THE GIGANTIC LAND-TORTOISE OF RODRIGUEZ. 57 regards the relative lengths of the several segments of the cervical column. In nearly all the cervical vertebree two forms may be distinguished, of which one is still more slender than the other; but, with this exception, there is a great uniformity of structure or want of variation.
No portion of the atlas has been preserved. Of the second vertebra we have two specimens, both of young individuals. It is provided with a moderately high neural crest, extending along the whole length of the bone in one specimen, and being limited to the anterior half in the other.
The ¢hird vertebra has a condyle in front and a glenoid cavity behind, and scarcely any indication of a median crest of its neural arch.
The fourth vertebra is biconvex, with the centrum much compressed, and passing into a sharp hemal crest which runs along the whole of its length. There is no neural crest.
The fifth vertebra has a glenoid cavity in front and a condyle behind; a median neural ridge is slightly indicated, and, in very old individuals, accompanied by a pair of other ridges diverging in the direction of the posterior zygapophyses. The centrum is compressed into a sharp ridge, highest on the posterior half of its hamal surface. There is a deep impression inwards and backwards of each of the anterior zygapophyses. ‘The first of the vertebre, of which the measurements are given below, belonged to an unusually large individual (Plate XI. figs. B).
The sixth cervical vertebra (Plate XII. fig. B) has a glenoid cavity in front and a broad condyle behind, which, by a deep median notch, is divided into two lateral halves. Its dorsal surface is flat, without crest, whilst on its visceral surface a rather high crest is evenly continued along nearly the whole length of the vertebra. As mentioned in the preliminary notice, Ann. & Mag. N. H. 1873, vol. xi. p. 397, this vertebra is distinguished by the perforation of the neural arch by a large ovate foramen (a) on each side, close to the anterior apophyses. There is no doubt that these two foramina are closed by membrane in fresh specimens, and are the result of the absorption of the osseous substance caused by the pressure of the posterior zygapophysis of the preceding vertebra when the neck is so much bent that the fifth and sixth vertebre stand nearly at a right angle. I am confirmed in this view by the fact that the natural bend of the cervical column is strongest at this place, which is opposite to the front margin of the carapace. However, although a very deep impression is constantly found immediately behind the anterior zygapophyses, which are perpendicular to the longitudinal axis of the centrum, the actual perforation of the neural arch has not always taken place. It is absent in all the vertebre of less than 55 millims. length; and out of twelve vertebre above that size I found it to occur five times. The seventh, biconcave vertebra (Plate XIII. fig. B), is distinguished from that of the Aldabra and Galapagos Tortoises by the entire absence of the high prominence on the dorsal surface of the neural arch. On the other hand, the median crest at the opposite I
58 THE RACES OF THE MASCARENES. surface of the centrum is as much developed as, or even more than, in any of the species named. A lateral impression behind the anterior zygapophyses is only slightly indicated. Measurements of three specimens will be given below, to show the extent of variation in this vertebra. Specimen ¢ belonged to a very large individual, perhaps to the same of which a fifth vertebra has been preserved ; and from a:comparison with corresponding vertebre of other gigantic Tortoises we may assume that this individual had a shell about 43 feet long. As in the seventh, so in the ezghth vertebra (Plate XIV. fig. B), the neural crest is either entirely absent or reduced to a pair of low ridges starting from the centre and diverging towards the posterior zygapophyses. In old individuals it is much less distinct than in young ones. ‘The diversity between the stout and slender forms is well marked, as may be seen from a comparison of the subjoined measurements. 3rd vert. 4th vert. 5th vert. 6th vert. 7th vert. 8th vert. Slen- Stout. Slen- Stout. Slender. Stout. Slender. Stout. Os, WY & Slender. Stout, der. der. —_—_—_o FY , mmmmmm.mmmm.mm,mm.mm.mmmm.mm.mmmmmmmm mm.mm,mm. Length of centrum... 53 46 85 52 115 80 70 56 110 69 66 71 106 63 58 67 36 30 Depth of centrum in the middle ...... UB 185920 el 34 21 17 18 PAR aes Mo} fs) 50 25 26 44 20 20 Horizontal width of middle of centrum... 11 18 11 OF on 22) 21S aL a2 25 18 7 22 30 19 15 29 16 18 Width of anterior conylorteesethcciteceMeyAbaig}Ko) Width of anterior glenoid cavity ...... a ef wi A 46 28 25 28 386 26 23 28 55 31 31 Widthofposterior condyleine ss): ais ee .. 18 18 (broken) 20 18 17 (broken) 26 23 27 chy, Ae We 29 eae Widthofposterior glenoid cavity.... 16 16 .. + Soap Ree eee Lf SL a ne 65 34 36 Distance of outer margins of anterior zygapophyses ...... 28.27 33 23 45 35 29 30 42 31 30 33 62 40 29 57 28 34 Distance of outer margins of posterior zygapophyses ...... 26) “24527 22 44 29 26 30 (broken) 25 24 27 HO Bi 383 72 30 40 Of the dorsal and caudal divisions of the vertebral column only a small number of disjointed vertebre have been collected. They are of too general a type to require any further remark; and it is quite impossible to form an idea as to the number of the caudal vertebrae.
Limb bones.—In the scapulary (Plate XXIV. fig. C) we notice the acute angle at which the scapula and acromium meet; the body of the scapula proper is slender, trihedral in form, with its anterior side flat or but little convex. The coracoid becomes ankylosed to the scapula at an early age of the animal ; and the neck above its glenoid
THE GIGANTIC LAND-TORTOISE OF RODRIGUEZ. 59 surface is very much constricted. It must be mentioned that there are in the collection a right and a left scapula, evidently belonging to the same individual, which differ remarkably from all the others in not having the coracoid ankylosed, although the individual appears to have been fully adult, and in having the body of the scapula proper uniformly compressed. It is difficult to account for this apparently individual aberration. The measurements of this bone are the following, a and 6 being more stoutly built than c and d:— a. b. C. d.
millim.millim.millim.millim.
Length of scapula measured from the suture with the coracoid... 205 146 168 135 CircumferenceimitsmanddleAeris.seekhGscanadadotascadeeowtnevase68434.238 Longitudinal diameter of glenoid cavity ................eeceeeeeee eee 50 35 38 32 Whenetheof.coracoid: 22.6 40.2hn ace ssess re foemeiiteg See eiwenaalene 73 60 66 51 GihedtestNACLOLCOLACOIG,dantonsdatsasdinnsemsienearadhedscoronas70(?)5445 Meme PURO Lea CLONMIMMG SiN V.Uay.cio 5 dai vc sac aren choses ane sos ses adnoes eas aende 73 55 65 51 On comparing with each other a considerable number of humeri (Plate X XV. fig. C), they can be arranged in two groups:—l, a slender form, with distinctly trihedral shaft; to this belong specimens some of which are nine inches in length; other, smaller ones clearly show by the porous structure of their epiphyses that they belonged to young or half-grown individuals. 2. The second form of humerus has a shorter and stouter shaft, and is rather flattened from the front backwards than trihedral. The specimens of this form do not exceed five inches in length; yet they are clearly bones of old individuals in which growth had ceased. As we know that in all the other gigantic Tortoises, with the life-history of which we are acquainted, the male greatly exceeds in size the female, it appears to be almost certain that the large bones belonged to the male and the smaller to the female sex. In other respects the various humeri exhibit the following characteristics. The head is nearly entirely raised above the level of the summit of the radial tuberosity; the ulnar tuberosity is raised high above the level of the head, from which it is separated by a deep groove; the ulnar margin of the shaft is much curved, the opposite margin being nearly straight. The canal for the bloodvessels, on the radial edge of the bone close to the elbow-joint, is perfectly bridged over, perforating the substance of the bone from the front to the hinder side. The following measurements are taken from an individual supposed to have been a full-grown male (a), half-grown male (6), and adult female (c) :— a. b. Ce millim.millim.millim.
Lengthofthehumerus(measuredinastraightlinefromthesummitof thesheadtothemiddleofthetrochlea)©....cccsccdesevsndicsteecetsdseccstsons210120108 Circumference of the narrowest part of the shaft ......0..0.. 0... ccc ceeeeee eee 75 41 44 moneesuadiametenolthieheades..)..5.Bicyjeadavesdveceagedstsngnoateaseecaaeesve39222] Sihontestadiameteriofthehead|2.2.asencncyseussvasescsaeesoscncssiaveselaveecees381919 Hxtremesbreadthbetweenthecondyles.......sceccscenesacecsesovecrssevesesact«643937 12
60 THE RACES OF THE MASCARENES. The bones of the forearm show similar differences as regards their proportions of length and width as the humerus, differences which probably are likewise due to sex. The ulna (Plate XXVIII. figs. F & G) is a comparatively straight bone, with its radial edge but slightly emarginate; the radius (Plate XXVIII. fig. E) is quite straight and remarkably slender. The ulna of an apparently not quite full-grown male individual (fig. F) has a length of 118 millims., and a width of 14 millims. in its narrowest part ; that of an adult female (fig. G) shows only a length of 64 millims. and a width of 11 millims. The longest radius in the collection is 105 millims. long, with a circumference of 29 millims. Of the carpus and remaining leg-bones nothing has been preserved. Pelvis (Plate XXVI. fig. C, and Plate XXVIII. fig. A)—In the pelvis the sutures disappear at an early age; the most important point of its structure is the great width of the symphysial bridge between the obturator foramina, which are comparatively narrow. In the majority of the individuals the transverse diameter of this bridge is much greater than the vertical, although the latter is increased by the development of a lower median crest. In very young individuals (that is, in individuals in which the horizontal diameter of the pelvis does not exceed a length of three inches) this characteristic feature of all flat-headed Tortoises is hidden by the development not only of a lower but also of an upper crest. Although the iliac bones are slender, like the remainder of the skeleton, yet the longitudinal diameter of the pelvis does not much exceed the horizontal one. In other respects, only the ilium shows a decidedly characteristic form: from its middle, which is considerably constricted, three ridges arise, running up towards the extremity of the bone, which thus assumes a conspicuously trihedral shape, with three nearly equidistant spinous processes. The two pelves, of which the following measurements are taken, are nearly perfect ; fragments of much larger ones are in the collection. millim. millim. Longest inner vertical diameter of pelvis from summit of ilium to symphysis......... 106 80 Longest inner horizontal diameter of;pelvis: Gi.5...-..eeeei eee eee ee eee 98 84 Shortest inner horizontal diameter of pelvis (between ilio-pubic prominences) ...... 68 55 Longest diameter of foramen obturatoriu:...c0.c.cseses oer ee eee eee eee 26 21 Width of ‘symphysial bridge. :.2.4:.:sc:¢se..0+ueceseeaeee jee eeeeene eee cee ee eee eee een 22 18 Depthof‘symphysialbridge:06220.fc.c4c00.aweescecscetecnenteenseteeeeeeteeaeee2213 Least breadth of posterior portion of ossa ischii..........- Iesiceciclee sesoceae cues scomeseetseett 49 38 TeengthOfOgMWys3%.acsvieynsweebenofbncelensdcoeseeneaeREaEeeOEEeeeteeee9370 Least:
breadth of 08:10.
..4%..0S na.
«0s cassecstanen vd verse coe cooneer aes eer eee eeeneiaso act eeeets 21 15 Femur (Plate X XIX. fig. C).—Although a considerable number of specimens of this bone have been collected, they are, singularly enough, all of middle or small size, and show scarcely any diversity of structure. The shaft is slender, nearly straight, irregularly subtetrahedral, and about as broad in front as behind. The head has an oval form, and is slightly raised above the level of the summit of the larger trochanter, from
THE GIGANTIC LAND-TORTOISE OF RODRIGUEZ. 61 which it is separated by a deep groove. ‘The larger and lesser trochanters are separated from each other by a more or less deep smooth indentation. I add the measurements ofthreespecimens:— a. b. C.
millim.millim.millim.
Wenethiy sweets crass yes tenn ay reas 125 104: 95 Least circumference.................. 45 44 39 Of the lower leg no part deserves particular description. To the kindness of M. Bouton I am indebted for a specimen of the right tid¢a of extraordinary dimensions (Plate X XVII. fig. D). It is 200 millims. long, and has in the middle a circumference of 76 millims. The largest specimen in the “‘ Transit-of-Venus” collection is 170 millims. long, with a circumference of 57 millims.
Our two largest specimens of fibula measure 140 millims. with a circumference of 41 millims., and 115 millims. with a circumference of 30 millims. Such of the remains of the foot as have been found do not show any peculiarity deservingofnotice.
62 THE RACES OF THE GALAPAGOS.
TorTOISEs indigenous to the Galapagos archipelago may be recognized by the following characters :—The nuchal plate is constantly absent; the posterior margins of the two gular plates are convergent, meeting at a more or less obtuse angle, never forming a straight, or nearly straight, transverse line. Neck and legs long. The shell black. One of the scutes on the inner side of the elbow is conspicuous for its size, much larger than those surrounding it. i In the skull, the crown is flat; the palate moderately concave; the front part of the intermaxillary truncated, elevated. The fourth cervical vertebra biconvex. The symphysial bridge between the foramina obturatoria of the pelvis is flat, broader than deep. Osseous carapace very thin. Nuchal vertebre and limb-bones elongate. It is evident from this diagnosis that the Galapagos Tortoises are differentiated from the Aldabra races by the same important structural characters as those of the Mascarenes, to which, on the other hand, they are most closely allied in every respect, except in the presence of a double gular plate. The various modifications of the form of the carapace and of the sculpture of the scutes which we have noticed in the forms of the islands of the Indian Ocean we find repeated in those of the opposite end of the globe. Porter's and DaRrwin’s statements that the various islands are inhabited by distinct species are fully borne out by my observations. Unfortunately, with the exception of the species collected by Commander Cookson, we do not possess positive and exact information as regards the localities whence our examples were obtained; but Portsr’s accounts are sufficiently detailed to enable us to relegate, with more or less certainty, some of the species before us to the places of their nativity. James Island yielded Tortoises of the broad, circular type; and therefore either 7. elephantopus or T. nigrita came from that island, probably the former. ‘There can be no doubt that we have, in T. ephippium, the species inhabiting Charles Island; 7. microphyes and T. vicina we know now to be indigenous to Albemarle, and 7’. abingdonit to Abingdon Island. Possibly other specific forms still exist; but those of Chatham and Indefatigable (and probably Charles) Islands have been extinct for a long time, and not even remains of them are known to exist in collections. As has been said in the “ Introduction,” the fate of all
TESTUDO ELEPHANTOPUS. 63 these animals seems to be sealed: what has been accomplished in the Mascarenes has commencedintheGalapagos.
The specific characters are apparent in mature individuals only. In young examples, which are rather common in collections, the distinctive characters, external or osteological, are incompletely developed, so that it is, at present, extremely difficult and somewhat hazardous to refer very young individuals (up to about 15 inches in length) to the species to which they belong. This resemblance of young examples cannot be used as an argument against the distinctness of the various species, as generally, in Vertebrates as well as Invertebrates, specific characters are not developed before a certain period, which varies exceedingly even in groups nearly related to one another. 1.
TESTUDO ELEPHANTOPUS.
The Tortoise to which Haruan (Journ. Ac. Nat. Sc. Philad. v. 1825, p. 284) gave this name was only 21 inches long over the curvature, or about 17 inches in a straight line, and therefore a young animal. A reference to the measurements and figure given by Haran shows clearly that he had an animal with the broad form of the body and with a posteriorly truncated sternum—characteristics by which a small series of examples before me are distinguished, and more especially one individual of nearly the same size as that described by Haruan.
Dumérin and Bisron (Erpétol. Génér. ii. p. 115) identify Harian’s example with one deposited by Quoy and Garmarp in the Paris Museum under the name of Testudo nigra. This specimen is still smaller than Haruan’s, and of an age at which the specific characters are not yet developed ; and therefore there is no evidence whatever to show that this identification by Dum&éri and Bisron is correct; and as long as it is uncertain to which of the specific forms the young “ 7. nigra” should be referred, the name had better be disused altogether. DuMm&rit and Brpron associate with this young specimen another of large size, distinguished by its broad form, smooth plates, and posteriorly excised sternum, but without giving any convincing proof that these two examples are of the same species. I have not seen an example agreeing in all points with that large example ; and it may possibly be another species distinct from those described here. The materials which I refer to 7. elephantopus are the following :— 1. An adult male example: a perfect skeleton with carapace, but without epidermoid plates. The carapace is 31 inches long. History of the specimen unknown; purchased of a dealer in Paris for the Oxford Museum, and kindly lent to me by Professor Rot- LESTON,F.R.S.
2. An immature female example: a perfect skeleton with carapace, but without epidermoid plates. The carapace is 28} inches long. Hab. Galapagos Islands. Property of the Royal College of Surgeons. Notes on this example by Professor OwrEN in Descript. Catal. Osteol. Ser. R. Coll. Surg. i. 1853, p. 194. no. 1011. 3. Carapace, without epidermoid plates, of an immature male example, 23 inches long. History unknown. Property of the Free Public Museum, Liverpool.
64 THE RACES OF THE GALAPAGOS.
4. Carapace, with epidermoid plates, of a young example, 18 inches long. Sex and history unknown. Property of the Free Public Museum, Liverpool. 5. A perfect skeleton with carapace, 154 inches long, obtained at Colon, and presented by Captain E. M. Leeps (S.S. ‘Tasmanian ’) to the author, and now in the British Museum.
Carapace.—In the largest example (specimen No. 1) (Plate XXX. fig. A), which has been prepared into a skeleton, the outlines of the epidermoid plates can be clearly traced. It is a fully adult male which, to judge from the condition of the bones, had ceased to grow a long time before its death; the dorsal portion of the shell is extremely thin, in some parts quite transparent. There is almost a total absence of anterior declivity of the first dorsal scute, its front margin being but very little below the level of the highest point of the carapace. The sides of this fore part of the carapace are expanded, not contracted as in 7. ephippium. 'The sternum is 243 inches long, and 23 inches broad between the lateral margins of the abdominal plates. It is deeply concave; and when the animal rested on the ground it touched it with the sides of the sternum, which are thicker than the remainder of the carapace, and on a transverse terminal callosity produced by the reverted posterior margin of the sternum, which is straight, truncated, without excision.
Another male example (specimen No. 3) agrees in every respect with the preceding, except in the sexual characters being much less developed, the specimen being only 23 inches long, and therefore much younger. ‘The first dorsal scute is more declivous towards the front, the concavity of the sternum less deep, and its terminal callosity only indicated by the very porous and rough surface of the bone. In young examples (15 to 18 inches long) the concentric striz are numerous, but not deeply cut; and in this respect the present species is intermediate between 7. nigrita and 7. ephippiwm. The posterior end of the sternum is nearly truncate, the hind margin of each anal plate being obtusely rounded, and the plates being separated by so shallow a notch that, evidently, with advancing age the sternum would have assumed the same truncate shape which we find in the adult specimens. It remains to add the principal measurements of the specimens examined :— Length of carapace. Width of carapace. Sternum. Caudal plate. Instr.line.Overcurve.Instr.line.Overcurve.Length.Width.Length.Width. Spec. no. in. in. in. in. in. in. in. in.
Leese aos. 31 373 26 40 245 23 7. AOE 283 363 23 35 223 19 aie oh eer ee 23 274 18 29 183 163 13 33 Ai ROO ae 1S 22 123 192 14 12 13 3 gu aaa eli 18} 113 19 124 11 14 23 Osteology.—The skull of an adult example of Testudo elephantopus (specimen No. 1, fig. A of Plates XLIJ.—XLIV.) is distinguished by a very short snout and a singularly