Chapter I: Front Matter (1)

JOURNAL
OF
THE PROCEEDINGS
OF
THE LINNEAN SOCIETY.

ZOOLOGY.

VOL. IV.

LONDON:
LONGMAN, GREEN, LONGMANS AND ROBERTS,
AND
WILLIAMS AND NORGATE.
1860.

PRINTED BY TAYLOR AND FRANCIS,
RED LION COURT, FLEET STREET.

LIST OF PAPERS.

Page

GARNER, ROBERT, Esq., F.L.S.
On the Shell-bearing Mollusca, particularly with regard to
Structure and Form 35

HANLEY, SYLVANUS, Esq., F.L.S.
On the Linnean Manuscript of the 'Museum Ulricæ' 43

HUXLEY, Prof. T. H., F.R.S., F.L.S., F.G.S., Professor
of Natural History, Government School of Mines.
On the Dermal Armour of _Jacare_ and _Caiman_,
with Notes on the Specific and Generic Characters of recent
_Crocodilia_ 1

SALTER, S. J. A., Esq., M.B., F.L.S., F.G.S.
On the Moulting of the Common Lobster (_Homarus vulgaris_)
and Shore Crab (_Carcinus mænas_) 30

SANDWITH, Hon. H., M.D., C.B., Colonial Secretary of
the Mauritius.
On the Habits of the "Aye-Aye" (_Cheiromys madagascariensis_,
L., Cuv.) 28

WALKER, FRANCIS, Esq., F.L.S.
Catalogue of the Dipterous Insects collected at Makessar, in
Celebes, by Mr. A. R. Wallace, with Descriptions of New
Species 90

WALLACE, A. R., Esq.
On the Zoological Geography of the Malay Archipelago 172

INDEX 185

JOURNAL OF THE PROCEEDINGS

OF THE

LINNEAN SOCIETY OF LONDON.

On the dermal armour of _Jacare_ and _Caiman_, with notes on the Specific and Generic Characters of recent _Crocodilia_. By T. H. HUXLEY, Esq., F.R.S., F.L.S., Prof. of Nat. History, Gov. School of Mines.

[Read Feb. 17th, 1859.]

In the course of a recent investigation into the nature of the singular extinct reptile, _Stagonolepis_, I was led to inquire somewhat minutely into the character of the exoskeleton, or dermal armour, of the existing _Crocodilia_. To my surprise, I found that very little detailed information on the subject was to be obtained from the standard repertories of Comparative Anatomy, or even from the special monographs on Crocodilian structure and classification; but I was still more astonished to discover, among whole genera of recent _Crocodilia_, an exoskeleton possessed of characters such as have been universally supposed to be peculiar to long extinct forms of the order, and whose existence in any recent species has hitherto, so far as I can ascertain, been completely overlooked.

The attempt to discover the limits within which this remarkable exoskeleton is to be found, led me to look, more critically than I had previously done, into the arrangement and specific characterization of the recent _Crocodilia_. I have thereby arrived at results which, imperfect as they are, may be of service by leading others to inquire into the exact characters of species not at present within my reach; and I therefore propose to preface my account of the peculiarities of the exoskeleton in two of the genera of recent Crocodiles with some remarks on the classification of the group, and with a few notes upon the characters of the species and the limits of the genera.

Everyone is acquainted with the great improvement effected in this branch of Herpetology by Cuvier, who divided the Crocodiles, which he regarded as constituting only a single genus, into the three subgenera _Alligatores_, _Crocodili_, and _Longirostres_. Subsequent writers have admitted these highly natural subdivisions; but there has been a constant tendency to raise their rank. The genus _Crocodilus_ has become the order _Crocodilia_; the subgenera _Alligatores_, &c., have been elevated into families; Dr. Gray has shown that the _Alligatores_ must be divided into three genera, and that there are at least two genera of _Crocodili_; and, while one of Cuvier's species of _Longirostres_ has been suppressed, the group is very generally retained with a changed name (_Gavialis_), a very important addition having been made to it in the _Crocodilus Schlegelii_ of Müller and Schlegel.

Unless the considerable materials contained in the British Museum, the Hunterian collection, the collection of Dr. Grant, and the Christchurch Museum at Oxford had been freely placed at my disposal, I should have been wholly unable to acquire the information contained in the following pages. It is only right, therefore, that I should take this opportunity of offering my thanks to my friends Dr. Gray, Prof. Quekett, Dr. Grant, and Dr. Rolleston for the many facilities they have liberally afforded me.

The recent species of the order _Crocodilia_ are divisible into three families, which correspond with the original subgenera of Cuvier, and may be termed the _Alligatoridæ_, the _Crocodilidæ_, and the _Gavialidæ_.

I. In the ALLIGATORIDÆ the teeth are strong and unequal, and the posterior ones differ greatly in shape from the anterior. The anterior pair of mandibular teeth, and the fourth pair (or the so-called canines) are received into pits in the margins of the premaxilla and maxilla; while the mandibular teeth behind these pass inside, and not between, the maxillary teeth. The mandibular symphysis does not extend back beyond the level of the fifth tooth, and often not nearly so far. The line of the premaxillo-maxillary suture on the palate is straight, or convex forwards. The wide posterior nares look downwards, and are situated forwards on the palate.

This family embraces three genera, readily distinguishable by osteological characters--_Alligator_, _Caiman_, and _Jacare_.

Genus 1. ALLIGATOR.

Dental formula, 20-20/20-20. 9th maxillary tooth the largest of its series. The snout is very broad, flattened, and rounded at the end. There is an indistinct longitudinal interorbital ridge; and there are two short ridges along the line of junction of the prefrontal and lachrymal bones. The aperture of the external nares is divided into two parts, by the prolongation forwards of the nasal bones. The supra-temporal fossæ are well-marked and open, though not large. The vomers do not appear in the palate. The feet are well webbed. The dorsal bony scutes are not articulated together; and there are no ventral scutes.

This genus contains only one species, the well-known _Alligator Mississipiensis_, or _lucius_, which is exclusively North American.

Cuvier (Oss. Foss. ed. 4. vol. ix. p. 211) gives the appearance of the vomer in the palate as a general character of the _Alligatores_; but this bone is not visible in the palate of any of those _Alligatores_ which Cuvier would have referred to his _A. lucius_ or _A. palpebrosus_, and which form the genera _Alligator_ and _Caiman_ as here defined. The vomers are in fact as slender and delicate as in the Crocodile, and extend only between the level of the tenth maxillary tooth anteriorly and the descending processes of the prefrontal posteriorly.

What may be called the median nares, or the arch formed by the postero-lateral part of the vomer and the anterior and superior lamina of the palatine bone on each side (which would constitute the posterior boundary of the posterior nares, if the palatine and pterygoid bones gave off no inferior or palatine processes), are situated nearly on a level with the twelfth tooth, or with the palato-maxillary suture.

Genus 2. CAIMAN.

Dental formula 20-20/22-22 (Natterer). The face is without median or transverse ridges, but it is sharply angulated along a line which extends from the orbit forwards along the sides of the snout. The anterior nasal aperture is undivided in the dry skull. The vomers do not appear in the palate. The supra-temporal fossæ are obliterated, the circumjacent bones uniting over them. The webs of the feet are rudimentary. The dorsal scutes are articulated together by lateral sutures and anterior and posterior facets; and there is a ventral shield, consisting of similarly articulated scutes.

Natterer[1] has described three species of _Caiman_--_C. palpebrosus_, _C. trigonatus_, and _C. gibbiceps_. The Caimans abound chiefly in tropical South America; but they are found as far north as Mexico, a specimen of _C. palpebrosus_ in Dr. Grant's collection coming from that country.

Genus 3. JACARE.

The snout is broad, and rounded at the end[2]. Each prefrontal bone is traversed close to its anterior extremity by the ends of a strong transverse ridge, which then curve round and pass forwards on the lachrymal and maxillary bones, to subside opposite the ninth tooth. The anterior nasal aperture is not divided by bone. The vomers, separated by a longitudinal suture, appear in the palate between the premaxillaries and the palatine plates of the maxillaries. The temporal fossæ, though not large, are open. The webs of the feet are small. The dorsal scutes are articulated together, as in the preceding genus; and there are similarly-articulated ventral scutes. There are 18-20 teeth on each side, above and below; and the fourth tooth in the upper jaw is the largest. The mandibular symphysis extends back nearly to the fifth tooth.

In a skull of _Jacare_ (_fissipes?_), 19 inches long, in the British Museum, I find that part of the vomer which is visible in the palate to be a rhomboidal plate, somewhat truncated anteriorly, and rather more than 1-1/2 inch long and 1 inch wide. Its anterior end comes within 3/8ths of an inch of the posterior margin of the anterior palatal foramen. Its posterior margin reaches to the level of the eighth tooth. The visible portion of each vomer is only its anterior end, which forms a thick and solid wedge-shaped plate, broader in front than behind, and articulating by a rough anterior and outer face with the premaxilla, by an obliquely ridged posterior and outer face with the maxilla, and by its internal face with its fellow. Its upper, rounded surface projects but little into the nasal passage. 2-1/4 inches behind its anterior end, the posterior and upper extremity of the vomer passes into a thin and narrow plate of bone, whose plane is at first inclined at an angle of 45° to that of the anterior part of the bone, but gradually becomes vertical; as it does so it deepens, until, 3 inches behind the anterior extremity, the vomer is a thin vertical plate of bone, 5/8ths of an inch deep, which articulates below with the palatine plate of the maxilla, and, about 1 inch behind this, with the palatine plate of the palatine bone. The upper edge of this plate nowhere extends to one-third of the height of the nasal chamber. It gives off a horizontal process outwards, which, gradually increasing in width, inclines downwards until it comes into contact, first, with the inner surface of the maxilla, and, 3/4ths of an inch behind this, with the nasal plate of the palatine bone. In front of its junction with the maxilla, the horizontal plate of the vomer presents a long free edge, concave externally; and this bounds the median nares internally and posteriorly. Throughout its junction with the maxilla, the horizontal plate is parallel-sided; but after it joins the palatine bone, it gradually narrows posteriorly, in consequence of the gradual increase in width of the palatine, and ends almost in a point, 6-1/4 inches behind its anterior end. The posterior edge of the vertical plate is extremely thin, and 7/8ths of an inch deep. It articulates with the anterior end of the vertical plate of the pterygoid, while the straight inferior edge articulates throughout with the palatine plate of the palatine bone. The vomers terminate midway between the median nares and the descending process of the prefrontal. The median nares are bounded entirely by the vomer and the maxilla. They correspond with the nasal face of the palato-maxillary suture, but are rather behind its palatine face, and they are about on a level with the interval between the tenth and eleventh teeth. If the anterior edge of the palatine bone bounded them, they would be a little behind the twelfth tooth. The posterior nares, 2-1/8 inches wide, by 7/8ths of an inch long, look altogether downwards, are completely divided by a bony septum, and have the form of a rhomboid with its narrowest side posterior. They are surrounded by a strong raised ridge, incomplete only at the anterior and outer angles of the rhomboid.

Five species of _Jacare_ are enumerated by Natterer--_J. fissipes_, _J. sclerops_, _J. nigra_, _J. punctulata_, and _J. vallifrons_. They have met with only in South America.

[1] "Beitrag zur näheren Kenntniss der Sudamerikanischen
Alligatoren," 'Annalen des Wiener Mus.,' Band i.

[2] According to Natterer, the dental formula of _J. nigra_ and
_J. fissipes_ is 18-18/18-18, of _J. sclerops_ 19-19/20-20, of _J.
callifrons_ and _J. punctulata_ 20-20/18-18.

* * * * *

II. In the family of the CROCODILIDÆ the teeth are usually strong and very unequal in size, and there is always a considerable difference between the anterior and the posterior teeth. The two anterior mandibular teeth are received into pits in the premaxilla; but the canines pass into grooves (which may be converted into fossæ) situated at the junction of the premaxilla and maxilla. The other mandibular teeth are received between the maxillary teeth. The symphysis of the lower jaw does not extend beyond the level of the seventh or the eighth mandibular tooth. The premaxillo-maxillary suture may be either straight or strongly convex backwards. The divided vomers do not appear in the palate. The posterior nares look more or less backwards, and are transversely elongated. The supra-temporal fossæ are always open, and the feet are distinctly webbed. The dorsal scutes are not articulated; and there are no ventral scutes.

Two genera, _Crocodilus_ and _Mecistops_, are distinguishable in this family.

Genus 4. CROCODILUS.

The teeth are always strong and very unequal, the strongest in the upper jaw being the tenth. The mandibular symphysis does not extend beyond the level of the sixth tooth. There are usually six cervical scutes, in two rows, or forming a rhomb, and separated by a distinct interval from the tergal scutes. There are 18 or 19 teeth above, and 15 below, on each side.

1. _Crocodilus vulgaris._

As Cuvier has remarked, it is extremely difficult to find good distinctive characters for all the species of this genus. My first difficulty was to ascertain the precise characters of that species which has been misnamed _vulgaris_, inasmuch as I could find neither in the British Museum, nor in the Museum of the Royal College of Surgeons, any _authentic_ skeleton or skull of this, the so-called Nilotic Crocodile. This difficulty subsisted up to the time that the chief statements contained in the present essay were laid before the Linnean Society; but since then I have been enabled, by Dr. Gray's permission, to examine the skull of a small stuffed specimen, brought to this country from Egypt by Sir Gardner Wilkinson, and to study the splendid entire skeleton of a _Crocodilus vulgaris_ in the Christchurch Museum at Oxford, presented to that Institution by the gentlemen who shot it on the Nile, and set up with great care under the auspices of my friend Dr. Rolleston, Lee's Reader in Anatomy and Curator of the Museum. Fortunately the entire skin has been preserved; so that this is the most complete record of the hard parts of any individual crocodile with which I am acquainted, besides being, so far as I am aware, the only authentic entire skeleton of _Crocodilus vulgaris_ in this country. I subjoin the chief points of interest which I noted in my brief examination of this valuable specimen:--

Inches.

The total length of the skeleton is 114
" " " skull 16
Between the outer edges of the posterior ends of
the quadrate bones 8-3/4
From the snout to the middle of the canine notch 2-3/4
Transverse diameter of snout opposite 10th tooth 4-7/8
Long axis of orbit 2-1/4
Short axis of orbit 1-5/8
Interorbital space opposite the middle of the orbit 1-3/4
Anterior edge of the orbit from end of snout 10-1/2
Syncipital[3] area in length, about 2-1/2
" " in breadth anteriorly 3-3/4
" " " posteriorly 4
Supra-temporal fossæ, wide 7/8
" " long 1-1/8
Least width of parietal 7/16
Total length of mandible 20-1/2
Its greatest depth 3
Length of cervical region (or anterior 8 vertebræ) 10-1/2
" dorso-lumbar region 27
" sacral " 3-3/4
Length of humerus 7-1/2
" ulna 5-1/4
" fore foot, extreme length 6
" femur 8-1/2
" tibia 6
" hind foot, extreme length 9-1/4

From the above measurements it will be seen that the skull is somewhat slender. Behind the canine groove it widens to the tenth tooth, which is 5-3/4 inches behind the end of the snout. It retains about the same diameter to the twelfth tooth, and then slowly widens again,--a sudden increase in size, to the extent of half-an-inch, taking place opposite the posterior margin of the orbit, owing to the flanging-out of the jugal. On the whole, however, there is a slow and even increase in breadth, from the canine groove to the ends of the ossa quadrata. The nasal aperture is pyriform, its wider end being forwards, and its narrow posterior extremity, into which the pointed ends of the nasal bones project, attaining the level of the first tooth behind the canine groove.

[3] By this term I denote that squarish flat area bounded by
the postfrontal and squamosal bones laterally, by the occiput
posteriorly, and by a line joining the outer angles of the
postfrontals anteriorly.

On the left side there is only a pit for the reception of the anterior mandibular tooth, while on the right side this pit is converted into a complete foramen. On the upper face of the skull, the premaxillo-maxillary suture runs vertically upwards through the canine groove, and then passes obliquely backwards to a point 5 inches behind the end of the snout. The anterior part of this suture lies in a strong ridge, which is continued downwards and forwards on the premaxilla to the level of the fifth tooth, a groove separating it from the margin of the nasal aperture. Posteriorly this ridge dies away, but a curved irregular elevation, convex inwards, arises opposite the tenth tooth. It is wholly confined to the maxilla, not extending on to the nasals.

There is a distinct, rough, irregular elevation, bounded on its outer side by a sharp groove, which extends back to the orbit, on the lachrymal bone. The profile of the skull is convex as far as the posterior boundary of the nostril, and very slightly concave from that point as far as the twelfth tooth. It then passes back as a straight, slightly ascending line, only interrupted by the lachrymal ridge, to the margin of the occiput. The inferior margin of the maxilla is convex downwards as far as the canine groove, whose lower end is indicated by a deep sinuation. It then becomes convex again, the crown of the curve being at the ninth and tenth teeth, and its posterior end sweeping into a concavity whose summit is at the twelfth tooth. Behind this the edge of the maxilla is only slightly convex. The inferior contour of the jugal bone is very concave; but the articular end of the quadrate bone descends to the level of the edge of the ninth alveolus.

The orbits have a sort of heart-shape, their apices being turned forwards, and their more convex sides inwards.

The supra-temporal fossæ are half-moon-shaped, their straight sides being external and so inclined that, if prolonged, they would decussate upon a line joining the anterior margins of the orbits.

On the palatine surface of the skull, the premaxillo-maxillary suture runs backwards from the canine groove, as far as the level of the middle of the second alveolus behind the groove (or that of the seventh tooth), which point it reaches at about the junction of the middle with the inner third of the palatine plate of the maxilla. The suture then turns abruptly forwards until it reaches the level of the anterior margin of the alveolus of the sixth tooth, when it bends suddenly inwards to meet its fellow. The whole suture, therefore, has the form of a W. The vomers are completely hidden.

The posterior nares look downwards and backwards; their aperture is, from the incompleteness of the septum, single, and has a transversely elongated crescentic form. It measures 1-1/8 inch in width by 3/8ths antero-posteriorly. The basi-sphenoid is seen for about 1/8th of an inch on the base of the skull behind it, bounding the sides of the eustachian tube. The dental formula is 18-18/15-15. The fourth and tenth teeth are largest in the upper jaw, the first and fourth in the lower. The eight posterior teeth on each side in the upper jaw, and the five posterior in the lower, have a marked constriction between the short crown and the fang of the tooth. There are deep interdental pits for the reception of the mandibular teeth between the third and fourth, and fourth and fifth teeth above, and between the succeeding teeth from the sixth to the thirteenth.

The hyoidean cornua are very strong curved bones, the chord of whose arc measures 3-1/2 inches. They are concave inwards, convex outwards, concave posteriorly, convex anteriorly; they are flattened from side to side below, but they end above in subcylindrical styloid extremities.

In the ninth vertebra the neurocentral suture passes just above the base of the parapophysis; it traverses the parapophysis in the tenth and eleventh vertebræ, while in the twelfth the parapophysis suddenly rises to the root of the diapophysis, and the suture lies far below it. The centra of the dorsal vertebræ, as far as the thirteenth inclusive, have hypapophyses. The diapophyses of the ninth vertebra pass almost horizontally outwards, but are a good deal inclined backwards. In the succeeding vertebræ up to the fourteenth or fifteenth, the diapophyses are, in addition, inclined upwards, the upward inclination being most marked in the tenth, eleventh and twelfth vertebræ. From the fifteenth vertebra onwards, the transverse processes pass almost directly outwards, without either upward or backward inclination. The span of the transverse processes is greatest in the eighteenth and nineteenth vertebræ, in which the distance between the extremities of these processes is 7-1/4 inches, a length about equal to that of the longest vertebral rib.

The rib of the ninth vertebra is terminated by a single long and slender semicartilaginous process which does not unite with the sternum. Each of the vertebral ribs from the tenth to the seventeenth vertebræ inclusively, on the other hand, is united with the sternum, or its continuation, by two such semicartilaginous costal elements, which may be respectively termed sternal and lateral. The sternal elements of the ribs of the tenth and eleventh vertebræ are united with the sternum proper; those of the next five vertebræ are connected with its median backward prolongation, while those of the seventeenth vertebra are attached to the processes into which this prolongation divides posteriorly.

The sternal costal elements are very broad and flat, and though the lateral ones are less so, they are wide and expanded. The lateral costal pieces of the eleventh to the sixteenth vertebræ inclusively, give attachment to very large and flat, triangular, _processus uncinati_. Those of the twelfth are 3-3/4 inches long and 1-3/8 inch wide at their widest part. The transverse processes of the twentieth vertebra bear rudimentary ribs. The centrum of the thirteenth vertebra is 1-3/4 inch long, and the vertebra is 3-3/4 inches high from the lower edge of the centrum to the summit of the neural spine. The centra of the vertebræ retain nearly the same length to the twentieth caudal; but behind this vertebra they are shorter, as are the anterior dorsal vertebræ. The first caudal vertebra is provided with two styliform bones, which represent the chevron bones of the other caudal vertebræ, but are not united below.

The dorsal scutes have the arrangement which his often been described. They are separated (except perhaps the median rows) by integumentary spaces, neither overlapping nor uniting by sutures; and there are no ventral scutes.

Among the osteological characters which have been detailed, the peculiarities of the tergal armour, the proportions of the skull, combined with the characters of the ridges upon its surface, and the form of the premaxillo-maxillary suture amply suffice to diagnose this species. Even in the small skull, only 5-1/2 inches long, lent to me by Dr. Gray, the characteristic features of the species are well exhibited, although age appears to give rise to many differences. Thus the posterior margin of the external nostrils does not extend so far back as in the adult, and the facial is smaller in proportion to the syncipital region, whose anterior and posterior transverse dimensions are very nearly equal. The orbits are proportionally larger, the interorbital space more excavated; and the outer straight margins of the supra-temporal fossæ are parallel with the longitudinal axis of the skull. Still more important differences are visible on the palatine face of the skull. The premaxillo-maxillary suture reaches back, indeed, to the line of the seventh tooth; but it forms an even curve whose summit is in the middle line. The aperture of the posterior nares, again, has a totally different form from that which it assumes in the adult. It is somewhat heart-shaped, with its apex forwards, measures 1/4 inch long by 3/16ths at broadest, and looks altogether downwards, while its anterior margin is situated far more forward in the palate than that of the adult.

2. _Crocodilus biporcatus._

This, the best-known Crocodile, is a very well-marked species, characterized (beside the peculiarities of its dermal armour) by a comparatively slender skull, similar in shape to that of _C. vulgaris_, and, like it, without any sudden enlargement immediately behind the canine groove; and by the strong ridge which arises on each lachrymal bone close to the anterior edge of the orbit, and is continued forwards on to the line of junction of the nasal and maxillary bones, so that the naso-maxillary suture traverses the axis of the ridge, and then CURVES outwards, descending towards the alveolus of the tenth tooth. The premaxillo-maxillary suture is W-shaped; and its salient angles reach backwards even to the level of the posterior margin of the seventh alveolus.

3. _Crocodilus Americanus_ (_acutus_, Cuv.)

has the slenderness of snout (even more marked) and the form of the premaxillo-maxillary suture of the preceding species; but it is at once distinguished from this and all other Crocodiles (except _C. rhombifer_) by the marked longitudinal and transverse convexity of the middle of the face, which gives the profile a totally different aspect from that of the other species, which are flat or concave in this region.

4. _Crocodilus Journei_

is another unmistakeably distinct and very remarkable species. The descriptions and figures given by Graves, Bory de St. Vincent, and Duméril and Bibron, of the unique specimen of this Crocodile to the Bordeaux Museum, would alone have compelled me to differ entirely from the view taken by Dr. Gray of the affinities of this species. These observers agree in stating that _Crocodilus Journei_ has six cervical scutes, arranged as in the other Crocodiles, and, as Graves says, "separated by an interval of four inches" from the commencement of the tergal scutes, whence it is obviously impossible that it can be a _Mecistops_. But, in addition to this, I had the good fortune to find, among the recent additions to that excellent osteological collection which Dr. Gray has gradually formed at the British Museum, the skull of a Crocodile obtained from a dealer in Paris, and labelled by him "Croc. de l'Orinoke." I at first imagined this Crocodile to be a _Mecistops_; but on careful investigation it turned out to be no other than the skull of a _Crocodilus Journei_, somewhat larger than the Bordeaux specimen, but, as the subjoined measurements will prove, agreeing with it in all its proportions:--

Inches.

Length from end of snout to end of ossa quadrata 22-1/2
Breadth between outer margins of ossa quadrata 9-3/4
---- at the level of the anterior margins of the orbits 5-1/2
---- at the tenth tooth 3-1/2
---- at the end of the snout 2-3/4
---- of the interorbital space 1-3/4
Length of mandibular symphysis 5

Now Duméril and Bibron expressly state that the length of the head of _C. Journei_ equals 2-1/2 times its greatest transverse diameter, that the width of the jaws at the anterior margins of the orbit equals one-fourth the length of the head, and that at the tenth tooth it equals one-sixth the length of the head; and these are as nearly as possible, it will be observed, the relations of the same dimensions in the above list.

In the specimen in the British Museum there are eighteen teeth on each side above, and fifteen below. The Bordeaux specimen is stated to have the same dental formula, except that there are sixteen teeth in the left ramus of the mandible. The fourth and tenth maxillary teeth are stated by Graves to be as large again as the others; and the corresponding alveoli have these proportions to one another in the British Museum specimen. In fact, there can be no doubt that this skull is that of a true _Crocodilus Journei_.

But its general characters at once prove the close affinity of _C. Journei_ with the other true Crocodiles, from which it differs only in its elongated and gradually tapering skull, and in the more backward extension of the mandibular symphysis[4], which attains the level of the posterior margin of the sixth tooth.

In this character, and in the extreme slenderness of the snout, there is doubtless an approximation to _Mecistops_; but _Crocodilus Journei_ is sharply separated from that genus by the characters of its teeth, and by those of its dermal armour.

[4] The greater proportional length of the symphysis is noted by
Duméril and Bibron.

5. _Crocodilus bombifrons_ (_palustris?_).

All the species of _Crocodilus_ which I have hitherto mentioned have, in common, the backward curvature of the premaxillo-maxillary suture to the level of the seventh tooth. But there is a species of Crocodile, about whose proper specific name I am by no means clear, in which this suture passes straight across the palate, or may even be a little convex forwards.

And not only do the skulls of this species exhibit this approximation to those of the _Alligatoridæ_, but they resemble them still further in their rounded snouts, their great width immediately behind the canine groove, and in the fact that, in young specimens, one or the other canine may be received into a pit instead of into a groove[5].

In the Hunterian Collection there are seven skulls, varying in length from 5-1/4 inches up to 16 inches, in none of which does the crown of the premaxillo-maxillary suture extend beyond a line joining the sixth pair of teeth. In all there are two short ridges (convergent in young specimens, nearly parallel in old ones) upon the lachrymal bones, which end before reaching the anterior limits of those bones. They all have an oblique ridge on the upper jaw above the tenth tooth; and the snout attains the width which it has opposite this tooth immediately behind the canine groove. In the British Museum there are five middle-sized skulls with the same characters; but two of these have a pit on one side of the upper jaw, and a groove on the other, and one has something between a pit and a groove on each side.

Dr. Gray, has in his 'Catalogue[6],' mentioned the peculiar transverse disposition of the premaxillo-maxillary suture in his _Crocodilus bombifrons_; and on examining the two crania thus named in the British Museum collection, one of which is 20 and the other 21 inches long, I can discover no distinguishing character between them and those already described. There can be no doubt then, I think, that these constant and well-marked characters, exhibited by fourteen skulls which vary in length from 5-1/4 to 21 inches, prove the existence of a distinct species of Crocodile, which I would provisionally term _bombifrons_.

[5] In a skull of this species 14-1/2 inches long, in the British
Museum, the vomers are completely excluded from the palate, and
their anterior ends do not extend for an eighth of an inch beyond
the palatine part of the palato-maxillary suture, which lies on a
level with the anterior margin of the twelfth alveolus. Each vomer
is 2-3/8 inches long, and presents the same general form as that
of _Jacare_; only the anterior division is but a very small, flat
and thin plate, not a quarter of an inch long. The boundary of the
median nares is formed in equal proportions by the vomer and the
palatine, and is opposite the fourteenth tooth. The hinder end of
the vomer articulates with the end of the descending process of
the prefrontal.

[6] 'Catalogue of the Tortoises, Crocodiles, and Amphisbænians in
the Collection of the British Museum,' 1844, p. 59.

I believe that this species has been constantly confounded with _biporcatus_, from which it may be at once distinguished by the direction of the premaxillo-maxillary suture, and by the shape of the snout behind the canine groove. I have found these distinctions to hold good at all ages; but the last-mentioned difference is far more marked in middle-aged than in either young or old specimens.

All the skulls named _Crocodilus palustris_ which I have seen are referable either to _C. biporcatus_ or to _C. bombifrons_. With respect to the _C. palustris_ of Lesson and Duméril and Bibron, the latter authors consider it to be only a variety of _C. vulgaris_. Their description would, however, apply very well to _C. bombifrons_, as I have defined it above; and they expressly state ('Erp. Générale,' t. iii. p. 113) that all their specimens (twelve in number and varying in length from 30 centimetres to more than 3 metres) came from the East Indies or the Seychelle Islands. Now, Duméril and Bibron enumerate only three Asiatic Crocodiles--_C. biporcatus_, _C. palustris_, and _C. galeatus_, the last of which was only known to them by description; so that all the numerous Asiatic crocodiles which passed through their hands belonged either to _C. biporcatus_ or _C. palustris_. On the other hand, all the skulls of crocodiles from Asia which I have met with (amounting to at least twenty) are either those of _C. biporcatus_ or of the species which I have called _bombifrons_; so that I suspect the latter title will turn out to be a synonym of _palustris_.

6. _Crocodilus rhombifer._

I have not been able to obtain any skull of this species, which, according to Cuvier's account and figures ('Oss. Fossiles,' t. ix. p. 102), resembles _C. Americanus_ in the great convexity of its nasal region, but differs from it in the greater breadth of the skull, and in the strong converging preorbital ridges, which appear to be limited to the lachrymal bones. If the figures are to be I trusted, however, there are other very important distinctive characters about the cranium of this species; for Cuvier's, fig. 2, pl. 331, which gives a view of the palate, shows the premaxillo-maxillary suture forming a nearly straight transverse line.

* * * * *

There remain several species of _Crocodilus_ whose skulls I have not been able to examine, and of which no sufficient descriptions exist. Of these, (7.) _C. galeatus_ and (8.) _C. Gravesii_ (_planirostris_) would appear to be very distinct forms. (9.) _C. marginatus_ is considered by Duméril and Bibron to be only a variety of _C. vulgaris_; and they take the same view of (10.) _Crocodilus suchus_. Professor Owen, however, has figured the cranium of an Egyptian mummy under this name ('Monograph on the Reptilia of the London Clay,' Pal. Soc., 1850). In the under-view of this skull (tab. i. fig. 2), the junction of the premaxilla and the maxilla in the palate seems to be broken away; but on the left side, the palatine process of the maxilla is entire, as far as the level of the anterior margin of the sixth tooth, and there is not a trace of a suture behind this point. Are there, then, two or more species of Crocodile in Egypt, as Geoffroy St.-Hilaire supposed?

With regard to the distribution of the species of _Crocodilus_, _C. vulgaris_, _C. marginatus_, and _C. suchus_(?) appear to be exclusively African; all the crocodiles from other parts of the Eastern hemisphere, which I have met with, belong, as I have stated above, either to _C. biporcatus_ or _C. bombifrons_, both of which species are found in the Ganges. _Crocodilus galeatus_ appears to be peculiar to Siam. _Crocodilus Americanus_ and _C. rhombifer_ are undoubtedly American. _C. Journei_ has been supposed to be African; but such positive evidence as exists tends rather to prove it to be an American species. Thus Bory de St. Vincent states that the Bordeaux specimen is "suspected to have come from America;" and, as I have said, the skull in the British Museum is labelled "from the Orinoko."

_Crocodilus Gravesii_ (_planirostris_) is supposed by Bory de St. Vincent to have been brought from the Congo; but its real origin is not known.

Genus 5. MECISTOPS.

The cranium is elongated, and the snout slender and Gavial-like. There are eighteen slender and subequal teeth above, and fifteen below, on each side. The mandibular symphysis extends back to the level of the seventh tooth. The cervical scutes are arranged in two transverse rows, each of which contains two scutes; and there is no space left between the posterior row and the tergal series.

This excellent genus, as established by Dr. Gray, includes Cuvier's _Crocodilus cataphractus_ (which Dr. Gray considers to be the young of a species whose full-grown form was discovered by Mr. Bennett in West Africa), _Crocodilus Journei_ and _Crocodilus Schlegelii_. As I have endeavoured to show, however, _C. Journei_ is a true crocodile; and, as I shall point out below, Müller and Schlegel have satisfactorily proved _C. Schlegelii_ to be a Gavial. Consequently _Mecistops_ is at present represented by only one species, which must be called _M. cataphractus_ if _M. Bennettii_ of Gray is really the adult of the form which Cuvier described.

* * * * *

III. In the family of the GAVIALIDÆ, the snout is always very long and slender; the teeth are for the most part slender, sharp-edged, and subequal. The two anterior mandibular teeth pass into grooves, one of which lies on each side of a beak-like prominence of the premaxillæ, which carries the two anterior upper teeth. The canines are received into grooves. The mandibular symphysis extends back to at least the fourteenth tooth, and is partly formed by the junction of the splenial bones. The premaxillo-maxillary suture is always strongly convex backwards. The posterior nares are situated more forward than in the _Crocodili_. The temporal fossæ are large. The feet are strongly webbed. The dorsal scutes are not articulated; and there are no ventral scutes.

I distinguish two genera in this family, _Rhynchosuchus_ and _Gavialis_.

Genus 6. RHYNCHOSUCHUS.

There are twenty teeth above, and eighteen or nineteen below, on each side; the mandibular symphysis extends to the fifteenth tooth. The posterior teeth of the upper jaw, and almost all those of the lower jaw, are received into interdental pits; the orbital margins are not raised; and the premaxillæ are hardly at all expanded. The premaxillo-maxillary suture does not reach the third tooth behind the notch.

I propose the name _Rhynchosuchus_ to indicate that generic type which is at present represented by the solitary species called by Müller and Schlegel _Crocodilus_ (_Gavialis_) _Schlegelii_, and admirably described and figured by them in their essay, 'Over de Krokodilen van der Indischen Archipel,' in the 'Verhandelingen over de natuurlijke Gesch. der Nederl. overzee. Bezittingen,' 1839-1844. Under the title _Crocodilus_ (_Gavialis_) _Schlegelii_ (p. 18), they say--"The Gavial from Borneo, when compared with the Indian one, is principally distinguished by the following characters:--

"1. By its stronger form and better developed limbs.

2. By its much less slender head and snout, which last does not narrow so suddenly in front of the eyes as in _G. Gangeticus_.

3. By the smaller number of teeth, of which there are twenty above and eighteen below on each side, while _G. Gangeticus_ has 28/26 or 27/25; furthermore, the teeth are stouter, less curved, and less sharp, and are disposed more perpendicularly, and the ninth tooth of the upper jaw (reckoning from the front) is considerably larger and stronger than the others; whence it follows that, just as in the true Crocodiles, the snout at the level of this tooth exhibits a lateral projection.

4. By the shorter symphysis of the under jaw.

5. By the absence of the swollen nasal prominence (neusklep), which characterizes the Gangetic Gavial.

6. By the less expanded form of the tabular upper surface of the hinder part of the skull.

7. By the very slight production of the edges of the orbit.

8. By the large eyes.

9. By the presence of a number of small nuchal shields, while _G. Gangeticus_ has but one pair.

10. By the strongly developed keels of the dorsal scutes.

11. By the much larger scales on the under parts and on the legs of the animal.

12. By the different colours with which it is variegated."

* * * * *

These authors further point out that the vomers appear for a small space in the posterior part of the palate, that the opercular or splenial bones join in the symphysis of the lower jaw, and that the cervical and dorsal scutes form one continuous shield; and they represent the two anterior mandibular teeth passing in grooves on either side of the end of the premaxilla. In fact, they fully and completely establish the fact that their new species belongs to the _Longirostres_ of Cuvier, or to the Gavials of later writers.

Under these circumstances, it is somewhat surprising to find the deliberate conclusions of these careful investigators set aside in the following brief passage:--

"This Bornean species (_C. Schlegelii_) was, in fact, originally
described as a new species of Gavial; but the nasal bones, as in the
fossil from Sheppey, figured in t. ii. 15, extend to the hinder
border of the external nostril."--_Owen_, _Fossil Reptilia of the
London Clay, Crocodilia_, p. 15: 1850.

Müller and Schlegel give remarkably clear and beautiful figures of the skull of their Gavial; and these show at once that the nasal bones do not "reach the hinder border of the external nostril," but meet the premaxillaries at a point very distant from that border, viz. opposite the ninth tooth. Even did the nasal bones reach the posterior boundary of the nostril, such a character would not outweigh those derived from the relations and number of the teeth, the structure and extent of the mandibular symphysis, and the disposition of the dermal scutes,--all of which are so clearly and definitely set forth by Müller and Schlegel, that it seems difficult to understand how any one who had consulted the original memoir could have overlooked them.

It was possible, however, that Müller and Schlegel, notwithstanding their great opportunities, might have erred in their statements; and I therefore gladly seized the opportunity of testing their description by comparing it with an authentic skull of the species in question, from New Guinea, in the collection of the British Museum.

I have found the statement of Müller and Schlegel minutely accurate in almost all points; and there cannot be the slightest doubt, not only that the Schlegelian crocodile is one of the _Gavialidæ_, but that it forms a distinct generic type in that family, as different from _Gavialis_ as _Caiman_ is from _Jacare_, or _Mecistops_ from _Crocodilus_.

* * * * *

The following are the most important measurements of the skull of _Rhynchosuchus Schlegelii_ in the British Museum collection:--

Inches.

Length from the end of the premaxilla to that of
os quadratum 23
Breadth from outer edge of one os quadratum to that
of the other 8-3/4
Breadth across the face in front of the orbits 4
" at the 9th tooth 2
" at the 5th tooth 1-1/2
" at the 3rd tooth 1-3/4
" of the beak-like curved process which carries
the two anterior teeth 1
Mean width of lower jaw from symphysis to extremity 1-5/8
Length 12
No tooth measures transversely more than 3/16

The face is very smooth; but a slight longitudinal groove runs down on each side from the anterior margin of the orbit for about two inches. Anteriorly to this point the snout retains a nearly even diameter as far as the ninth tooth, in front of which it suddenly narrows a little, retaining nearly the same dimensions to the fourth tooth, where it widens a very little, and then suddenly narrows to the terminal beak. The lower jaw does not expand at all at its extremity. The nasals join the premaxillaries opposite the ninth tooth, and the splenial bones, in the lower jaw, end opposite the tenth mandibular tooth, as the figures of Müller and Schlegel show. The vomers appear between the inner edges of the palatines posteriorly, as a thin bony band 1-3/8 inch long by 1/8 inch wide, which tapers at each end and is divided by a longitudinal suture. The ninth tooth of the upper jaw is stronger than the rest.

The only point in which the description of Müller and Schlegel seems to me to be incomplete[7] is with regard to the disposition of the teeth. They say--"The teeth of _C. Schlegelii_, as regards their form and development, more nearly resemble those of the true Crocodiles; but in the way in which the teeth of the two jaws are opposed, there is the most complete resemblance between our species and the Gangetic Gavial,--both which species differ from all other crocodiles in the circumstance that when the mouth is shut, all the teeth of the under jaw project outside the lateral margin of the upper jaw" (_l. c._ p. 22).

What I find is this:--The anterior teeth of both the upper jaw and the mandible are long, slender, sharp-edged, and slightly curved. The posterior eleven, on each side, in the upper jaw, are short, straight, conical, and constricted below their crowns. There are deep interdental pits between the ten posterior mandibular teeth, into which the opposed teeth of the maxilla are received when the jaws are closed. All the mandibular teeth, except the two anterior and the fourth pair, pass into like pits in the upper jaw. The anterior eight teeth on each side of the upper jaw pass straight down outside the lower jaw. In the Gangetic Gavial the relations of the teeth of the two jaws appear to me, as I shall state below, to be very different.

_Rhynchosuchus Schlegelii_ inhabits the inland lakes of Borneo, and is found in New Guinea.

[7] Or it is possible that the _Rhynchosuchus_ from New Guinea,
which I have examined, is specifically distinct from the Bornean
form.

Genus 7. GAVIALIS.

There are twenty-seven or twenty-eight teeth in the upper, and twenty-five or twenty-six in the lower jaw. The mandibular symphysis extends to the twenty-third or twenty-fourth tooth. The lateral teeth of both jaws are, all but the very hindmost, directed obliquely downwards (or upwards), forwards or outwards, and are not received into interdental pits. The anterior margins of the orbits are raised. The premaxillæ and the end of the mandible are greatly expanded. The premaxillo-maxillary suture reaches the level of the fourth tooth behind the canine notch.

The only true _Gavialis_ is the well-known _G. Gangeticus_ from the East Indies. In this 'Gavial,' or 'Garrhial,' the vomers are slender bones which do not extend further forwards than the level of the twenty-second or twenty-first tooth, and have but a very short and slender representative of the anterior flattened division of the bone in _Jacare_; posteriorly they extend back to the level of the descending processes of the prefrontals. In a skull 25 inches long the vomers have a length of about 4 inches, extending as they do a little further forward than the palato-maxillary suture. The median nares are opposite the twenty-fifth tooth.

* * * * *

All the _Crocodilia_ which I have enumerated are provided with two perfectly distinct kinds of dermal armour,--the one consisting of plates of horn, produced by a modification of the superficial layer of the epidermis; the other composed of discs of bone marked by a peculiar pitted sculpture on their outer surfaces, and developed within the substance of the dermis. To the former I shall apply the term "scales;" the latter are what I have denominated "scutes."

All recent _Crocodilia_ have both scales and scutes in the dorsal region of the body, the scutes underlying, and having the same general form as the scales. In all, the ventral region of the body is also covered with scales which have a very definite shape; but in no recent Crocodilian which I have examined, save those species which are included in the genera _Caiman_ and _Jacare_, are there any scutes in the ventral region.

Again, in the genera _Alligator_, _Crocodilus_, _Mecistops_, Rhynchosuchus, and _Gavialis_, the edges of the scutes, except those of the two median longitudinal rows, are hardly ever united by sutures, nor do the posterior margins of those in each transverse row overlap the anterior margins of the succeeding row. At any rate, there is no flat, bevelled, articular facet on the outer surface of the anterior margin of a scute, for articulation with the inner surface of the posterior margin of its predecessor. In the genera _Caiman_ and _Jacare_, however, the lateral edges of all the scutes of the dorsal and ventral shields are united by serrated sutures; and the anterior end of the outer face of each is provided with a well-marked smooth facet, which is overlapped by the smooth under-surface of the scute in front of it.

I first noticed the remarkable structure of the dermal armour of these _Alligatoridæ_ in the skin of a _Jacare_ (_sp. incerta_), wanting the end of the tail, but which must have belonged to an animal between five and six feet in length. It had long been in my possession; but I had never before had occasion to study its characters minutely.

The horny scales, which had the appearance of thin tortoise-shell, could be readily peeled off (especially by the aid of a little caustic potash); and then the white surface of the subjacent bony scute upon which they were modelled came into view. It is to be understood, however, that the inner surface of the scale corresponded only in its general form with the outer surface of the scute; for it did not dip into the pits with which the latter is sculptured. These are in fact filled by the dry dermis which extends over and encloses the scute, a very thin layer (bearing the rete mucosum) being interposed between it and the scale; so that the pitted sculpture does not come out well until the scutes have been boiled.

The _dorsal_ scutes are both carinated and angulated. By the application of the former term, I mean to indicate that, along a median or submedian longitudinal line, their substance is more or less elevated, so as, in many cases, to form a very prominent crest. This crest always subsides before it reaches the anterior margin of the scute, though it may extend beyond the posterior margin. Its highest point is always behind the centre of the scute, and is devoid of sculpture. The sculpture however seems to radiate from this point, inasmuch as it consists, on the greater part of the scute, of distinct pits, which are usually round towards the centre, but towards the periphery become ovals with their long axes directed towards the point in question.

The smooth inner surfaces of the scute shelve towards a depression which corresponds with the external ridge, under which the sides of the scute seem to meet in an angle. This may be called the 'angulation' of the scute. From before backwards, the inner surface of the scute is a little convex. The scute is thickest in the middle; posteriorly, it thins off to an edge and overlaps its successor; anteriorly, its outer surface is bevelled off at an acute angle with the inner, so as to give rise to a smooth shelving surface--wide from side to side, narrow from before backwards--forming the 'articular facet,' which is overlapped by the inner surface of the posterior edge of the preceding scute. I have termed this the 'articular facet;' but it must not be supposed that there is anything like a true joint between the opposed facets of the overlapping and overlapped scutes; on the contrary, they are at once separated and connected by the dermal connective tissue.

The posterior margin of the articular facet is separated by a deep transverse groove, divided by little partitions into as many pits, from the rest of the sculptured surface; but there is no trace of any suture dividing the scute into two portions. The lateral margins of each scute are united by serrated sutural edges with those which lie next to them in the same transverse row; so that each row forms a nearly solid flat bony bar, composed, in the middle of the back, of as many as ten distinct scutes. The outer edges of the outermost scutes only, thin off and exhibit no sutural serration, inasmuch as they are not directly connected with ANY other scutes.

The median line of the back corresponds in general with the suture between the two middle scutes of each transverse row; so that the scutes are disposed symmetrically on either side of that line. Furthermore, the anterior part of the inner surface of each of the two middle scutes is connected by ligament with the extremity of the spinous process of a vertebra; at least, this is the case in the dorsal, lumbar, sacral, and anterior caudal regions.