, but it remains stationary and appears very small in large specimens.

The organ of the tortoises is essentially of the same type as that of the crocodiles, but it is nowhere directly attached to the pelvis or to any other skeletal part. The whole organ, when withdrawn, lies in a ventral, long recess of the wide outer cloacal chamber, and its crura extend so far back as to form the continuation of the ventral and lateral walls of the recessus which is continued into the neck of the urinary bladder. Its orifice and those of the seminal ducts are enclosed by the walls of the deep groove which runs along the underside of the organ. This is always of considerable size, surprisingly large in Trionyx. The clitoris is small, sometimes tiny.

The sexual act is extremely prolonged in Chelonians and still more so are the preliminaries, but in crocodiles it is the deed of a few seconds. Lizards and snakes insert only one side.

There remains the question whether the unpaired organ of the crocodiles and tortoises, which is the prototype of the mammalian organ in every essential point, and the paired organs of the lizards and snakes, are to a certain extent homologous organs in so far as they can both be derived from the same indifferent condition. With this view we assume that originally the protrusible walls of the outer cloacal chamber became specialized into a right and left imperfect intromittent organ, that subsequently, in lizards, those hemipenes were shifted back towards the tail and were henceforth bound to develop separately, while in the crocodiles, tortoises, mammals and birds the two primitive lateral evertile flaps approached each other towards the ventral anterior side of the cloaca, and that this led to a fusion, beginning probably at the basal part, which at the same time was farther withdrawn from the surface and secured the reception of the sperma from both vasa deferentia into one canal. This hypothesis has been objected to by Boas, but accepted by Gegenbaur (p. S38) after having been rejected on p. 533 of his Lehrbuch. The Fat bodies belong at least physiologically to the generative system. They are placed outside the peritoneum. In lizards they appear as two masses in the pelvic region, the black peritoneal lining covering only their dorsal side. They consist of a network of arteries and connective tissue, the meshy spaces of which are filled with " fat "; they each receive an artery from the femoral vessel which enters them in the inguinal region; the veins collect into the abdominal. In snakes the fat bodies are very long, extending from the cloaca to the liver. Tortoises seem to have only traces of them, but in Sphenodon and in crocodiles they resemble those of lizards. - The peculiar organ suspended from the right abdominal wall of crocodiles, variously mentioned as mesenteric gland or body, or fatty spleen, by Butler, is possibly related to the same category. The fat bodies of reptiles are sometimes vaguely alluded to as hibernating bodies; like the fat bodies which are attached to the generative glands of Amphibia they do not become reduced during the eventual hibernation but are largest before the pairing season, by the end of which they are exhausted, looking reddish or grey after the loss of their stores of fat and probably other important contents The Embryonic Development. Fertilization of the egg always takes place internally, and the egg containing a large amount of food-yolk is of course meroblastic. It is sufficient to mention that many lizards, some chameleons and many snakes (not Sphenodon, geckos, crocodiles and Chelonians) retain their, in these cases very thin-shelled, eggs in the oviducts until the embryo is ready to burst the egg-membrane during the act of parturition or immediately after it. Such species are usually called ovoviviparous, although there is no difference between them and other viviparous creatures, for instance the marsupials. The majority of reptiles are oviparous and the egg is enclosed in a strong parchment shell, with or without calcareous deposits. Only gas exchange can take place between such an egg and the outside, and it loses by evaporation, whilst in the batrachian egg various other exchanges are easy through the thin membrane. The salamander embryo, within its thin egg-membrane, even grows to a size many times larger than the original egg, it does not only breathe, but it is also nourished through the gills, and by some means or other the waste products are partly eliminated without filling the bladder. The amphibia are born as larvae and live as such for a long time, often in a most imperfect condition. Nothing of all this applies to the reptile, which leaves the egg as a perfect little imago. A great amount of yolk supplying the material, and a large " bladder " to receive the waste products and to act as respiratory organ, have made this possible. That the allantois and the amnion behave precisely in the same way in the mammals with their much reduced yolk, only testifies to the superior value of these organs, and after all there is no difference in this respect between a monotreme and a reptile. These two organs seem to have come into existence with the reptiles and constitute the most reliable diagnostic feature between higher and lower vertebrates. All reptiles, birds and mammals have a navel, a feature unknown and impossible in Batrachia and fishes. A few remarks on these important embryonic organs may not be superfluous, especially concerning their possible origin.

Whilst the urinary bladder of the Batrachia remains within the body throughout the embryonic stage, this organ undergoes in the higher vertebrates, reptiles, birds and mammals, considerable modifications, and it assumes, henceforth as Allantois, new important functions besides that of being the receptacle of the embryonic urine. The development of the Allantois is in intimate causal connexion with that of the Amnion. All the Allantoidea are also Amniota and vice versa, but the term Amniota is preferable, since the basal portion of the Allantois remains in the adult as the urinary bladder, as an organ henceforth equivalent to and homologous with that of the Anamnia. The primary feature seems to be the allantois which leaves the body cavity, remains without the amniotic folds, even after these have enclosed the body within the amniotic bag, and then spreads nearly all over the inner side of the egg-shell. Having thus come into the closest possible contact with the atmospheric air, the vessels of the allantois can exchange their carbon dioxide for oxygen and the allantois becomes the respiratory organ of the embryo. Herewith stands in direct correlation the complete absence of any internal and of external gills in the embryonic reptiles. The blood vessels of the allantois are fundamentally the same as those of the batrachian bladder, namely, branches from the pelvic arteries (later hypogastrics) and veins which return from the base of the bladder to the abdominal wall and thence to the liver.

In the normal reptilian egg, surrounded by its non-yielding shell, space is absolutely limited, and whilst the yolk is being diminished and increased secretion of urine distends the bladder, this soon protrudes out of the body cavity proper into the extra-embryonal coelomatic space between the true amnion and the false amnion or serous membrane. It fills this space so far as the yolk-sac allows it. It seems reasonable to suppose that this growth of the allantois has been one of the causes of the caudal amniotic fold; the sinking of the embryo into the space of the diminishing yolk-sac is no doubt another cause, but the fact remains that the amnion is the chief hindrance to the closing of the body-wall at the region of the future navel.

The life-histories of embryonic development are the domain of the embryographers. They are the imperfect accounts of the ways and means (often crooked and blurred, owing to short cuts and in adaptation to conditions which prevail during the embryonic period) by which the growing creature arrives at those features which form the account of the anatomical structure of the adult. Comparative anatomy, with physiology, alone lead through the maze of the endless embryonic vagaries and afford the clues for the reconstruction of the real life-history of an animal and its ancestry. For detail the reader is referred to numerous papers quoted in the list of literature, and to the various text-books, above all to the Handbuch d. vergleichenden Entwicklungsgeschichte d. Wirbelthiere, edited by O. Hertwig, Berlin.

Authorities On Anatomy: Bibliography.-The appended list of papers (many with shortened titles) represents but a fraction of the enormous literature dealing with the anatomy of reptiles. Special stress has been laid upon the more recent publications. A great amount of information, general and detailed, is contained in Bronn's Klassen u. Ordnungen d. Thierreichs, the three volumes concerning reptiles having been written by C. K. Hoffmann (Leipzig, 1878-1890) E. D. Cope's Crocodilians, Lizards and Snakes of North America, U.S. Nat. Mus., Washington, 1900; H. Gadow's " Amphibia and Reptiles," vol. xiii. of The Cambridge Natural History (London, 1901); above all in C. Gegenbaur's Vergleichende Anatomic d. Wirbelthiere (Leipzig, 1898-1901).

Skeletal.-J. F. v. Bemmelen, " Schaedelbau v. Dermochelys coriacea," Festschr. f. Gegenbaur (1896); E. Gaupp, " Morphologie d. Schaedels," Morpholog. Arbeiten (1894), iv. pp. 77-128, pls.; ibid. (" Problems Concerning the Skull "), Anat. Ergebn. (1901), x. pp. 847moo'. W. K. Parker," Skull of Lacertilia," Phil. Trans. 170 (1880), pp. 5956 4 0, pls. 37-45; "of Tropidonotus," ibid. (1879), 16 9, PP.385-417, " Crocodilia," Trans. Zool. Soc. (1885), xi. pp. 263-310, pls.; " Chamaeleons," ibid. (1885), xi. pp. 77-105, pls. 15-19.; F. Siebenrock, " Kopfskelet d. Scincoiden, Anguiden u. Gerrhosauriden," Ann. Nat. Hofmuseum (Wien, 1892), vii. 3. Of the enormous, still increasing, literature concerning the homologies of the auditory ossicles, a few only can be mentioned; the papers by Kingsley and Versluys contain most of the previous literature: W. Peters, several most important papers in Monatsber. Ak. Wiss. (Berlin, 21st Nov. 1867, 5th Dec. 1867, 7th Jan. 1869, 17th Jan. 1870, 15th Jan. 1874). H. Gadow, " Modifications of the First and Second Visceral Arches, and Homologies of the Auditory Ossicles," Phil. Trans. 179 (1888), B. pp. 451-485, pls. 71-74; " Evolution of the Auditory Ossicles," Anat. Anz. (1901), xix. No. 16. J. Versluys, " Mittlere u. aussere Ohrsphare d. Lacertilia u. Rhynchocephalia," Zool. Jahrb. Anat. (1898), 12, pp. 161-406, pls. (most exhaustive and careful); ibid., " Entwickl. d. Columella auris b. Lacertiliern," ibid. (1903), 18, pp. 107-188, pls. J. S. Kinzslev, "The Ossicula auditus," Tufts College Studies, No. 6 (1900). E. Gaupp, " Columella auris," Anat. Anz. (1891), vi. p. 107. T. H. Huxley, " The Representatives of the Malleus and Incus of the Mammalia in the other Vertebrata," P.Z.S., 1869. W. K. Parker, " Struct. and Development of Crocodilian Skull," Trans. Zool. Soc. (1883), xi., especially pls. 68 and 69. H. Gadow," Evolution of the Vertebral Column of Amphibia and Amniota," Phil. Trans. (1896), 136, pp. 1-57 (with a list of ninety-three papers). G. B. Howes and H. H. Swinnerton, Development of the Skeleton of Sphenodon," Trans. Zool. Soc. (1901), xvi. pp. 1-86, pls. 1-6. G. A. Boulenger, Catalogue of Chelonians, Rhynchocephalians and Crocodiles, Brit. Mus. 1889; Cat. of Lizards (3 vols., 1885-1887); Cat. of Snakes (3 vols., 18 931896); these volumes contain a great body of osteological observations, ignored by most compilers of anatomical text-books; " Osteol. of Heloderma, and Vertebrae of Lacertilia," P.Z.S., pp. 109-118 (1891). L. Calori, " Skeleton of Varanus, Lacerta," Mem. Acc. Sci. Instit. Bologna (8, 1857, and 9, 1859). E. D. Cope, " Osteology of Lacertilia," Proc. Am. Phil. Soc. (1892), 30, pp. 185-221; " Degeneration of Limbs and Girdles," Journ. Morph. (1892), vii. pp. 223-244. E. Ficalbi, Osteologia del Platidattilo (Pisa, 1882). A. Goette, " Beitrage z. Skeletsystem," Arch. micr. Anat. (1877), 14, pp. 502-620. A. Gunther, " Anatomy of Hatteria," Phil. Trans. (1867), 1 57, pp. 59562 9, pls. S. Orlandi, " Note anatomiche s. Macrosincus," Atti S. Lig. (Geneva, 1894), v. 2; " Skelet d. Scinc. Anguid. Gerrhosaurid," Ann. Naturhist. Hofmus. (1895), x. pp. 17-41; " Skelet d. Agamidae," Sitzb. Ak. Wiss. Wien (1895), 104, pp. 1089-1196. F. Siebenrock, " Skelet v. Brookesia," Sitzb. Ak." Wiss. Wien (1893), 102, pp. 71-118; " Skelet v. Uroplates," Annal. Naturhist. Hofmuseum (1892), vii. pp. 517-536, 1893; " Skelet d. Lacertiden," Sitzb. Ak. Wiss. Wien (1894), 102, pp. 203-292. C. Smalian, " Anat. d. Amphisbaenid," Zeitschr. wiss. Zool. (1885), 42, pp. 126-202. A. Voeltzkow, " Biolog. u. Entwickl. von Crocodilus," Abh. Senckenb. Ges. (1899), 26, pp. 1-150, 17 pls. E. A. Case, " Osteology and Relationships of Protostega," Journ. Morph. (1897), xiv. pp. 21-60. H. Goette, "Entwickl. des Carapax d. Schildkroeten," Zeitschr. wiss. Zool. (1899), 66, PP. 4 0 -434, pls. O. P. Hay, " Morphogeny of Chelonian Carapace," Amer. Nat. (1898), 3 2, pp. 9 2 9-94 8. G. Baur, " Morphol. Unterkiefer d. Rept.,"Anat. Anz. (1896), xi. pp. 410-415. M. Furbringer, " Brustschulterapparat and Schultermuskeln. Reptilien," Jena Zeitschr. (1900), 34, PP. 215-718, pls. 13-17 (with a list of many titles of papers concerning reptiles; and a new, unsatisfactory classification of the whole class). C. K. Hoffmann, " Becken d. Amphib. u. Reptil.," Niederl. Arch. f. Zool., iii. E. Mehnert, " Beckenguertel d. Emys lutaria," Morph. Jahrb. (1890), 16, pp. 537-571, pl.; " Os hypoischium, &c. d. Eidechsen," Morph. Jahrb. (1891), 17, pp. 123-144, pl. W. K. Parker, " Shoulder Girdle and Sternum," Roy. Soc. London, 1868. A. Rosenberg, " Development of Skeleton of Reduced Limbs," Zeitschr. wiss. Zool. (1873), 2 3, pp. 116-170, pls. A. Sabatier, " Comparaison des ceintures et des membres ant. et post," Mem. Ac. Montpellier (1880), xix. C. Gegenbaur, Untersuch. z. verg. Anat., " I. Carpus u. Tarsus " (1864), II. " Schulterguertel " (1865) (the most important monographs). A. Banchi, " Parafibula," Monitore Zool. Italiano (1900), xi. No. 7 (A nodule ][ between femur and fibula in Lacerta). G. Baur, " Carpus u. Tarsus d. Reptil.," Anatom. Anzeig. iv. No. 2. G. Born, " Carpus u. Tarsus d. Saurier," Morph. Jahrb. (1876), 2, pp. 1-26, pl. A. Carlsson, " Gliedmassenreste bei Schlangen," Svensk. Vetensk. Ac. Handlingar, ii. (1886). A. Johnson, Development of Pelvic Girdle," Q.J.M.S. (1883), 2 3, pp. 399-411. G. Kehrer, " Carpus u. Tarsus," Ber. Naturf. Ges. (Freiburg, i. 1886). W. Kuekenthal, " Entwickl. d. Handskelets'des Crocodiles," Morph. Jahrb. (1892), 1 9, pp. 4 2 -55. H. F. Sauvage, " Membre anterieur du Pseudopus," Ann. Sci. Nat.-Zool. 7. art. 15 (1878). A. Stecker, " Carpus u. Tarsus bei Chamaeleon," Sitzb. Ak. Wiss. (1877), 75, 2, pls. R. Wiedersheim, Gliedmassenskelett, Schulter u. Beckenguertel (Jena, 1892). K. Baechtold, Uber die Giftwerkzeuge der Schlangen (Tubingen, 1843). A. Duges, " Venin de l'Heloderma," Jubil. Soc. Biol. (1899), PP341 37. D. F. Weinland, " On the Egg-tooth of the Snakes," Proc. Essex Institute (Salem, 1856); and in Wurttemb. Jahresheft. Verein vaterl. Naturk. (1856). G. S. West, " Buccal Glands and Teeth of Poisonous Snakes," P.Z.S. (1895), pp. 812-826, pls. 44-46.

Tegumentary.-A. Batelli, "Bau der Reptilienhaut," Arch. mikr. Anat. (1880), 17, pp. 346-361, pls. J. E. V. Boas, " Wirbelthierkralle," Morph. Jahrb. (1894), xxi. pp. 281-311, pls. A. Haase, " Bau d. Haftlappen bei den Geckotiden," Arch. Naturg. (1900), 61, pp. 3 21 -345, pls. R. Keller, " Farbenwechsel d_ Chamaeleons," Arch. ges. Physiol. (1895), 61, pp. 123-168. C. Kerbert, " Haut der Reptilien," Arch. mikr. Anat. (1876), 13, pp. 205-262. F. Maurer, Epidermis and ihre Abkoemmlinge (Leipzig, 1895). F. Schaefer, " Schenkeldruesen d. Eidechsen," Arch. Naturg (1902), 68, pp. 27-64, pls. F. Todaro, Ricerche f. nel labor. di anat. norm. di Roma (1878), II. 1. F. Toelg, " Drusenartige Epidermoidalorgane d. Eidechsen u. Schlangen," Arb. Zool. Inst. Wien (1904), 15, pp_ 119-154, pls.

Nervous System.-J. F. Bemmelen, "Beitr. Kenntniss d. Halsgegend bei Reptilien Mededeel," Natura Artis Magistra (Amsterdam, 1887). L. Edinger, " Zwischenhirn d. Reptilien," Abh. Senckenb. Ges. (1899), 20, pp. 161-197, pls. J. G. Fischer, " Gehirnnerven d. Saurier," Abhandl. Naturwiss. Verein, Hamburg, II. (1852), pp. 115-212 (with many excellent illustrations). M. Furbringer, " Spinooccipital Nerven," &c., Festschr. f. Gegenbaur, iii. (1896). S. P. Gage, " Brain of Trionyx," Proc. Am. Micr. Soc. (1895), xvii. pp. 185-222. E. Gaupp, " Anlage d. Hypophyse b. Sauriern," Arch. mikr. Anat. (1893), 4 2, pp. 569-680. Giuliani, " Struttura d. midolla spinale d. Lacerta viridis," Ric. Lab. di Anat. Roma, ii. J. Grimm, " Ruckenmark v. Vipera berus," Arch. Anat. Phys. (1864), pp. 502-511, pl. 12. C. L. Herrick, " Brain of Certain Reptiles," Journ. comp. Neurol. (1891), i. pp. 1-36, iii. (1893), pp. 77-106, 119-140, with many plates. 0. D. Humphry, " Brain of Chelydra," Journ. comp. Neurol. (1894), PP. 73116. H. v. Jhering, Das peripherische Nervensystem (4to, Leipzig, 1873), pls. St G. Mivart and R. Clarke, " Sacral Plexus of Lizards, &c.," Trans. Linn. Soc. Zool. i. (1877), pp. 5 1 3-53 2, pls. 66, 67. H. F. Osborn, " Origin of the Corpora callosa," Morph. Jahrb. xii. pp. 530-543. H. Rabl-Rickhard, " Centralnervensystem d. Alligator," Zeitschr. wiss. Zool. (1878), xxx. pp. 336-373, Pis. 19 and 20. " Python," ibid. (1894), Iviii. pp. 6 94-7 1 7, pl. 41. G. Ruge, " Peripher. Gebiet. d. N. facialis " (masticator muscles, &c.), Festschr. f. Gegenbaur (1896), iii. L. Stieda, " Centralnervensystem d. Emys," Zeitschr. wiss. Zool. (1875), xxv. pp. 361-408.

Sense Organs.-R. Hoffmann, " Thraenenwege d. Vogel u. Reptil.," Zeitschr. f. Naturw. (Nat. Verein Sachsen u. Thiiring., 1882). C. Rose, " Nasendriise u. Gaumendrisen d. Crocodils," Anat. Anz. (1893), viii. pp. 745-751. C. Ph. Sluitez, " Jacobson's Organ v. Crocodilus," Anat. A nz. (1892), vii. pp. 540-545. O. Seydel, " Nasenhohle u. Jacobson's Organ d. Schildkroten," Festschr. f. Gegenbaur (1896), ii. B. Solger, " Nasenwand u. Nasenmuschelw. d. Reptil.," Morph. Jahrb. (1876), i. pp. 4 6 7-494, P l. E. Beraneck, " Parttalauge d. Rept.," Jen. Zeitschr. (1887), xxi. pp. 374-410, pls.; ibid., Anat. Anz. (1893), No. 20. P. Francotte, " L'Oil parietal, &c. chez les Lacertiliens," Mem. couronne Ac. Belgique (1898), 55, No. 3. H. W. de Graaf, Structure and Development of the Epiphysis in Amph. and Rept. (Leiden, 1886; written in Dutch). W. B. Spencer, " Presence and Structure of the Pineal Eye in Lacertilia," Q.J.M.S. (1886), 27, pp. 165-237, 7 pls. H. Strahl u. E. Martin, " Entwickl. d. Parietalauges b. Anguis u. Lacerta, " Arch. f. Anat. u. Phys. (1888), pp. 146-165, pl. io. A. Dendy, " Development of Parietal Eye of Sphenodon," Q.J.M.S. (1899), 4 2, P p. 1-87 and pp. 111-153, 13 plates. H. Miller, Schriften z. Anat. u. Physiol. d. Auges, edit. O. Becker (Leipzig, 1872). E. Ficalbi, " Palpebralapparat d. Schlangen u. Geckonen," Att. Soc. Tosc. Pisa, 'ix.' C. K. Hoffmann, "Anatomie d. Retina d. Amph. Rept. u. Vogel. Niederl.," Arch. Zool. (1875), iii. M. Borysiekiewicz, Retina v. Chamaeleo vulgaris (Leipzig, 1889), 7 pls. M. Weber, " Nebenorgane d. Auges d. Reptil.," Arch. f. Naturg. (1897), 43. E. Clason, " Gehororgan d. Eidechsen," Anatom. Studien (Leipzig, 1873). C. Hasse, " Gehororgan d. Krokodile," &c., ibid; "Gehororgan d. Schildkroeten, von Tropidonotus natrix," ibid. G. Retzius, Gehororgan d. Wirbelthiere, i. (Stockholm, 1881).

Muscles.-O. C. Bradley, " Muscles of Mastication of Lacertilia," Zool. Jahrb. Anat. (1902), 18, pp. 475-4 88. M. Firbringer, " Vergleich. Anatomie d. Schultermuskeln," Jena Zeitschr. (1873), vii. PP. 2 37-3 20; (1874), vii. pp. 175-280; (1900), xxx. pp. 215-718; Morph. Jahrb. (1875), i. pp. 636-816; Knochen u. Muskeln d. schlangendhnlichen Saurier (Leipzig, 1870), H. Gadow, " Bauchmuskeln d. Crocod. Eidechs. Schildkroeten," Morph. Jahrb. (1882), vii. pp. 57-100, pl.; " Myologie d. hinteren Extremitaet d. Reptilien," ibid. (1882), vii. pp. 327-466, pls. G. M. Humphrey, " Muscles of Pseudopus," Journ. An. Phys. (1872), vii. G. Killian, " Ohrmuskeln d. Crocodile," Jen. Zeitschr. (1890), xxiv. pp. 632656, pl. F. Maurer, " Ventrale Rumpfmuskulatur d. Reptil.," Festschr. f. Gegenbaur (1896), i. St G. Mivart, " Muscles of Iguana," P.Z.S. (1867), p. 766; " of Chamaeleon," ibid. (1870), p. 850. N. Rosen, " Kaumuskeln d. Schlangen u. Gif tdruese," Zool. Anz. (1906), 28, pp. 1-7. A. Sanders, " Muscles of Platydactylus," P.Z.S. (1870), p. 413; " of Liolepis," ibid. (1872), p. 154; " of Phryrosoma," ibid. (1874), P. 71; F. Walther, " Visceralskelett u. Muskulatur b. Amph. u. Rept.," Jen. Zeitschr. (1887), xxi. pp. 1-45, pls.

Respiratory System.-F. E. Beddard, " Trachea and Lungs of Ophiophagus bungarus," P.Z.S. (1903), pp. 3 1 9-3 28. G. Butler, " Suppression of one Lung in various Reptiles," ibid. (1895), p. 691.. S. H. Gage, " Pharyngeal Respiration in the Soft-shelled Turtle," Proc. Am. Ass. Adv. Sci. (1884), pp. 316-318; and Amer. Nat. (1886), xx. pp. 233-236. J. Henle, Vergl. anat. Beschreibung d. Kehlkopfes (1839). F. Siebenrock, " Kehlkopf u. Luf troehre d. Schildkroeten," Sitzb. Ak. Wien (1899), 108, pp. 5 6 3-595, pls. G. Tornier, " Kopflappen u. Halsluf tsaecke bei Chamaeleonen," Zool. Jahrb. Anat. (1904), 21, pp. 1-40, pls. D. Bertelli, " Pieghe dei reni primitivi nei Rettili. Contributo allo sviluppo del diaframma," Atti Soc. Toscan (Pisa, 1896), 15, (1898), 16. I. Bromann, Entwicklung d. Bursa omentalis and aehnlicher Recessbildungen (Wiesbaden, 1904). G. Butler, " Subdivision of Body-cavity in Lizards, Crocodiles and Birds," P.Z.S. (1892), PP. 45 2 -474, 4 pls.; " Subdivision of Body-cavity in Snakes," ibid. (1892), PP. 477497, pl. 6; " The Fat Bodies of the Sauropsida," ibid. (1889), p. 602, pls. 59-60. F. Hochstetter, Scheidewandbildungen in d. Leibeshohle der Krokodile, Voeltzkow, Reise in Ostafrika, vol. iv. pp. 141-206, pls. 11-15 (Stuttgart, 1906).

Vascular System.-F. E. Beddard, various papers on vascular system of Ophidia and Lacertilia, P.Z.S. (1904); " Notes on Anatomy of Boidae," ibid. (1903), pp. 107-121. F. E. Beddard and P. C. Mitchell, " Structure of Heart of Alligator," ibid. (1895). A. Greil, " Herz u. Truncus arteriosus d. Wirbelthiere Reptilien," Morph. Jahrb. (1903), 31, pp. 123-310, pls. O. Grosser and E. Brezina, " Entwickl. Venen d. Kopfes u. Halses bei Reptil.," Morph. Jahrb. (1895), pp. 28 9-3 2 5, pls. 20 and 21. F. Hochstetter, several important papers on vascular system of reptiles, Morph. Jahrb. (1891, 1892, 1898, 1901); ibid., " Blutgefa,ss-System," 0. Hertwig's Entwickl. d. Wirbelthiere (Jena, 1902); " BlutgefaessSystem d. Krokodile," Voeltzkow, Reise Ostafrika (Stuttgart, 1906, iv.). A. Langer, "Entwickl. Bulbus cordis bei Amph. u. Rept.," Morph. Jahrb. (1894), pp. 40-67. J. Y. Mackay, " Arterial System of Vertebrates, homologically considered," Memoirs and Memoranda in Anatomy (London and Edinburgh, 1889), i. B. Panizza, Sopra it sistema linfatico dei rettili (Pavia, 1833). C. Roese, " Vergl. Anat. d. Herzens d. Wirbelthiere," Morph. Jahrb. (1890), 16, pp. 27-96, pls. A. Sabatier, Etudes sur le cceur et la circulation centrale (Paris, 1873); " Transformat. du systeme aortique," Ann. Sc. Nat. Ser. (1874), 5, J. 19. H. Watney, " Minute Anatomy of Thymus," Phil. Trans. (1882), 173, pp. 1063-1123, pis. 83-95.

Urino-genital System.-J. E. V. Boas, " Morphol. d. Begattungsorgane d. Wirbelth.," Morph. Jahrb. (1891), xvii. pp. 171-287, pl. 16. J. Budge, " Das Harnreservoir d. Wirbelthiere," Neu Vorpommern, Mittheil. 7_(1875), pp. 20-128, pl. W. R. Coe and B. W. Kunkel, " Reproduce. Org. of Aniella," Amer. Natural. (1904), 3 8, pp. 4 8 7-49 0. H. Gadow, " Cloaca and Copulatory Organs of the Amniota," Phil. Trans. B. (1887), pp. 5-37, pls. 2-5. K. Hellmuth, "Kloake u. Phallus d. Schildkroeten u. Krokodile," Morph. Jahrb. (1902), 30, pp. 582-613. F. v. Moeller, " Urogenitalsystem d. Schildkroeten," Zeitschr. wiss. Zool., 6 5, pp. 573-598, pls. F. W. Pickel, " Accessory Bladders of Testudinata," Zool. Bull. (1899), ii. pp. 291-301. F. Schoof, Zur Kenntniss d. Urogenitalsystems d. Saurier. Arch. f. Naturg. (1888), 54, P. 62. P. Unterhoessel, " Kloake u. Phallus d. Eidechsen u. Schlangen," Morph. Jahrb. (1902), 3 0, PP541-581.0. Schmidtgen, " Clorke and ihre Organe bei SchildkrOter," Zool. Jahrb. (1907), PP. 357-412, P 1.3 2, 33. (H. F. G.) IV. Distribution In Space This zoo-geographical review deals only with modern reptiles. We begin with a survey of the faunas of some of the most obvious land-complexes which bear close resemblance to the now classical " regions " of P. L. Sclater and A. R. Wallace. None of these " regions " has definable frontiers, and what acts as a bar to one family may be totally ignored by another. According to the several orders of reptiles the world is mapped out in very different ways. The African fauna does not stop at the Suez Canal, nor even at the Red Sea; there is a transitional belt noticeable in the countries from Syria to Arabia, Persia and India. To the north, Indian influence extends right into Turkestan, or vice versa; the Central Asiatic fauna passes into that of India. On the Chinese side prevailing conditions are still almost unknown; Wallace's line is more or less rigidly respected by Trionychidae, hooded Elaps, vipers and Lacertidae, while it has not the slightest influence upon crocodiles, pit vipers, Varanidae, Agamidae, &c. In the western hemisphere we have a grand illustration of the interchange of two faunas and of the fact that it is neither a narrow strait nor an equally narrow isthmus which decides the limitation of two regions. Central America and the Antilles form one complex with S. America. The nearctic region ends at the edge of the great Mexican plateau, which itself is a continuation of the north continent. Many nearctic forms have passed southwards into the tropics, even into faroff S. America, but the majority of the southerners, in their northern extension, have been checked by this plateau and have surged to the right and left along the Pacific and Atlantic tropical coastlands. The present writer happens to have made a special study of this part of the world (cf. " The Distribution of Mexican Amphibians and Reptiles," P.Z.S., 1905, pp. 191-294); the N. and S. American faunas have therefore been more fully treated in the following review of the various faunas. No doubt others can be treated in a similar manner, but the physical features between N. and S. America are unique, and the results are closely paralleled by those of the fauna of birds. The narrow and long neck of the isthmus of Panama (once no doubt much broader) is no boundary; if the meeting of N. and S. had taken place there, that narrow causeway would be crowded, and this is not the case.

New Zealand.-The only recent reptiles are Sphenodon (q.v.), which testifies to the great age of these islands; about half a dozen Scincidae of the genus Lygosoma, members of a cosmopolitan family; and some few geckos, e.g. Naultinus, of a family of great age, world-wide distribution and with exceptional facilities of distribution.

Australian Region. - Of crocodiles only C. johnstoni in N. Australia and Queensland; C. porosus on the N. coast, and occurring on various Pacific islands, as far E. as the Fiji Islands. Tortoises are represented only by the pleurodirous Chelydidae, e.g. Chelodina; they are absent in Tasmania and on the Pacific islands. New Guinea possesses the aquatic Carettochelys, sole type of a family.

The bulk of the Lacertilian fauna is composed of skinks, geckos, agamoids and Varanidae, with the addition of a small family which is peculiar to the region, the Pygopodidae. A peculiar type, Dibamus, inhabits the borderlands, namely, New Guinea, the Moluccas, Celebes and the Nicobar Islands; and, finally, a single iguanoid, Brachylophus, is common in the Fiji Islands; how it came there, or how it survived its severance from the American stock, is a mystery. The skinks are in this region more highly developed and more specialized than in any other part of the world; they exceed in numbers the geckos, which generally accompany the skinks in their range over the smaller islands of the Pacific; in these islands members of these two families represent the whole of the Lacertilian fauna. The Australian agamoids are chiefly peculiar and partly much differentiated forms (e.g. Moloch and Chlamydosaurus), but some have distinct affinities to, or are even identical with, Indian genera. The Varanidae are also closely allied to Indian species.

Of snakes, amounting to about one hundred species only, we note about one dozen Typhlopidae, and of Pythoninae simply Python, and the Boine Enygrus on the islands from New Guinea to Fiji. There are but surprisingly few innocuous colubrine snakes, scarcely a dozen, and all belonging to Indian genera. The bulk of the snakes belong to the poisonous Elapinae, all of genera peculiar to the region, e.g. Acanthophis, Pseudechis, Notechis. Such a preponderance of poisonous over harmless snakes is found nowhere else in the world. Tasmania is tenanted by poisonous snakes only. In Australia 'we meet, therefore, with the interesting fact that, whilst it is closely allied to S. America, but totally distinct from India by its Chelonians, its lizards and colubrine snakes connect it with this latter region. With regard to the other Ophidians, they have their nearest allies partly in India, partly in Madagascar, partly in S. America; and the character of the Australian snake fauna consists chiefly in its peculiar composition, differing thereby more from the other equatorial regions than those do among themselves. Wallace's line marks the boundary between India and Australia only as far as Chelonians are concerned, but it is quite effaced by the distribution of lizards and snakes. Thus in New Guinea lizards of the Indian region are mixed with Pygopodidae, and an island as far E. as Timorlaut is inhabited by snakes, some of which are peculiarly Indian, whilst the others are as decidedly Australian. The islands N. of New Guinea and of Melanesia are not yet occupied by the Ophidian type, and only species of Enygrus have penetrated eastwards as far as the Low Archipelago, whilst the Fiji Islands and the larger islands of Melanesia have sufficiently long been raised above the level of the sea to develop quite peculiar genera of snakes.

Indian Region. - Of Crocodilia C. palustris, the " mugger " or marsh crocodile, and C. porosus; Gavialis gangeticus; Tomistoma schlegeli in Borneo, Malacca and Sumatra. Of tortoises Platysternum megacephalum, type of a family from Siam to S. China; many Trionychidae and Testudinidae, mostly aquatic; whilst the terrestrial Testudo is very scantily represented. One species which is common in the Indian peninsula (T. stellata) is so similar to an African species as to have been considered identical with it; the Burmese tortoise is also closely allied to it, and the two others extend far into western-central Asia. Thus this type is to be considered rather an immigrant from its present headquarters, Africa, than a survivor of the Indian Tertiary fauna, which comprised the most extraordinary forms of land tortoises. Wallace's line marks the E. boundary of Trionyx; species of this genus are common in Java and Borneo, and occur likewise in the Philippine Islands, but are not found in Celebes, Amboyna or any of the other islands E. of Wallace's line. Agamidae are exceedingly numerous, and are represented chiefly by arboreal forms, e.g. Draco (q.v.) is peculiar to the region, Ceratophora and Lyriocephalus exclusively Ceylonese; terrestrial forms, like Agama and Uromastix, inhabit the hot and sandy plains in the N.W. and pass uninterruptedly into the fauna of western-central Asia and Africa. The Geckonidae, Scincidae and Varanidae are likewise well represented, but without giving a characteristic feature to the region by special modification of the leading forms except the gecko Ptychozoon homalocephalum in Malaya. The Lacertidae are represented by one characteristic genus, Tachydromus - Ophiops and Cabrita being more developed beyond the limits assigned to this region. Finally, the Eublepharidae and Anguidae, families whose living representatives are probably the scattered remains of once widely and more generally distributed types, have retained respectively two species in W. India, and one in the Khasi Hills, whilst the presence of a single species of chameleon in S. India and Ceylon reminds us again of the relations of this part of the fauna to that of Africa.

The Indian region excels all the other tropical countries in the great variety of genuine types and numbers of species of snakes. Boulengeri recognizes 267 species, i.e. about one-fifth of the total number of snakes known. India is the only country in the world possessing viperine, crotaline and elapine poisonous snakes (their proportion to harmless snakes being about i: io), e.g. Vipera russelli, the " daboia " (see Viper); Lachesis, e.g. gramineus, an arboreal pit viper; Naja tripudians, the cobra; Bungarus coeruleus, the " krait "; Callophis; and Hydrophinae along the coasts of the whole region. Several sub-families and families are peculiar to the region: the Uropeltidae with Rhinophis in southern India, and Uropeltis confined to Ceylon; Ilysiidae in Ceylon and Malay Islands, elsewhere only in S. America; the opisthoglyphous Elachistodon westermanni of Bengal; the Homalopsinae, with many species from Bengal to N. Australia; further the Amblycephalidae; Xenopeltis unicolor, sole type of a family; and the Acrochordinae, a sub-family of aglyphous Colubridae, ranging from the Khasi Hills to New Guinea. Of other Colubridae, we notice numerous Tropidonotus, Coronella and Zamenis, the latter one of the most characteristic types of the warmer parts of Eurasia. Tree-snakes, e.g. Dipsas and Dendrophis, are common. Of other families we note a great number of Typhlopidae, of which T. braminus occurs even on Christmas Island. Lastly various species of Python, but no Glauconiidae, the only family not represented in the Indian region, which claims the Uropeltidae, Xenopeltidae and Amblycephalidae as peculiar to itself.

Gunther remarks that to this region Japan has to be referred. This is clearly shown by the presence of species of Ophites,Callophis, Trimeresurus s. Lachesis, Tachydromus, characteristically Indian forms, with which species of Clemmys, Trionyx, Gecko, Halys, and some Colubrines closely allied to Chinese and Central Asiatic species are associated. Halys is a central Asiatic pit viper. The few reptiles inhabiting the northern part of Japan are probably of palaearctic origin.

THE African Continent. - Of crocodiles, C. vulgaris in the E., C. cataphractus and Osteolaemus tetraspis in the W. There are many Chelonians, especially small land tortoises of Testudo, and with Cinyxis which is peculiar to this continent; the freshwater Clemmys only in the N.W. corner; several genera of the pleurodirous Pelomedusidae, Pelomedusa galeata, which is equatorial and southern, with an outlying occurrence in the Sinai peninsula, and Sternothaerus with several tropical and southern species; of Trionychidae the tropical Cycloderma and Cyclanorbis peculiar to the country, and the large Trionyx triunguis which ranges from the Senegal and Congo into the Nile system with its big lakes, but occurring also in Syria.

Of Lacertilia the geckos and skinks, and the typically old world families of Lacertidae and Varanidae are well represented; also Amphisbaenidae; Gerrhosauridae and Zonuridae, peculiar to Africa and Madagascar; a few Eublepharinae and a few of the so-called Anelytropidae in West Africa. But the most important feature of this Lacertilian fauna is the almost universal distribution of chameleons in numerous and some highly specialized forms, Chameleon and Rhampholeon. We note the entire absence of Iguanidae and of Anguidae, the latter represented by Ophisaurus only in the north-western corner.

Of snakes only one sub-family is peculiar, the Rhachiodontinae with the sole species Dasypeltis scabra, the egg-swallowing snake. Many Typhlopidae and Glauconiidae, but no Ilysiidae; large pythons, Eryx in the N., and a boa, Pelophilus fordi in the W. of Africa. Of poisonous snakes there is an abundance, notably the Viperinae have their centre in this continent; besides Echis, which is also Indian, there are peculiar to the continent Bitis, the puffadder, Causus, Atractaspis, Cerastes, and Atheris which is an arboreal genus, all 'of which see under Viper. The pit vipers are entirely absent. Elapinae are numerous, e.g. hooded cobras like Naha haje and Sepedon the " ringhals." Many opisthoglyphous tree snakes and a considerable number of innocuous colubrines, e.g. Lycodon, Psammophis and Coronella or closely allied genera all also in India, but Coluber-like forms and Tropidonotus are very scantily represented, chiefly in the N.

On the whole the reptilian fauna of Africa is not rich, considering the huge size of the continent, but this may be accounted for by the great expanse of desert in the N. half and of veld in the S. Lastly, the enormous central forests are still scarcely explored.

Madagascar and certain other islands have a fauna which is as remarkable for its deficiencies as it is for its present forms. The following well-defined groups are absent: Trionychidae and Chelydidae; Agamidae, Lacertidae, Anguidae, Amphisbaenidae, Varanidae and Eublepharinae; all the Viperidae and Elapinae, so that this large island enjoys perfect absence of poisonous snakes, not counting the practically harmless opisthoglyphous tree snakes; there are further no pythons and no ilysias.

The actual fauna consists of: Crocodilus vulgaris, which is said to be extremely abundant; of Chelonians, Pelomedusa galeata and 1 The same authority enumerates 536 species of reptiles for British India, i.e. about one-sixth of all the recent species of reptiles (Fauna of British India, edit. W. T. Blanford, London, 1890).

Sternothaerus, both also in Africa, Podocnemis, which elsewhere occurs in South America only, and several Testudinidae; of these Pyxis is peculiar to Madagascar, while Testudo has furnished the gigantic tortoises of Aldabra, the Seychelles, and recently extinct in Mauritius and Madagascar. Of lizards are present a few Gerrhosauridae and Zonuridae, both African types; the remarkable occurrence of two iguanid genera Chalarodon and Hoplurus, both peculiar to the island; skinks, many geckos, and Uroplates, sole type of the Uroplatinae and an abundance of chameleons, of the genera Chameleon, with Ch. parsoni, the giant of the family, and the small species of Brookesia, a genus peculiar to Madagascar. Of snakes we note Typhlopidae and Glauconiidae, and the remarkable occurrence of Boinae, two of the genus Boa (Pelophilus), one of Corallus on the main island and Casarea on Round Island. There are opisthoglyphous mostly arboreal snakes, and the rest are innocuous colubrines, some few with Indian and African affinities, e.g. Zamenis s. Ptyas, more with apparently S. American relationship, or at least with resemblance in taxonomic characters.

An analysis of this peculiarly compound and deficient fauna gives surprising results, namely, the almost total absence of affinity with the Indian region, close connexion with Africa by the possession of Gerrhosauridae, Zonuridae, Chameleons and Pelomedusidae; lastly, the presence of several tree boas, of Podocnemis and of Iguanidae, i.e. families and genera which we are accustomed to consider as typically neo-tropical. Peculiar to Madagascar, autochthonous and very ancient, is only Uroplates. Ancient are also the tortoises, chameleons, geckos, boas, typhlops, gerrhosaurids and zonurids. The absent families may be as ancient as the others, but most of them, notably Varanus, lacertids and agamids are of distinctly northern, palaeotropical origin, and we can conclude with certainty that they had not spread into S. Africa before Madagascar and its satellites became severed from the continent.

Europe And Temperate Asia. - The present reptilian fauna of this vast area is composed almost entirely of the leavings of those groups which are now flourishing with manifold differentiations under more genial climes, in Africa and India. Fossils, none too numerous, tell us that it was not always thus, since crocodiles, alligators and long-snouted gavials, all the main groups of chelonians, iguanoids, &c., existed in England, the crocodilians persisting even towards the end of the Tertiary period.

There are no crocodiles now in the Eurasian sub-region, excepting small survivors in the Jordan basin, on the borderland of Africa; but the Yang-tse-Kiang is inhabited by an alligator, A. sinensis, while all its congeners are now in America. This finds, to a certain extent, a parallel in Trionyx, of which one species lives in the Euphrates basin, likewise borderland, and another, T. maacki, in rivers of N. China, e.g. in the Amoor. Of other Chelonians we note several species of Testudo, two of them European; Emys europaea, chiefly in Europe, with the other species E. blandingi in the eastern United States; and a few species of Clemmys, a truly periarctic genus.

Of Lacertilia we exclude the chameleon. Of geckos Hemidactylus turcicus extends from Portugal to Karachi; Platydactylus facetanus is at home in most S. Mediterranean countries; Teratoscincus is peculiar to the steppes and deserts of Turkestan and Persia; other geckos in the transitional region from Asia Minor to India. Of Lacertae we have Anguidae, Agamidae, Lacertidae, Amphisbaenidae and Scincidae, most of them in Europe represented by but one or two species. Thus Blanus cinereus in Mediterranean countries, Asia Minor and Syria, represents the Amphisbaenidae which are found nowhere else in Europe or Asia, but plentiful in Africa and both Americas. Of the Anguidae, Anguis fragilis is peculiar to Europe, Ophisaurus apus in S.E. Europe, another in Indo-Burman countries, with the rest of the species in N. America. Of Scincidae few in Europe, e.g. Chalcides s. Seps s. Gongylus, others from Asia Minor eastwards, e.g. Scincus, and Ablepharus in Turkestan. Agamidae do not occur in Europe but they exist in considerable numbers from Asia Minor and Turkestan to China, with Phrynocephalus peculiar to central Asia. Lastly, the Lacertidae, of which several species of Lacerta, Psammodromus, Acanthodactylus in Europe, but the majority in Africa and warmer parts of India; in a similar manner the Manchurian forms are related to Chinese.

The total number of palaearctic snakes amounts to about sixty, the majority living in the Mediterranean countries and in W. Asia. One Typhlops in the Balkan peninsula and in W. Asia, in Persia also Glauconia; Eryx jaculus extends into Greece from S.W. Asia as sole representative of the Boidae. Several vipers, the common viper, V. berus, from Wales to Saghalien Island, V. aspis, V. latastei and V. ammodytes in S. Europe; a pit viper, Ancistrodon, e.g. halys, in the Caspian district, thence this genus through China and again in N. America. Echis extends N. into Turkestan. The Indian cobra ranges N. to Transcaspia and far into China. All the other snakes belong to the aglyphous and opisthoglyphous Colubridae; of the latter Coelopeltis is peculiar to S. Europe and S.W. Asia; Macroprotodon cucullatus to S. Spain, the Balearic Islands and N. Africa; Tephrometopon peculiar to Turkestan and neighbouring countries; none extending into E. Asia. Of the aglyphous colubrines the most characteristic genus is Zamenis incl. Zaocys, very widely spread and including more species than any other palaearctic genus; several species of the wide-ranging genus Tropidonotus, besides Coluber, with. Rhinechis scalaris in S.W. Europe. There are, besides, other genera, especially in the debatable countries of S.W. Asia, Persia and Afghanistan, and speaking generally the colubrines show less affinity to African than to Indian forms, just as we should expect from the prevailing geographical conditions. If it were not for the N.W. corner of Africa and portion of its N. coast, the European fauna would have very little in common with Africa.

North America. -Of this huge continent only the United States and Mexico come into consideration, since N. of 45° latitude reptilian life is very scarce. The area, however, with these restrictions, is larger than the Indian and Malay countries, and larger than the Australian region. Yet the fauna is comparatively poor, very poor indeed, if it were not for Mexico and the Sonoran province, which seems to be the ancient centre of distribution of much of the present typically N. American fauna.

Characteristic of the area is the abundance of Chelonians and Iguanidae, to which Tejidae have to be added in the S.; equally characteristic is the complete absence of Pleurodirous Chelonians, of Chameleons, Agamidae, Lacertidae, Varanidae and Viperinae. The fauna is composed as follows: Crocodilia, with Crocodilus americanus and Alligator mississippiensis in the S. Of Chelonians the Chelydridae, peculiar to the E. half but for the reappearance of a species of Chelydra in Central America; many Cinosternidae likewise almost peculiar to the area; of Testudinidae an abundance of freshwater forms, notably Chrysemys, and Emys in common with Europe, whilst terrestrial tortoises are extremely scanty, namely one species of Testudo, T. Polyphemus, the gopher, and two of Cistudo, C. carolina; lastly, two Trionyx in the whole of the Mississippi basin and thence N. into Lake Winnipeg, 51° N. Lacertilia: Geckos are very scarce; N. America has received only Sphaerodactylus notatus from the Antilles into Florida, and Phyllodartylus tuberculosus into California from the Pacific side of Mexico; Eublepharinae are absent. Of Iguanidae we have a typically Sonoran set, e.g. Crotaphytus, Holbrookia, Uta, Phrynosoma, Sceloporus, and a S. set of which only Anolis extends out of the tropics. It is significant that only a few species of Sceloporus and Phrynosoma extend into the United States, although far N.; of the large genus Anolis only A. carolinensis enters Texas to Carolina. Sceloporus may be called the most characteristic genus of Sonoraland and Mexico. Of the tropical family of Tejidae only Cnemidophorus, with many species in Mexico, a few in the adjoining N. states, and with C. sexlineatus over the greater part of the Union. Anguidae: Ophisaurus ventralis in the United States; the other species in the Old World. Diploglossus peculiar to mountains of Mexico. Gerrhonotus, the main genus, centred in Mexico, but G. coeruleus ranges from Costa Rica along the Pacific side right into British Columbia, the most northern instance of a New World reptile.

Xenosaurus grandis of Mexican mountains is the monotype of a family, and the same would apply to Heloderma (H. suspectum, the Gila monster of the hottest lowland parts of Arizona and New Mexico; and H. horridum of Mexico) if it were not for Lanthanotus of Borneo. Scincidae: of this cosmopolitan family America possesses the smallest number, and it is significant that the number of species decreases from N. to S.; Eumeces from Minnesota and Massachusetts through Mexico, with many species, and Lygosoma s. Mocoa laterale from S.E. and Central States to Mexico. Xantusiidae, a small family, is composed of a N. or Sonoran and a S. or Central American-Antillean group; e.g. Xantusia of the deserts of Nevada and California. Aniella, monotype of a family of California to El Paso, Texas, i.e. peculiar to Sonoraland, Amphisbaenidae with Rhineura in Florida and the marvellous Chirotes in Lower California and the Pacific side of Mexico; the other members of this family are tropical so far as America is concerned.

Snakes: of Typhlopidae only Anomalepis mexicana, peculiar to Nuevo Leon; of Glauconiidae several extending N. into Texas and Florida. Boinae continue N. as the arenicolous Lichanura of Lower California and Arizona, and the likewise arenicolous Charina bottae which extends from California to the state of Washington; the other members of the family are all tropical, extra-regional. Of Viperidae only pit vipers occur, but of them rattlesnakes cover the whole of the habitable area; Ancistrodon, without a rattle, e.g. the moccasin snake and the water viper, has other species in central and E. Asia. Of Elapinae, far into the E. United States only the genus Elaps with a few species, of which E. fulvius, the commonest, ranges from S. Brazil far into the S. and E. states. A few opisthoglyphous, terrestrial, snakes just enter the United States from Mexico, e.g. Trimorphodon. Of aglyphous colubrines species of genera like or resembling Tropidonotus, Coronella and Coluber, including Pityophis and Spilotes, are abundant, the latter being very characteristic; Ischnognathus a.nd Contia, Ficimia and Zamenis likewise are clearly nearctic, or Sonoran.

The Greater Antilles have essentially neotropical, i.e. Central American and S. American affinities, but there is also some Sonoran infusion. - There is Crocodilus americanus; no Chelonians are natives except one or two Chrysemys. Of Lacertilia, geckos are abundant; of Iguanidae several arboreal forms, notably the large Iguana, and Metopoceras of Haiti, and Cyclura, both peculiar; of Anguidae Celestus, peculiar, but closely allied to Diploglossus; of Xantusiidae the peculiar genus Cricosaura s. Cricolepis. Of Amphisbaenidae Amphisbaena itself occurs in Puerto Rico and on the Virgin Islands. Of Tejidae only Ameiva, not Cnemidophorus. Snakes: a Typhlops in Puerto Rico; of boas Epicrates, Ungalia and Corallus, the latter re-occurring in Madagascar. Absent are: Viperidae, Elapinae and Opisthoglyphs; of aglyphous colubrines the Central American genera Urotheca, Dromicus, Drymobius and Leptophis; the genera of distinctly northern origin.

South And Central America. - The fauna is very rich. It is advisable first to mention those groups which are either confined to Central America (including the hot lowlands of Mexico), e.g. the Dermatemydidae, Eublepharinae, Anelytropsis and the aglyphous colubrines: Urotheca, Dromicus, Drymobius, Leptophis, Rhadinea, Streptophorus, or which, from their N. centre have sent some genera into Central America, or beyond into the S. continent: e.g. Chelydra rossignoni, ranging from Guatemala to Ecuador; one Cinosternum extending into Guiana; Testudo tabulata, the only terrestrial tortoise of S. America, besides the gigantic creatures of the Galapagos Islands; a. few Eublepharinae reaching Ecuador; of Anguidae Gerrhonotus coeruleus, extending S. to Costa Rica; of Scincidae, Mabuia and Lygosoma, which extend far into S. America, and the same applies to the Amphisbaenidae. Immigrants from the N. are probably also the Iguanidae, although they have found a congenial home in the S. countries, where they are now represented by an abundance of genera and species, e.g. Laemanctus and Corythophanes of Mexico, Anolis, Iguana, Basiliscus, Ctenosaura, Polychrus, Hoplurus, Chalarodon. Amongst snakes the following appear to be of N. origin: Boidae (with the Pythonine Loxocaemus bicolor in Mexico), in spite of their great development of boas and anacondas in the S.; certainly Crotalinae, of which only one species, C. terrificus, is found in S. America; further, some aglyphous colubrines, which have sent a few species only into Central, and still fewer into S. America,* e.g. Tropidonotus, Ischnognathus, Contia,* Ficimia, Coluber, Spilotes, Pityophis, Coronella* and Zamenis. After these numerous restrictions we should expect the genuine autochthonous fauna of the S. American continent to be very scanty, especially if we remember those important Old World groups which are absent in America, e.g. Varanidae, Lacertidae, Agamidae and chameleons, and that Central and S. America have no Trionychidae. The oldest S. American reptilian fauna is composed as follows. It is the only part of the world which possesses Chelydidae in abundance, e.g. of Chelys the Matamata, Hydromedusa, and of Pelomedusidae, Podocnemis, which re-occurs in Madagascar. Crocodilia are represented by Crocodilus americanus and C. moreleti in the N. and by about five species of Caiman. Of Lacertilia geckos are rather few, mostly in the N.W. of the continent, more numerous in Central America and the Antilles. The Tejidae are clearly a neotropical family, with several dozen genera in S. America; of all these, only Ameiva and the closely allied Cnemidophorus extend through and beyond Central America: Ameiva into the E. and W. hot lands of Mexico and into the Antilles, Cnemidophorus through Mexico far into most of the United States with a few species. Of snakes there is an abundance. Typhlopidae and Glauconiidae are well represented. Of aglyphous colubrines many genera, some of these extending northwards into Mexico, but not to the Antilles, e.g. Atractes, Tropidodipsas, Dirosema, Geophis, Xenodon. Opisthoglypha are very numerous in genera and species both in S. and Central America, whence many of the arboreal forms extend into the hot countries of Mexico, while a few terrestrials have spread over the plateau and thence into the United States, none entering the Antilles; such typical neotropical genera are Himan- !odes, Leptodira, Oxyrhopus, Erythrolamprus, Conophis, Scolecophis, Homalocranium, Petalognathus, Leptognathus. Most of the Amblycephalidae are neotropical, the others in S.E. Asia. Of Elapinae only the genus Elaps occurs, but with many species. Of the Crotalinae, Lachesis is the essentially neotropical genus, with many species, some of which enter the hot lands of Mexico, e.g. L. lansbergi s. lanceolatus, a very widely distributed species, the only pit viper which has entered the Lower Antilles.

The above survey of the world shows that but very few of the principal families of reptiles are peculiar to only one of the main regions." The occurrence of some freak, constituting a little family or sub-family by itself in some small district, and therefore put down as peculiar to a whole wide region, cannot be much of a criterion, e.g. Rhachiodon, Elachistodon, Acrochordinae, Uroplates, Xenosaurus, Heloderma, Aniellidae, Dibamus, Anelytropidae, Platysternum. They are not characteristic of large' countries, but rather local freaks. Quite a number of very ancient families have such a wide distribution that they also are of little critical value, notably the peropodous snakes, which have survivors in almost any tropical country; such cosmopolitans are also geckos and skinks.

A difficulty which is ever present in such zoogeographical investigations is the uncertainty as to whether our zoological families and sub-families and even genera are genuine units, or heterogeneous compounds, as for instance the Anelytropidae, of which degraded skinks there is one in Mexico, two others in W. Africa. Heloderma in Mexico and Lanthanotus in Borneo are both without much doubt descendants of some Anguid stock, but when we now combine them, in deference to our highest authority, as one family, we thereby raise the tremendous problem of the present distribution of this family. Boas and pythons are likewise not above suspicion, cf. some boas in Madagascar and the python Loxocaemus in Mexico. The opisthoglyphous colubrines are almost certainly not a natural group, not to speak of numerous genera of the aglyphous assembly. To avoid arguing in a circle, such doubtful units had better be avoided whilst building hypotheses.

G. Pfeffer has recently endeavoured to show by an elaborate careful paper (" Zoogeographische Beziehungen Siidamerikas," Zool. Jahrb., Suppl. viii., 1905), " that nearly all the principal groups of reptiles, amphibians and fishes had formerly a universal or subuniversal distribution, and that therefore it is not necessary to assume a direct land connexion of S. America with either Africa or Australia, with or without an Antarctic." Many cases of such a former universal distribution are undoubtedly true, but the question remains how the respective creatures managed to attain it.

For true characterization of large areas we must resort to the combination of some of the large wide-ranging families, and equally important is the absence of certain large groups; both to be selected from the following table.

1 Including the related Dermatemydidae and Cinosternidae. a With an exception.

Entering, or in the borderland.

4 Mediterranean countries.

Rhineura; formerly wider distribution.

In Asia.

Deductions from this table show, for instance, that Australia is quite sufficiently characterized by the possession of Chelydidae and Varanidae; Madagascar by the presence of chameleons and Pelomedusidae. On the other hand, the separation of the whole of Africa from Asia, or the diagnosis of the palaearctic " region," would require the combination of several positive and negative characters.

Chelonians are very diagnostic, expressed by the following combinations of families: America as a whole: Chelydridae and Cinosterridae and Dermatemydidae.

N. America: Chelydridae and Trionychidae, but only E. of the Rockies.

S. America: Chelydidae and Pelomedusidae.

Africa: Trionychidae and Pelomedusidae.

Madagascar: Pelomedusidae and Testudinidae.

India and Eurasia: Trionychidae and Testudinidae.

Australia: Chelydidae only.

That the Chelonians are regionally so very diagnostic that their main families are still in rational agreement with the main divisions of land, is perhaps due, first, to their being an ancient group; secondly, to their limited means of distribution (none across the seas, omitting of course Cheloniidae, &c.); and lastly, to their being rather indifferent to climate. Note, for instance, Trionyx ferox from the Canadian lakes to the Gulf of Mexico, Cinosternum pennsylvanicum from New York to New Orleans. It may be taken for certain that wherever a Testudo occurs as a genuine native, it has got there by land, be the locality the Galapagos, Aldabra, Madagascar or some Malay islands. The Trionychidae reveal themselves as of periarctic origin, being debarred from Australia, Madagascar and the neotropical region (alleged from Eocene Patagonia). Testudinidae are cosmopolitan, excluding Australia, and practically also the Antilles; and Testudo is most instructive with its almost similar distribution; but something has gone wrong with this genus in America, where it flourished in mid-Tertiary times.

Pleurodira are less satisfactory than they appear to be from a merely statistical point of view. The Pelomedusidae, being known from European Trias and from nearctic cretaceous formations, may have had a world-wide distribution; but Chelydidae may well have centred in an antarctic continent. Chelydridae were periarctic and have disappeared from Eurasia; N. American offshoots are the Cinosterridae and Dermatemydidae, the latter now restricted to Central American countries.

Crocodilia, probably once universal, afford through the Chinese alligator an instance of the original intimate connexion of the whole holarctic region, paralleled by many other animals which now happen to be restricted to E. Asia and to eastern N. America.

Lacertilia are less satisfactory for short diagnoses. America alone combines Iguanidae and Tejidae: - N. America: Iguanidae, Anguidae, Tejidae (and Rhineura in Florida).

S. America: Iguanidae, Anguidae, Tejidae and many Amphisbaenidae.

Africa and Madagascar: Chameleons and Zonuridae and Gerrhosauridae.

Madagascar: Chameleons and Iguanidae.

India: Varanidae, Agamidae and Lacertidae, all of which also in Africa.

Australia alone has Pygopodidae.

The Lacertilia are now distributed upon principles very different from those of the tortoises. According to the lizards the world is divided into an E. and a W. half. The W. alone has Iguanidae and Tejidae, the E. alone that important combination of Varanidae and Agamidae. Further subdivision is in most cases possible only by exclusion, e.g. exclusion of Lacertilia and chameleons from Australia; of Varanidae and Agamidae from Madagascar. Lizards are rather susceptible to climatic conditions, infinitely more than water tortoises.

As regards Ophidia, America has Crotalinae and Elapinae, but no Viperinae. Eurasia and India alone combines Viperinae, Crotalinae and Elapinae. Africa, Viperinae and Elapinae but no Crotalinae. Australia only Elapinae. Madagascar none of these groups.

The Viperinae must have had their original centre in the palaearctic countries, and they have been debarred only from Australia and Madagascar. Both vipers and pit vipers are still in Asia, but true vipers are absent in America, with their fullest development now in Africa, whilst pit vipers went E., covering now the whole of America, and having developed the rattlesnakes in Sonoraland. The Elapinae are undoubtedly of Asiatic origin; they have overrun Africa, were too late for Madagascar, but early enough for Australia, where they are only poisonous snakes; and only one genus, Elaps, has got into, or rather, has differentiated in America, in the S. of which it is abundant.

Opisthoglypha are useless for our purpose; they are cosmopolitan, with the exception of Australia, but probably they have one ancient centre in S. America, and another in the old world.

Amblycephalidae afford another of those curious instances of apparent affinity between S.E. Asia and Central America; paralleled by Pelamis bicolor, which ranges from Madagascar to Panama, while all the other Hydrophinae belong to the Indian Ocean and the E. Asiatic seas. Aglyphous Colubrines show undoubted affinity between N. America and Eurasia; the whole group is absolutely cosmopolitan, and many of the genera, e.g. Coluber, Tropidonotus and Coronella, have proved their success by having acquired an enormous range. Snakes have comparatively few enemies, and they possess exceptional means of distribution. It is rare for a terrestrial species to have such a wide range as Crotalus terrificus, from Arizona to Argentina, or as the India cobra, which, like the tiger, is equally at home in Malay islands, Manchuria and Turkestan.

The tortoises divide the habitable world into a S. and a N. world, much as do the anurous Batrachians; the lizards split it into an E. and a W. hemisphere. The poisonous snakes, the most recent of reptiles in their full development and distribution, allow us to distinguish between Australia, America and the rest of the world.

(H. F. G.)

Bibliography Information
Chisholm, Hugh, General Editor. Entry for 'Reptiles'. 1911 Encyclopedia Britanica. https://www.studylight.org/​encyclopedias/​eng/​bri/​r/reptiles.html. 1910.