Transcriber's Notes:

This text uses the symbols for male (♂) and female (♀). These symbols may not be correctly displayed by some e-reader devices. An upside-down Y has been represented as [Y].

Some presumed printer's errors have been corrected. In particular, punctuation has been normalized and entries in the Index were altered to match the main text.

THE SNAKES OF EUROPE

UNIFORM WITH THIS VOLUME

THE LIFE OF CRUSTACEA
BRITISH FRESHWATER FISHES
THE OX AND ITS KINDRED
THE LIFE OF THE MOLLUSCA

THE
SNAKES OF EUROPE
BY
G. A. BOULENGER
LL.D., D.Sc., Ph.D., F.R.S., F.Z.S.

WITH FOURTEEN PLATES AND FORTY-TWO FIGURES IN THE TEXT

METHUEN & CO. LTD.
36 ESSEX STREET W.C.
LONDON

First Published in 1913

PREFACE

There is no work in the English language dealing with the Reptiles of Europe. I have therefore endeavoured to supply this desideratum, so far as the Snakes are concerned, by drawing up in a concise form an account of what is known of their characters, their distribution, and their life-histories. Professor Sordelli, of Milan, having kindly acceded to my request to reproduce some of the beautiful figures drawn by him for the work published in collaboration with the late Professor Jan under the title of “Iconographie Générale des Ophidiens,” I have been able to supplement my descriptions with illustrations which leave nothing to be desired from the point of view of accuracy. A few drawings have been made specially for this book by Mr. J. Green. I have further to acknowledge the permission given by the Trustees of the British Museum, the India Office, and the Zoological Society, to reproduce a few figures from previous publications of which I am the author.

In order to render this little book more useful, the account of the Snakes of Europe has been preceded by an Introduction summarizing what is known of Snakes generally.

I have purposely avoided overburdening a work of this kind, which aims at concision, with bibliographical references and synonymic lists. I am sure my readers will be thankful for being spared this display of erudition. Whenever I have had to compile, and to trespass on ground that is not my own, I have been careful to draw only from the writings of the most trustworthy authorities. The descriptions of the species are based on the collection in the British Museum, which has been considerably increased since the publication of the Catalogue of Snakes (1893-1896). I have also had access to Monsieur F. Lataste’s rich private collection, now under my care, and Dr. R. Gestro has kindly entrusted to me for study the collection of Italian Snakes in the Genoa Museum. I am indebted to Dr. L. W. Sambon for the chapter on Parasites, which he has written at my request.

To all who have helped me I beg to tender my hearty thanks.

G. A. B.

CONTENTS

LIST OF PLATES

THE SNAKES OF EUROPE

INTRODUCTION

CHAPTER I
DEFINITION AND CLASSIFICATION

Snakes, Ophidia—regarded by some authorities as an order of the class Reptilia, by the author as a sub-order of the order Squamata, which includes besides the Lizards, Lacertilia, the Chameleons, Rhiptoglossa, and the extinct Dolichosauria and Mosasauria—may be defined as greatly elongate scaly Reptiles without limbs, or with mere vestiges of the hind pair, without movable eyelids, without ear-opening, with elongate, deeply forked tongue retractile into a basal sheath, with transverse vent and paired copulatory organs, and with the two halves of the lower jaw independently movable, connected at the symphysis by an elastic ligament.

The latter character alone distinguishes them from all Lizards, but no single Lizard possesses all the others in combination.

In their most highly developed form these Reptiles are adapted for rapid reptation and for swallowing prey much exceeding their own calibre; hence the bones of the skull, on which a prehensile function devolves, are loosely attached to the cranium by ligamentous elastic tissue, or articulated in such a manner as to permit a wide buccal expansion; whilst the absence of a sternum and the mobile attachment of the ribs allow a corresponding dilatation of the body as the prey descends into the digestive canal.

The fatal venom which many of these Reptiles possess has so impressed the mind of men, even the scientific, that for a long time snakes were primarily divided into poisonous and non-poisonous, a classification in which the more important characters, derived from the general structure, and especially from the skull, were subordinated to the physiological. Such a system was far from reflecting natural relationships. Besides, as our knowledge progressed, drawing a distinction between poisonous and harmless snakes became more and more difficult, so many snakes previously regarded as harmless proving to be poisonous in various degrees—at least enough to paralyze the small prey on which they subsist, if not to be of serious danger to man.

In the division into families, as followed in this work, the presence or absence of a poison organ is left out of consideration. Further, in this as in many other groups of the animal kingdom, external characters do not furnish trustworthy indications for higher divisions, and the definitions of the families are therefore based exclusively on osteological characters. For those who wish to name snakes with facility, the key which concludes the chapter on External Characters will, however, remedy this defect, and suffice for the identification of all the European species without any reference to their anatomy. Many attempts have been made to furnish an easy criterion for the distinction of harmless from poisonous snakes, but the characters hitherto suggested with this object can only be applied successfully to the small number of representatives in a limited area. Thus, in Southern Australia it might be stated that all snakes showing the regular nine large shields on the upper surface of the head are dangerous to man, whilst those with small shields or scales are harmless; but in most parts of Europe this criterion would have to be reversed. In some countries the shape of the pupil might be used for the purpose, in others the size of the ventral shields, or the presence or absence of a loreal shield, between the nasal and the preocular, and so on. But when we have to deal with the snakes of the whole world, about 2,000 species, of which nearly one-third are poisonous to a greater or less degree, every attempt at a definition of the two categories without regard to the dentition breaks down. Only those who have made a study of the snakes of the world can make a guess from the general appearance as to an unknown form being poisonous or not, and even they may sometimes feel embarrassed, unless the dentition be examined; the mistakes which have occasionally been made by some experienced herpetologists are proof sufficient of the fallacy of external characters for this purpose.

The Ophidia are divided into nine families, the first, third, seventh, and ninth of which have representatives in Europe:

I. No transverse (ectopterygoid) bone; pterygoid not extending to quadrate or mandible; no supratemporal; nasals in contact with prefrontals; coronoid present; vestiges of pelvis.

Maxillary loosely attached to lower surface of cranium, toothed; lower jaw edentulous; a single pelvic bone

1. Typhlopidæ.

Maxillary bordering mouth, forming a suture with premaxillary, prefrontal, and frontal, toothless; pubis and ischium present, latter forming a symphysis

2. Glauconiidæ.

II. Transverse bone present; both jaws toothed.

A. Coronoid present; nasals in contact with prefrontals.

1. Vestiges of pelvis; supratemporal present.

Supratemporal large, suspending quadrate

3. Boidæ.

(Subfamilies: Pythoninæ, Boinæ.)

Supratemporal small, intercalated in the cranial wall

4. Ilysiidæ.

2. No vestiges of pelvis; supratemporal absent

5. Uropeltidæ.

B. Coronoid absent; supratemporal present.

1. Maxillary horizontal; pterygoids reaching quadrate or mandible.

Nasals in contact with prefrontals

6. Xenopeltidæ.

Nasals not in contact with prefrontals

7. Colubridæ.

Three series: A. Aglypha (subfamilies: Acrochordinæ, Colubrinæ, Dasypeltinæ); B. Opisthoglypha (Homalopsinæ, Dipsadomorphinæ, Elachistodontinæ); C. Proteroglypha (Hydrophiinæ, Elapinæ).

2. Maxillary horizontal, converging posteriorly towards palatine; pterygoid not reaching quadrate or mandible

8. Amblycephalidæ.

3. Maxillary vertically erectile perpendicularly to transverse bone; pterygoid reaching quadrate or mandible

9. Viperidæ.

(Subfamilies: Viperinæ, Crotalinæ.)

The technical terms employed in the above synopsis will be found explained and illustrated by figures in the chapter on the Skeleton.

No serial arrangement can express the affinities of the various groups as conceived by the classificator; a diagram therefore follows to show the author’s views as to their interrelationships, and possibly their phylogeny. Leaving aside the Typhlopidæ and Glauconiidæ, which should be regarded as burrowing types independently derived from some Ophidian form less specialized than any with which we are at present acquainted, and probably without direct relationship to the Lizards, the family Boidæ, and more especially the Pythons, claim the position of ancestral group, from which all other snakes may have been derived.

Further remarks on this subject in the chapter on Dentition.

It is to be regretted that paleontology cannot help us at present as concerns the lines of evolution, the comparatively few fossil Ophidians known, from the Lower Eocene upwards, the remains of which can be identified with some measure of certainty, being either non-poisonous types (Boidæ, Ilysiidæ, Palæophiidæ, Colubridæ) or Viperidæ (Viperines from the Miocene of France and Germany, Crotalines from the Miocene of North America). The vertebræ from the Puerco Eocene of America, on the limit between the Cretaceous and Eocene periods, described as the oldest snake remains, Helagras, Cope, are stated to approach the Lacertilian type.

Whether the vertebræ named Symoliophis, Sauvage, from the chalk of France, and Coniophis, Marsh, from the Laramie Cretaceous of North America, are Ophidian, as claimed by their describers, or Dolichosaurian, cannot be decided without further material.

CHAPTER II
EXTERNAL CHARACTERS—INTEGUMENT

The form varies enormously, worm-like in some, comparatively short and heavy, elongate and more or less slender, or extremely gracile and almost filiform, in others. In this respect our common Grass-snake occupies a central position, and for this reason is termed a moderately slender form, anything above or below this standard being described as comparatively short or elongate. Our shortest and stoutest European Snakes are the Vipers, especially Vipera ursinii; our longest and slenderest, the Coluber and Zamenis, especially Zamenis dahlii. These extremes in both directions are, however, far surpassed by many exotic snakes, as we find on comparing, for instance, one of the African Puff-adders (Bitis), with certain Oxybelis and Leptognathus from Tropical America. The body may be somewhat rigid, as in some burrowing and ground snakes, not unlike in appearance to our Slow-worm and other limbless Lizards; or extremely flexible, as in many Pythons and Boas and in the Tree-snakes generally. This flexibility may be accompanied by a vertical compression of the body in relation with an arboreal existence, whilst sluggish snakes, such as most of the Viperidæ, may be remarkable for the flattening of the body, which they may further increase when basking in the sun or in order to assume a more formidable appearance on the approach of an enemy. This power of flattening out the whole or the anterior part of the body is possessed by many snakes, poisonous as well as harmless, and reaches its highest degree in the Cobras of India and Africa, the expanded anterior part being known as the “hood,” from the Portuguese name “Cobra di capello.”

Thoroughly aquatic snakes are often short and heavy, but some of the marine forms, or Hydrophids, may be extremely slender, with the posterior part of the body compressed. In some of these Sea-snakes the gracility of the anterior part, or “neck,” as it has been called, contrasts very strikingly with the great girth of the body towards the tail, and suggests a limbless Plesiosaur.

The tail, the part of the body behind the transversely cleft vent, is most frequently about one-fourth or one-fifth of the total length; but it may be much shorter, even reduced to a mere stump, as in the Typhlops, or, at the opposite extreme, enter for one half in the length of the snake, as in the African Xenurophis. This organ may taper gradually to a fine point; or end abruptly, as if mutilated; or terminate in a horny spine, such as we see in some of the Typhlops or in the Australian Death-adder, Acanthophis, or in a series of horny segments which are vibrated like a rattle, as in the well-known Crotalus of America, to which we shall refer again at the end of this chapter. In some of the burrowing Uropeltidæ, the very short tail is obliquely truncated, with indurated shields above, and acts as a trowel. And, finally, the marine snakes of the subfamily Hydrophiinæ are distinguished by a strongly compressed, oar-shaped tail, with rounded vertical outline. In a few forms, arboreal or aquatic, the tail is more or less prehensile.

Males generally have a longer tail than females, and the genital organs, which are lodged in its base, cause a swelling of that region which contrasts with the more gradually tapering extremity of the female, thus affording a means of distinguishing the sexes externally in the majority of snakes.

The rudimentary hind limbs of Boid snakes, to be mentioned further on in the description of the skeleton, terminate in a claw-like horny spur, which appears on each side of the vent in the male, and sometimes also, though less distinctly, in the female. These spurs are probably of use in facilitating the pairing, an explanation which appears the more plausible from the fact that the snakes provided with them have the copulatory intromittent organs destitute of the erectile spines which are present in most others.

The head varies in shape as much as the body. Although never actually compressed, except in the rostral region, it may be very narrow and elongate, whilst in the opposite extreme it may be strongly depressed, and so broad behind as to be abruptly defined from the anterior part of the body, or “neck.” This feature is very marked in some of the Viperidæ, and this has given rise to the incorrect generalization that poisonous snakes are distinguished from the harmless by a broad and flat head, notwithstanding the fact that some of the most dangerous, such as the Mambas, Cobras, and Kraits, have a comparatively narrow or small head, not or but slightly defined behind, whilst, on the other hand, the very opposite condition obtains in not a few of the harmless Colubrids.

Leaving the Typhlopidæ and Glauconiidæ aside for the present, snakes have a wide gape, cleft far beyond the vertical of the eyes, with, when closed, one or two notches in front for the passage of the protrusible, bifid tongue. In most snakes this chink is in the lower border of the rostral shield, capping the tip of the snout, and allows free passage to the whole tongue; in the Hydrophids, or Sea-snakes, there are two notches in the lower border of the rostral shield, through which only the bifid end of the tongue can be protruded. The eyes, varying from minute to enormous, are usually free from the surrounding shields, and may move under a transparent cap like a watch-glass, which appears to represent the lower eyelid of Lizards. The view as to this homology is derived from our knowledge of various conditions in certain series of Lizards of the families Lacertidæ and Scincidæ, where we find a transparent disc appearing like a small window in the movable lower eyelid, gradually increasing in size so as to occupy the whole of the lower eyelid, which finally becomes fused with the rudimentary upper lid and loses its mobility. In Ilysia and in most of the Uropeltidæ, the transparent disc over the eye is confluent with a thick horny shield of which it occupies the middle.

The pupil is usually circular or vertical, rarely horizontal. In some forms it is difficult to decide whether it is round or vertically elliptic; in others, like the Boas and Vipers, for instance, it is decidedly vertical, and contracts to the same extent as a cat’s. In some Water-snakes, and in Sea-snakes generally, the round pupil may contract to a mere dot. The contraction of the pupil is independent on the two sides.

The snout, or the part of the head anterior to the eyes, may be short or long, rounded or pointed, depressed or compressed, sometimes projecting strongly beyond the mouth, turned up at the end, or terminating in one (Langaha) or two (Herpeton) long scaly dermal appendages. In some burrowing forms it is provided with a more or less trenchant horizontal or vertical edge. When the sides of the snout (loreal region) form an angle with the upper surface, the angle is termed the “canthus rostralis,” which may be intensified by the loreal region being concave.

The deep pits which are sometimes present on the lips or between the nostril and the eye (loreal pit) will be alluded to further on under Sensory Organs.

The nostrils are either lateral, or, in the aquatic forms, directed upwards, sometimes entirely on the upper surface of the snout.

Most snakes have a longitudinal groove on the chin (mental groove) to allow for the distension caused by the lateral movements of the rami of the lower jaw.

In the Typhlopidæ, the head passes gradually into the vermiform body, and the small mouth is situated on the under surface of the projecting snout; the head so resembles the extremely short tail, and the mouth is so similar in shape and position to the vent, which is close to the posterior extremity of the snake, that such creatures are often believed by non-critical observers to have a head at each end. The eyes are very small, and covered over by the semi-transparent head-shields, or they may be completely concealed. There is no mental groove. It is much the same with the Glauconiidæ, which have, however, a somewhat less abbreviated tail. In both, the nostrils often open on the lower side of the snout, which may be excavated so as to appear hooked in profile, or may be provided with a sharp cutting horizontal edge.

Snakes are covered with epidermal folds in the form of scales and shields, the shape and arrangement of which affords important characters for their classification. Dermal ossifications are absent.

The scales on the body are usually elliptic or lanceolate and imbricate, forming straight longitudinal and oblique transverse series, and they are replaced on the belly and under the tail by transverse shields mostly corresponding in number with the series of scales, and also with the vertebræ. The body of the Typhlopidæ and Glauconiidæ is uniformly covered with polished, closely adherent, rounded, overlapping, sub-equal scales, without even an indication of ventral shields. In some of the Acrochordinæ, aberrant aquatic Colubrids, the scaling consists, above and beneath, of small juxtaposed, sometimes spinose granules, the skin being suggestive of the shagreen of sharks. In the marine snakes of the subfamily Hydrophiinæ, the ventral shields are often absent or merely indicated, and the scales are mostly juxtaposed or feebly imbricate, sometimes tetragonal or hexagonal, and occasionally studded with spinose tubercles. In the more typical Ophidia the imbricate scales may be long and narrow or short and broad, with every intermediate step between the two extremes; smooth or furnished with a longitudinal ridge or keel, or even several keels; nearly equal in size or with the median or outer series more or less enlarged, the longitudinal series in odd, rarely in even number; instead of running in longitudinal series parallel with the axis of the body, as is the rule, they are sometimes disposed obliquely, and among those in which we meet with this peculiarity several genera are further remarkable in having some of the oblique lateral scales furnished with a serrated keel, to which we shall again allude in the chapter on Habits, when dealing with the rustling sounds produced by certain snakes. The number of longitudinal series of scales on the body varies from 10 (Herpetodryas) to nearly 100 (Python, Boa); in the European species from 17 (Contia modesta) to 50 (Eryx jaculus). The scales are sometimes furnished near the end with one or two shallow impressions, termed “apical pits,” which afford indications for the distinction of genera and species; unless of a lighter or darker colour, as is often the case, these pits are not always easy to see, except in a strong light and with the aid of a powerful magnifying glass.

The ventral shields, also called “gastrosteges,” usually occupy the whole width of the belly; but they may be much narrower—in Eryx, for instance. They are sometimes bent at an angle on the sides, and this angle may even form a sharp keel, accompanied by a notch in the posterior border, corresponding to the keel, as in several of the more arboreal genera of Colubrids. The shields under the tail, termed subcaudals or “urosteges,” are sometimes similar to the ventrals, but more often disposed in pairs; in certain species or individuals some of the subcaudals are single, and the others paired. When the number of subcaudals is given in the descriptions, each pair is reckoned as one, and the conical or spine-like shield which caps the end of the tail is not included. These numbers afford important characters for the definition of species, and sometimes also for the distinction of sexes. The subcaudals are nearly always much fewer than the ventrals, but the difference is often not so great in the males as in the females, the tail of which is usually shorter in proportion to the body. It is noteworthy that in many species, if the number of subcaudals (C.) be added to that of the ventrals (V.), the total is nearly the same in the male as in the female, however much the respective numbers may differ when taken separately. The following figures may be given by way of example, taken from British specimens:

Although this rule is by no means universal, and does not apply at all to some species, it will be found to hold good in many cases, and is of interest in showing that the changes that have taken place in the vertebral column (the vertebræ corresponding in number to the shields), according to the sexes, have been by a modification of the character of the segments about the anal region, a conversion of trunk vertebræ into caudals, or vice versa. In dealing with certain species—of Vipers, for instance—it is important, for systematic purposes, to keep the counts of shields distinct for the two sexes.

The shield which covers the vent, the anal shield, is either single or divided into two.

Some snakes have the head covered with scales or small tubercles similar to those on the body, but in the great majority the lepidosis is in the form of large symmetrical juxtaposed shields, the shape, proportions, and number of which furnish some of the most important characters for the distinction of genera and species. These head-shields belong to two primarily different types, from each of which all further modifications may be regarded as derived by alteration in shape or by disintegration. The first type is that shown by the Typhlopidæ and Glauconiidæ, which is explained by the figure on the next page.

The rostral, which is usually the largest of the head-shields, extends to the upper surface of the head, of which it may occupy the greater part. In the Glauconiidæ, the ocular usually borders the mouth.

As may be seen by a comparison of the first figure with the second, the arrangement of the head-shields is essentially different from that which prevails in the Colubrids and the majority of other snakes.

The second type is exemplified by the head of a member of the genus Zamenis.

In the descriptions, temporals 2 + 3 means two superposed temporals in the first row, three in the second. The internasals and the temporals, and the loreal and the preocular, are sometimes absent, and the prefrontal or the internasal may be single. One or two large shields are in rare cases present behind the parietals, and are called occipital.

Fig. 1—Head of Typhlops braminus. (From “Fauna of British India”)

f, Frontal; ip, interparietal; l, labial; n, nasal; o, ocular; p, parietal; po, preocular; prf, prefrontal; r, rostral; so, supraocular.

A breaking up into smaller shields takes place in many snakes. In the Pythons, for instance, the frontal may be divided into two by a longitudinal cleft, and separated from the prefrontals by small shields. In some Vipers, such as V. berus and V. ursinii, in which the frontal and parietals, though reduced in size, usually preserve their primitive condition, the former is normally separated from the supraocular by a series of small shields, and the internasals and prefrontals are broken up; in these snakes the small shield or shields behind the rostral are termed “apical,” and those on the upper edge of the snout are termed “canthals.” The shield which, in Vipers, separates the rostral from the nasal is called “naso-rostral.” Allusion has been made above to the scaly dermal appendages which terminate the snout in certain genera. Some Viperidæ are furnished with horn-like erect spines above the eyes or at the end of the snout, which add greatly to their sinistral appearance.

Fig. 2—Head of Zamenis ventrimaculatus. (From “Fauna of British India”)

cs, Chin-shields (anterior); cs´, chin-shields (posterior); f, frontal; in, internasal; l, loreal; la, labial (upper); la´, labial (lower); m, mental; n, nasal; p, parietal; pf, prefrontal; pro, preocular; pto, postocular; r, rostral; sbo, subocular; so, supraocular; t, temporals (first row); t´, temporals (second row); v, first ventral.

The periodical shedding of the outer layer of the epidermis in a single piece, including even the covering of the eye, is one of the most striking peculiarities of snakes, although paralleled in the Lizards of the family Anguidæ, to which our British Slow-worm belongs. The skin becomes detached at the lips, and is turned inside out from head to tail, without any sort of laceration when the snake is in good health. These exuviæ are transparent, but often carry a certain amount of pigment, especially those of the Vipers, in which the characteristic dark markings are perfectly visible; they usually exceed the length of the reptile, owing to stretching. In Sea-snakes the epidermis is cast piecemeal, and sloughing is a longer operation than in ordinary snakes.

In Rattlesnakes each piece of the rattle, or “crotalon,” in which the tail terminates, represents a retained portion of the sloughed epidermis. This remarkable appendage looks like a number of horny rings, but it consists in reality of hollow, bell-like pieces, similar to the terminal one, or “button,” each with a circular constriction, in which the incurved free edge of the following piece fits, thus keeping the pieces together without impairing the mobility necessary to produce the rattling sound for which the apparatus is intended. At each exuviation one bell-shaped horny piece is added. The number of segments in the rattle is, therefore, not an index to age, as formerly believed; nor is it to the number of exuviations, for whilst segments are being added at the base of the apparatus the terminal ones break off and are lost. A Crotalus sixteen months old may have six pieces to the rattle if there have been six exuviations and no loss. No rattle appears ever to comprise more than about twenty pieces, even in old specimens. The size of the terminal button shows whether it was formed at birth or at any later period, no growth taking place in the horny tissue.

So far as trustworthy records are concerned, the largest snakes known, the Malay Python reticulatus and the South American Anaconda, Eunectes murinus, reach a length of 25 to 30 feet. Measurements of skins must be accepted with caution, as a skin may easily be stretched to once and a half its real length; in estimating the exact length from such a stretched skin, it is necessary to deduct the interstitial spaces showing between the scales, and about one-fourth of the scale to allow for the overlap. The smallest snake known is 4 inches long (Glauconia dissimilis). The largest European snake (Coluber quatuorlineatus) is reported to reach a length of 8 feet; the smallest (Typhlops vermicularis) does not exceed 14 inches.

Key to the Identification of the European Snakes from External Characters only

I. Eyes minute, under the head-shields; mouth small, inferior; body vermiform, covered with uniform scales above and beneath; vent close to the end of the body, the extremely short tail ending in a small spine

Typhlops vermicularis.

II. Eyes very small, with vertical pupil; upper surface of head covered with small scales; ventral shields much narrower than the body; tail short, ending obtusely; subcaudals single, or mostly single; scales smooth or feebly keeled, in 40 to 50 rows

Eryx jaculus.

III. Eyes small, moderate, or large; ventral shields at least nearly as broad as the body; tail tapering to a point; subcaudals paired.

A. Pupil round; upper surface of head with nine large shields; no upper labial in contact with the parietal; anal shield usually divided.

1. Dorsal scales strongly keeled, with paired apical pits; a single anterior temporal.

a. Nostrils lateral; internasals broadly truncate in front.

Scales in 19 rows; normally 1 pre- and 3 postoculars; usually 7 upper labials, third and fourth entering the eye; ventrals 157-181; subcaudals 50-88

Tropidonotus natrix.

b. Nostrils directed upwards; internasals much narrowed in front.

Scales in 19 rows; normally 2 pre- and 3 or 4 postoculars; suboculars sometimes present; usually 8 upper labials, fourth or fourth and fifth entering the eye; ventrals 160-187; subcaudals 48-79

Tropidonotus tessellatus.

Scales in 21 (rarely 19 or 23) rows; normally 1 or 2 pre- and 2 postoculars; usually 7 upper labials, third and fourth entering the eye; ventrals 147-164; subcaudals 46-72

Tropidonotus viperinus.

2. Dorsal scales smooth or feebly keeled; normally a single loreal.

a. Two or three superposed anterior temporals (very rarely one); nostril usually between two nasals.

α. A subocular below the preocular.

* Scales smooth, in 17 or 19 rows.

Two upper labials entering the eye; preocular not in contact with the frontal; scales with two apical pits; ventrals more or less distinctly angulate laterally, 160-230; subcaudals 87-131

Zamenis gemonensis.

Two upper labials entering the eye; preocular usually in contact with the frontal; scales with a single apical pit; ventrals very distinctly angulate laterally, 205-218; subcaudals 98-132

Zamenis dahlii.

** Scales in 23 to 29 rows (usually 25 or 27), with two apical pits.

Upper labials usually separated from the eye by a series of suboculars; preocular in contact with the frontal; scales smooth; ventrals very distinctly angulate laterally, 222-258; subcaudals 77-107

Zamenis hippocrepis.

Two upper labials entering the eye; preocular not in contact with the frontal; scales feebly but distinctly keeled; ventrals not angulate laterally, 195-234; subcaudals 56-90

Coluber quatuorlineatus.

Two upper labials entering the eye; preocular not in contact with the frontal; scales smooth or faintly keeled; ventrals not or but very obtusely angulate laterally, 172-214; subcaudals 50-80

Coluber dione.

β. No subocular; scales smooth, or faintly keeled on the posterior part of the body.

* Ventrals more than 200; scales with two apical pits.

Snout obtuse; rostral broader than deep; scales in 21 or 23 rows; ventrals distinctly angulate laterally, 212-248; subcaudals 60-91

Coluber longissimus.

Snout obtuse; rostral broader than deep; scales in 25 or 27 rows; ventrals not angulate laterally, 222-260; subcaudals 68-90

Coluber leopardinus.

Snout pointed, strongly projecting; rostral deeper than broad, wedged in between the internasals; scales in 25 to 29 rows; ventrals not angulate laterally, 201-220; subcaudals 48-68

Coluber scalaris.

** Ventrals not more than 200; scales mostly with a single apical pit.

Rostral at least as deep as broad, often wedged in between the internasals; usually 7 upper labials, third and fourth entering the eye; scales in 19 (rarely 21) rows; ventrals 153-199; subcaudals 41-70

Coronella austriaca.

Rostral broader than deep; usually 8 upper labials, fourth and fifth entering the eye; scales in 21 (rarely 19 or 23) rows; ventrals 170-200; subcaudals 49-72

Coronella girondica.

b. A single anterior temporal; nostril in a single nasal; scales smooth, with single apical pits, in 17 rows; ventrals 150-191; subcaudals 53-78

Contia modesta.

3. Scales longitudinally grooved in the adult, in 17 or 19 rows; two loreals; canthus rostralis strongly marked; frontal very narrow, in contact with the preocular; ventrals 160-189; subcaudals 68-102

Cœlopeltis monspessulana.

B. Pupil vertical or vertically subelliptic (sometimes appearing round in Macroprotodon).

1. Scales smooth, mostly with single apical pits; upper surface of head with nine large shields.

Frontal 11⁄2 to 2 times as long as broad; loreal separated from the eye by the preocular; one upper labial usually in contact with the parietal; scales in 19 to 23 (rarely 25) rows; ventrals 153-192; anal divided; subcaudals 40-54

Macroprotodon cucullatus.

Frontal 11⁄4 to 11⁄2 times as long as broad, much shorter than the parietals; loreal entering the eye; scales oblique, in 19 or 21 rows; ventrals 186-222; anal divided; subcaudals 48-73

Tarbophis fallax.

Frontal 11⁄4 to 11⁄2 times as long as broad, nearly as long as the parietals; loreal entering the eye; scales oblique, in 19 or 21 rows; ventrals 203-235; anal entire; subcaudals 54-70

Tarbophis iberus.

2. Scales keeled, with two apical pits; anal shield entire.

a. No pit between the nostril and the eye; upper head-shields small, if present; nasal separated from the rostral by a naso-rostral; eye separated from the upper labials by suboculars.

α. Snout not turned up at the end; supraocular usually extending posteriorly beyond the vertical of the posterior border of the eye; frontal and parietal shields usually well developed; usually a single series of scales between the eye and the upper labials.

Snout obtusely pointed, flat above, or with the canthus slightly raised; rostral usually in contact with a single apical shield, rarely with two; 6 to 9 upper labials, usually 7 or 8; scales in 19 rows, rarely 21; ventrals: ♂ 120-135, ♀ 125-142

Vipera ursinii.

Snout pointed, with raised canthus; rostral in contact with a single apical shield; 8 or 9 upper labials; scales in 21 rows, rarely 19; ventrals: ♂ 130-148, ♀ 130-150

Vipera renardi.

Snout truncate or broadly rounded, flat above or with slightly raised canthus; rostral in contact with two apical shields, rarely with one; 8 or 9 upper labials; scales in 21 rows, rarely 19 or 23; ventrals: ♂ 132-150, ♀ 132-158

Vipera berus.

β. Snout usually more or less turned up at the end or produced into a scaly dermal appendage; supraocular not extending posteriorly beyond the vertical of the posterior border of the eye; frontal and parietals often absent or very small; 2 or 3 series of scales between the eye and the upper labials; 9 to 13 upper labials; scales in 21 or 23 rows, rarely 19 or 25.

Snout simply turned up, the raised portion bearing 2 or 3 scales; rostral not more than once and a half as deep as broad; ventrals: ♂ 134-158, ♀ 141-169

Vipera aspis.

Snout simply turned up or produced into a small appendage, the raised portion with 5 or 6 (rarely 3) scales; rostral 11⁄2 to 2 times as deep as broad; ventrals: ♂ 125-146, ♀ 135-147

Vipera latastii.

Snout produced into an appendage covered with 10 to 20 scales; rostral not reaching the summit of the rostral appendage; ventrals: ♂ 133-161, ♀ 135-163

Vipera ammodytes.

γ. Snout not turned up at the end; supraocular narrow or broken up into several small shields; upper surface of head with small, usually keeled scales; two or three series of scales between the eye and the upper labials; scales in 23 to 27 rows, usually 25; ventrals: ♂ 151-177, ♀ 153-180

Vipera lebetina.

b. A pit between the nostril and the eye; upper surface of head with 9 large shields; nasal in contact with the rostral; third upper labial entering the eye; scales in 23 rows; ventrals 149-174; subcaudals 31-44

Ancistrodon halys.

CHAPTER III
COLORATION

In dealing with the coloration, we have first to distinguish between the colour and the markings. The former is very often highly variable among snakes of the same species, to say nothing of the changes which may take place with age or with the condition of the individuals, whether before or after exuviation; it is not unusual to find among specimens from the same locality a great range of variation, from greyish-white to brown, or red, or black, as, for instance, in our Common Viper. The latter afford more important characters, and often furnish valuable indications for the distinction of species; but even the disposition of the markings is subject to great individual variations, more likely to mislead than to help the inexperienced student in the discrimination of species. It is therefore always advisable to resort in the first instance to structural characters for the purpose of specific identification, and to fall back on coloration only as a means of confirmation. If we were to be guided by colour and markings alone, how could we believe that an adult four-lined Coluber quatuorlineatus is of the same species as the handsomely spotted Coluber sauromates; and yet, if we compare the young of these two snakes we find them to be absolutely identical in their markings, and, in the absence of any structural differences, we are forced to conclude that they only represent two forms of the same species, of which the latter is the more primitive.

It is nevertheless a fact that, with a few exceptions, the markings, however variable they may be, are reducible to certain fundamental patterns to which the innumerable variations may be traced back, and their derivation followed and scientifically explained. Let us consider, for instance, another species of Coluber, highly variable in its markings: C. leopardinus, of which the typical form, so called from having been the first described and named, is not by any means to be regarded as the most primitive.

First, we must take for granted that the markings of all such snakes, whether consisting of spots, stripes, or bars, start from a regular arrangement, which may be theoretically represented by four paired longitudinal series on the head and body: (1) Dorsal series (D); (2) Dorso-lateral (DL); (3) Lateral (L); (4) Ventro-lateral (VL). The first starts from the middle line of the head, and is continued along the spine; the second occupies the space between the first and third, which originates at the tip of the snout, passes through the eye, and is continued on the temple and along the side of the body; the fourth follows the lower lip, and extends along each side of the belly. Bearing this in mind, we find that the variety of C. leopardinus named schwoederi, with a vertebral series of paired spots, is to be regarded as the most primitive, from which we can derive, on the one hand, the true leopardinus by imagining a transverse fusion of the spots of series D into a single row, some of the spots often actually revealing, in their biscuit shape, their dual origin; whilst, on the other hand, confluence of the paired spots of the same series into two longitudinal stripes produces the variety named quadrilineatus (see [Plate VII].). In this particular instance, the paired series D has fused into a single streak on the head, and the series L appears to have departed from its primitive course to extend on the upper surface of the head, both in front of and behind the eye.

Many snakes show an interocular band extending from lip to lip, through the eyes, across the interorbital region. In others the lateral stripe L may bifurcate in front of the eye, an upper branch extending across the snout, through transverse fusion of series D and DL, and it may also bifurcate in like manner on the temporal region, fusing with the corresponding marking on the other side to form a W-shaped figure. The pattern of markings on the upper surface of the head is, however, often very complicated, and hence difficult of explanation.

As a second example of the derivation of patterns, we may mention Vipera aspis, which varies enormously as to its mid-dorsal markings, forming, in different individuals or even on different parts of the body, single or paired spots, a zigzag band, or transverse bars; all these are derived from the paired spots of series D. Each pair of spots may fuse and form transversely oval or elliptical spots or bars, or the spots may assume an alternate disposition from which, through confluence, the zigzag or sinuous band results. Thus, spotted and striped patterns may be traced to a common origin, however fundamental the difference between them appears at first sight. If the elements of the four series, D, DL, L, and VL, unite transversely with each other, and also with the spots on the ventral surface, we obtain ringed forms such as the Coral-snakes. That the black nuchal collar of our common Grass-snake is actually formed by the fusion of the spots of three originally distinct series has been proved by tracing the development of the markings in the embryo.

In various species a pair of light streaks extends along the back, bordering the D area, without interfering with the other markings, as we see, among European snakes, in some specimens of Tropidonotus natrix and viperinus, and Vipera berus.

Although it sometimes happens that a definite system of markings prevails throughout a genus, such as the annulate form in the South American Elaps, this is far from being universally the case; many closely allied species, or individuals of the same species, may be distinguished by very different patterns. Even on the same individual we may find two opposite types of markings without any transition, as in two Central American species of widely different genera, Polyodontophis annulatus and Zamenis mexicanus, in which the anterior part of the body is annulate or barred, and the rest longitudinally striped.

It is also a remarkable fact that very often the two sides of the body are not alike in their markings, appearing as if formed of the union, on the median line, of the right and left halves of two individuals. Thus it may happen, in annulate forms, that some of the annuli are broken exactly in the mid-dorsal and mid-ventral lines, and that the halves do not correspond in number on the two sides. In the handsome South American Lachesis alternatus, which derives its specific name from the two series of large C-shaped, dark, light-edged markings which adorn its back, these markings are not always alternating, as is the rule; but some may lie opposite to each other and back to back, this being due to the fact that the numbers of the markings do not correspond on the two sides. In one specimen I count twenty-four of these markings on the left side, and twenty-seven on the right. This shows that great importance cannot be attached to the number of the markings, for systematic purposes. In fact, in some Coral-snakes, Elaps fulvius for instance, the number of annuli may vary from twelve to fifty-two, with every gradation between the extremes. The bilateral asymmetry to which we have alluded produces the chess-board arrangement of the ventral spots in many snakes.

Among the markings which call for investigation as to their meaning, we must allude to the presence, in some Colubrids, of a small, light, dark-edged spot, or of a pair of light dots close together, in the middle of the parietal shields or on each side of the suture between these shields, which correspond in their position to the parietal organ of many Lizards. May not this marking be in some way correlated with sensory organs, like the apical pits on the scales of the body? And what is the explanation of such bizarre signs as the spectacle or the eye-spot on the hood of the Indian Cobra? At present it is as inexplicable as the lugubrious emblem on the thorax of the Death’s-head Moth. It cannot be suggested that it is a warning mark intended to terrify intruders, for when the Cobra is at rest the hood is folded, and the characteristic marking is not displayed; whilst as soon as it is aroused, and the hood expanded, it faces its enemy in such a way that the spectacle, or ocellus, is not to be seen.

First among the most brilliantly coloured snakes, of which there are many, stand the Coral-snakes, Elaps, of America, mostly annulate with red, yellow or white, and black. This striking coloration obtains also in diverse harmless snakes inhabiting the same part of the world, and this coincidence has been adduced in favour of the theory of mimicry, correlated with that of natural selection, which accounts for the resemblance as being of advantage to a harmless species, which is thus mistaken for one notorious for its deadly poison, and advertised as such by its brilliant colours (warning coloration). But other poisonous and much more dangerous snakes are not, as a rule, endowed with brilliant colours. It is true that these also may have their mimics: the Krait, Bungarus cæruleus, and Lycodon aulicus, in India, the Pit-viper, Ancistrodon himalayanus, and Psammodynastes pulverulentus, in the Himalayas and Assam, are good examples of such cases. On the other hand, there are equally striking instances of what one would regard as mimics if they only occurred together; thus, there is no better case of general resemblance between a poisonous and a harmless snake than we find in the Indian Cobra and the Coluber corais of tropical America, where Cobras are absent, or between a Viper and the Boid Enygrus asper, from New Guinea, where no Vipers exist.

Without attempting to offer any suggestion to account for the similarity of markings which prevails in certain parts of the world, attention may be drawn to the predominance of longitudinal dark and light stripes in the Indo-Malayan representatives of the American Elaps, shared by many innocuous snakes of similar form inhabiting the same region, and to the striped tails common to various Colubrids of Madagascar, as if the snakes of a district had agreed to conform to certain fashions in dress.

It is further noteworthy, in relation to the theory of warning coloration, that many Uropeltids, innocent burrowing creatures living underground or concealed under stones or rotting tree-trunks in the forests of Southern India and Ceylon, hardly ever showing themselves in daylight, are among the most striking for their bright yellow or red and black markings. We may point out at the same time the very marked resemblance in form and coloration between the Uropeltid Melanophidium bilineatum, and the Apodal Batrachian Ichthyophis glutinosus, both occurring together in Southern India.

The colour of snakes often harmonizes with their surroundings. Thus, many Tree-snakes, Boid, Colubrid, or Viperid, are of a bright green, like the foliage in which they are concealed. On the other hand, other Tree-snakes are not green, or only some specimens are green, as in the genera Dendraspis and Dispholidus. Desert-snakes are of the yellowish or reddish colour of the sand or rock on which they live, and in species whose range extends over different districts the desert individuals are paler, without or with less distinct markings, as compared to their fellows among other surroundings. In addition to their markings, some snakes are adorned with a metallic iridescent gloss, due to a fine striation of the scales.

The iris is often metallic, gold, bronze, or copper-red, and the black streaks of the head sometimes extend over it.

Although, unlike many lizards, snakes are unable to rapidly alter their colours, some produce a semblance of this phenomenon when inflating their neck or body; this is due to the presence of dark and light markings or of a bright pigment in the interstitial skin, which is not seen when the scales overlap. Thus, in the Indian Tree-snake Dryophis mycterizans the skin between the green or brown scales in the anterior part of the body is black and white, producing a striped pattern when the neck is inflated; the skin of the same region is bright vermilion in the Malay Tropidonotus subminiatus; many more examples could be quoted. The spectacle marking on the hood of the Indian Cobra involves the scales as well as the interstitial skin.

As a rule there are no sexual differences in colour. Yet these are so marked in our Common Adder that the sex of a specimen can nearly always be recognized by the coloration. This is, however, the exception, even in the genus to which the Adder belongs. A nuptial dress is unknown in snakes.

A special livery for the young is rather exceptional, but very often the new-born is more vividly coloured than its parents, and in many black varieties the young is similar to the typical form. Some green Tree-Boids (Chondropython and Corallus caninus) are not green, but yellowish, cream-colour, or pinkish, when young, the green appearing around the white spots, which are the remains of the ground colour, and gradually spreading over the whole body. Conversely, the young of a variety of the Pit-viper Lachesis wagleri, common in the Malay Peninsula, is green, and the adult black and yellow. In the young of Grayia ornata, a West African Water-snake, the markings of the young are to those of the adult like positive and negative in photography, the white bars, forked on the sides, which extend across the black back of the former being gradually transformed into black bars on a light ground in the latter; in such a case it is impossible to decide whether the dark or the light parts are to be considered as the ground colour.

That the skin of many snakes contains soluble colouring matter of a special kind is well known, green snakes, such as Dryophis prasinus and Lachesis gramineus staining the spirit in which they are preserved. Chemists have not yet paid attention to this question, which requires investigation.

Melanism is frequent in snakes, and sometimes affects all individuals in the same locality. It seems undesirable to bestow varietal names on such aberrations, as is so frequently done by systematists, any more than we should in the case of albinos. Melanism may be produced in two ways: by an extension of the black markings, which invade the whole surface, as in the males of Vipera berus; or by a general darkening of the ground colour and of the markings, as in the females of the same species. In the latter case, the markings reappear under certain lights or after a prolonged sojourn in spirits. Sometimes, as in Zamenis gemonensis, the uniform black colour appears only as the snake approaches the adult condition, the young having the normal livery.

Partial albinism is rare; perfect albinism, characterized by absence of black pigment in the eye, rarer still. Cases have been observed, among European species, in Tropidonotus natrix and tessellatus, in Coluber longissimus, and in Coronella austriaca.

CHAPTER IV
SKELETON

The typical Ophidian skull is characterized by a solidly ossified brain-case, with the distinct frontals and the united parietals extending downwards to the basisphenoid, which is large and produced forward into a rostrum extending to the ethmoidal region. The nasal region is less completely ossified, and the paired nasals are often attached only at their base. The occipital condyle is either trilobate and formed by the basioccipital and the exoccipitals, or a simple knob formed by the basioccipital; the supraoccipital is excluded from the foramen magnum. The basioccipital may bear a strong, curved ventral process or hypapophysis (in the Vipers).

The prefrontal is situated, on each side, between the frontal and the maxillary, and may or may not be in contact with the nasal; the postfrontal, usually present, borders the orbit behind, rarely also above, and in the Pythons a supraorbital is intercalated between it and the prefrontal.

Fig. 3—Skull of Python amethystinus. (From British Museum Catalogue of Snakes)

an, Angular; ar, articular; bo, basioccipital; bs, basisphenoid; cor, coronoid; c.a, columella auris (stapes); d, dentary; eo, exoccipital; epg, ectopterygoid (transverse); f, frontal; m, maxillary; n, nasal; p, parietal; pl, palatine; pm, premaxillary; prf, prefrontal; pro, proötic; pg, pterygoid; ptf, postfrontal; q, quadrate; so, supraoccipital; sor, supraorbital; sp, splenial; ste, supratemporal; tu, turbinal; v, vomer.

The premaxillary is single and small, and as a rule connected with the maxillary only by ligament. The paired vomer is narrow. The palatine and pterygoid are elongate and parallel to the axis of the skull, the latter diverging behind and extending to the quadrate or to the articular extremity of the mandible; the pterygoid is connected with the maxillary by the ectopterygoid or transverse bone, which may be very elongate, and the maxillary often emits a process towards the palatine, the latter bone being usually produced inwards and upwards towards the anterior extremity of the basisphenoid. The quadrate is usually large and elongate, and attached to the cranium through the supratemporal (often regarded as the squamosal). In rare cases (Miodon, Polemon) the transverse bone is forked, and articulates with two branches of the maxilla. The quadrate and the maxillary and palatopterygoid arches are more or less movable to allow for the distension required by the passage of prey, often much exceeding the calibre of the mouth. For the same reason, the rami of the lower jaw, which consist of dentary, splenial, angular, and articular elements, with the addition of a coronoid in the Boidæ and a few other small families, are connected at the symphysis by a very extensible elastic ligament.

The hyoid apparatus is reduced to a pair of cartilaginous filaments situated below the trachea, and united in front.

There are various modifications according to the genera. A large vacuity may be present between the frontal bones and the basisphenoid (Psammophis, Cœlopeltis); the maxillary may be much abbreviated and movable vertically, as in the Viperidæ; the pterygoids may taper and converge posteriorly, without any connexion with the quadrate, as in the Amblycephalidæ; the supratemporal may be much reduced, and wedged in between the adjacent bones of the cranium; the quadrate may be short or extremely large; the prefrontals may join in a median suture in front of the frontals; the dentary may be freely movable, and detached from the articular posteriorly.

Fig. 4—Skull of Typhlops lumbricalis. (From British Museum Catalogue of Snakes)

Lettering of the bones as in Fig. [3]

Fig. 5—Skull of Glauconia macrolepis. (From British Museum Catalogue of Snakes)

Lettering of the bones as in Fig. [3]

The deviation from the normal type is much greater still when we consider the degraded, worm-like members of the families Typhlopidæ (Fig. [4], p. [43]) and Glauconiidæ (Fig. [5]), in which the skull is very compact and the maxillary much reduced. In the former this bone is loosely attached to the lower aspect of the cranium; in the latter it borders the mouth, and is suturally joined to the premaxillary and the prefrontal. In both the tranverse bone and the supratemporal are absent, but the coronoid element is present in the mandible.

Fig. 6—Skull of Tropidonotus natrix. (From British Museum Catalogue of Snakes)

Lettering of the bones as in Fig. [3]

The principal modifications of the skull in the European genera may be contrasted as in the following synopsis:

I. Quadrate articulating with the cranium, supratemporal absent; mandible much shorter than the skull, with coronoid bone; maxillary small, on lower aspect of cranium; pterygoids not extending to quadrate; nasals forming long sutures with the premaxillary, prefrontals, and frontal

Typhlops.

II. Quadrate suspended from the supratemporal; mandible at least as long as the skull; pterygoids extending to quadrate or mandible.

A. Mandible with coronoid bone; nasals in sutural contact with frontals and prefrontals; transverse bone short, not projecting much beyond cranium; maxillary not half as long as mandible, which is not longer than skull (to occiput)

Eryx.

Fig. 7—Skull of Zamenis gemonensis. (From British Museum Catalogue of Snakes)

B. No coronoid bone; nasals isolated.

1. Maxillary elongate, not movable vertically.

a. Maxillary half as long as mandible.

Supratemporal half as long as skull, projecting far beyond cranium; mandible much longer than skull

Tropidonotus.

Supratemporal not half as long as skull, projecting far beyond cranium; mandible much longer than skull

Zamenis.

Fig. 8—Skull of Coluber longissimus. (From British Museum Catalogue of Snakes)

Supratemporal not half as long as skull, projecting but slightly beyond cranium; mandible much longer than skull

Coluber.

Supratemporal not half as long as skull, not projecting beyond cranium; mandible not longer than skull

Coronella, Contia.

b. Maxillary not half as long as mandible, which is longer than skull; supratemporal not half as long as skull, projecting beyond cranium.

Fig. 9—Skull of Coronella austriaca. (From British Museum Catalogue of Snakes)

Quadrate longer than supratemporal; maxillary much longer than quadrate, nearly straight in front of prefrontal; a large vacuity between the frontal bones and the basisphenoid

Cœlopeltis.

Quadrate not longer than supratemporal; maxillary little longer than quadrate, strongly curved in front of prefrontal

Macroprotodon.

Quadrate longer than supratemporal; maxillary little longer than quadrate, nearly straight in front of prefrontal

Tarbophis.

Fig. 10—Skull of Vipera lebetina. (From British Museum Catalogue of Snakes)

Lettering of the bones as in Fig. [3]

2. Maxillary much abbreviated and erectile; supratemporal not half as long as skull; mandible much longer than skull; basioccipital with a strong process.

Maxillary bone solid

Vipera.

Maxillary bone hollowed out

Ancistrodon.

The vertebræ number 130 to 500—in the European forms 147 (Vipera ursinii) to 330 (Coluber leopardinus).

The vertebral column consists of an atlas (composed of two vertebræ) without ribs; numerous precaudal vertebræ, all of which, except the first or first three, bear long, movable, curved ribs with a small posterior tubercle at the base, the last of these ribs sometimes forked; two to ten so-called “lumbar vertebræ” without ribs, but with bifurcate transverse processes (lymphapophyses) enclosing the lymphatic vessels; and a number of ribless caudal vertebræ with simple transverse processes. When bifid, the ribs or transverse processes have the branches regularly superposed.

The centra have the usual cup-and-ball articulation, with the nearly hemispherical or transversely elliptic condyle at the back (procœlous vertebræ), whilst the neural arch is provided with additional articular surfaces in the form of pre- and post-zygapophyses, broad, flattened, and overlapping, and of a pair of anterior wedge-shaped processes called zygosphene, fitting into a pair of corresponding concavities, zygantrum, just below the base of the neural spine. Thus the vertebræ of snakes articulate with each other by eight joints in addition to the cup-and-ball on the centrum, and interlock by parts reciprocally receiving and entering one another, like the joints called “tenon-and-mortice” in carpentry. The precaudal vertebræ have a more or less high neural spine which, as a rare exception (Xenopholis), may be expanded and plate-like above, and short or moderately long transverse processes to which the ribs are attached by a single facet. The centra of the anterior vertebræ emit more or less developed descending processes, or hæmapophyses, which are sometimes continued throughout (Fig. [11], A), as in Tropidonotus, Vipera, and Ancistrodon, among European genera.

Fig. 11—Posterior Precaudal Vertebræ of Lioheterodon (A) AND Heterodon (B). (From British Museum Catalogue of Snakes)

a, Back view; b, lower view; c, side view.

In the caudal region, elongate transverse processes take the place of ribs, and the hæmapophyses are paired, one on each side of the hæmal canal. In the Rattlesnakes the seven or eight last vertebræ are enlarged and fused into one.

No snake shows any rudiments of the pectoral arch, but remains of the pelvic are found in the Typhlopidæ, the Glauconiidæ, the Boidæ, and the Ilysiidæ. In the first these vestiges are reduced to a single bone (ilium?) on each side; in the second they consist of ilium, pubis, and ischium, the latter forming a ventral symphysis, and a rudimentary femur; whilst in the third there is a long ilium, attached to the lower branch of the first bifurcate transverse process of the lumbar vertebræ, bearing three short bones, the longest of which, regarded as the femur, terminates in a claw-like spur which, in males at least, usually appears externally on each side of the vent.

CHAPTER V
DENTITION

In the most generalized snakes—those which show the nearest approach to lizards—teeth are present not only on the rami of both jaws, but also on the premaxillary bone, on the palatines, and on the pterygoids. A reduction of the dentition takes place in various genera, in which the teeth of either the upper or the lower jaw, and of the palatines or pterygoids, or both, may be absent, and the premaxillary is devoid of teeth in the great majority, including all European representatives, of the Ophidia.

In the egg-eating snakes of the genera Dasypeltis and Elachistodon the dentition is very much reduced, in accordance with the peculiar régime, and this deficiency is compensated by the development on some of the anterior thoracic vertebræ of long, tooth-like processes (hypapophyses) directed forwards, and capped with a remarkably dense, vitreous tissue simulating enamel, the function of these tooth-like processes being to break the shell of the egg within the gullet, where none of its contents are lost, the shell being afterwards rejected through the mouth in the form of a pellet.

With the exception of the worm-like Typhlopidæ, which are provided with a few teeth in the upper jaw only, European snakes have teeth on the maxillary, palatine, pterygoid, and dentary bones. Unless the maxillary be strongly abbreviated and modified in connexion with the poison apparatus, as in the Viperidæ, the teeth in the jaws as well as on the palate form single longitudinal series; they are elongate, conical, with or without a sharp posterior edge, more or less recurved, acutely pointed, sometimes needle-like, and directed backwards, as behoves their function, which, in addition to attack and defence, is to prevent the retrogression of the prey in the act of prehension and deglutition. A notable exception occurs in the genus Iguanognathus, from Sumatra, all the teeth having spatulate crowns ribbed along the outer side. Unfortunately, nothing is known as to the food of this remarkable snake. The teeth are coated with a thin layer of enamel. It was held, for a time, that the glossy outer coating was only due to a denser structure of the dentine. As in all living Reptiles with the exception of the Crocodiles, the teeth are not implanted in true sockets, but simply ankylosed to the bone on which, when detached, their slightly enlarged base, or rather the bony tissue on which it rests, leaves a shallow impression, or pseudo-socket. In the process of biting or feeding, some of the teeth are frequently lost, and are readily replaced by others lying in reserve in the gum at the inner side, and becoming fixed to the bone soon after a vacancy occurs. Such replacement teeth, of different grades of development, form several series, so that in a snake like our common Tropidonotus the mouth may contain four times as many teeth as are functional, without reckoning different earlier stages of tooth germs which escape ordinary observation, being placed vertically one above the other.

Three types of teeth, connected by every intermediate step, are distinguished: the solid, the grooved, and the canaliculated or tubular, so-called “perforated”; the third, as we shall explain, being only a further modification of the second. In the grooved tooth, a sulcus runs along the anterior or outer surface, its object being to convey into the wound the secretion of a poison gland. It varies in depth according to the species, and may be so slight as to escape detection without a very strong magnifying glass. In some the sulcus may be very deep and wide, forming a canal round which the tooth folds to the extent of its borders nearly meeting; from this condition the so-called “perforated” fang is derived through the complete fusion of the borders of the tooth, and the obliteration of the line of union except at each extremity. The structure of such a fang may be best understood by imagining a tooth, lined all round with the same layer of dentine and enamel, being flattened out in a vertical plane and then folded over, the outer edges coalescing on the front median line in such a way that the inner wall of the tooth is in reality the anterior surface, and the outer wall the posterior surface, of the ordinary tooth.

Grooved teeth, with open canal, are situated either at the anterior extremity (Proteroglyphs) or at the posterior extremity (Opisthoglyphs) of the maxillary bone, usually followed or preceded by a series of solid teeth, which in some cases may likewise show a more or less distinct groove. Such may also be present on the teeth of the lower jaw, as in the European Cœlopeltis, in some specimens of which a faint groove is visible on the outer side with the aid of a strong lens.

The tubular fangs of the Viperidæ are inserted on the posterior extremity of the much abbreviated and erectile maxillary bone, which bears no other teeth. The Proteroglyphs (Cobras, Coral-snakes, Sea-snakes) and the Solenoglyphs (Vipers, Pit-vipers, Rattlesnakes) may be regarded as the diverging extremes in the development of the poison apparatus, both culminating in forms with tubular fangs, the former as derived directly from the Aglyphs (harmless snakes), the latter from the Opisthoglyphs, likewise evolved out of the Aglyphs. That the insertion of the poison fangs of the Viperidæ is really on the posterior extremity of the maxillary bone is evident from the condition of the bone in its recumbent position, especially in the African Viper, Causus, which in several respects departs less markedly from the Colubrid type than our European Vipers.

The poison fangs of the Viperidæ appear to be movable, folding in the mouth when at rest, and erected, or even thrust forward, when ready to act. This, however, is simply due to the mobility of the maxillary bone, to which they are ankylosed as in all other snakes. There are normally two equally-developed fangs, close together and side by side, to each maxillary, followed by several replacement fangs loosely attached behind them, usually in two series of four. When the two fangs are in situ, they of course both function in the act of biting, although only one is in relation with the single poison duct; often, however, there is only one fang in position, either the right or the left, the place of the other being indicated by a shallow socket which will soon be filled by one of the posterior reserve fangs moving forward and becoming ankylosed to the bone. Snake-charmers who extract the poison fangs of the snakes they use for their performances have therefore to renew the operation frequently, unless they amputate the bone on which the fangs are inserted, an injury which the creature does not long survive.

The dentition of the snakes in which the maxillary bone is not movable vertically falls under three divisions: the Aglyphs, in which the teeth are all solid; the Opisthoglyphs, in which one or more (usually two) of the hindermost teeth are provided with a groove; and the Proteroglyphs, in which grooved or canaliculated teeth are situated in front, followed or not by solid teeth. Beyond these three principal divisions, the dentition furnishes important characters for the classification, although that importance has sometimes been over-estimated. The maxillary teeth may be equal in length (Isodonts), or the anterior the longer (Lycodonts), or the posterior the longer, increasing gradually in size (Coryphodonts) or abruptly, without (Syncranterians) or with a diastema, or break, in front of them (Diacranterians). These categories are, however, so completely connected as to preclude their use in taxonomy beyond helping to define genera. The number of maxillary teeth and the relative proportions and disposition of the mandibular teeth also afford useful generic characters.

The European genera may be arranged as follows, according to the dentition:

I. Teeth few, disposed in a transverse series in the upper jaw only

Typhlops.

II. Teeth in both jaws and on the palatines and pterygoids.

A. A series of solid teeth along the maxillary; no grooved teeth.

1. Anterior maxillary and mandibular teeth longest; 9 or 10 maxillary teeth

Eryx.