THE GIGANTIC LAND-TORTOISES OF MAURITIUS. 49 one (No. 39953), both of the right side, have the shaft rather smooth, subtetrahedral, with a large tuberosity on the posterior surface in the middle of the length. The semilunar facet for the articulation with the humerus is comparatively narrow (40: 21). 6. A third radius (No. 39952) agrees with the first two in the shape and smoothness of the shaft, which, however, is a little more slender. The tuberosity occupies also the same place; but the semilunar articular facet is conspicuously broader and shorter (37:23).
ce. The fourth radius (No. 39951) is of the left side of a large individual, and distinguished by rough crests and protuberances, especially in its lower fourth. Its shaft is more compressed, and the tuberosity on its hinder surface placed above the middle and connected by a strong ridge with the lateral facet for the articulation with the ulna. The semilunar articular facet has nearly the same form as in the first type (39: 21).
Measurements.
No. 39950. No. 39953. No. 39952. No. 39951.
millim. millim, millim, millim.
Mote WO PACIUSS 1.2 sg. ata sce aaincssetvaie Seas Shoah 127 97 125 182 Circumference in its narrowest part............... 53 32 47 52 Longitudinal diameter of semilunar facet ...... 40 24, 37 39 Transverse diameter of the same .................. 21 13 23 21 Pelvis.—The two collections contain nine perfect specimens, and fragments of as many other individuals, referable to three distinct types :— a. Seven out of the nine perfect. specimens belong to a type (Plate XXVI. fig. A and Plate XXVIII. fig. C) which is distinguished by the short vertical and long horizontal diameter, thus affording sufficient ground for associating it with that low carapace which has been referred to T. triserrata. 'The male and female specimens of this type can be clearly recognized by the great difference in size. The ossa ilii are short and broad, far apart, so that the vertical diameter of the pelvis does not much exceed, or may be even shorter than, the horizontal. A strong ridge rises from the middle of the outer surface of the bone and ascends to the anterior spine. The lower part of the pubic bones is strongly inclined downwards, concave above in the male, and slightly convex in the female, and emits laterally a strong process, dilated at its end and more or less obliquely directed outwards. The posterior part of the ossa ischii is of considerable width, sometimes concave above, always with a trenchant symphysial crest beneath ; in one female specimen (No. 39935) it develops on each posterior corner a hamate process pointing outwards. The obturator foramina are comparatively narrow, sometimes scarcely wider than the symphysial bridge between them, which is perfectly flat, without median ridge.
In two of the male specimens (both in Mr. Newton’s collection) all the ridges and crests are much more prominent, the edges sharper, every impression and hollow deeper H
50 THE RACES OF THE MASCARENES.
than in the other specimens. They evidently belonged to very old individuals; and the atrophied condition of the bone is no doubt due to age. The individual to which one of these pelves belonged is represented also by a humerus showing the same peculiarity. The female pelvis is scarcely half the size of the male.
b. Of the second type (Plate XXVI. fig. B, and Plate XXVIII. fig. B) two perfect male examples are in the British Museum, and two fragments of a female and male in Mr. Newton’s collection. It is readily distinguished from the first type by its considerably longer and narrower ossa ilii, the longitudinal diameter of the pelvis much exceeding in length the horizontal. On the outer surface of the ilium scarcely a trace of a ridge is visible in the upper dilated portion. ‘The lower part of the pubic bones is gently inclined downwards and slightly concave above, and emits laterally a long, strong, nearly styliform process, obliquely directed outwards. ‘The posterior part of the ossa ischii is of moderate width, deeply concave above, with a trenchant symphysial crest beneath, which, expanding towards behind, forms a large triangular tuberosity. The upper posterior corner of each side is developed into a smaller tuberosity, pointing upwards. The obturator foramina are of moderate width, considerably wider than the symphysial bridge between them, which is provided with a median ridge above. This type propably belongs to 7. inepta.
c. The specimens of the third type are in a fragmentary condition, and were sent to me by M. Bouton with the scapula and femur. They came evidently from a deposit different from that in which the Dodo bones and the other Mauritian Tortoises were found, as they are not stained with the brown peat-colour, nor show any traces of rootlets of plants having been attached to them. Their surface is of a dirty white; and many of the edges are worn, as if the bones had lain in running water between pebbles. The four fragments belong to four individuals, each consisting of an acetabulum with the neighbouring parts; two have nearly the entire ilium preserved, one an obturator foramen with symphysial bridge, all the lateral pubic process. The posterior ischiadic tuberosity and the anterior pubic dilatation are lost. These pelves resemble in every respect, as far as we can see, the first type; but the os ili is narrower, though considerably less so than in the second type. The lateral pubic process is slightly thickened at the end.
They belong to the species Testudo leptocnemis, which is characterized chiefly by its femur, and of which we have noticed the scapula above.
THE GIGANTIC LAND-TORTOISES OF MAURITIUS. ol Measurements. T. triserrata. T. inepta. T. leptocnenis. oe N#* No. 39932. No. 39934. No. 39935. 75.10.20.1. No. 39931. Bl. B2. B3. 3 3 ? ? 3 3 mm. mm. mm. mm. mn. mm. mm. mm. mm. Longest inner vertical diameterofpelvis(fromsummit of ilium to symphysis) ... 126 135 95 90 156 140 Longest inner horizontal diameter of pelvis \.........+.. 128 140 90 84 122 102 Shortest inner horizontal diameter (between ilio-pubic PLOMMMIMENCES) A wis. f.c0eseve 89 82 60 60 94 72 Longest diameter of foramen OUMULATOTIUM: ..s.cc. 2s. ..cn. 35 32 20 19 4.0 36 ape we) Width of symphysial bridge = 26 27 22 16 30 22 er Breadth of posterior portion GMOSSAISCHIT 60... 0.5:6e0ns: 64. 55 34 38 65 50 Wenethiot os tht ......5....0003 112 110 75 73 130 113 120 105 Least width of os ili ...... 34 28 22 19 28 21 27 —_2 Femur.—Thirteen out of fifteen specimens are perfectly similar; and having been found in the “ Mare aux Songes” with the carapaces of T. triserrata and T. inepta, we may infer that the femurs of these two species do not differ from each other. a. All these specimens (Plate X XIX, fig. A) are rather stout, with a nearly straight subtetrahedral shaft; its distal portion is much dilated, consequently its outer margin much curved. ‘The ovoid head of the femur is separated from the trochanters by a deep and broad cavity, out of which arises a more or less indistinct smooth groove, dividing the trochanters. The head itself does not rise above the level of the summit of the larger trochanter. b. The two remaining femurs (Plate XXIX. fig. B) offer chief evidence of the existence of another species, for which I have proposed the name Testudo leptocnemis. They were found together with the scapule and pelves described above, and differ from the preceding species by the slender shaft and much less dilated condyles. The shaft is less distinctly tetrahedral, the anterior side being somewhat narrowed and very convex. ‘The summits of the ovoid head and larger trochanter are on the same level; and the hollow between head and trochanters is deep and broad; also the groove between the trochanters is distinct. * Newton collection. 9) -_
52 THE RACES OF THE MASCARENES.
Measurements.
T. triserrata. T. inepta. T’. leptoenemis.
No. 389954. No. 39958. No. 39959. No. 39955. B7. B8.
millim. millim. millim. millim. millim. millim.
Hensthvof femur.e.s 12s 152 133 116 152 135 140 Least circumference of femur 75 Al 56 75 63 60 Width of condyles ............ 62 56 45 59 48 50 Tibia.—Two examples of this bone are in the collection of the British Museum. Although they do not much differ in size, the ossification of the apophyses of the smaller one (No. 39960) is in a comparatively much more backward stage than in the larger example (No. 39961) (Plate XX VII. fig. D), which is 125 millims. long, with a circumference of 50 millims. in its narrowest part. A still larger specimen, 140 millims. long, is in Mr. Newton’s collection.. The articular surface is divided into two portions for the reception of the condyles, of which the inner is deeply concave, the outer nearly flat. The shaft is slightly bent. The distal end develops a prominent inner malleolus, separated from the opposite part of the articular surface by adeep hollow. Fibula.—Only one fibula (No. 39962) (Plate XXVIII. fig. H) is preserved; it is of the right side of a very large individual, evidently the same of which the radius (No. 39951) has been described. It is a straight slender bone, 134 millims. long, nearly cylindrical in the middle, and gradually svidening below. Its proximal end presents an oblique concave facet to the articular surface of the outer condyle, its distal end a subquadrangular convex surface to the tarsus. Its lower and upper portions and its hinder surface are provided with rough tubercular crests and protuberances, which is evidently but incidental to the age of the individual.
B. THe Gigantic LAND-ToRTOISE OF RODRIGUEZ.
The scanty historical evidence preceding the final extinction of the Rodriguez Tortoise has been referred to in the Introduction.
Remains of this Tortoise had been discovered and had reached Europe many years ago; but no particular attention was paid to them. M. J. DEsJARDINS, one of the first explorers of the fauna of Mauritius, sent a bone of a Tortoise, found, in 1786, in a cave in Rodriguez, with some remains of the Solitaire to Paris *, where they were examined by Cuvier and BLAINVILLE, who erroneously stated them to have been recently found under a bed of lava in Mauritius}. Another Mauritian naturalist, C. TEL- FAIR, in searching, in 1832, for bones of the Solitaire in Rodriguez, succeeded in obtaining “numerous bones of the extremities of one or more large species of Tortoise,” which were presented to the Zoological Society of London, and exhibited at one of the meetings {. These bones were still in the possession of the Society three or four years *Proc.Comm.Zool.Soc.ii.p.111.StricklandandMelville,‘TheDodo,’pp.51-53. t Edinb. Journ. Nat. Sci. iii. p. 30. £ Proc. Zool. Soc. 1833, p. 31.
THE GIGANTIC LAND-TORTOISE OF RODRIGUEZ. 53 before the publication of Strickland and Melville’s memoir on the Dodo (1848); but no further attention being paid to them, they were lost. Another portion of TELFAaIR’s collection was presented by him to the Andersonian Museum at Glasgow, where they arestillpreserved.
Some well-preserved bones, kindly sent to the writer by M. Bouton, of Port Louis, in 1872, satisfactorily proved that the Tortoise of Rodriguez is distinguished from all its congeners by well-marked characters (Ann. & Mag. Nat. Hist. 1873, xi. p. 397); but it was only when these remains were supplemented by those preserved in the Andersonian Museum at Glasgow and intrusted to me by the curators of that institution for examination, and when, finally, the extensive series collected during the Transit-of-Venus Expedition arrived, that our knowledge of its specific characters became tolerably complete. No further important additions can be expected from Rodriguez, with the exception of the small bones of the foot and caudal vertebrae; and these will be but of small value, unless they be found in their natural connexion. With the aid of the carapaces brought home by Mr. SiaTER, we are now enabled to recognize the Rodriguez ‘Tortoise in some carapaces which reached Europe in the last century, probably during the lifetime of the species, and which we find noticed by the following herpetologists :— 1. Scnorrrr (Histor. Testud. 1792, p. 103, pl. 22. fig. B) has reproduced a sketch of a Tortoise 22 feet long, which was communicated to him by Vosmaer, who examined the specimen which then was in “‘ Museo Principis Arausionensis in the Hague. This seems to have been a male, with a carapace very similar in form to that of the male described below ; its front and hind margins, being still provided with the epidermoid scutes, have an undulated outline. Schoepff was informed by Vosmaer that the carapace had been brought from the Cape of Good Hope; and expressing himself uncertain whether it should be considered a distinct species, or a sexual, local, or individual variety of the Tortoise described by Perrault, he named it ‘“ Testudo indica Vosmaeri.” 2. Dum&rit and Brsron recognized Schoepft’s Tortoise in a skeleton with complete carapace in the Paris Museum. The description of the specimen, whose shell measured 75 centims. over the curvature, again perfectly agrees with our male specimen, and supplies a detailed account of the outer epidermoid covering. The authors adopt the binomial term, “ Testudo vosmeri,” which, of course, supersedes that proposed by myself (Zestudo rodericensis, Ann. & Mag. Nat. Hist. 1878, xi. p. 397). By the singular resemblance of the general form of the male of this species to that of some of the Galapagos Tortoises, they were led into the error of supposing that 7. vosmeri came from the Galapagos Islands (Erpétol. Génér. ii. p. 140).
3. A second specimen, probably a young female, likewise in the Paris Museum, and without known history, was considered by the French herpetologists a distinct species, Testudo peltastes (ibid. p. 138). This description agrees in every respect with our youngcarapacesfromRodriguez.
54 THE RACES OF THE MASCARENES.
The Rodriguez Tortoise differs from the Mauritius and Galapagos Tortoises by the more slender build of all the various parts of its skeleton; its neck must have been capable of still greater flexion, as is evidenced by the deep postapophysial impressions or actual perforations of the cervical vertebre. Although careful comparative measurements show beyond doubt that this Tortoise had longer limbs and a longer neck than even some of the Galapagos Tortoises, yet, taking also into consideration the extreme thinness and fragility of its carapace, we must infer that this general slenderness of the bones must have been partially due to the same cause, probably a diminished supply of the calcareous salts, or a diminished power of assimilation of them. The bones collected by the naturalist of the ‘Transit-of-Venus Expedition belonged to several hundred individuals; and there are in some cases as many as forty specimens of one and the same bone in the collection; yet no variation in structure equivalent to that observed among the Galapagos, Aldabra, and Mauritius Tortoises could be detected, so that evidently in this small island there was room for one species only. ‘The only variation which is worth recording is one which can be explained as a sexual difference, the female having been of a smaller size and somewhat stouter form than the male, as is the case in the other Gigantic Tortoises.
Bones far exceeding in size the majority of their kind are not rare, and prove that the Rodriguez Tortoise was quite equal in bulk to Testudo elephantina, many (probably male) individuals having had a carapace 44 feet in length. From Duméril and Bibron’s descriptions we learn that the scutes were perfectly smooth or nearly so, and that the shell of the adult was black, whilst the young were of a lighter brown colour, the sternum being dotted with yellow.
DescriptionoftheOsseousRemains.
Skull (Plate XXIII. fig. C).—The skulls of twenty-four individuals of various ages, in a more or less imperfect condition, have been collected. They show very little variation beside that which is simply due to age. Two of the largest and most perfect are respectively 4) and 4§ inches long, measured from the intermaxillary to the occipital condyle, and 35 and 3 inches broad at the widest part, viz. between the tympanic processes. 1. ‘The frontal region is very broad and flat transversely, sloping in a more or less strong curve towards the nostril, its greatest width (in front of the postfrontals) being as much as, or not much less than, one half of the greatest width of the skull. 2. Only the foremost part of the parietals forms a flat surface, the remainder being compressed into a crest that passes into the long and narrow occipital spine, which rises scarcely above the level of the upper surface of the skull. 3. The tympanic case, with the mastoid, is produced backwards, the hind margin of the paroccipital forming a rather strong curve. 4. There is a shallow impression or hollow in front of the occipital condyle. 5. On the front margin of the temporal fossa, corresponding to the suture between parietal and tympanic, and directly in front of the foramen caro-
THE GIGANTIC LAND-TORTOISE OF RODRIGUEZ. 55 tidis extern, there is a large, prominent, concave, rough tuberosity for the insertion of a portion of the temporal muscle; exteriorly a narrow and deep groove separates this tuberosity from the zygomatic arch; inside of it runs a sharp ridge in an oblique direction from the tympanic condyle to the upper limit of the temporal fossa. This ridge separates the foramina by which the nerves and blood-vessels that leave the cerebral cavity by the foramen sphenoidale (portions of the nervus quintus and carotis cerebralis) issue from the substance of the bone. ‘These external foramina vary somewhat in number and position; but there are generally two behind and a smaller one in front of the ridge. 6. Tympanic cavity large, but somewhat constricted in the interior portion of its entrance; the outer tympanic rim has a rather irregular outline, not well defined in front, and with a deep and rather broad notch in the posterior part of its circumference for the passage of the Eustachian tube. 7. The columella is attached to and rests upon a long, nearly straight and rather sharp ridge, which runs from the notch mentioned to the stapedial foramen. 8. The posterior wall of the inner tympanic cavity, which in fresh examples is formed by cartilage and an open space in the preserved skull after maceration, is but limited in extent, about one seventh of the area of the tympanic opening. 9. The front margin of the intermaxillary projects beyond that of the frontal, the nasal opening being about as wide as long. 10. Although the position of the choane is advanced forwards, yet on the palatal view of the skull the ‘greater portion of them may be seen uncovered by the alveolar lamelle of the maxillaries; the front extremity of the vomer which separates the choane is very narrow and slender. 11. The foremost part of the intermaxillary is hollowed out below and vertically bent downwards to form the truncated beak. The suture between the intermaxillary and vomer is immediately behind the inner angle of the alveolar edges of the maxillaries. 12. Palatal region rather deeply concave and broad, a longitudinal median crest being scarcely indicated. The distance between the foramina palatina is much less than that between these foramina and the anterior extremity of the vomer. Outer pterygoid edge rather elevated and sharp. 13. Anterior surface of the tympanic pedicle with a moderately deep impression. 14. Lower jaw with a double alveolar ridge; its symphysial portion simply vertical, without a backward expansion of the lower margin of the bone.
Carapace (Plates XXI. and XXII.).—Tolerably complete carapaces of five individuals were found; unfortunately they are of more or less young age, the two largest being only 165 and 21 inches long. Four of them are very similar to one another, whilst the fifth and largest, though presenting the same specific characteristics as the others, differs considerably in shape. The carapace is distinguished by its extreme thinness and fragility, the lateral portions being only 1 millim. thick. Although the weather must have had some effect upon the preservation, exposure to it cannot have been the only cause, and can only slightly have deteriorated the surface, as the lines of impression of the epidermoid plates are distinctly preserved. Another characteristic of
56 THE RACES OF THE MASCARENES.
this species is the extraordinary width of the anterior and posterior openings, the sternum being very short, with both lobes contracted.
The normal shape of the carapace of the young (Plate XXII.) is a regular oval with subvertical sides, a moderate declivity in front, and a strong declivous expansion behind. No part of the margin is visibly reverted; and the vertebral plates are scarcely raised in their central portion. There is no nuchal plate; and the gular and broad caudal were undivided.
In two out of the four normal specimens the sternum is slightly concave, in the two others nearly flat. The anterior lobe is broader than long, truncated in front, or more or less tapering. The posterior is still shorter, with a very broad subtruncate or somewhatroundedhindermargin.
With regard to the fifth specimen (Plate XXI.), we may assume as almost certain that it was of the male sex, and rather more advanced in age than the others. The front portion of the dorsal shield is unfortunately broken} away and lost; but the perfectly level central profile of the three middle vertebral scutes shows clearly that there was no declivity of the outline in front; and the lateral margins of the front opening appear to have been somewhat reverted. The hind portion of the dorsal carapace is much produced backwards; and its marginal plates are expanded in a quite remarkable manner, being almost horizontal. The anterior portion of the sternum is much constricted, tapering in front, with concave lateral margins. Posterior lobe short and truncated as in the other specimens. The postgular plates approach so close to the abdominals that the intervening portion of the pectorals is reduced to a narrow strip. The following are the principal measurements of three of the specimens :— Length of carapace. Width of carapace. Sternum. Caudal plate.
Instr.line.Overcurve.Instr.line.Overcurve.Length.Width.Length.Width. Spec. in. in. in. in. in. in. in. in.
Deer eork 163 204 10 22 aT 9 13 A Oeicaineeanes 16 19 94 20 11 8? 14 4, Sass eset 21(?) 24.(?) lli 203 15 103 13 43 Cervical vertebre.—All the cervical vertebre are of a comparatively slight build, with thin walls for the spinal canal, with simple and low crests, and with the articular processes but slightly diverging.
Of the majority a considerable number of specimens have been collected; and we have sufficient means of judging of the amount of variation in each vertebra, also of comparing it with the corresponding bone in the other allied species; but, unfortunately, we do not possess a perfect series of the vertebrae of one and the same individual ; and although some more or less perfect series may be selected as apparently belonging to the same individual, such connexion between them is by no means a matter of absolute certainty: yet it would have keen of some importance tv obtain correct data as