Ga Maries Phe saat SATAN SENT Se WP WAIVER ER Pune Use Ne e666 genet haee a+ ir Sint aad to ACE pecans ig Mires hee Tet oe ea nt aint a cata aaa la Sp. é ac S : a oe 5 YY: : 2 Ne: : S £4: Ae = SLE = Yl dn EO ki? = 2 AYE 2 = SKE? aan z a 2 5 2 z JLILSNI NVINOSHLINS Sa luvad (150 B RARI ES SMITHSONIAN INSTITUTION NOLLNLILSNI _NVINOSHLINS Sa luvae Zz é x « ae Line, 7. = GUY, 2 = z = QZ, = = JA gO? : g a La? : fer & 2 Foe 2 = 2 7 Fae 2 = = = eS = >" = = ARI ES. SMITHSONIAN INSTITUTION. NOILNLILSNI_NVINOSHLINS.” _LIBRARI ES SMITHSONIAN _INSTITUT| ta a a a uu = ud , a & = = a = Ps Ll, A ze | = a A al < = < = = i ar ah 3 = fs) 2 fe) 2 S = g = = : a Zz a : S ILILSNI NVINOSHLINS S3IY¥vVugI7 LIBRARIES HT cine lenge veba SCI RU SrSIMt INGUMSL NTS $3 uv = 5 = 6 = 6 5 ee : = a= ¢ a o ON = a = 5 S 2 5 2 NX 5 > % = > Fy a — = DW \S res — a sei ae . —_— 2 Fe = o = a = a, 5 mn Zz B Zi oD z a eee ARI ES, SMITHSONIAN — es NDT ESM NMINGSHIINS Se fuyvud Mul B RAR J ES Tn sOn sf: Zz z g "g “io g 2 NN = = = ‘S = “Zz Xs = Zz 3 * = S S o z= L ° QW Net x : fe) : SO LD an n ? a KB Q Wee = 2 ra S iE 2X = \ = = = = = = z = bas = itp) a Zz w A | eR IENVINOSHIINS Sa lYvyegiq_ re aidbear ie pogaiae vamelmian ean 5° LuV 2 ht oe lu ve uw = ASS us 2 ON: d = pil : : 2gN = i Ses LY a = Cc RAS Cc Ss « Yr oc Y oc ONS a Z A = S : LS e é a aa eS 5 v2 Sa ey z a i ARI ES_ SMITHSONIAN” INSTITUTION NOILNLILSNI NVINOSHLIWS ~S3 lyvud vale LIBRARI SSMU Sei Mash INSULIN z : fit Poe o GY ty, = Q wo \ 2 ow ° thy wo 5 $Y: E > We 5 2 5 GG Gr: E = WAN = 3 EG LS: = by ee = ae KS = 2 47 FF = mec a o a S\N 2 m 8 Cae, = 7) z no an = no = no LILSNI_ NVINOSHLINS, Sa fuvugit sk! B RARI ES, oMITHSONIAN INSTITUTION NOIMALLISNI eS NVINGSHTINS ae I Biv < = DS = Sie = Bic = Yp,, = = Wy, z = z * i Gy 2 = PD: B 5 fe~fin ON B B Ne ra Li Ye @ B ff = = = LY z = RN \S =» = = S > = ww alee Sin ee a 2 AR! ES SMITHSONIAN = INSTITUTION "NOLLALILSNI_ NVINOSHLIWS” $3 luv 11_LIBRARI me ie wo = Zz z2 c ! @ - z - & = Yl, % = a= = a ag = = ay pa PS : e = me S a S a rs) = Ss} a ro) = 3 5 Z a = a = “NVINOSHLINS ~Saluyar JILSNI NVINOSHLINS S31YVvudit ole IBRARI ES (SMITHSONIAN SINS TITUTION NOLO ETESN UGA ie is z = 3S = ° = S 7 o = Ss Be os = 4 = ES E a Sew 5 ae 5 > = = EYWeZG = cS 3 E x = = “GEG; oe = ae E m D m 2 tify m 2 m a = n = i¢p) = no = ©? SMITHSONIAN > NOILNLILSNI NVINOSHLINS S34 luvud Mtl BRARI So UNA EN wo z g = = ed = < = = = = = Zz = > = 5 = 5 ee on Ss 5 oS 2 3 5S ne se Oo 2 g = 2 = Zz iE Zz E > — > = = S > = ISNI NVINOSHLIWS VSIAldadvaudl i LIORARIES OMVIILIFMOUNTIAIN INOTIITULIUIN Peake IN VPI Vi inte tw wwe tio UU S i -1ES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS SaluvudIT_ LIBRARIES SMITHSONIAN Salyvydl) INSTITUTION INSTITUTION saiuyvydl LB Ly “ey INSTITUTION INSTITUTION “ BAS S3IuYvVYsIT Ssaluvuygl INSTITUTIO Z z ee a a ae z Z = = = BEEZ, | = Zz S| Zz N = 5 = 4p, 2 = S x 5S WSS g : 8 UW? g ‘2 3 z we z2 = PRS Ley aia an Zz i z= = RNS _NVINOSHLINS “LIBRARIES Meat le Vs eS UN ea Mere CI 2) 2 z ul Z iS Z i a Z = fz = a x a = A = =. cuet A < el < GC WwW < = ae 4 Cae A e a SNS 5 Wo - ca S = oe a 2 i a r 3 ee =I 2 i il RIES SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLIWS Sa1uvuaiy PUB RAR LES SM ITBSONIAN INST re ie ie fo) ine ro) = SY 9° Ae = 2 = a We 5 = 5! ro = yee 5) Bu Es 2 iS \ = = z a = e = a = 2 nm 2 Le z o = on = Z = on ILSNI NVINOSHLINS S3IYVUYEIT LIBRARIES NOILNLILSNI NVINOSHLINS S31uVud: Peay aks MB fp os SMITHSONIAN _ INSTITUTION ee SMITHSONIAN YS “ a, RIES SMITHSONIAN INSTITUTION NOILOLILSNI NVINOSHLINS S3IYVusit Z bof NVINOSHLINS S31uVvda NVINOSHLINS S3J1uvuglt NVINOSHLIWS SMITHSONIAN SMITHSONIAN NVINOSHLIWS SMITHSONIAN 4 LIBRARIES SMITHSONIAN _INSTITUTIC ee ce ul Z ts a ul a _ we Lani OS Sr + neve oa Pe. : =| Ey th a < = < A <5 A a Oy: cc c i oe a @ 3 : : : : : : =; e Zz S 2 x5 2 a 2 ILSNI_NVINOSHLINS SZIYVYGIT LIBRARIES SMITHSONIAN_INSTITUTION NOILNLILSNI Sa1uvud! Ae 5 £ 5 Dy 5 = 5 wo = wo = tgp wo — o S = Ee = ‘Golf 4 = \, = a a] hfe, rs] \S 5 F 2 LL e : ss = a = a Be ee eS = 2. ANE RIES SMITHSONIAN INSTITUTION NOILONLILSNI NVINOSHLINS S3IMVuaIT_ SMITHSONIAN INSTITUTIC n” mae 2 nv ie n a p22) z= * = ONT = 4; <= = < = < os Zz = Ys, F 4 z = z 2 E 2,0 & = E z - ae z 2 5 2 5 2 Bae ILSNI_NVINOSHLINS S3IYVYEIT_ LIBRARIES SMITHSONIAN INSTITUTION _NVINOSHLIWS S31uVud a ws A os z e < wl ws (7) ot wn ees n = = 0 =a we sca a = oc | et =o < Action | +” Orper—OPHIDIA, remarks on Genus—Patzxoruis TypHzus 22 3) 3? 55 poreatus bP) 3) Parrryx, remarks on a rhombifer depressus 22 Pleurodont Lizard (?) toliapicus (?) longus Part I. Page 61 62 66 67 67 70 71 INDEX TO THE SPECIES DESCRIBED IN THE MONOGRAPH ON THE REPTILIA OF THE LONDON CLAY, &c. Alligator Hantoniensis, Wood, Mid. Ho. Part II, p. 42, tab. viii, fig. 2. Chelone antiqua, Kenig (?). Part I, p. 10. Chelone breviceps, Owen, Lond. Cl. Part I, p. 10, tabule i, ii. » convexa, Owen, Lond. Cl. Part I, p. 21, tab. vii. » crassicostata, Owen, Lond. Cl. Part I, pp. 27 and 42, tabs. xi, xii, xiii, xi 4, and xiii B. a cuneiceps, Owen, Lond. Cl. Part I, p. 33, tab. xv, ae declivis, Owen, Low. Ho. Part I, p. 30, tab. xiv. » Harvicensis, Woodward (2). Part I, p. 25. », latiscutata, Owen, Lond.Cl. Part I, p. 20, tab. vi. » longiceps, Owen, Lond. Cl. Part I, p. 16, tabs. ili, iv, and v. » planimentum, Owen, Lond. Cl. Part I, p. 25 and 40, tabs. ix, x, and xa. » subcarinata, Bell, Lond. Cl. Part I, p. 37, tab. viii a. 5 subcristata, Owen, Lond. Cl. Part I, p. 24, tab. viii. » trigoniceps, Owen, Mid. Ko. Part I, p. 31, no plate. Crocodile de Sheppy (:) Cuvier. Part II, pp. 29 and 31. Crocodilus champsoides, Owen, Lond. Cl. Part II, p. 31, tab. ii, fig. 2?, and tab. iii. 55 Hastingsie, Owen, Mid. Eo. Part II, pp. 37 and 44, tabs. vi, vii, viii, fig. 1, ix and xii, figs. 2, 5. 5 Spenceri, Buckland. Part II, p. 29. a5 toliapicus, Owen, Lond. Cl. Part II, p. 29, tab. ui, fig. 1, tab. v, figs. 1, 2, 3, 5, 6, and 12, tabs. ii, 114, fig. 1, and tab. iy. INDEX. Emys bicarinata, Lond. Cl., Bell. Part I, p. 73, tabs. xxv, xxvi. » Comptoni, Lond. Cl., Bell. Part I, p. 71, tab. xx. » Conybeari, Owen, Lond. Cl. Part I, Sup. I, p. 77, tabs. xxviiia and xxviiis. » erassus, Owen, Mid. Eo. Part I, p. 76, tab. xxvii. s, Delabechii, Lond. Cl., Bell. Part I, p. 74, tab. xxviii. » de Sheppy, Cuv. (2?) Part I, p. 67, tab. xxiv. 33 3 H. v. Meyer. (?) Part I, p. 10. », levis, Bell, Lond. Cl. Part I, p. 70, tab. xxii. » Parkinsoni, J. E. Gray. Part I, p. 10. ,, testudiniformis, Owen. Part I, p. 67, tab. xxiv. Gavialis Dixoni, Owen, Mid. Eo, Part II, p. 42, tab. x. Paleophis (?) longus, Owen, Low. Eo. Part III, p. 66, tab. xiv, figs. 85-37, 45, 46. 3 porcatus, Owen, Mid. Eo. Part ITI, p. 61, tab. xiv, figs. 13—15, 18, 20, 21. 5 toliapicus, Owen, Lond. Cl. Part III, p. 63, tabs. xv and xvi. 55 typhus, Owen, Mid. Eo. Part III, p. 56, tab. xi, figs. 5—8, tab. xiv, figs. 1—3, 7—9, 16, 17, 26, 27, 28 (figs. 6, 10—12) (?). Paleryx depressus, Owen, Mid. Eo. Part III, p. 68, tab. xiii, figs. 37, 38. - rhombifer, Owen, Mid. Ko. Part III, p. 67, tab. xiii, figs. 29—82. Platemys Bowerbankii, Owen, Lond. Cl. Part I, p. 66, tab. xxiii. Part II, Sup. I, p. 1, tab. xxix, figs. 1, 2. es Bullockii, Qwen, Lond. Cl. Part I, p. 62, tab. xxi. Testudo plana, Kdnig. Part I, p. 27. Trionyx Barbar, Owen, Mid, Eo. Part IT, p. 50, tab. xvia. », circumsuleatus, Owen, Mid. Eo. Part I, p. 59, tab. xix 8, figs. 1—3. » Henrici, Owen, Mid Eo, Part I, p. 46, tab. xvi. », imcrassatus, Owen, Up. Ho, Part I, p. 51, tabs. xvii, xvili, and xix. 5 marginatus, Owen, Mid. Eo. Part I, p. 55, tab. xix* and tab. xixB, figs. 4— 6. ,, planus, Owen, Mid. Eo. Part I, p. 58, tabs. xixc and xix p, fig. 6. » pustulatus, Owen, Lond. Cl. Part I, p. 60, tab. xix B, figs. 7—9. » Yivosus, Owen, Mid. Eo. Part IL, p. 56, tab. xviii. » Sp., Owen. Part I, Mid. Ho., tab. xix p, figs. 3—5. » Sp., Owen. Part I, Mid. Eo., p. 61, tab. xix p, fig. 7. MONOGRAPH ON THO Osis di Teeth Da OF THE LONDON CLAY, AND OF THE BRACKLESHAM AND OTHER TERTIARY BEDS. PART LI. Paces 1—76; Puarns I—XXVIII, and VIIda, Xa, XIITa, XIIIe, XVIa, XVITIa, XIX*, XIXz, X1Xc, XTX». CHELONIA (Cuaztons, &c.). BY PROFESSOR OWEN, D.C.I.., F.RS., F.LS., F.G.S., &e. AND PROFESSOR BELL, F.RS., F.L.S., F.G.S., &c. Issued in the Volume for the Year 1848. LONDON: PRINTED FOR THE PALHZONTOGRAPHICAL SOCIETY. 1849. MONOGRAPH ON THE FOSSIL REPTILIA OF THE LONDON CLAY. OrvER.—CHEHLONIA. Family—M ar ina. Genus— CHELONE. THE majority of the Fossil Chelonians of the Eocene tertiary deposits, defined or described in my ‘ Report on British Fossil Reptiles,’ belonged to the marine division of the order, and to the genus Chelone ; and as the species of this genus depart least from the ordinary reptilian type in the modification of the bones of the trunk, composing the characteristic thoracic-abdominal case of the order, I propose to commence with them those descriptions of the Chelonian reptiles which fall to my share of the present Monograph. In order to facilitate the comprehension of the descriptions and figures of the fossil Chelonians, a brief notice is premised of the composition and homologies of the carapace and plastron, or roof and floor, of that smgular portable abode, with which the reptiles of the present order have been endowed in compensation for their inferior powers of locomotion or other modes of escape or defence. In the marine species of the Chelonian order, of which the Chelone mydas may be regarded as the type, the ossification of the carapace and plastron is less complete, and the whole skeleton is lghter than in those species that live and move on dry land: but the head is proportionally larger—a character common to aquatic animals,— and being incapable of retraction within the carapace, ossification extends in the direction of the fascia, covering the temporal muscles, and forms a second bony covering of the cranial cavity: it is interesting to observe, however, that this accessory defence is not formed by the intercalation of any new bones, but is due to exogenous growth from the frontals (11), parietal (7), postfrontals (12), and mastoids (s, see T. I, TUE EX ND): The bony carapace is composed externally of a series of median and symmetrical pieces (fig. 1, ch, si—s11, py), and of two series of unsymmetrical pieces (p/1—s, 71—12) on each side. The median pieces have been regarded as lateral expansions of the summits of the upper vertebral (neural) spines,* the median lateral pieces as similar * Cuvier, Lecons d’Anatomie Comparée, tom. i (1799), p. 212. 2 FOSSIL REPTILIA OF THE LONDON CLAY, developments of the vertebral ribs (pleurapophyses),* and the marginal pieces as the homologues of the sternal ribs (heemapophyses).t I must refer the reader to my Memoir, communicated to the Royal Society, for the facts and arguments which have led me to regard these pieces, as dermal ossifications, homologous with those that support the nuchal and dorsal epidermal scutes in the crocodile. Most of the bony pieces of the carapace are, however, directly continuous, and connate,{ with the obvious elements of the vertebra, which have been supposed exclusively to form them by their unusual development; the median pieces have accordingly been called “vertebral plates,” and the medio-lateral pieces “ costal plates.” I retain the latter name, although with the understanding and conviction that they are essentially or homologically distinct parts from the vertebral ribs or pleurapophyses with which they are connate and more or less blended. But, with regard to the term “vertebral” plate, since the ribs (coste) are as essentially elements of the vertebra as the spimous processes themselves, I have been in the habit, in my Lectures, of indicating the median series by the term “neural plates,’ which term has the further advantage of removing any ambiguity from the descriptions that might arise from their being mistaken for the superincumbent epidermal shields, which are likewise called “ vertebral plates” in some English works.) The term “marginal” is retained for the osseous plates forming the periphery of the carapace; but the median and symmetrical ones, which seem also to begin and end the “neural” series, are specified, the one by the term “nuchal plate,” the other by that of ‘* pygal plate.” The “ neural plates” are numbered as in the classical Monograph of Bojanus.|| In the subjoined woodcut of the carapace of the loggerhead turtle (Chelone caowanna) (fig. 1), ch is the nuchal plate ; s\ to s11 the neural plates ; pl) to pls the costal plates ; and m\ to m12 the marginal plates. The carapace is impressed by the superimposed epidermal scutes or shields, which consist of a median series, called “ vertebral scutes” v1 to v5; * Tbid. p. 211. Rathké has recently supported this determination by arguments drawn from the mode of development of the carapace. See ‘ Annales des Sciences Naturelles,’ Mars, 1846; and ‘ Ueber die Entwickelung der Schildkroten,’ 4to, 1848, where he says, p. 105 :—‘‘ Ausser den Rippen und den horizontal liegenden Tafeln, zu welchen sich die Dornfortsiitze des zweiten und der sechs folgenden Riickenwirbel ausbilden, dienen bei den erwachsenen Schildkroten zur Zusammensetzung des Riickenschildes noch eine ) > oder mehrere Knochenplatten,”’ viz. the ‘marginal plates.” I have shown how Rathké was deceived by over-estimating the character of connation, in my ‘ Observations on the Development of the Carapace and Plastron of the Chelonians,’ which conduct to a different conclusion to that at which Cuvier and Rathké have arrived. (Philosoph. Transactions, 1849.) + Geoffroy, Annales du Museum, tom. xiv (1809), p. 7. t This term is used in the definite sense explained in my work on the ‘Archetype of the Vertebrate Skeleton’ (8vo, V. Voorst, p. 49), as signifying those essentially different parts which are not physically distinct at any stage of development; and in contradistinction to the term “confluent,” which applies to those united parts which were originally distinct. § See Griffiths’s translation of Cuvier, vol. ix, Synopsis of Reptilia, p. 6—“ fifth vertebral plates prominent.” || Anatome Testudinis Europe, fol. 1821, tab. ii and iy. CHELONIA. and of a lateral series of “ costal scutes ;” there is also a peripheral series of “ marginal scutes” corresponding with and impressing the mar- ginal plates. The nuchal plate (cA) remarkable for its breadth in all Chelonia, and usually sends down a ridge from the middle line of its under surface, which is attached by ligament to the summit of the neural arch of the first dorsal vertebra. The first true neural plate, si, is much narrower, and is connate with the summit of the neural spine of the second dorsal vertebra; the succeeding vertebral neural plates, s2 have the same relations with the succeeding neural spines, but the ninth, tenth, and eleventh, like the nuchal (c/) and pygal (py), plates are independent ossifications in the substance of the derm. The costal pieces of the carapace are supra-additions to is 88, co Carapace of the Loggerhead Turtle (Chelone caouwnna). eight pairs of pleurapophyses or vertebral ribs, those, viz. of the second to the ninth dorsal vertebree inclusive. The slender or normal portions of the ribs project freely for some distance beyond the expanded and connate portions (“ costal plates” of the carapace), along the under surface of which the rib may be traced, of its ordinary breadth, to the neck and head, which liberates itself from the costal plate to articulate to the interspace of the two contiguous vertebral bodies, (centrums), to the posterior of which such rib properly belongs. The woodcut (fig. 2) illustrates this structure: ch shows the inner side of the nuchal plate; ¢1 is the first rib, articulated to the fore part of the body of the first dorsal vertebree; p/1 is the first rib of the carapace (the second rib of the dorsal series), connate with the first costal plate; p/2 to pls, are the succeeding ribs and costal plates of the carapace. Inner view of carapace of the Loggerhead Turtle (Chelone caovanna). The heads of the ribs articulate to 4 FOSSIL REPTILIA OF THE LONDON CLAY. the interspaces between their own vertebral body, and that of the preceding vertebra. The tenth vertebra supports a short pair of ribs in Chelone and in Hmys, but not in Trionyx; and this vertebra is commonly reckoned as a “lumbar” one. The eleventh and twelfth vertebre have short and thick ribs, which abut against the iliac bones, and they are regarded as forming the sacrum. The remaining vertebre belong to the tail, and are “caudal.” The costal plates articulate with each other, and with the neural plates by fine dentated sutures. The free extremities of the ribs are implanted into sockets of those marginal plates which are opposite to them. ‘The Ist, 2d, 3d, and 10th, are not so articulated in the loggerhead turtle. But all the marginal plates articulate with each other, and with the nuchal (c/) and pygal (py) plates by sutures. The osseous basis of the plastron consists of nuime pieces, one single and sym- metrical, the rest in pairs. The median piece, s, is the exfosterna/; the anterior pair, es, is the epzsternal ; the second pair, 4s, the hyosternal; the third pair, ps, the hyposternal; and the posterior pair, zs, the ziphisternal. With regard to the nature or homologies of these bones, three views have been Fig. 3. taken. The one generally adopted, on the authority of Cuvier, Bojanus, and Geoffroy St. Hilaire, is, that the nine bones of the plastron are subdivisions of a vastly expanded sternum, or breast-bone; the second view is, that these subdivisions of the sternum are enlarged by combination with ossifications of the in- tegument ;* and the third view, in which Rathke stands alone, is, that they are exclusively dermal bones, and have no homologues in the endoskeleton of other vertebrata.t Since this opinion is given as the result of that celebrated embryologist’s observations on the de- velopment of the Chelonian reptiles, I have tested it Bones of the plastron of the Loggerhead by a series of similar researches on the embryos and arte Cele capadie2): young of the Chelone mydas and Testudo indica, and have been led by them to conclusions distinct from any of the three theories above cited. The sternum, like the carapace, is, without doubt, a compound of connate, endo- skeletal and exoskeletal pieces; but the endoskeletal parts are not exclusively the homologues of the sternum. For the details of the observations, and the special arguments on which these conclusions are founded, I must refer to my paper in the ‘Transactions of the Royal Society,’ 1849; the homologies of the endoskeletal parts of the plastron will require a brief illustration here from comparative anatomy. * Peters, Observationes ad Anatomiam Cheloniorum, 1838. + Ueber die Entwickelung der Schildkroten, 4to, 1848, p. 122. CHELONIA. 5 Geoffroy St. Hilaire, whose views are generally adopted, was guided in his deter- mination of the parts of the plastron by the Big. 4. analogy of the skeleton of the bird: which analogy may be illustrated by the subjoined dia- grams of corresponding segments of the thorax of a bird (fig. 4) and of a tortoise (fig. 5). In both figures cis the centrum or vertebral body ; as the neural arch and spine ; compressed in the bird, depressed and laterally expanded, accord- ing to Geoffroy, in the tortoise ; p/ the pleura- pophysis, or vertebral rib, expanded in the tortoise, and with its broad tubercle articu- lating with the expanded spine; 4, /’ in fig. 5, answers to / in fig. 4, and is the hamapo- physis (sternal rib, or ossified cartilage of the rib); A, ds in fig. 5, is As in fig. 4, i. e. exclusively a sternum, with the hs Thoracic segment of the skeleton of a Bird. entosternal piece, /s’, developed hori- zontally in the tortoise, and vertically in the bird. The primd facie simplicity of this view has imposed upon most comparative anatomists: and _ yet there are other vertebrate animals more nearly allied to the Chelonia than birds, and with which, therefore, comparison should have been insti- tuted before general consent was —— ae yielded to the Geoffroyan hypothesis. Thoracic cadens of the skeleton of a Tee If, e. g. we take the segment of a crocodile’s skeleton (fig. 6) corresponding with that of the tortoise (fig. 5), the comparison will yield the following interpretation: in both figures ¢ is the centrum: zs the neural arch and spine, with d the diapophysis; se a median dermal bony plate (connate with zs in the tortoise) ; pl the pleurapophysis ; sc sc lateral dermal bony plates (connate with p/ in the tortoise); 4, /’ in fig. 5, answers to /’ in fig. 6, an intercalated, semi-ossified piece between p/ and / in the crocodile; 4, ds in fig. 5, answers to 4, the heemapophysis in the crocodile; and As in Thoracic segment of the skeleton of a Crocodile. fig. 5, exclusively represents /s, the sternum in the crocodile. 6 FOSSIL REPTILIA OF THE LONDON CLAY. Such a comparison, in my opinion, guides us to a truer view of the homologies of the thoracic-abdominal bony case of the Chelonians, especially with regard to the lateral or parial pieces of the plastron, than the comparison exclusively relied on by Geoffroy St. Hilaire. The Plesiosaurus, by its long and flexible neck, small head, expanded coracoid and pubis, and flattened bones of the paddles, comes much nearer to the turtle than the crocodile does; and its abdominal ribs, or hemapophyses, are more developed than in the crocodiles; a comparison of the ventral surface of the skeleton, such as that figured by Dr. Buckland, in his ‘ Bridgewater Treatise,’ vol. ii, pl. 18, fig. 8, will show how clearly those abdominal ribs would correspond with the hyosternals and hyposternals of the turtle, if they had coalesced together at their middle parts, leaving their outer and inner extremities free. With regard to the marginal pieces m1—wm12, figs. 1 and 2, although the comparisons illustrated by figs. 4, 5, 6, show that they answer rather to the intercalated piece /’ in the crocodile than to the entire sternal rib 4 in the bird; yet the phenomena of their development demonstrate that they are exclusively bones of the dermal skeleton, retaining their freedom from anchylosis with the endoskeletal elements, like the nuchal, pygal, and last three neural plates (ch, py, s9, 510, and s11, fig. 1). This insight into their true nature teaches why they do not correspond in number with the vertebral ribs or pleurapophyses (p/i—p/s, fig. 2). In the loggerhead turtle, for example, the first three and the tenth (m1, m2, m3, and m0) have no corresponding pleur- apophyses articulating with them; and if even ¢1 be supposed to correspond to m3, there are no rudiments of ribs answering to m1 and m2. The marginal plates are not constant in number; the Chelone mydas has two less than the Chelone caouanna has. Some species of Zrionyxr (Cryptopus, Dum. and Bibron) have a greater number, but of smaller and less regular size, confined to the posterior part of the limb of the carapace; in other species of Zrionyx (Gymnopus, Dum. and Bibron), and in Sphargis, the marginal part of the carapace retains its embryonic condition in all Chelonia, as a stratum of cartilaginous cells in the substance of the derm, forming the thickened, flexible border of the carapace. The rudiments of the hyosternals and hyposternals have originally the form of sternal or abdominal ribs; extend transversely, and rise at their outer extremities to join those of the first and sixth pair of vertebral ribs, completing the heemal, or inferior vertebral arch, without the interposition of any of the marginal pieces, which are merely applied to the outer sides of the heemapophysis or sternal ribs. The expansion of the parts of the plastron, especially in the fresh-water and land tortoises, is due chiefly to the ossification of a layer of cartilage-cells in the substance of the derm, which ossified plates are connate with the more internal elements of the plastron, representing the sternum and sternal ribs. In the following descriptions of the fossil Chelona, the terms ‘ entosternal, episternal, hyosternal, hyposternal,’ and ‘ xiphisternal,’ will be used as absolute designations of the combined endoskeletal and exoskeletal bones of the plastron, without implying assent to the hypothesis that first suggested those names to Geoffroy St. Hilaire. CHELONTA. 7 The scapular and pelvic arches, and the bones of the extremities of the Chelonia, are described and figured in the ‘Ossemens Fossiles’ of Cuvier ;* where, also, the figures of the modifications of the carapace and plastron, in the fresh-water and land tortoises, will suffice for the purpose of ulterior comparisons with the fossils described in the present work, if they be understood according to the homologies above discussed, and which are illustrated by the figures 1 and 2 of the carapace, and fig. 3 of the plastron of the Chelone caouanna. With regard to the more immediate subjects of the present Monograph, it must be admitted that the important generalizations of Cuvier and Dr. Bucklandt have been confirmed, but not materially extended, by subsequent observations on the remains of reptiles of the Chelonian order. Cuvier, after admitting that his results in regard to the tortoises were not so precise as those relating to the crocodiles, sums up his chapter on the fossil Che/onza in the following words: ‘‘ Toutefois nous avons pu nous assurer que les tortues sont aussi anciennes dans le monde que les crocodiles ; qu elles les accompagnent généralement, et que le plus grand nombre de leurs débris appartenant a des sous-genres dont les espéces sont propres aux eaux douces ou a la terre ferme, elles confirment les conjectures que les os de crocodiles avoient fait naitre sur l’existence diles ou de continens nourissant des reptiles, avant qu'il y ait eu des quadrupedes vivipares, ou du moins avant qu’ils aient été assez nombreux pour laisser une quantité de débris comparable a ceux des reptiles.’ Dr. Buckland also states, in general but precise terms, that ‘ the Chelonian reptiles came into existence nearly at the same time with the order of Sawrians, and have continued coextensively with them through the secondary and tertiary formations unto the present time. Their fossil remains present also the same threefold divisions that exist among modern Che/onia into groups. respectively adapted to live on land, in fresh water, or the sea.”’§ The remains of sea turtles (Chelone) have been recognised in the Muschelkalk, the Wealden, the lower cretaceous formation at Glaris, and the upper chalk-beds at Maestricht. Figures of Chelonites, as that in the Frontispiece to Woodward's ‘Synoptical Table of British Organic Remains,’ and in Konig’s ‘Icones Sectiles’ (pl. xviii, fig. 232, a and 4), have been published; but no true marine Chelonian, from Eocene strata, had been scientifically determmed prior to the communication of my Paper on that subject to the Geological Society of London.|| All the Chelonites from Sheppey, described and figured in the last edition of Cuvier’s ‘Ossemens Fossiles,’ for * Tom. vy, pt. 2, pl. xii and xiii. + Bridgewater Treatise (1836), p. 256. ft Ossemens Fossiles, 4to, tom. y, pt. u, p. 249. § Bridgewater Treatise, p. 256. || Proceedings of the Geological Society of London, vol. iii, pt. ii, p. 570, December 1, 1541. 8 FOSSIL REPTILIA OF THE LONDON CLAY. example, are referred to the fresh-water genus Hmys; and the statement in the earlier edition of the ‘Ossemens Fossiles,’ that the greater part of the remains of Chelonian reptiles belong to the fresh-water or terrestrial genera, is repeated. The aim of the Memoir, communicated to the Geological Society in December, 1841, was to show that the conclusion deduced by Cuvier, from an imperfect carapace from Sheppey, which might probably have belonged to a species of Hmys, had been unduly extended to other Chelonites, which undoubtedly belonged to the marine genus Chelone ; and that this genus was represented, in the Eocene strata, by at least six species; the remains of five of which were from the London Clay at Sheppey, and those of a sixth were tolerably abundant in the cliffs near Harwich. In the carapace of the fossil Chelonian from Sheppey, communicated by Mr. Crowe, of Faversham, to Cuvier, and figured in the ‘ Ossemens Fossiles’ (tom. v, part 2, pl. xv, fig. 12), the author of that great work conceived that all the characters of the genus Enys were perfectly recognisable. He points out the proportions of the neural plates, which are as long as they are large; and in the figure they are represented of nearly a quadrate form, and not rhomboidal. The fifth neural plate in the fragment figured (probably the eighth) is separated from the sixth (ninth) by a point, which is made by the mesial ends of the fifth (probably the seventh) pair of costal plates; a structure which Cuvier says slightly recalls what he had observed in the Jura Aiys of Soleure.* But Cuvier admits that the neural plates (plaques vertébrales) are narrower than those of existing Hmydes ; and that the equal breadth of the ribs is a character common to the Chelones with the Hmnydes. Now, in reference to the carapace figured by Cuvier, it is to be observed, that the — margins are wanting; and that the broad conjoined portions of the costal plates are not longer than they might have been, had the fossil belonged to a turtle (Chelone) ; and, consequently, that there is no proof that they were united together by suture throughout their whole extent, as in the Hmydes ; but that they might have terminated in narrow tooth-like processes, as in the Chedones. The narrowness of the neural plates is a character which, with their smoothness, undoubtedly approximates the fossil to the Chelones; and, without imtending to affirm that the fossil in question does not belong to the family Hmydide, which unquestionably existed at the time of the deposition of the Sheppey clay, its determimation appears to me to be much less decisive than might be inferred from the remarks m the ‘Ossemens Fossiles.’ * Tom. cit., p. 234. This structure is not, however, peculiar to the genus Hmys ; in the carapace of the Chelone caouanna, in the Museum of the Royal College of Surgeons, the seventh neural plate is separated from the eighth by the junction of the expanded extremity of the seventh rib on one side with that of the opposite rib, and the eighth neural plate from the ninth by the same modification of the eighth pair of ribs. A similar modification may also be seen in the carapace of the Trionyx Henrici, T. XVI. CHELONIA. 9 Mr. Parkinson describes the plastron of a Sheppey Chelonite,* in which the hyosternal and hyposternal pieces are not united, but leave a vacancy in the middle, which he conjectured may have been filled up by membrane. This specimen must have belonged to a specimen at least four inches in length, exclusive of the head and neck. But Cuvier supposes that it may, nevertheless, have belonged to an Emys; and that the vacancy of the bony sternum merely indicated the nonage of the individual.t The grounds on which Cuvier refers to the genus mys, the imperfect and dislocated carapace and plastron of M. Bourdet’s Sheppey Chelonite,t are not detailed; but it is evident that the hyposternals in that specimen are in contact at the posterior moiety of their median margins only; and that the margins recede anteriorly, leaving a median interspace; which, as the plastron is nearly a foot in length, can hardly be attributed to the immature state of the individual. And if, as Cuvier supposes, this specimen belongs to the same species as those in the collections of Messrs. Crowe and Parkinson, the same objection to their belonging to a fresh-water tortoise holds good , as to the one figured by M. Bourdet. The question of the reference of these Eocene fossils to the fresh- or sea-water families of the Chelonian order, seems to me to admit of the safest determination by examining the crania of the Sheppey Chelonites; since the differences in the extent to which the temporal fossze are protected by bone, and in the proportions in which the bones enter into the formation of that covering, are strongly marked in the genera Emys and Chelone. But here Cuvier appears to have been unusually biassed in favour of the Emydian nature of the Sheppey fossils; for in reference to the cranium, figured by Mr. Parkinson, the affinities of which to the turtle’s skull will be presently pointed out, Cuvier observes : “elle est probablement aussi d'une Emyde, bien qu’elle participe des caracteres de Tortues de Mer, par la maniére dont le parietal recouvre sa tempe ; mais nous avons vu que /’ Himys expansa differe trés peu de Tortues de Mer a cet égard, et la partie antérieure de la téte fossile ressemble d’avantage & celle d’une Emyde qu’ a celle d’une Chelonée, surtout par le peu de largeur de l’intervalle des yeux.’ Now the most striking difference between the temporal bony vault of the Hmys expansa and that of any known species of Chelone, is seen in the diminutive size of the post-frontals in this exceptional case among the Hmydes, as contrasted with their large size and actual extension over the temporal fossze in the Che/ones :—and this difference is accompanied by a proportional diminution in the breadth of the parietals in the true marine turtles. * Organic Remains, vol. ili, p. 268, pl. xviii, fig. 2. + Ossemens Fossiles, tom. v, pt. ii, p. 235. t Tom. cit., pl. xv, figs. 14-15. § Tom. cit., p. 235. bo 10 FOSSIL REPTILIA OF THE LONDON CLAY. But the figure in Parkinson’s work gives clearly the latter character ; whence also we may infer that it agreed more with the Che/ozes also in the size of the postfrontals ; although the anatomy of the skull is too obscurely delineated to demonstrate this fact. The following important affinities are, however, unquestionably indicated in Parkinson’s figure :—/irst, the large size of the orbits, which are nearly six times greater than those of the Himys expansa; secondly, their more posterior and lateral position; and f/irdly, the greater breadth of the mterorbital space: in all which characters the Sheppey fossil closely resembles the true Chelones, and differs from the only known species of ys (Podocnemys) expansa, in which the temporal openings are protected by a bony roof. That fresh-water tortoises have left their bony cuirasses in the Sheppey clay, will be subsequently shown; but the evidence of the genus Hmys, adduced by Cuvier, is incompetent to prove their existence; and, it may be affirmed, that of the fossils cited by the founder of Paleontology, some, with great probability, and others with certainty, are referable to the marine genus, Che/one. Without further discussing the question as regards these evidences, I shall proceed to describe the specimens from Sheppey which I have myself had the opportunity of examining; and shall commence with those which belong undoubtedly to the marine family. CHELONE BREVICEPS. Owen. Tabule I and II. Proceedings of the Geological Society, December 1, 1841; Report on British Fossil Reptiles, Trans. British Association, 1841, p. 178. Syn. Emys Parxrnsonu. J. E. Gray. — ope Sueppey. H.v. Meyer (?). CHELONE antiaua. Kenig (?). The first of the Chelonites, which led me to the recognition of this species, was a nearly perfect cranium from Sheppey (Tab. I, figs. 1—4), wanting only the occipital spine, and presenting a strong and uninterrupted roof, extended posteriorly from the parietal spine on each side (7, 7), over the temporal openings to the mastoids (8, 8); and formed anteriorly by a great development of the posterior frontals (22). This unequivocal testimony of the marine genus of the fossil, is accompanied by similar evidence afforded by the large size and lateral aspect of the orbits, the posterior boundary of which extends beyond the anterior margin of the parietals; and by the absence of the deep emargination which separates the superior maxillary from the tympanic bone in the fresh-water tortoises, and especially in the Podocnemys expansa. In general form, the skull of the present species of Sheppey Che/one resembles that of the Chelone mydas, BRONGN.: but it is relatively broader; the prefrontals (14) are CHELONIA. 11 less sloping, and the anterior part of the head is more vertically truncate. The orbits are relatively larger, and extend nearer to the tympanic cavity. The frontals (11) enter into the formation of the orbits in rather a larger proportion than in Chelone mydas. In the Chelone caouanna™ they are wholly excluded from the orbits. The trefoil shape of the occipital tubercle is well marked (fig. 4) ; the depression in the basioccipital, bounded by the angular pterygoid ridges, is as deep as in most true turtles (fig. 3); the lateral borders of the expanded parietals are united by a straight suture along a great proportion of their extent to the large postfrontals. These proportions are reversed in the Podocnemys expansa, in which the similarly expanded plate of the parietals is, chiefly united laterally with the squamosal and tympanic bones. In other fresh-water tortoises the parietal plate in question does not exist. The same evidence of the affinity of the Sheppey Chelonite in question to the marine turtles, is afforded by the base of the skull (fig. 3); the basioccipital (1) is deeply excavated; the processes of the pterygoids (24), which extend to the tympanic pedicles, are hollowed out lengthwise: the palatal processes of the maxillary and palatine bones are continued backwards to the extent which characterises the existing Chelones; and the posterior or internal opening of the nasal passages, is, in a pro- portional degree, carried further back in the mouth. The lower opening of the zygomatic spaces is wider in the present Sheppey Chelonite, than in Podocnemys expansa. The external surface of the cranial bones in the fossil is roughened by small irregular ridges, depressions, and vascular foramina, which give it a wrinkled or shagreen-like character. The following are dimensions of the specimen described : Inches. Lines. Length of cranium from the occipital condyle : : . 2 9 Breadth of cranium across the malars (26) —. : : 0 2 7 Antero-posterior diameter of orbit . 6 ; é : 1 0 The lower jaw, which is preserved in the present fossil, hkewise exhibits two characters of the marine turtles; the dentary piece (32), e. g. forms a larger proportion of the lower jaw than in the land or fresh-water tortoises. The joint of the rami is completely obliterated at the symphysis, which is not longer or larger than in Chelone mydas. The species represented by this fossil, which is preserved in the British Museum, and by a very similar one in the Hunterian Collection, is selected for the first of the Eocene Chelonians to be described in the present Monograph, because it is one of the few with which the characters of the carapace and plastron can with certainty be associated with those of the cranium. * Ossem. Fossiles, tom. y, pt. ui, pl. xi, fig. 2. 12 FOSSIL REPTILIA OF THE LONDON CLAY. In the rich collection of Sheppey fossils, belonging to J. 8. Bowerbank, Esq. F.R.S. there is a beautiful Chelonite (Tab. II, figs. 1, 2) including the carapace, plastron, and the cranium, which is bent down upon the fore part of the plastron; and which, though mutilated, displays sufficient characters to establish its specific identity with the skull of the Chelone breviceps just described. Both the carapace and plastron present the same finely rugous surface externally as the cranium; in which character we may perceive a slight indication of affinity with the genus Zrzonyz. The carapace (T. II, fig. 1) is long, narrow, ovate, widest at its anterior half, and tapering towards a point posteriorly; it is not regularly conver, but slopes away, like the roof of a house, from the median line (fig. 3), resembling, in this respect, and its general depression, the carapace of the turtle Chelone mydas. There are preserved the nuchal plate (fig. 1, c/) with ten of the neural plates (~1—vz10), only the eleventh and pygal plates being wanting. The eight pairs of costal plates (y/i—wp/s) are also present, with sufficient of the narrower tooth-like extremities of the six anterior pairs of ribs, to determine the marine character of the fossil, which is indicated by its general form.* The nuchal plate (fig. 6, c/) is of a transversely oblong form, with the anterior margin gently concave. Its antero-posterior diameter, or length, is ten lines; its transverse diameter, or breadth, is two inches. The lateral margins are bounded by two lines meeting ata slight angle; to the anterior one, the first of the marginal plates, mi, is attached; the posterior line bounds part of the vacant interspace between the first costal plate (/i), and the anterior marginal plate. The presence of this plate would prove for the genus Chelone as against Zrionyx, were the characters of the cranium, the impressions of the vertebral scutes, and the sternum wanting. The nuchal plate in the Himydes is hexagonal, and nearly as long as it is broad. The Chelonite from the tertiary beds near Brussels, figured by Cuvier,t has the nuchal plate of nearly the same form as the present specimen from Sheppey. The neural plates in the Chelone breviceps are as narrow as in the Che/ones generally; and as in the Brussels Chelonite above cited. The first neural plate (si, fig. 1) is four-sided; the rest, to the eighth (ss), are hexagons of a more regular figure than in the existing Cielones, and are articulated to more equal shares of the contiguous alternate costal plates (/1—p/s). The first costal plate (/1) is directed more outwards, does not incline backwards, as in recent C/elones, and its anterior angle is less truncated than in them. (See Howwlseps 3.) The length of the second costal plate (p/2) is one inch, nine lines; more than half of the narrow terminal extremity of the connate rib is preserved; the proportions of * In an Hmys with a carapace seven inches in length, the corresponding extremities of the ribs would have been united together by the laterally-extended ossification. + Ossemens Fossiles, tom. y, pt. 2, pl. xv, fig. 16. CHELONIA. 13 the remaining costal plates correspond with those of the Chelone mydas, and Chel. caouanna. The last pair of costal plates (p/s) articulates with the eighth, ninth, and tenth neural plates, but does not overlap or supersede any.of them. Not any of the costal plates articulate with those of the opposite side, so as to interrupt the series of vertebral plates, as in the carapace of the Chelone caouanna (fig. 1, p. 3),as in Mr. Crowe’s Sheppey Chelonite, figured by Cuvier (tom. cit. pl. xy, fig. 12); and as is shown in the view of the concave surface of the Brussels species (tom. cit. pl. xv, fig. 16). The ninth neural plate (fig. 1, s9) is the narrowest, as in the Chelones, and as in the Brussels Chelonite, figured by Cuvier, in loc. cit. pl. xii, fig. 8, instead of being suddenly expanded, as in most Hinydes. The tenth neural plate (s10) expands to a breadth equal with its length; the eleventh and pygal plates, as already observed, are wanting in the fossil. The vertebral or median ends of the costal plates present a modification of form, corresponding with that of the interspaces of the neural plates to which they are articulated. Only the first pair (p/1) present that form which characterises all but the last pair in the existing Chelones, and in the Brussels Chelonite ; viz., a straight line with the posterior angle cut off; the rest being terminated by two nearly equal oblique lines, meeting at an open angle, as shown in Tab. II, fig. 1, p/2—yp/7. This character would serve to distinguish the Chelone breviceps, if only a portion of the carapace, including the vertebral extremity of a rib, were preserved. The free extremities of the ribs are thicker in proportion to the costal plates, than in the Chelone caouanna, or the Chel. mydas ; and more resemble, in this respect, those of the Chel. imbricata, the species characterised by the size and beauty of the horny scutes, commonly called “ tortoise-shell.” More or less complete impressions of the five horny vertebral scutes (vi—vs), and of four costal scutes on each side of the vertebral ones, show the forms and proportions of these characteristic parts, and especially of the median series, notwithstanding they were among the soluble and perishable elements of this ancient turtle of the Thames. The hexagonal vertebral scutes are characterised by the near equality of their sides, and the angle of about 100°, at which the two outer sides meet. The anterior border of the first vertebral scute, v', has crossed and impressed the nuchal plate, c/, near its anterior border; this scute has covered the rest of the nuchal plate, and more than half of the first neural plate. The second vertebral scute, 2”, includes the rest of the first neural plate, the whole of the second, and almost the whole of the third neural plate. The third vertebral scute, 0’, includes the hind border of the third neural plate, with the whole of the fourth and fifth neural plates. The fourth vertebral scute includes the sixth and seventh, and very nearly the whole of the eighth neural plates, and the outer angles of this scute terminate over the suture between the sixth and seventh costal plates. 14 FOSSIL REPTILIA OF THE LONDON CLAY. The plastron of the Chelone breviceps (Tab. I, fig. 2), although more ossified than in existing Chelones, yet presents all the essential characters of that genus. There is a central vacuity left between the hyosternals (4s) and hyposternals (ps); but these bones differ from those of the young Amys in the long pointed processes which radiate from the two anterior angles of the hyosternals (4s), and the two posterior angles of the hyposternals (js). | The xiphisternals (vs) have the slender elongated form, and oblique union by reciprocal gomphosis with the hyposternals (4s), which is characteristic of the genus Chelone. The posterior extremity of the right episternal (es) presents the equally characteristic, slender pointed form. With these proofs of the modification of the plastron of the present fossil according to the peculiar type of the marine Chelones, there is evidence, however, that it differs from the known existing species in the more extensive ossification of the component pieces: thus the pointed rays of bone extend from a greater proportion of the margins of the hyosternals and hyposternals; and the intervening margins do not present the straight line at right angles to the radiated processes. In the Chelone mydas, and Chel. caouanna (fig. 3, p. 4), for example, one half of the external margin of the hyosternal and hyposternal, where they are contiguous, are straight, and intervene between the radiated processes, which are developed from the remaining halves, while in the Chelone breviceps, about a sixth part only of the corresponding external margins are similarly free, and there form the bottom, not of an angular, but a semicircular interspace. The radiated processes from the inner margins of the hyosternals and hyposternals, are characterised in the Chelone breviceps by similar modifications, but their origin is rather less extensive ; they terminate in eight or nine rays, shorter, and with intervening angles more equal than in existing Chelones. ‘The xiphisternal piece, zs, receives in a notch the outermost ray or spine of the inner radiated process of the hyposternal, as in the Chelones, and is not joined by a transverse suture, as in the Lmydes, whether young or old. Subjoined are dimensions of the plastron of Mr. Bowerbank’s fossil : Inches. Lines. Shortest longitudinal diameter of hyosternal and hyposternal pieces 2 Transverse diameter of ditto Total length of plastron Oy = ob on om The bones of the scapular arch, especially the coracoid, Cuvier has shown to afford © distinctive characters of the natural families of the Chelonia ; but the Eocene Chelonites described by Cuvier, did not yield him this opportunity of thus testing their affinities. In the Chelone breviceps here described, the left coracoid (52, fig. 2) is preserved in nearly its natural position; it is long, slender, symmetrical; cylindrical near its humeral CHELONIA. 15 extremity ; flattened, and gradually expanded from its humeral third, to its sternal end, which is relatively somewhat broader than in the Chelone mydas and Chelone caouanna. Inch. Lines. Its lengthis . , 6 : : 1 6 Breadth of sternal end : A : 0 7 The characters thus afforded by the cranium, carapace, plastron, and by one of the bones of the anterior extremity, prove the present Sheppey fossil to belong to a true sea turtle; and at the same time most clearly establish its distinction from the known existing species of Chelone. On account of the shortness of the skull, especially of the facial part and of that which intervenes between the orbit and ear, compared with the breadth of the skull across the mastoids, I have proposed to name this extinct species, Chelone breviceps.* By the characteristic shape of the median extremities of the costal plates of the carapace, I have been able to determine some fragmentary Chelonites which have afforded better ideas of the size of the species represented by Mr. Bowerbank’s more complete but immature specimen of Chelone breviceps. A portion of the carapace of the Chelone breviceps, including the fourth, fifth, sixth, and part of the third and seventh neural plates, with a considerable proportion of the third, fourth, fifth, and sixth costal plates, is preserved in the museum of Mr. Robertson, of Chatham. The characters of the rugous surface of these bones, and of the equal- sided angles by which the costal plates articulate with the neural plates, do both, and especially the latter, point out the species to which the present fragment belongs. It has formed part of an individual double the size of the specimen above described, and figured from Mr. Bowerbank’s collection, and therefore it had a carapace sixteen inches in length. Although the costal plates have been continued further along the ribs than in the younger example, the more complete state of the sixth rib, in Mr. Robertson’s specimen, shows that they retained their longitudinally-striated, tooth-like extremities, which, in the sixth rib, is two thirds of an inch in length; the length of the expanded part being four inches, and its breadth one inch nine lines. The internally prominent part of the rib is much less developed than in Chelone planimentum, and Chelone crassicostata, afterwards to be described. The right hyosternals and hyposternals are present, and they likewise preserve the character of the Chel. dreviceps in their rugous surface and minor breadth, as compared with those parts in the Chelone longiceps, the extinct species next to be described. Besides the specimens above described, on which the present extinct species of turtle * Proceedings of Geological Society, December 1, 1841, p. 570. Report on British Fossil Reptiles, Trans. Brit. Association, 1841, p. 178. 16 FOSSIL REPTILIA OF THE LONDON CLAY. has been established, remains of the Chelone breviceps are preserved in the Hunterian Museum, and in that of my esteemed friend and coadjutor, Professor Bell, $.R.S. I know no other locality of the species than that of Sheppey, in Kent. CHELONE LONGICEPS. Owen. Tab. III, IV, and V. Proceedings of Geological Society of London, December 1, 1841, p. 572. Report on British Fossil Reptilia, Trans. British Association, 1841, p. 177. The second species of Chelone, from the Eocene clay at Sheppey, which I originally - recognised and defined by the fossil skull, Tab. HI, differs more from those of existing Chelones by the regular tapering of that part into a prolonged pointed muzzle, than does the Chelone breviceps by its short and anteriorly-truncated cranium. The surface of the cranial bones is smoother than in the Chel. breviceps ; whilst their proportions and relations prove the marine character of the present fossil as strongly as in-that species. The orbits (Tab. III, figs. 1 and 2, 0,) are large; the temporal fossze (ib. fig. 3, 7) are covered principally by the posterior frontals (fig. 2, 12); and the osseous shield completed by the parietals (7), and mastoids (8), overhangs the tympanic (28), ex- occipital (2), and paroccipital (4) bones. The compressed spine (3) of the occiput is the only part that projects further backwards. The palatal and nasal regions of the skull afford further evidence of the affinities of the present Sheppey Chelonite to the true turtles. The bony palate (fig. 3) presents, in an exaggerated degree, the great extent from the intermaxillary bones to the posterior nasal aperture which characterises the genus Chelone ; and it is not perforated, as in the soft turtles (Zrzonyx), by an anterior palatal foramen. The extent of the bony palate is relatively greater than in the Chelone mydas, and the trenchant alveolar ridge is less deep; the groove for the reception of that of the lower jaw is shallower than in the Che/one mydas, or the extinct Chel. breviceps, arising from the absence of the internal alveolar ridge, in which respect the Chel. longiceps resembles the Chel. caretta. The Chelone longiceps is distinguished from all known existing Chelones by the proximity of the palatal vomer (13, fig. 3), to the basisphenoid (5), and by the depth of the groove of the pterygoid bones (24), and in both these characters in a still greater degree from the Trionyxes; to which, however, it approaches in the elongated and pointed form of the muzzle, and the trenchant character of the alveolar margin of the jaws. The following are dimensions of the skull described : Inches. Lines. Length of the skull . : 2 ; c : : : 4 0 Breadth of ditto across the zygomat 2 6 Anterc-posterior diameter of orbit . 1 2 CHELONIA. 17 In a second example of the skull of Chelone longiceps, two of the middle neural plates, and the corresponding costal plates of the right side, portions of vertebre, with the right xiphisternal piece, humerus and femur, are cemented together, and to the cranium by the petrified clay. (T. IV, fig. 1.) The neural plates (s2, s3) are flat and smooth; the entire one measures one inch two lines in length, and nine lines across its broad anterior part :—this receives the convex posterior extremity of the preceding plate in a corresponding notch. A small proportion, about one sixth, of the anterior part of the external margin, joins the second costal plate (p/2); the remaining five sixths of the outer margin forms the suture for the vertebral end of the third costal plate (p/3). In this respect, the Chel. longiceps resembles the existing Chelones ; and differs, as well as in the smooth and flattened surface of the vertebral plates, from the Chelone breviceps. The length of the third costal plate, in the fragmentary example here described, is three inches ; the impression of the commencement of the narrow portion, formed by the extremity of the coalesced rib, is preserved. The marginal indentations of the vertebral scutes are not half a line in breadth. The transverse impression between the first and second vertebral scute crosses the first neural plate, nine lines from its posterior extremity ; the second neural plate is free, as in other Chelones, from any impression, being wholly covered by the second vertebral scute. The expanded ribs are convex at the under part, slightly concave at the upper part in the direction of the axis of the shell; they slope very gently from the plane of the neural plates, about half an inch, for example, in an extent of three inches; thus indicating a very depressed form of carapace. The xiphisternal bone (as), like that of Chel. breviceps, is relatively broader than in the existing turtles, and both the internal and external margins of its posterior half are slightly toothed. A part of the notch by which it was attached to the hyposternal remains upon the broken anterior extremity of the bone. It measures one inch two lines across its broadest part ; its length seems to have been three inches and a half. The humerus presents the usual characters of that of the Chelones ; its length is two inches three lines; its breadth across the large tuberosities ten lines. The radius and ulna extend in this Chelonite from beneath the carapace into the right orbit; the radius is one inch and a half in length; the ulna one inch, three lines in length; portions of vertebrze adhere also to the mass, the state of which indicates that the animal had been buried in the clay before the parts of the skeleton had been wholly disarticulated by putrefaction. A mass of Sheppey clay-stone supporting the ninth-and tenth neural plates, and the expanded portions of the sixth, seventh, and eighth costal plates of the right side, exhibits the characters of the marine turtles in the great relative expansion of the 3 18 FOSSIL REPTILIA OF THE LONDON CLAY. tenth neural plate ; and the tooth-like continuation of the rib from the posterior angle of the eighth costal plate (pls, T. IV, fig. 2). These portions of the carapace, from their smooth surface, the impressions of the horny scutes, the form of the vertebral ends, and the concavity of the upper surface of the costal plates, evidently belong to the same species as the fossil last described. A similar mass of Sheppey clay-stone, in Mr. Lowe’s collection, supports a larger proportion of the hinder part of the carapace, including the sixth, seventh, eighth, ninth, and tenth neural plates, part of the fifth neural plate, more or less of the last four pairs of costal plates, with the impressions of the third and fourth ribs of the _ right side; the impression of apparently the whole of the free, slender, termination of the third rib is preserved, and also that of the fifth rb, confirming the generic characters indicated by the skull. The smooth outer surface of the bones of the carapace, the forms of the neural plates, and the concomitant modification of the commencement of the costal plates articulated therewith, concur to establish the specific distinction from the Chelone breviceps, and indicate the specimen to belong to the present species, Chelone longiceps. The seventh, eighth, and ninth neural plates progressively decrease in size; and the ninth presents a simple, quadrangular, oblong form ; the tenth neural plate suddenly expands, and has apparently a triangular form, but its posterior border is incomplete. The indications of the comparative flatness of the carapace of the Chelone longiceps, (in this respect, as mm the elongated and pointed form of the skull, approaching the genus Zrionyx,) which were derived from an examination of the foregoing fragments, and particularly of the portion preserved with the cranium on which the species is founded, are fully confirmed by the almost entire carapace which, subsequently to the publication of my ‘Report on British Fossil Reptiles,’ where the present species is first noticed, I have had the opportunity of examining in the collection of Mr. Bowerbank. This carapace, as compared with that of the Chelone breviceps in the same collection, presents the followmg differences :—it is much broader and flatter. The neural plates are relatively broader; the lateral angle from which the intercostal suture is continued, is much nearer the anterior margin of the plate—the Chelone longiceps, in this respect, resembling the existing species of turtle (see fig. 1, p. 3). The costal plates are relatively longer; they are slightly concave transversely to their axis on their upper surface, while in Chel. breviceps they are flat. The external surface of the whole carapace is smoother; and although it is as depressed as in most turtles, it is more regularly convex; not sloping away by two nearly plane surfaces from the median longitudinal ridge of the carapace. The following minor differences may be noticed in the two Sheppey Chelonites : the nuchal plate of the Chel. longiceps (Tab. V, fig. 1, ch) is more convex at its middle part, and sends backwards a short emarginate process to join the first neural CHELONTA. 19 plate (s1) ; in which it resembles the Chel. mydas. Both posterior angles of the first neural plates are produced, and truncate to articulate with the second pair of costal plates; and the second neural plate is quadrangular. In a portion of another carapace of the Chelone longiceps the second neural plate (s2) is pentangular, the right anterior corner being produced, and truncate to join with the first costal plate of the right side; the left posterior corner of the first neural plate (s1) bemg produced, and truncate, to articulate with the second costal plate of the left side. This structure I believe, however, to be an individual variety. But the characters of the species are exemplified in more constant modifications of the carapace. The succeeding neural plates to the seventh inclusive (s3—s7) are hexagonal, with the anterior lateral border much shorter than the posterior lateral border, as in Chelone mydas, and not of equal extent, as in Chelone breviceps; they become more equal in the seventh and eighth neural plates, which also decrease in size; the nth plate (s9) is very small, quadrangular, and oblong, as in Mr. Lowe’s fragment. Only a small portion of the tenth neural plate is preserved in Mr. Bowerbank’s beautiful specimen. The impressions of the horny scutes are deeper, and the lines which bound the sides of the vertebral scutes (vi—v4) meet at a much more open angle than in the Chel. breviceps, in which the vertebral scutes have the more regular hexagonal form of those of the Chel. mydas. Their relations to the neural plates are nearly the same as in Chel. breviceps. The plastron (Tab. V, fig. 2) is more remarkable than that of the Chel. dreviceps for the extent of its ossification ; the central cartilaginous space bemg reduced to an elliptical or subquadrangular fissure. The four large middle pieces /yosternals (hs) and /yposternals (ps), have their transverse extent relatively much greater as compared with their antero-posterior extent, than in the Chel. breviceps; and this might be expected, in conformity with the broad character of the bony cuirass indicated by the carapace. The median margins of the /yosfernals (hs) are developed in short toothed processes, along their anterior three fourths; the median margins of the Ayposternals (ps) have the same structure along nearly their whole extent ; the intermediate space between the smooth or edentate margins of the opposite bone is ten lines; the expanded end of the long coracoid (52) is seen projecting into this space. The xiphisternals (vs) are relatively broader than in Chel. dreviceps, or in any of the existing turtles; and are united together, or touch each other, by the toothed processes developed from the whole of their median margins. ‘The entosternal piece is broad, flat on its under surface, and is likewise dentated at its sides. The outer surface of each half of the plastron inclines, as in the Chelone mydas, towards a submedian longitudinal ridge. The breadth of the plastron, in the specimen figured (fig. 2), along the median suture, uniting the hyosternals and hyposternals, is six inches: the narrowest antero- posterior diameter of the conjoined hyosternals and hyposternals is two inches nine lines. 20 FOSSIL REPTILIA OF THE LONDON CLAY. The breadth of the plastron, at the junction of the xiphisternals with the hyposternals, is two inches six lines. The posterior part of the cranium is preserved in Mr. Bowerbank’s specimen (fig. 1), withdrawn beneath the anterior part of the carapace; the fracture shows the osseous shield covering the temporal fossee; and the pterygoids remain, exhibiting the deep groove that runs along their under part. It is most satisfactory to have found that the two distinct species of the genus Chelone, determined, in the first instance, by the skulls only, should thus have been confirmed by the subsequent comparison of their bony cuirasses ; and that the specific differences, manifested by the cuirasses, should be proved by good evidence to be characteristic of the two species founded on the skulls. Thus the portion of the skull preserved with the carapace first described (Tab. II, figs. 2 and 3), served to identify that fossil with the more perfect skull of the Chelone breviceps (Tab. I), by which the species was first indicated. And, again, the portion of the carapace adhering to the perfect skull of the Chelone longiceps (Tab. IV, fig. 1) equally served to connect with it the nearly complete osseous buckler (Tab. V, fig. 1), which, otherwise, from the very small fragment of the skull remaining attached to it, could only have been assigned conjecturally to the Chel. longiceps ; an approximation which would have been the more hazardous, since the Chelone breviceps and Chelone longiceps are not the only turtles which swam those ancient seas that received the enormous argillaceous deposits of which the Isle of Sheppey forms a part. CHELONE LATISCUTATA. Owen. Tab. VI. Proceedings of the Geological Society of London, December 1, 1841, p. 574. Report on British Fossil Reptiles, Trans. British Association, 1841, p. 179. A considerable portion, measuring three inches in length, of the bony cuirass of a young turtle from Sheppey, including the first to the sixth neural plates (T. VI, fig. 1, s1—s6), with the corresponding pairs of costal plates (y/i—p/6), and the hyosternal (fig. 2, 4s) and hyposternal (ps) elements of the plastron, most resembles that of the Chelone longiceps in the form of the carapace, and especially in the great transverse extent of the above-named parts of the plastron: it differs, however, from the Chel. longiceps, and the other known fossil Chelonites, in the greater relative breadth of the vertebral scutes (v2, v3), which are nearly twice as broad as they are long. The central vacuity of the plastron is subcircular; and, as might be expected, from the apparent nonage of the specimen, is wider than in the Chel. longiceps ; but the toothed processes given off from the inner margin of both hyosternals and hyposternals are small, sub-equal, regular in their direction, and thus resemble those of the Chel. longiceps ; the slender point of the episternal (s) is preserved in the interspace between CHELONIA. 21 the hyosternals. Both hyosternals (Zs) and hyposternals (ps) are slightly bent upon a median longitudinal prominence of their under surfaces. The length of the third costal plate (p/3) is one inch seven lines; its antero- posterior diameter or breadth, six lines: in the form of the vertebral extremities of the costal plates, and of the neural plates to which they are articulated, the present fossil resembles the Chel. longiceps ; but the fifth neural plate is more convex, and is crossed by the impression dividing the third vertebral scute (v3) from the fourth, which impression crosses the suture between the fifth and sixth neural plates in both Chelone longiceps and Chelone breviceps. Whether, m the progressive change of form, which the vertebral scutes may have undergone in the growth of this young turtle, as during the growth of the young loggerhead turtle (Chelone caouanna), by an increase of length, without corresponding increase of breadth, the impression between the third and forth vertebral scute, might also retrograde to the interval between the fifth and sixth neural plates, I am uncertain, having only had the opportunity of comparing the scutes of the young and old loggerhead turtles, not the skeletons. The change in the lateral angles of the vertebral scutes, resulting from the elongation of the scutes themselves, in the loggerhead, would be similar to that in the Chelone longiceps, as compared with the Chel. latiscutata, on the hypothesis that the latter is the young of the former ; but in my present uncertainty I prefer to indicate the specimen in question, by the definite name proposed in my original Memoir; its description as a distinct species being more likely to attract the attention of Collectors to similar specimens, and to enable them to identify such. Figure 3, T. VI, gives the degree of convexity of the carapace, and the double curve of the plastron produced by the prominence of the principal heemapophyses /s and ps. The left scapular arch (51) is exposed in this view- CHELONE CONVEXA. Owen. Tab. VII. Proceedings of the Geological Society of London, December 1, 1841, p. 575. Report on British Fossil Reptiles, Trans. British Association, 1841, p. 178. The fourth species of Chelone, indicated by a nearly complete cuirass, from Sheppey, holds a somewhat intermediate position between the Chelone breviceps and the Chelone longiceps ; the carapace being narrower, and more convex than that of Chel. longiceps ; broader and with a more regular transverse curvature than in the Chelone breviceps. Although the specimen is equal in size to either of the two with which it is here compared, the costal plates hold an intermediate length, which shows that this character is not due to a difference depending upon age. The fossil in question includes the first to the eighth neural plate inclusive ; the first plate (s1) expands behind, and both posterior angles are truncated to articulate with the second costal plates (p/2). The second neural plate (s2) is quadrate, half as long again as broad, and the second pair of costal plates articulate with this, as 22 FOSSIL REPTILIA OF THE LONDON CLAY. well as with the first and third plates, as in the Chel. longiceps (Tab. V, fig. 1). The tooth-like extremity of the connate rib is preserved on the right side. The fourth costal plate (p/4) is two inches four lines im length, nine lines in breadth ; the angle at which the expanded part contracts to the extremity of the connate rib is well shown on the right side. The third to the eighth neural plates expand anteriorly, and have the anterior angles cut off to articulate with the costal plates in advance ; they diminish m size very gradually, and the antero-lateral borders, formed by the above-named truncated angles, do not increase in length as in the corresponding plates in the Chelone longiceps. The vertebral scutes (v2, v3, v4) resemble more in form those of the Chel. longiceps than of Chel. breviceps ; but, notwithstanding that the whole carapace is narrower than in Chel. longiceps, the vertebral scutes are broader ; and the lines which converge to the lateral angle have a more marked sigmoid curvature. i Chel. convexa. Chel. longiceps. 3 Inches. Lines Inches. Lines. The length of the second vertebral scute is 1 8 1 8 Breadth 5 : : set: A 2 6 2 2 The two succeeding scutes (v3 and v4) more rapidly diminish in size than in either the Chel. breviceps or longiceps, and the transverse impression between the third and fourth vertebral scute crosses the lower third of the fifth neural plate, as in Chelone ‘latiscutata. All the scutes have left deeper and rather wider impressions than in the preceding species. The second to the fifth costal plates inclusive, are more equal in length than in the existing Chelone mydas or Chel. caowanna, and in this character the present species more resembles the Chel. imbricata. The distinction of the present from the previously described fossils, already manifested in the structure of the carapace and the form of the vertebral scutes, is more strikingly established in that of the plastron (Tab. VII, fig. 2), which, in its defective ossification, resembles the same part in the existing species of Chelone. All the bones, but especially the xiphisternals (vs), are more convex on their outer surface than in other turtles, recent or fossil. The central vacuity is greater than in any of the above-described fossil species. The internal rays of the hyosternals come off from the anterior half of their inner border, and are divided into two groups: the lower consisting of two short and strong teeth, projecting inwards towards the extremity of the entosternal (s) ; while the rest extend forwards along the inner side of episternals (es). The same character may be observed in the corresponding processes of the Ayposternals (ps), which are limited to the posterior half of their inner border. The external radiated process of the hyosternals (/s) arises from a larger proportion of the outer margin, than in the Chel. mydas ; but from a somewhat less proportion than in Chel. breviceps. CHELONIA. 23 The external process of the hyposternal (ys) is relatively much narrower than in the Chel. breviceps (T. I, fig. 2), and, @ fortiori, than in Chel. longiceps (Tab. V, fig. 2). The straight transverse suture by which the hyosternals and hyposternals of the same side are joined together, is much shorter than in the other fossil Chelones ; and is similar in extent to that in Chel. mydas ; but the following differences present themselves in the plastron of the Chelone convera, as compared with that of the Chelone mydas. The median margin of the hyosternals forms a gentle curve, not an angle: that of the hyposternals is likewise curved, but with a slight notch. The longitudinal ridge on the external surface is nearer the median margin of the Ayosternals and hyposternals and is less marked than in the Chelone longiceps ; especially in the hyposternals, which are characterised by a smooth concavity in the middle of their outer surface. The suture between the /yosfernals and hyposternals is nearer to the external, transverse, radiated process of the hyposternals. The median vacuity of the sternal apparatus is elliptical in the Chel. convera, but square in the Chel. mydas. The characteristic lanceolate form of the episternal bone (s) in the genus Cheloze, is well seen in the present fossil. The entosternal element of the plastron is sub- circular, or lozenge-shaped ; and generally broader than it is long in the Emydians. The true marine character of the present Sheppey Chelonite, so well given in the carapace and plastron, is likewise satisfactorily shown in the small relative size of the entire femur (65) which is preserved on the left side, attached by the matrix to the left xiphisternal. It presents the usual form, and slight sigmoid flexure, characteristic of the Chelones ; it measures one inch in length. In an Hinys of the same size, the femur, besides its greater bend, is one inch and a half in length. A Chelonian cranium from Sheppey, two inches five lines in length, in the museum of Professor Bell (T. VI, fig. 4), and a second of the same species from the same locality, two inches nine lines in length, in the museum of Fred. Dixon, Esq., F.G.S., of Worthing, belong to the same species, and differ from the cranium of the Chelone Greviceps, in the more pointed form of the muzzle, and the less rugose character of the outer surface of the bones; they equally differ from the Chelone longiceps in the less produced, and less acute muzzle, and the more rugose surface of the bones. The parietals (7) are bounded anteriorly by a semicircular line, not by a semioval one, as in Chel. longiceps, or by an angular one, as in Chel. breviceps. The frontals (11) enter into the formation of the orbits, as in both the foregomg species. The orbits are subcircular, as in Chel. longiceps, not subrhomboidal with the angle rounded off, as in Chel. breviceps. The postfrontals (12) are large, and form a slight projection at the back part of the supraorbital ridge.. The tympanic cavity is larger in proportion than in the Cheloue longiceps. The palate is traversed by a deep median, longitudinal groove, between which and the shallower grooves on the inner sides of the alveolar borders, are two well-marked, diverging, longitudinal prominences. The 24. FOSSIL REPTILIA OF THE LONDON CLAY. bony palate is longer than in Chelone breviceps, shorter than in Chel. longiceps. The symphysis of the lower jaw (T. VII, fig. 3) is longer or deeper than in the Chelone breviceps, but is convex below from side to side, and not flattened as in the Chelone planimentum. All the specimens of Chelone convexa, which I have been able to determine, are from the London clay of Sheppey. CHELONE SUBCRISTATA. Owen. ‘Tab. VIII. Proceedings of the Geological Society of London, December 1, 1841, p. 576. Report on British Fossil Reptiles, Trans. British Association, 1841, p. 179. The fifth species of Chelone from Sheppey, distinguishable by the characters of its carapace, approaches more nearly to the Chelone caouanna in the form of the vertebral scutes (vl—v4), which are narrower in proportion to their length, than im any of the previously described species; but the Chelone subcristata is more conspicuously distinct by the form of the fifth and seventh neural plates (ss, s7), each of which supports a short, sharp, longitudinal crest; a similar crest is developed from the contiguous ends of the second and third neural plates (s2, s3); the middle and posterior part of the nuchal plate (c/) is raised into a convexity, as in the Chel. longiceps ; but not into a crest. The keeled structure of the above-cited neural plates is more marked than in the third and fifth neural plates of Chelone mydas, which are raised into a longitudinal ridge. The neural plates in the present carapace have the ordinary, narrow, elongated form of those in the true Chelones. The nuchal plate (c/) has the middle of its hinder border produced backwards, instead of bemg emarginate, as in the Chel. breviceps (T. I, fig. 1, cf). The first neural plate in the Chelone subcristata (T. VIII, s1) resembles that in the Chelone convexa, but is narrower in proportion to its length; the second (s2) is also quadrangular, as in Chel. convexa, but is narrower; the third to the seventh likewise differ from those in Chel. convexa only by being narrower; but the eighth and ninth neural plates are relatively smaller than in any of the before-described fossils, and resemble those of existing Chelones. The expanded plate is more elevated, and is bent down-on each side, with the middle part forming an obtuse longitudinal ridge. A part of the contiguous portion of the first (y/1) and the second (p/2) costal plates are raised into a slight convex eminence on each side; the surface of the remaining pairs of ribs is flat in the axis of the body, but they are more convex transversely to that axis, and in the direction of their own length, than in the other Chelonites. The whole outer surface of the bones of the carapace is as smooth as in the Chel. longiceps and Chel. convewa. Length of carapace from the first to the eighth neural plate inclusive : Ch. subcristata. Ch. breviceps. Ch. longiceps. Ch. convexa. Inches Lines. Inches Limes. ~ Inches. Lines. Inches. Lines. 7 4 5 6 i) 9 5 8 CHELONIA. 25 The length of the present fossil carapace, to the tenth neural plate, inclusive, is nine inches. The breadth between the ends of the third costal plates, in a straight line, is six inches six lines. The succeeding costal plates more gradually decrease in breadth, than in the Chel. longiceps and Chel. convexa ; and the entire carapace more resembles in form that of the Chel. mydas, and Chel. caowanna. The epidermal scutes are defined by deep impressions, and as wide, relatively, as in the Chel. mydas and Chel. convera. The length of the second vertebral scute is two inches one line ; its breadth is two inches two lines; the length of the fourth vertebral scute is two inches three lines; and its breadth one inch eleven lines, and, at its posterior margin, only nine lines. This scute is narrower than in Chel. caouanna, or any of the previously described fossil species; the outer angles are less produced than in the Chelone caouanna. Sufficient of the plastron is exposed in the present fossil to show by its narrow elongated xiphisternals (vs), and by the wide and deep notch in the outer margin of the conjoined hyosternals and hyposternals (4s and ps), that it belongs to the marine Chelones. The xiphisternals are articulated to the hyposternals by the usual notch or gomphosis; they are straighter and more approximated than in the Chel. mydas and Chel. caouanna. The external emargination of the plastron between the hyosternals and hyposternals, differs from that of the recent turtles in being semicircular, instead of angular; the Chel. subcristata approaching, in this respect, to the Chel. breviceps. The shortest antero-posterior diameter of the conjoined hyosternals and hyposternals is two inches seven lines. The length of the xiphisternal is two inches six lines; the breadth of both, across their middle part, is one inch three lines. The name proposed for this species indicates its chief distinguishing character, viz., the median interrupted carina of the carapace, which may be presumed to have been more conspicuous in the horny plates of the recent animal, than in the supporting bones of the petrified carapace. CHELONE PLANIMENTUM. Owen. Tab. IX and X. Proceedings of the Geological Society of London, December, 1841, p. 576. Report on British Fossil Reptiles, Trans. British Association, 1841, p. 178. Syn. Cuetons Harvicensrs, Woodward (°). The skull of a large Che/one (T. TX) from the Eocene clay near Harwich, m Professor Sedgwick’s collection at Cambridge, resembles, in the pointed form of the muzzle, the Chel. longiceps of Sheppey; but differs in the greater convexity and breadth of the cranium (fig. 2); and the more abrupt declivity of its anterior contour (fig. 3), and from other Chelones by the broad expanse of the inferiorly-flattened symphysis mente (fig. 1). 4 26 FOSSIL REPTILIA OF THE LONDON CLAY. The osseous roof of the temporal fossee, and the share contributed to that roof by the postfrontals (T. IX, figs. 2 and 3, 12), distinguish the present, equally with the foregoing Chelonites, from the Lmys (Podocnemys) eapansa, and, a fortiori, from other genera and species of the fresh-water families (Hmydide and Trionicide). In the oblique position of the orbits (fig. 3, 0), and the diminished breadth of the interorbital space (fig. 2), the present Chelonite, however, approaches nearer to Zrvonyx and “mys than do the previously-described species. But the sides of the face converge more rapidly towards the muzzle. Its most marked and characteristic difference from all existing Chelones is shown by the greater antero-posterior extent, breadth, and flatness of the under part of the symphysis of the lower jaw, whence the specific name here given to the species. The posterior border of the symphysis is defined by a regular semicircular curve, and the rami of the jaw have completely coalesced. Since at present there is no means of identifying the well-marked species, of which the skullis here described, with the Chelonite figured in the frontispiece to Woodward's ‘Synoptical Table of British Organic Remains,’ and alluded to, without additional description or characters, as the Cheionia Harvicensis, in the additions to Mr. Gray’s ‘Synopsis Reptilium’ (p. 78, 1831); and since the extensive deposit of Eocene clay along the coast of Essex, like that at the mouth of the Thames, contains the relics of more than one species of ancient British turtles,* I prefer indicating the one here established by a name having reference to its peculiarly distinguishing character, rather than to associate arbitrarily the skull, which gives the true specific distinction, with the ill-defined carapace to which the vague name of //arvicensis has been applied; more especially as the fossil carapace to which the present skull more probably belongs, from the circumstance under which it was discovered, also presents well- marked, and readily-recognisable specific characters. This carapace (T. X) is also contained in the museum of Professor Sedgwick, and is understood to have formed part of the same individual turtle as the skull (T. IX) on which the species, Chel. planimentum, was founded. In general form this carapace differs from that of the existing Chelones, in being less contracted and pointed posteriorly than in the Chelone mydas and Chel caouanna, and more contracted posteriorly than in the Chel. imbricata. In the proportion which the pleurapophyses (true ribs), bear to the superimposed costal plates, (p/41—s) it resembles Chelone mydas, and Chelone caowanna, more than it does the Chel. imbricata. But the pleurapophyses are more prominent and distinct from the costal plates throughout their entire length, than in the Chel. mydas or Chel. caowanna, and present an obtuse angular ridge towards the cavity of the abdomen. The five posterior pairs of ribs of the carapace (p/4—yp/s) are preserved, with part * Sir C. Lyell alludes to the Chelonites of Harwich in his ‘ Elements of Geology ¥ ‘This formation is well seen in the neighbouring cliffs of Harwich, where the nodules contain many marine shells, and sometimes the bones of Turtles.” (Vol. ii, p. 337.) CHELONIA. o7 . of the first three on the left side, and one of the coracoids (52) showing the rather sudden and considerable expansion of its sternal or mesial half. The interval between the free extremities of most of the ribs, is about equal to twice and a half the breadth of each extremity; but the interval between the seventh (plz) and eighth (p/s) rib, measured, like the others, at the terminal border of the costal plates, is equal to thrice the breadth of the free part of the seventh rib. In this respect the Chelone planimentum resembles the Chel. mydas more than it does the Chelone caouanna, i which the interval between the free extremities of the seventh and eighth ribs is less than that between the sixth and seventh. The length of the costal plate of the fourth rib is twice that of the eighth rib, as in the Chelone caouanna ; in Chel. mydas it is more than twice as long; in Chel. imbricata it is only one third longer. The marginal pieces in the Chelone planimentum seem to have been narrow or slender in proportion to their length. The following admeasurements show that, in the large proportionate size of the head, the Chelone planimentum corresponds with the existing turtles : Inches. Lines. Length of the cranium . : : ; . : A 5 5 6 Depth of ditto. 0 é ; : : : : 0 4 0 Breadth of ditto . 3 3 5 : . . : 6 5 0 Length of the carapace . 5 : ‘ : 3 5 oe le 6 Greatest breadth of ditto : : 3 ; a i Haus ahs} 0 Tab. IX and X satisfactorily illustrate the characteristic forms and proportions of the unique specimen in the Cambridge Museum ; the carapace is figured of half the natural size. CHELONE CRASSICOSTATA. Owen. Tab. XI and XII. Trstupo Purana. Kénig. ‘Icones Sectiles,’ Pl. XVI, fig. 192? That the extinct species of Eocene turtles attained larger dimensions than those given above, is proved by a fossil skull from the Harwich clay, in the collection of Professor Bell, which gives the followig dimensions : Inches. Lines. Total length of the cranium . 5 : : 0 : 8 0 Its greatest breadth : ; : : ; 6 0 The antero-posterior extent of the symphysis menti . 3 0 The vertical diameter of the orbit . 1 9 do. do. of the nostri é ¢ 0 9 This skull differs from that of the Chelone planimentum in the minor depth of the maxillary bone below the orbit (compare T. IX, fig. 3, with T. XI, fig. 2, 21), in the more acute and attenuated muzzle; but especially in the minor breadth and the different configuration of the posterior margin of the symphysis of the lower jaw (compare T. IX, 28 FOSSIL REPTILIA OF THE LONDON CLAY. fig. 1, with T. XI, fig. 3). With regard to the comparative anatomy of the bones of the skull, and the pattern of the scutation of the upper surface of the cranium, I regret that the state of the specimen in Professor Bell’s collection does not permit the deduction of other distinctive characters which such parts of the cranial organization so satisfactorily afford. A great proportion of the osseous parietes is wanting; but the cast in the hard matrix of the wide lateral cavities (12, 12), which were over-arched by the expanded postfrontal and parietal bones, indicates the prominence of the postfrontals at the upper and outer angle of the orbits. The orbits (or) appear to have been more ovate and less circular than in the Chelone planimentum ; and the sides of the orbital part of the skull do not converge so rapidly towards the muzzle, but meet at a more acute angle. That a second species of turtle, distinct from the Chelone planimentum, has left its remains in the Harwich clay, is very decisively demonstrated by the almost complete carapace in the British Museum, the inner surface of which is represented, on the scale of six inches to a foot, in T. XII. This carapace, both by its general contour, by the relative length of the costal plates to one another, and by their relative breadth to the adherent pleurapophyses beneath, more resembles the carapace of the Chelone imbricata than that of the other known existing species of turtle; and, as the peculiar characters of the Chelone imbricata are exaggerated, it differs in a proportional degree from the Chelone planimentum. ‘These characters are seen in the great breadth of the pro- minent inferior part of the ribs, and of the free extremity of the rib (p/i—p/s), as compared with the total breadth of the costal plate. The intervals between the free extremities, where the expanded plate terminates, are not equal to the breadth of the proper ribs ; in the Chelone imbricata they very slightly exceed the breadth of the free ends of the ribs. This character in the fossil, by which it is so markedly distinguished from the Chelone planimentum, and most other species, has suggested the name Chelone crassi- costata, or thick-ribbed turtle, which is proposed for the present species. The last pair of ribs of the carapace (T. XII, y/s) are remarkably short and thick, and are curved backwards on each side the broad terminal neural plates which they almost touch. In this character the Chel. crassicostata resembles the Chel. imbricata, and differs from the Chel. caouanna (fig. 2, p. 3), and from Chel. mydas. The subequality of length of the costal plates is another character by which the Chel. crassicostata resembles the Chel. imbricata, and differs from the Chel. mydas, the Chel. caouanna, as well as from the Chel. planimentum. In T. XII, as in the other figures, cd is the nuchal plate, p/i the first rib of the carapace (the second free pleurapophysis or vertebral rib), p/2 to p/s the remaining ribs of the carapace and costal plates ; s9, s10, and py are the terminal neural plates and pygal plate, which, like the nuchal plate, are developed in the substance of the integument, without becoming attached to the subjacent spinous processes of the vertebrae. The debris of the neural arches of the intermediate eight vertebree of the CHELONTA. 29 carapace are preserved in the interspaces of the beginnings of the ribs and costal plates in this beautiful Chelonite. It forms part of the Fossil Collection in the British Museum. A carapace of a smaller individual of Chelone erassicostata, from the Harwich coast, with the character of the broad and inwardly-prominent ribs strongly marked, is likewise preserved in the choice collection of my esteemed friend Professor Bell. One of the hyosternal bones, inclosed in the same nodule of clay, testifies to the partial ossification of the plastron in this species by its coarsely-dentated border; and, at the same time, shows a specific peculiarity by the convexity of that surface which was turned towards the cavity of the thoracic-abdominal case. On the moiety of the nodule containing the carapace and exposing its under surface, the slender rudimental rib of the proper first dorsal vertebra is preserved, in connexion with the first expanded rib of the carapace. Besides the specimen of Chelone crassicostata from Harwich, figured in T. XII, there is a mutilated carapace of a young Chelone, from the same locality, in the British Museum. ‘This specimen exhibits the inner side of the carapace, with the heads, and part of the expanded bodies, of four pairs of ribs, which indicate its specific agreement with the foregoing specimen, and demonstrate unequivocally its title to rank with the marine turtles. It is figured in Mr. Keenig’s ‘ Leones Sectiles’ (pl. xvi, fig. 192), under the name of Zestudo plana. A rare Chelonite from the hard Eocene clay apparently of Harwich, in the collection of my friend Frederick Dixon, Esq., F.G.S., of Worthing, shows the impressions from the under surface of the carapace, and also an instructive part of the under surface of the plastron itself. (T. XIII.) The proportions and degree of convexity of the under surface of the costal plates of the carapace (p/, p/) correspond with those parts in the Chelone crassicostata. The remains of the plastron include a great portion of the left hyosternal (As), left hyposternal (ys), and left xiphisternal (vs); the latter is articulated to the hyposternal by a notch, receiving a toothed process, and, reciprocally, near the upper part of a long oblique harmonia, between the outer border of the hinder angle of the hyposternal and the inner border of the upper half of the xiphisternal. The hyosternal is concave lengthwise, and is convex across on its under surface; the transverse linear impression, dividing the pectoral and abdominal scutes, crosses near its posterior border. The degree of concavity of the outer surface of this bone corresponds with the convexity of the upper and inner surface of the same bone in the specimen of the Chelone crassicostata from Harwich, in the Museum of Professor Bell; and it concurs with the characters of the costal plates in proving the present Chelonite to be of the same species. Impressions of the toothed mesial margin of the right hyosternal remain, and part of the toothed margin of the left hyposternal. 30 FOSSIL REPTILIA OF THE LONDON CLAY. The right coracoid (52) is exposed by the removal of the right hyosternal ; it differs in form from that preserved in the large specimen of Chelone planimentum, in Professor Sedgwick’s Museum, in expanding less suddenly at its sternal end, as compared with the coracoid of the Chelone mydas, or with that of the Chelone caouanna, which is somewhat broader than in the Chel. mydas; the coracoid of the Chel. crassicostata agrees with that of the Chel. planimentum in the greater degree of its expansion. At the anterior fractured surface of Mr. Dixon’s Chelonite, the long and slender columnar or rib-like scapula, is shown, extending from the under part of the head of the second costal rib downwards and outwards, for an extent of two inches, and then sending its acromial or clavicular prolongation at the usual open angle downwards and inwards to rest upon the episternal. The proportions of these parts of the scapular arch are quite those which characterise the genus C#elone, but they do not supply such marks of specific distinction as the coracoid element does. CHELONE DECLIVIS. Owen. Tab. XIV. The extinct turtle represented by this specimen, and indicated by the above term, bears the same relation to the Chelone convexa, which the Chelone longiceps* does to the Chelone latiscutata ;+ that is, it has the same general characters of the petrified parts of the carapace, but differs in the narrower proportions of the vertebral scutes (vl1—v4), and the more open angle at which their two lateral borders meet ; the vertebral angles of the costal scutes being correspondingly less acute. The specimen is from the Eocene deposits of Bognor, Sussex, and is preserved in the collection of Frederick Dixon, Esq. It consists of the seven anterior neural plates, and the corresponding seven pairs of costal plates (T. XIV), those of the right side having been broken away from their attachments to the neural plates, and bent upon the rest of the carapace at an acute angle with some slight separation of the sutures of the costal plates (fig. 2). The neural plates correspond in general form with those of the Chelone convewa, the hind ones being rather broader; the first (si) is crossed at its middle part by the impression dividing the first (v1) from the second (v2) vertebral scute; the second neural plate (s2) is an oblong four-sided one, with both ends of equal breadth. The third neural plate, s3, resumes the hexagonal figure with the broadest end, and two shortest sides at the fore part; and is crossed in its lower half by the impression dividing the second, v2, from the third vertebral scute, v3. The fifth neural plate (s5) is crossed by the next transverse impression nearer its lower border. The sixth and seventh neural plates retain the same form and proportions as in the Chelone convera, except a somewhat * Proceedings of the Geological Society of London, December 1, 1841, p. 572. + Ibid., p. 574. CHELONIA. 31 greater breadth, and have not their antero-lateral borders increased in length, as in the Chelone longiceps. The declination of the ribs from the neural plates, gives a greater degree of steepness to the sides of the carapace than in the Chelone convexa, and the impressions of the scutes have equal depth and breadth. The chief difference indicative of specific distinctions, lies in the form of those impressions; and the question is, whether, in the progress of growth which makes the longitudinal extent of two of the vertebral scutes in one specimen nearly equal to three, in another, so great a change could be effected in their shape as is shown in the specimen of Chelone convera ; in which it will be seen that the second vertebral scute (T. VII, v2), though more than one third shorter than in Chel. declivis (T. XIV, v2), is of the same breadth as that in the larger specimen, and that the rest differ in the same remarkable degree. CHELONE TRIGONICEPS. Owen. More than one of the old tertiary turtles (Chelone) are remarkable for the longitudinal extent or depth of the symphysis of the lower jaw. The turtles from the Eocene clay at Harwich have this character so strongly developed and the under surface of the symphysis so flattened, especially in one of the species, as to have suggested the “ nomen triviale” planimentum for it. The Chelone longiceps, if we may judge by the length of the upper jaw and bony palate, must have had a corresponding extent of the symphysis of the under jaw; and we may infer the same peculiarity from the straight alveolar borders of the maxillaries and their acute convergence towards the premaxillary bones in an allied species, Chelone trigoniceps, which I have described and figured in the Appendix to Mr. Dixon’s work on the ‘Fossils of Sussex,’ from a specimen which is in the collection of G. A. Coombe, Esq., and which was obtained from the Eocene clay at Bracklesham. Amongst the Chelonites which Mr. Dixon has obtained from the same formation and locality, are portions of the fore part of the lower jaw of four individuals of the genus Chelone, all exhibitmg the characters of the pointed form and great depth of the symphysis. One of these specimens agrees so closely in size and shape with the fore part of the upper jaw of the Chelone trigoniceps—fits, in fact, so exactly within the alveolar border, and so closely resembles that specimen in texture and colour, that, coming from the same formation and locality, and being obtained by the same collectors, I strongly suspect it to belong to the same species of Chelone, if not to the same individual. The known recent Chelones differ among themselves in the shape and extent of the bony symphysis of the lower jaw. Both the Chelone imbricata, and Chelone caouanna have this part deeper and more pointed than the Chel. mydas, but neither species has 32 FOSSIL REPTILIA OF THE LONDON CLAY. the symphysis so depressed or so slightly convex below as it is in the Bracklesham Chelones. These also differ amongst themselves in this respect. The symphysis which I have referred to the Chelone trigoniceps, is the broadest and flattest ; at its back part it shows a deep and broad genio-hyoid groove ; this is reduced to a transversely oblong foramen in Chelone mydas. The second species from Bracklesham, is indicated by the maxillary symphysis, the sides of which meet at a more acute angle, and it is narrower in proportion to its length, is more convex below, and more concave above, with the alveolar borders a little more raised, and the middle line less raised than in Chelone trigoniceps. (n this respect it is intermediate between the Chelone wnbricata, where the upper surface of the symphysis is more concave, and the Chelone caouanna, where it is flatter than in the Chelone trigoniceps. The fossil symphysis under notice, has also a smooth, transverse, genio-hyoid groove at its back part. It accords so closely in form with the end of the upper jaw of the Chelone longiceps, from Sheppey, that I refer it provisionally to that species. Two other specimens of the symphysis of the lower jaw, of rather larger size, appear to belong to the same species as that referred to the Chel. longiceps, by the characters of the concavity of the upper surface, the convexity of the lower surface, and the degree of convergence of the sides or borders of the symphysis. The larger of the two shows the genio-hyoid groove, and the nearly vertical outer side of the jaw, opposite the back part of the symphysis, and this shows no impression of the smooth fossa receiving the insertion of the biting muscles, whereas, in the Chelone trigoniceps, fig. 11, that fossa extends to the same transverse line or parallel with the back part of the symphysis. The very rare and interesting Chelonite in Mr. Coombe’s museum, was the first portion of the cranium of a reptile of this order that I had seen from the Eocene deposits at Bracklesham. It includes the bones forming the roof of the mouth, with portions of the bony nostrils and orbits, and the tympanic pedicles. The extremity of the upper jaw is broken off, but the straight converging alveolar borders clearly indicate the muzzle to have been pointed, as in the Chelone longiceps of Sheppey; and the muzzle being shorter, the form of the skull has more nearly approached that of a right-angled triangle. The whole cranium is broader and shorter, and the tympanic pedicles wider apart. The middle lie of the palate developes a somewhat stronger ridge; the orbits were relatively larger and advanced near to the muzzle: the malar bones are more protuberant behind the orbits, and their external surface inclines inwards as it descends from behind and below the orbit, to form the lower border of the zygoma, which it does not do in the Chelone longiceps. The upper surface of the fossil shows the palatines rising to form the vomer at the middle line, and the two small subcircular vacuities (occupied by membrane in the CHELONIA. 33 recent skull) between the palatines, prefrontals, and maxillaries; the anterior border of the temporal fossa, formed by the malar and pterygoid, is entire on one side, and shows that that vacuity was as broad as it is long. The olfactory excavations in the maxillaries are deep. The articular surface of the tympanitic pedicles closely accords with those of recent Che/ones. The very regular triangular form of the skull indicated by this fragment, has induced me to propose the name of Chelone trigoniceps for the species. CHELONE CUNEICEPS. Owen. Tab. XV. One of the most complete and instructive crania of the fossil turtles of our Eocene deposits is the subject of T. XV, the opportunity of describing and figuring which has been kindly afforded me by J. Toulmin Smith, Esq., F.G.S., of whose cabinet it forms part, and by whose skilful manipulation its variously configurated exterior has been disencumbered of the hard adherent clay. From the Chelone breviceps this specimen differs by its more prolonged and pointed muzzle; by the more sudden and sloping declivity of the prefrontal part of the cranium (fig. 1, 14); by the minor degree of rugosity of the surface of the bones ; and by the different disposition of the superincumbent horny scutella, which is dicated by their impressions. In the general arrangement of these impressions it accords better with the cranium of the Chelone longiceps ; but differs in the greater breadth of the skull as compared with its length; in the minor extent of the bony palate (fig. 3, 20, 21), the more advanced position of the posterior nostrils, and the greater length of the pterygoids (24). From’ the Chelone convewa it differs, in the greater relative breadth and flatness of the frontal bones, and of the whole interorbital platform (fig. 2, 11), in the downward slope of that part of the cranial profile, and in the more prominent convexities of the palatal processes of the maxillaries. From the Chelone planimentum it differs also, by the broader prefrontal part of the interorbital space, as compared with the transverse diameter of the back part of the skull: by the minor degree in which the frontal enters into the formation of the upper rim of the orbits; by the minor depth of the suborbital part of the maxillary and malar bones, and by a very different arrangement of the supracranial horny scutella. The basi-occipital (T. XV, figs. 3 and 4) is remarkable for the strong development of the tubercles for the insertion of the strong “ recti capitis antici,” and for the depth of the median groove between them; the semicircular fossa in front of these processes is bounded by a well-developed basi-sphenoidal ridge (5), the curve of which is deeper than in Chel. longiceps, but shallower than in Chel. breviceps. In the Chel. caouanna, in which the basi-occipital tuberosities are better developed than in the Che/. imbricata or Chel. mydas, they are bounded anteriorly by an angular or chevron-shaped ridge of the basi-sphenoid. The exoccipitals (2) form the usual share of the trilobate occipital - 9) 34 FOSSIL REPTILIA OF THE LONDON CLAY. condyle characteristic of the Chelonia. The paroccipitals (4) project backwards to a little beyond the posterior plane of the condyle, indicating an affinity to the Zrionycide. The inferior surface of the part of the tympanic to which they unite is concave. The parietals (fig. 2, 7) form together a large semielliptic, almost flattened, platform, relatively broader than m Chel. mydas, not convex, as in Chel. caouwanna; not in- dented by the mastoids, as in the Chel. longiceps, and not forming an angle between the frontals and postfrontals, as in the Chel. breviceps. The frontals (11) together form a pentagon, with the longest margin joining the parietals, the next in length con- verging to a point between the prefrontals, and the shortest borders joming the post- frontals. The postfrontals (12) and prefrontals (14) almost meet above the orbits, and exclude the frontals from entermg into the formation of its superior border. The Chel. mydas comes nearest to the Chel. cuneiceps im this particular ; whilst in the Chel. imbricata the frontals enter as largely into the formation of the upper border of the orbit as they do in the Chel. breviceps, Chel. longiceps, and Chel. conveva. The precise form of the termination of the prefrontonasals, the maxillaries, and premaxillaries cannot be determimed in the present specimen ; fortunately, the fracture of the anterior extremity of the skull has not extended to that of the bony palate. If this be bounded by a transverse line behind, drawn across the anterior border of the temporal fossze, the space included forms a right-angled triangle, and includes the whole of the posterior nostrils. In the Chel. longiceps the similarly defined space has the base shorter than the converging sides, and the posterior nasal aperture is behind the transverse line. The bony palate, also, of Chel. cuneiceps, instead of being pretty uniformly concave and even, as in Chel. longiceps and Chel. caouanna, is raised on each side between the middle line and the marginal alveolar plate into two convexities, as in Chel. mydas and Chel. imbricata; but the most prominent part of the palatal convexities (figs. 3 and 4, 21) is obtuse in Chel. cuneiceps, not sharp or angular, as in Chel. mydas and Chel. imbricata. The palatal part of the vomer (13) forms the median longitudinal groove dividing the convexities, which are formed by the palatal processes of the maxillary bones. The small part of the alveolar border of the maxillary which is entire terminates in a sharp edge, extending about four and a half lines below the level of the palate. The ridge of the palatines, which forms the anterior boundary of the posterior nostril, is not produced or bent below the level of the bony palate, as in Chel. caowanna, and as it is, although in a minor degree, in Chel. mydas ; and there is not that concavity between it and the oblique palatal tuberosity which exists in the Chel. mydas and Chel. imbricata. The pterygoids are more deeply (semicircularly) emarginate laterally than in any of the existing species of Chelones, and they are shorter in proportion to their breadth ; they bound internally the lower apertures of the temporal fossze, which are broader than they are long: in all the existing Chelones the opposite proportions prevail, CHELONIA. 33 and in Chel. imbricata especially the homologous apertures are twice as long as they are broad. The pterygoids, in the Chel. cuneiceps, develope a sharp ridge along their median suture; and short but well-defined processes at their anterior and outer angles. The channel or concavity upon the under part of the diverging portion of the pterygoid conducts obliquely into the temporal fossee in the Chel. mydas; in Chel. cuneiceps it leads directly forwards upon the under surface of the anterior part of the pterygoids exclusively, as in the Chel. imbricata and Chel. caouanna. In the Chel. mydas the malar approaches the mastoid very closely, and sometimes touches it by the posterior angle, thus separating the squamosal from the postfrontal ; the extent of the union between the squamosal and postfrontal is also shorter in the Chel. caouanna than in the Chel. imbricata. In the extent of that union (between 12 and 27) the Chel. cuneiceps resembles the Chel. imbricata, as do likewise the Chel. breviceps and Chel. longiceps. But the Chel. cuneiceps differs from all the recent species in the form of the squamosal (27), which is bent upon itself, forming a slightly curved linear eminence, where the lower and smoother part of the bone is bent, and, as it were, pressed inwards towards the tympanic (28), against which it abuts. This modification is natural, not the effect of accidental pressure upon the fossil. The lower border of the malar (26), which intervenes between the maxillary and squamosal, is sharp but convex, as in Chel. caouanna, not concave as in Chel. mydas, nor nearly straight, as in Chel. imbricata. But the concave curve of the inferior margin of the squamosal (27) most resembles that in Chel. imbricata. The antero-posterior extent of the mastoid (8) is less proportionally than in any of the recent Chelones, and it forms asmaller share of the upper border of the large meatus auditorius. The articular part of the tympanic descends below the squamosal further than in the recent turtles ; and its articular surface is more convex at its outer half, and more concave at its inner half; Chel. imbricata makes the nearest approach to the fossil in this respect. In the Chel. mydas and Chel. caouanna the articular surface is nearly flat. As the supracranial scutella have left unusually deep and well-marked impressions on this fossil skull, I have reserved their description, and the comparison of their different forms and proportions in the several fossil species, to this place. Three scutella occupy the median line of the upper surface of the cranium in the present species of Chelone, which, from the absence of any impression along the frontal and sagittal sutures, appear to have been single and symmetrical. The anterior and smallest answers to the “ frontal” scute (7/7); the next in size and position to the “sincipital” scute (sy); the hindmost and largest answers to the “ occipital” scute (0c), which is usually divided, and forms a pair in existing Che/ones. The frontal scute is long, narrow, hexagonal, broadest across the antero-lateral angles, from which the impressions extend outwards to the supraorbital margin, which divide the “fronto-nasal” scute from the “ supraorbital” scute (00). The sincipital scute is bounded on each side by a sigmoid curve, and both before 36 FOSSIL REPTILIA OF THE LONDON CLAY. and behind by an entering angle; it is broadest behind, and from the middle of the lateral border proceeds the transverse impression towards the back part of the orbit, which divides the “ supraorbital” scute (04) from the “ parietal” scute (ya). The occipital scute is bounded laterally by straight lines, which slightly diverge as they extend back- wards: there is no trace of an interoccipital scute. The parietal (ya) scute is the largest; impressions of five of its borders are preserved in the present fossil: the two exterior ones meet at an obtuse angle, a little above the middle of the meatus auditorius externus ; the antero-external border uniting with the postorbital scute (yo); the postero-external border with the external occipital scute (¢0). In the Chelone breviceps (T. I) the frontal scute is relatively larger than in the Chelone cuneiceps, and is nearly as broad as long. The sincipital scute is bounded laterally by two straight lines meeting at a very open angle, from which the transverse impression extends outwards between the supraorbital and parietal scutes. The straight limes bounding the sides of the occipital scute diverge from each other as they extend backwards more than they do in the Chelone cuneiceps. In the Chelone longiceps (T. II) a still more different pattern of the supracranial scuta- tion is presented. The occipital scutes (oc) are separated by an intervening interoccipital scute (70). The lateral borders of the sincipital scute are each bounded by three lines and two angles; the antero-lateral and postero-lateral angles being curved with the concavity outwards; and the transverse impression dividing the supraorbital scute (06) from the parietal scute (ya), proceeds from the middle of the intervening straight border of the parietal. The frontal scute (/r) is long and narrow, broadest behind, with its lateral borders gradually converging to a point anteriorly ; the impression dividing the supraorbital (04) from the frontonasal scute (/z) proceeds from the middle of that lateral border. Neither the division between the frontal and sincipital, nor that between the sincipital and interoccipital scutes are well marked. The Chelone convexa (T. VI, fig. 4), like the Chelone longiceps, has an interoccipital scute (70), and the sincipital scute (sy) has its sides bounded by three lines, of which the posterior one is curved with its concavity towards the occipital scute (oc), and so directed as to appear to form part of the posterior rather than the lateral- border ; the other two lines completing the lateral border and converging forwards, are divided or defined by a slight angle, from which the transverse impression proceeds outwards, which divides the supraorbital (04) from the parietal (ya) scutes. The frontal scute (fr) is a small hexagon, relatively wider than in Chel. longiceps or Chel. cuneiceps. The impression dividing the supraorbital (04) from the frontonasal (/z) scutes proceeds from the angle between the lateral and anterior sides of the frontal scute. The Chelone planimentum (TY. 1X) is peculiar, and differs from all the foregoing species by the forward extension of the occipital scutes which join the supraorbital scutes, and thus divide the sincipital scute (sy) from the parietal scute (pa); the sincipital scute CHELONIA. 37 is correspondingly encroached upon, as it were, and narrowed, its broadest part being nearer the anterior end, at the angle between its two straight lateral borders, from which angle the impression extends outwards that divides the occipital from the supra- orbital scute. The frontal scute (/r) is small and narrow, and the large supraorbital scutes meet in front of it at the middle line. They appear to be divided from the orbits by the encroachment of palpebral scutes (p/) upon the supraorbitary border. There appears to have been an interoccipital scute in the Chel. planimentum, as in the Chel. longiceps and Chel. convexa. Amongst existing Che/ones the interoccipital scute is constant only in the Chel. caouanna—the loggerhead of Catesby and Brown; but the sincipital scute in this species is vastly larger in proportion than in any of the fossils above described ; and it is further distinguished by the peculiar division of the supraorbital and parietal scutes. In the hawks-bill turtle (Chel. imbricata), the supracranial scutes leave as well- marked indentations upon the bones of the cranium as are seen in most of the fossil turtles, but the supraorbital scute is proportionably larger than in any of these, and the proportions and forms of all the other scutes are different. There are, also, two nasal scutes divided by a transverse groove from the frontonasals, which groove I have not yet met with in the corresponding part of any of the fossil Chelonian crania. The skull of the Chelone cuneiceps, here described, is from the London clay of Sheppy. CHELONE SUBCARINATA. Sell. Tab. VIII 4. The resemblance of this species to Chelone subcristata (p. 24, T. VIII) is so con- siderable, that it has not been without some hesitation that I have ventured to describe itas distinct. There are, however, certain characters by which it may be distinguished, and those of sufficient importance to be considered as specific. On comparing it with recent species, and even with most of the fossil ones from the same locality, there is a remarkable evenness in the arch of the carapace, which, with the exception of a slight carina on some of the posterior neural plates, to be hereafter mentioned, forms nearly a perfect arc of a circle, from the extremity of the costal plate of the one side to that of the other, without that flattening of the side which is seen in most other species. The nuchal plate (T. VIIL4, fig. 1, ci) has the posterior margin arched, and there is a short median process which goes to join the first neural plate (s1), in which respect it agrees with Chel. longiceps and with Chel. subcristata. 'This process is emarginate, to receive a slight triangular projection of the anterior margin of that plate. The first neural plate (s1) forms a parallelogram, the sides not being interrupted by any costal suture ; the posterior suture of the first costal plate (/1) extending to the second neural plate (sz). In this circumstance it differs from Chel. subcristata, longiceps, and convexa, and agrees with Chel. breviceps. This, however, may possibly be a variable character here, as 38 FOSSIL REPTILIA OF THE LONDON CLAY. it is in Chel. longiceps ; m one specimen of which, now before us, Professor Owen found that the articulation in question was to the anterior part of the second, instead of the posterior part of the first, neural plate; in other words, that the first neural plate was the isolated one instead of the second. ‘The remaining neural plates are hexagonal, becoming almost regularly shorter to the eighth ; the lateral angles meeting the costal sutures being nearly at the same distance from the anterior margin in each, and in no one at all approaching a regular equilateral hexagon, as in many of the neural plates in Chel. breviceps. The first three, and the anterior half of the fourth neural plates are flat; but on the posterior half of the fourth commences a low carina, which becomes highest on the posterior half of the sixth (sé), and anterior half of the seventh (s7). It thus differs from Chel. subcristata, in which there is a distinct. short, sharp, longi- tudinal crest (si) on the fifth and seventh neural plates, ‘‘ and a similar crest is developed on the contiguous ends of the second and third neural plates.” The ninth and tenth neural plates are wanting in the only specimen I have seen of the Chel. subcarinata. The first costal plate is flat (p/1), but the remaining ones, to the seventh inclusive, are slightly hollowed along the middle, being raised towards the anterior and posterior margins, where they are articulated to the contiguous ones. The whole surface of the bones of the carapace is less smooth than in most other fossil species, and conspicuously less so than in Chel. subcristata. In describing the forms of the vertebral scutes, (v1—v4), and of the costal ones as depending upon them, it is necessary, in order to arrive at any satisfactory comparison between these parts in different species, to bear in mind that a great change takes place in their outline during the growth of the animal; and that a vertebral scute, which, in a younger individual, has the middle of its outer margin exceedingly extended, so as to form a very acute angle, where the lateral margin of the costal scute joins it, and thus rendering it twice as broad as it is long, may in more advanced age have that angle very open, and having increased greatly in length, and scarcely at all in breadth from angle to angle, the length becomes greater than the breadth. Allowing, however, for this fact, there are doubtless considerable variations in this respect according to the different species, which are permanent and well marked. The first vertebral scute (v1) in the present species is quadrilateral, broader anteriorly ; the second and third (v2, v3) hexagonal, with the outer margins slightly waved, somewhat broader in the middle at the angles than at the anterior and posterior margins, the comparative breadth at that part being rather greater than in the corresponding scutes of Chel. subcristata, and much less so than in Chel. convexa, Chel. breviceps, or Chel. longiceps. The fourth vertebral scute (v4) is also hexagonal, but the portion posterior to the lateral angles is narrowed and produced backwards. The last of the series is fan-shaped. The outline of the costal scutes follows of course that of the vertebral ones. The plastron, in the specimen from which this description is taken (Pl. VIII, A, fig. 2), is more perfect than im that of almost any other fossil Chelonian I have seen. It CHELONIA. 39 agrees in its general form with that of Ciel. subcristata, but is less extensive, as regards its bony surface, than in Chel. longiceps or even than in Chel. breviceps. The entosternal bone (s) is somewhat wedge-shaped, with the anterior margin triangular, and a short winged process on each side of the anterior third of the bone extending outwards and backwards. The posterior extremity of the bone, and the winged processes are dentate. The episternals (es) are aliform, tending backwards and outwards, and inclosing between them the head of the entosternal (s), and the anterior processes of the hyosternal bones (4s). The latter have the anterior processes extending forwards on each side of the entosternal, approximating at their extremity the aliform processes of that bone. The median or internal processes nearly meet on the median line, and the den- tations are deep but slender; each hyposternal (ps) unites similarly with its fellow, and the posterior process extends backwards, in a long, narrow, triangular piece, uniting with the xiphisternal (vs), which latter forms a very elongated rhomb, the breadth of which is scarcely one fourth of its length, which in the present specimen is no less than two inches six lines. This form, with the elongation and narrowness of the posterior process of the hyposternal, gives to the hinder portion of the plastron in this species a narrower and more elongated outline than we find in almost any other; an approach to which is, however, indicated in the imperfect specimen of Chel. sudcristata figured in Plate VIII. The external notch, between the external process of the hyosternal and hyposternal, is deep and rounded. The central interspace is nearly quadrate, and about half as long again as it is broad. Inches. Lines. Length of the carapace as far as it is preserved : ; 9 5 Breadth of ditto from the extremity of the third costal plate on one side to that on the other . p : ‘ : c 7 4 Ditto, following the convexity of the carapace . : 2 9 3 Length of plastron from the anterior margin of the episternal to the extremity of the xiphisternal - : : 6 8 4 Breadth of ditto across the hyosternals . : : é : tf 0 The only specimen of this species which I have seen is from Sheppy, and is in the fine collection of J. S. Bowerbank, Esq., F.R.S. BE SUPPLEMENTAL REMARKS ON THE TURTLES FROM THE LONDON CLAY AT HARWICH. In the progress of the works now carried on in a part of the Harwich cliffs, with a view to the acquisition of the remains of the animal tissues and bone-earth which form the nodules that are ground up and used as manure, many remains of the Chelonian reptiles which formerly frequented the seas from which those Eocene tertiary strata have been deposited have been discovered. Mr. Colchester, of Little Oakley, Essex, who carries on large works of this kind for the “ Fossil Guano,” as it is termed, has transmitted to me a number of the nodules in question. The most intelligible and instructive of these I have marked from 1 to 10 consecutively, and shall notice them here in the same order. No. 1. Chelone planimentum. Thisis the half of an oval nodule of petrified clay, 20 inches in length, by 17 inches in breadth, exposing an irregular group of disarticulated bones of the carapace and other parts of the skeleton. The species is determined by a fragment of one of the costal plates with the connate rib. The plate measures 25 inches in breadth, the rib 8 lines, and forms the usual partial prominence from the even surface of the under part of the costal plate. Almost the whole of the very broad but short nuchal plate is recognisable: it measures 6 inches in transverse diameter, and only 13 inch in antero-posterior diameter. Part of the hyosternal bones, and the impression of the humerus are recognisable. No. 2 is the half of a nodule, 20 inches in length and 17 inches in breadth, exposing part of the plastron, and some other bones of the skeleton of the Chel. plani- mentum. It shows well the natural form of the under and outer part of the hyposternal bone, which is much more deeply excavated than in the Chel. crassicostata ; the lower portion of the bone is narrower in proportion to its length, and the xiphisternals are also in proportion longer and narrower than in that species. No. 3. Chelone planimentum. The half of an oval nodule, 17 inches in length and 13 inches in breadth. The fractured side exposing a cast of the inner surface of the carapace, which measures in length from the nuchal to the tenth neural plate inclusive 133 inches; and in breadth, across the third pair of costal plates from one end of the projecting rib to that of the opposite side, 11 inches. The anterior contour of the CHELONIA. Al carapace is well shown in this nodule, the marginal plates which join the nuchal plate being preserved. The free extremity of the rib attached to the third costal plate projects 1 imch 9 lines from that plate, and measures 7 lines in breadth, where it becomes free ; the breadth of the plate being nearly 2 inches. The transverse curve of the carapace is shown by this specimen to be much less than in the Chel. crassicostata. No. 4. Chel. planimentum. The nodule shows partly a cast of the outer surface of the carapace, with part of the carapace itself. The outer angles of the third and fourth vertebral scutes are here seen with the imner angle of the third costal scute. The outer angles of the vertebral scutes are more prominent than in Chel. declivis, Chel. subcristata, Chel. subcarinata, Chel. convexa, or Chel. longiceps ; they resemble most those in Chel. breviceps. The breadth of the third costal scute is 4 inches. The characteristic angular ridge, formed by the narrow connate rib, where it projects from the lower surface of the costal plate, is well shown in this specimen. No. 5. A nodule showing a cast of the under surface of the carapace seen from above, apparently of the Chel. planimentum. No. 6. A nodule, 10 inches long ‘by 9 inches broad, showing a still more imperfect cast of the under surface of the carapace, of apparently a younger specimen of the Chel. planimentum. No. 7. t. Impression of the fourth vertebral scute. . First costal plate. Second ditto. . Third ditto. . Fourth ditto. . Fifth ditto. . Sixth ditto. . Seventh ditto. 2. View of part of the dislocated right half of the carapace. The same letters and numerals signify the same parts. 5. Outline of the natural transverse curve of the carapace. LAT, Winkel del. cb lth, Dave Sor Titn sto the Ou: Skull of Chelone cunetceps, nat. size. Fig. : 1. Side view. tele * 2. Top view. ic = e 3. Base view. fi 4. Back view. | The letters and numerals indicate the same parts as in Tabs. Tand III. 3 3 Mika * TEXVE v Uh.“ to the. Ouecn Dinkel dol. ck lithe TAB. XVI: Trionyx Henrict. Fig. 1. Upper view of the carapace, wanting the nuchal plate, half nat. size. sj—s7. The first to the seventh neural plates inclusive. pli—pls. The eight costal plates of the left side. 2. Outline of the transverse curvature of the upper surface of the carapace. 3. The nuchal plate of probably the same species of Zrionye; but of another individual : half the natural size. TO VAL Exrxleben Un Fine by ip an TAB. XVI4. The carapace of the Trionyx Barbare, half nat. size. a, Shows the longitudinal contour of the middle of the upper surface, and fig. 2, the transverse curvature of the carapace. TEXVLA.. a Day £ Son, lith, to the Oucen is Mie = TAI exevall: Trionyx incrassatus. aa Fig. 1. Inside view of the carapace, wanting the nuchal plate, half nat. size. sl—s7. The first to the seventh neural plates, with the connate neural spines and arches. ieee The eight costal plates of the right side. . The place of attachment of the outer end of the first dorsal rib. c2—9. The portions of the consecutive dorsal ribs that become connate with the costal plates. 2. Transverse contour of the upper surface of the carapace. T. XVI. { 1 Fig. Ie bo TAB. XVIII. Trionyx incrassatus. Upper or outside view of the fore part of the carapace with the nuchal plate, half nat. size. ch. 'The nuchal plate. sl. The first neural plate. s2. The second neural plate. pli—pls. The three anterior costal plates of the left side. Under or inside view of the same specimen. ch. The nuchal plate. s}. The first neural plate connate with the neural spine of the second dorsal vertebra, of which part of the under surface of the centrum is here preserved and shown. s2. The second neural plate, with the connate neural arch of the third dorsal vertebra. pli—pl3. The three anterior costal plates of the right side. cl. The place of attachment of the outer end of the first dorsal rib. c2. The portion of the second dorsal rib that becomes connate with the first costal plate. 2’. The extremity of the rib that again becomes free. c3. The third rib attached to the second costal plate. c4. The fourth rib attached to the third costal plate. LE Ge. € al. Sby 72 ta leben. y Bopaycf, e & S ee TAB. XVIIIZ. Trionyx rivosus. 1. Outside view of hinder half of the carapace, nat. size. bo pl4s—pls. The fourth to the eighth costal plates inclusive. Inside view of the same specimen. pla—ypls. The fourth to the eighth costal plates inclusive. 5—9. The ribs (fifth to the ninth dorsal inclusive) connate with the above costal plates. sé. The neural arch confluent with the sixth neural plate. BAS N Day & Son tthe to the Ouecn From Nature on. Stone by J Firaleben , TAB. XIX. Trionyx incrassatus. The parts are figured half the nat. size. Fig. 1. Entosternal. 2. Episternal, under side. 2’. Ditto. upper side. Part of right hyosternal, and the hyposternal, under or outer side. nehs; Ditto, and ps, ditto, upper side. 4. Two views of the scapula (51), and connate acromial clavicle (58) ; 4” shows the osseous structure of the end of the scapula, nat. size. 5. The coracoid. 6, 6’. Two views of the ilium. 7, 7. Two views of the femur. 8, 8. Two views of an ungual phalanx, nat. size. 9. Under view of the sixth cervical vertebra; y, y, the hypapophyses. 9’. Upper view of the same vertebra; c, the anterior convex articular surface of the centrum; 2, the neural arch; 2, z, the zygapophyses (oblique or articular processes). 9”. Back view of the same vertebra, showing the double concavity of the articular surface. ‘2 Nat. size. Joe Dinkel Pee TAB. XIX+. Trionyx marginatus. Fig. 1. Upper view of the carapace, wanting the nuchal plate, half the nat. size. letters and figures indicate the same parts as in Tab. XVI. 2. Outline of the transverse curvatures of the upper surface of the carapace. The T_XIX+ i : iF ae. ga ewe : cae yi ; 25 no ake a rhs PUN inch SIMI SUI COT ONS! 2 Gvtsb OANA Wh — go TAB. XIXB. . Upper surface of the third costal plate, right side, of the Zrzonyx circumsulcatus, nat. size. . Articular margin of ditto; showing the natural curve and thickness of the plate. Peripheral margin of ditto; showing the groove. Upper surface of the third costal plate, right side, of the 7rionyx marginatus. . Sutural margin of ditto. Peripheral margin of ditto. . Upper surface of a fragment of a costal plate of the Zonya pustulatus, nat. size. . Under surface of ditto, showing the large and prominent adherent rib. . Peripheral margin of ditto. From. 2 on Stme by J Eraleben TXIXB. bo TAB. XIXC. Trionyx planus. . Upper view of the hind part of the carapace, half the nat. size. . Inside view of the same specimen ; p/5—p/s, the fifth to the eighth costal plates inclusive. . Peripheral border of the fifth costal plate, nat. size. . Peripheral border of the fifth costal plate of the Zionyz Barbare, showing the difference in their relative thickness. Dayk Son. LXIX.C. TAB. XIXD. Under view of the lower jaw of a turtle (Chelone), nat. size. (Hordwell Cliff). . Upper view of the same specimen. Upper view of the fourth cervical vertebra of a Zrionyx:; d, diapophysis. Nat. size. 4. Side view of the same specimen (Hordwell Cliff). On NI . The left os pubis of a Zrionyx, nat. size (Hordwell Cliff). . Part of the plastron of TZrionyx planus, half the nat. size; As, hyosternal; ps, hyposternal. . Right hyposternal of a Zrionyx, from Bracklesham, half the nat. size. (In the Collection of Frederick Dixon, Esq., F.G.S.) GND) ES Wo the Quce Hi Luph Son lit > S yy tone by . Ons TAB. XX. Emys Comptoni, nat. size. Fig. oh: i. fo) 1. Upper view of the carapace. oe 2. Under view of the plastron. | ae - oe 3. Side view, showing the marginal plates, 4—i7. js 4. Front view. : ; 5. Back view. TAB. XXI. Platemys Bullockii, half the nat. size. The plastron and some of the marginal plates, @ 2. s. es. As. Te: an. Entosternal. Episternals. Hyosternals. . Hyposternals. . XMiphisternals. . Intergular scute. . Gular scutes. . Humeral scutes. . Pectoral scutes. . Abdominal scutes. The impression between these and the pectoral scutes crosses the intercalated supernumerary bones between the hyosternals and hyposternals. Femoral scutes. Anal scute. Where the bones or scutes are in pairs, the figures or letters are placed on one of each pair. WA dei, lh TAB. XXII. Emys levis, two thirds the nat. size. Fig. 1. Upper view of carapace. 2. Under view of plastron. 3. Side view 4. Front view, showing the curvature and depth of the specimen. The letters and figures signify the same parts as in the preceding figures: /p, the accessory pieces of the plastron. T. XX. Rae A bor TAB. XXIII. Platemys Bowerbankit, half the nat. size. Fig s- 1. The mutilated carapace. 2. The plastron. The letters and figures indicate the same parts as in the preceding Table. q [hh] TAB. XXIV. Emys testudiniformis, nat. size. . Nuchal plate. . Upper surface of the third neural plate. . Under surface of ditto. . Upper surface of the fifth neural plate. . A posterior marginal plate. . Front view of a mutilated cuirass in which the carapace has been slightly depressed; the natural curve, across the middle, is indicated by the outline. Darkel del ct tith LAX | Day & Sen; Lith ta theOucar TAB. XXYV. Emys bicarinata, two fifths the nat. size. The letters and numbers on the carapace, the upper surface of which is figured, indicate the same parts as in the previous figures. The transverse contour of the carapace is given in outline above. q Vv » “hd ‘TAB. XXVI. Emys bicarinata, two fifths the nat. size. a The plastron; s, the entosternal piece. ae: 7 ; xs, the xiphisternals. . T.XXVT. 23 Nak: Wye: TABS XXVal: Emys crassus, two thirds the nat. size. Fig. 1. Outside view of the left hyposternal. 1’. Inside view of ditto. 2. Outside view of right hyosternal. 2’. Inside view of ditto. 3. Sutural border of hyosternal, nat. size. LXXVIL, . bas UR TR lhe Sas Nat. Size. Linkel, del. et eth, : Day Son, lth More luce TAB. XXVIII. Emys Delabechit, two fifths the nat. size. Upper or outside view of the carapace; the transverse curve is given in the outline above. pli—pls. The eight costal plates of the left side. . v2—v4. The second to the fourth vertebral scutes inclusive. T.XXVIN Dirk:eh, del et lith Day Sent, MONOGRAPH ON THE FOSSIL REPTILIA OF THE LONDON CLAY, AND OF THE BRACKLESHAM AND OTHER TERTIARY BEDS. PART I. Supriemenr No. 1. Paces 77—79; Puarrs XXVIIIa—XXVIIIb. CHELONIA (Emys). BY PROFESSOR OWEN, D.C.L., F.R.S., F.L.S., F.G.S., &c. Issued in the Volume for the Year 1856. LONDON: PRINTED FOR THE PALHONTOGRAPHICAL SOCIETY. 1858. SUPPLEMENT TO THE EOCENE CHELONTA. Emys CoNYBEARIT, Owen. T. XIII and XIV. The subject of the present description is the most complete specimen of fossil fresh-water Tortoise (mys) which has hitherto come under my observation. It was obtained, like the Amys Delabechii, Bell, which it most resembles, from the Eocene clay of Sheppey Island, and forms part of the large and instructive collection of Sheppey fossils belonging to J. S. Bowerbank, Esq., F.R.S. It consists of both carapace (T. XXVIII 4), and plastron (T. XXVIII B), giving the natural curves, depth, and periphery of the portable abode of the animal. Every constituent bone of the com- plex roof of this abode is preserved, and the impressions of every horny scute can be traced, uninterruptedly, upon its uninjured surface. The floor or “ plastron” is, unfortu- nately, not so entire, but its margins are unbroken, exhibiting the characteristic contour. The dimensions of this noble specimen of Eocene Hmys fall little short of those of the Emys Delabechit, the length of the carapace being 1 foot 63 inches, and its breadth 1 foot 3 inches. The forms and proportions of the nuchal (ch), pygal (py), and their connecting series of median neural plates (s 1 to s 9), are accurately shown in T. XXVIII 4: the eighth neural plate is obliterated, or has its place taken by the extension of ossification of the seventh and eighth costal plates towards the median line. The abraded and fractured state of the carapace of Hmys Delabechui* does not permit of a comparison of this particular. There is no trace of a median elevation or keel in Hmys Conybearii: the neural plates in Hmys bicarinatat are broader in proportion to their length than in the present species, and the series is not interrupted. * «Monograph on Fossil Reptilia of the London Clay,’ T. xxviii, Paleontographical Memoirs for 1848. fj Lbs, We XY. 11 78 SUPPLEMENT TO THE In the costal plates (T. XXVIII 4, pz 1—pi 8) the general form and proportions are sufficiently clearly represented; but it is worthy of special remark that the first (p7 1) has a small quadrate portion marked out on its inner or median border, by the impres- sion of a supplementary median scute, which I have not observed in other Emydians, recent or fossil; and should that supplementary scute be constant in the present species, such species might be indicated by a detached fossil first costal plate (pi 1). In Emydians generally the vertebral scutes are five in number; the costal scutes are four pairs: collectively, when they have been called “discal” scutes or plates, thirteen is the number. In the present extinct species they were fourteen in number, owing to the division of the first vertebral scute (v 1), or to the interposition of a small . transversely extended quadrate scute between it and the second vertebral scute (v 2). It will be fortunate if other specimens of the Amys Conybeari should show this cha- racter to be a constant one. In Lmys Delabechii the carapace is unluckily mutilated at the part requisite for determining whether the supplementary median scute existed. The second vertebral scute (» 2) in Amys Delabechii is as broad as it is long: in Himys Conybearw the breadth of the same scute is one third greater than its length. A similar difference of propor- tions, with more produced and acute lateral angles, characterises the third vertebral scute of Hinys Conybearti as compared with Himys Delabechit. The fourth vertebral scute in Himys Conybearii has the front half of its lateral margin wavy or crenate, not a simple sigmoid line, as in Hmys Delabechii. The breadth and depth of the scutal impressions are alike in both species. In Hydromedusa the number of discal scutes, or those inclosed by the marginal scutes, is fourteen, as in the present fossil; and, as it appears to me, by a similar transverse division of the first vertebral scute ; only the dividing line crosses the scute more anteriorly, so that the proportions of the front and hind divisions are reversed. Dr. Gray regards the front division as the homologue of the front marginal scutel called “nuchal” in other Chelydide and in Emydide, and characterises Hydromedusa as having “‘the nuchal plate large, placed behind the front marginal plate, like a sixth vertebral;” but if the carapace of Hydromedusa depressa* be compared with that of Chelodina oblonga,t and Sternotherus Derbianus,¢ the marginal series of scutes will be seen to be similar in number in the two species, and their condition to be essentially that which is defined in the characters of Sternotherus, as “‘Nuchal plate none.’§ The part called ‘nuchal plate” in Hydromedusa depressa answers to the front part of the first vertebral plate in Sternotherus and Chelodina: the different position of the dividing line in the Hmys Conybearii more strongly marks the true homological character of the first of the median series of discal scutes. I conclude, therefore, that, as in * «Catalogue of Shield Reptiles,’ Brit. Mus., 4to, p. 59. sf dbs tab: xxvis) Tb.) tabi xxi 1S Ibsypy ole HOCENE CHELONIA. 79 Sternotherus and Hydromedusa, the nuchal scute is absent in Hmys Conybearii, the number of marginal scutes being 12—12. In Emys nigricans the nuchal scute is absent, and the number of marginal scutes is as in Hmys Conybearii. The carapace of Hmys Conybeartw is moderately convex at the middle, and very slightly concave towards the margin, which is gently raised before and behind. The plastron, as in the section of Hmydide, including Hmys proper,* is solid, truncate before and notched behind (T. XXVIII 2), attached to the carapace by sutures of bone. The number and character of the sternal plates I have been unable to deter- mine: there is no transverse joint as in the Box-tortoises (Cistudo, Lutremys). The chief peculiarity of the plastron of the Hmys Conybearii is the concavity of its middle three fifths (T. XXVIII B, fig. 2)—a modification which is rarely seen save in some true Tortoises. This character, coupled with the divided first vertebral scute, would probably be deemed worthy of supporting a sub-generic distinction by some erpeto- logists. * Gray, tom. cit., 4, a, p. 15. It is called “sternum” in this work; but the ‘‘plastron”’ includes many other elements of the skeleton besides the sternum; to say nothing of the bony scutes, superadded to the sternum and abdominal ribs. ‘Tiga ae te . i Be 3 Se oy Li Ona nant Yi , i Te Mes OPW as, re We | PU ORS CU ieee eT aL dh a ie SOL Haein eet HORT tak ? p Pyne ae } if, Rae } papeey Lele hemo hive qcotaees A pi ievaaas | pS atonal Pea Ma anes ite Ro AGN. Gplbnloud wh Uy th Te. (altace ods peed cee | ; --—- ; “ i 7 f bel } i Stitt PIGLED FecORaCEh Re ay Bri § ‘, hey } { Se) bie oo Hat R Fite abe ra ’ ‘ | i ¥ , = to | : vie - Dil Cae a as ie TAB. XXVIII 4. Carapace of Hmys Conybearit, half nat. size. From the London Clay, Isle of Sheppey, Kent. In the Collection of J. S. Bowerbank, Esq., F.R.S. T XAVM A. mo m7 wiv mE Jo Winkel hth TAB. XXVIII B. Plastron, and (fig. 2) side view of carapace and plastron in outlme, of Emys Conybearii, half nat. size. From the London Clay, Isle of Sheppey, Kent. In the Collection of J. S. Bowerbank, Esq., F.R.S. T. XxXVIN B & : RSS ites Be MONOGRAPH ON ie FOSsth RER VITA OF THE LONDON CLAY, AND OF THE BRACKLESHAM AND OTHER TERTIARY BEDS. PART IL. Supprement No. 1. Paces 1—4; Prats XXIX. CHELONTA (Puatemys). BY PROFESSOR OWEN, D.C.L., F.R.S., F.L.S., F.G.S., &c. Issued in the Volume for the Year 1849. LONDON: PRINTED FOR THE PALHONTOGRAPHICAL SOCIETY. 1850. MONOGRAPH ON THE FOSSIL REPTILIA OF THE LONDON CLAY. SUPPLEMENT TO THE ORDER—CHELOMIA. Family— PALUDINOSA. Pratemys BowERBANKU (7). Tab. XXIX, figs. 1, 2. The evidence of species of Chelonia of the Fresh-water or Marsh-dwelling family, Paludinosa, has hitherto been derived only from such parts of the skeleton of the trunk as have been described, figured, and referred to the genera Platemys and Hmys, in Part I of the present Monograph, pp. 62-76. Since those pages were sent to press, Mr. Bowerbank has been so fortunate as to obtain from the Kocene clay at Sheppy the portion of fossil skull, of which two views are given of the natural size in T. XXIX, figs. 1, 2. If these figures, and especially the side view, fig. |, be compared with the corresponding view of the skulls, T. I, fig. 1; T. III, fig. 1; T. XV, fig. 1, or T. XI, fig. 2, a marked difference will be dis- cerned in the form and proportion of the orbit, which is smaller and more nearly circular in fig. 1, T. XXIX. But the bony chamber for the eyeball forms one of the characters by which the skull is distinguished in the marine and fresh-water families of the order Chelonia. The orbit, for example, is always much larger in proportion to the entire skull in the marine species, and commonly of the oval form, which is preserved in the beautiful fossil skull of the Chelone cuneiceps, T. XV, fig. 1; or with the upper and outer part even more produced and angular than is there represented. In the families HZwialia (Trionyx) and Paludinosa (Emydians), the orbit is not merely much smaller in propor- tion to the skull, it is circular, or nearly so, and not produced at the upper and outer angle. By this character, we are led to refer the fossil skull under description to the fresh-water division of the Chelonian order. Our choice between the Fluviatile or Paludinose families of that division is guided by the formation of the border of the orbit, and by the proportionate length and the form of the face or muzzle in advance of it. 1 FOSSIL REPTILIA OF THE LONDON CLAY. te) In the species of Himys (Podocnemys expansa) which I have selected for comparison, as offering upon the whole the nearest approach, which any Chelonian skull at my command gives, to the unique fossil in question, the malar bone (7, in Cuvier’s figure of the skull of Hmys expansa, pl. xi, fig. 9, of the ‘Ossemens Fossiles,’ tom. vy, pt. ii, 1825; 26 in the figure of the fossil, fig. 1, T. X XIX), becomes much contracted as it approaches the orbit, to which it contributes a small part of the posterior border. In the Chelones the malar bone forms a larger proportion of the orbital rim (see Cuvier, tom. cit., pl. xi, fig. 17), and contributes more to its under than its back part, which is chiefly formed by the characteristically large postfrontal 12 (g in Cuvier’s figs.); and this character was manifested in the ancient Eocene turtles as well as in the modern species, as may be seen by reference to the bones numbered 26 and 12, in T. I, fig. 1; T. XV, fig. 1, of the present work. The superior maxillary bone 21 (4 in Cuvier’s figs.) is longer in the Hmys, extends further back in the orbit, and is deeper at its posterior termination, than in the Chelones. In all these characters, derived from the bones entering into the formation of the orbit, the fossil under comparison agrees with the Amys, and, indeed, departs further from the Chelones than the Podocnemys expansa does, by the much smaller proportion in which the anteriorly contracted malar bone (26) contributes to the rim of the orbit. In the Zrionyces, the malar bone forms a larger proportion of the border of the orbit than in the Podocnemys expansa, and a fortiori, than in the fossil in question. The choice between the Fluviatile or Paludinose tribes of the fresh-water Chelonians, in the determination of this fossil, is better guided by the form and proportions of the skull anterior to the orbit. In the recent Zvionyces the muzzle is more acute, and in most of them more prolonged than in the Emydians, with which the fossil skull agrees in the shortness of the muzzle; whilst it departs further than most recent Emydians from the Zrionycid@, in the broad truncated character of its anterior termi- nation. There is also a very well-marked character of affinity to the Podocnemys eapansa, in the smooth and shallow canal which extends from the fore part of the orbit forwards to the border of the external nostril across the upper part or nasal process of the superior maxillary bone (21). This groove is very accurately represented in fig. 1, T. X XIX, in the fossil; itis rather broader in proportion to its length in the Podocnemys expansa ; but so far as it has depended upon the presence and arrangement of the facial scutes, it is decisive against the fossil having appertained to any species of soft turtle (Zrzonyx), in which such epidermal parts were entirely wanting. The marks of the supracranial scutes in the fossil are, as in some Emydians, too feebly and obscurely traceable to permit of a satisfactory comparison of their arrange- ment. The exterior surface of the prefrontal (16), frontal (11), postfrontal (12), and parietal (7) bones is subreticulate. The substance of the bones is thick and coarsely cancellous. The nasal bone is connate with the prefrontal, as in most modern Emydians ; in the proportion of this compound bone the fossil resembles more the ordinary CHELONIA. 3 Emydians (mys europea, e. g.) than it does the Podocnemys exapansa. The border of the prefronto-nasal bone forming the upper part of the nostril is thick and rounded ; as is also the lateral border of the same cavity formed by the maxillary. The lower part of this border of the maxillary shows the suture for the premaxillary, which must have presented similar proportions to the premaxillary of the Podocnemys expansa and other Emydians. The shape of the frontal (11), the proportion of the upper border of the orbit which it forms, and the course of its sutures with the contiguous bones, are clearly indicated in fig. 2, T. XXIX. The straight line formed by the suture between the frontal (11) and postfrontal (12) resembles that in the Podocnemys expansa ; it is bent or curved in the Chelones. To what extent the postfrontal (12) was continued backwards, whether so, as with the parietal, to roof over the temporal fossa, as in the Podocnemys expansd,* or, in a less degree, leaving that fossa open superiorly, as in the Emydians generally, is a question which will require for its determination a more perfect specimen than the fossil under description. The thickness, however, of the fractured posterior part of the postfrontal indicates that the bone had been broken not very close to its natural posterior border, on the supposition that this was free, as in the Emydians generally ; and the part of the suture of the postfrontal with the parietal which has been preserved, extends obliquely outwards and backwards, as in Po- docnemys expansa, not directly backwards, as in most of the Limydes with open temporal fossee. (Compare Cuvier, loc. cit., fig. 10 with fig. 14, the suture between g and /.) With respect to the parietal bones (7), these are too much mutilated to show more than the position and extent of the coronal suture. A few words may be perhaps expected relative to the difference which the fossil in question presents to the land-tortoises. In comparison with the skull of a Zestudo indica of corresponding dimensions with the fossil, the larger proportional size of the orbits distinguishes the skull of that terrestrial species almost as strongly as the same character does the skull of the marine turtles. But in addition to this, the malar bone forms a larger proportion of the back part of the orbit in the Zestudo, and the prefronto-nasal part of the skull is more bent down ; the suture between the frontals and prefrontals describes a curve convex forwards in the Testudo, whilst it deviates very little from a straight line in the fossil, and that little is convex backwards. The extent also of the upper surface of the postfrontals and parietals, so far as these are preserved in the fossil, is greater than the whole of those bones in the land-tortoise compared. Having been led by the foregoing comparisons to refer the fragment of the fossil skull (T. XXIX, figs. 1, 2) to the family Paludinosa, it is reasonable to conjecture that it may have appertaimed to some one of the large Emydians, which we already know to have left their carapaces in the Eocene clay of Sheppy. One commonly finds in * See Cuvier, loc. cit., pl. xi, fig. 1 a. z FOSSIL REPTILIA OF THE LONDON CLAY. the recent skeletons of Emydians, that any particular character of the exterior surface of the bones of the trunk is repeated on the upper surface at least of the bones of the head. This comparison, in the present instance, indisposes me to regard the fossil in question as having belonged to the Hmys levis, or to the Himys bicarinata, or to the Platemys Bullockwt with the punctate plastron. I should be rather led to select the Platemys Bowerbankiu from the character in question, as exhibited by the carapace and plastron described at p. 66. But, in provisionally registering the fossil skull in question under the name of Platemys Bowerbanku, 1 should wish to be understood as by no means vouching for the accuracy of the reference. The conjecture rests solely on the character above referred to, which is far from being decisive; and its only value is, that it happens to be the only one by which we can be guided at present in forming any opinion at all as to the specific relations of the fossil in question. TAB: XOXO CHELONIA. m Side view of the fore part of the skull of the Platemys Bowerbanki, nat. size. Upper view of the same fossil. Side view of a fractured tympanic bone of a large Turtle (Chelone), from Brackle- iy sham, showing the long and slender ossicle or ‘ columella’ (16) 2” situ ; nat. size. Extremity of the same tympanic bone, to which the ‘membrana tympani was © ainmelnet 7 nat. size. 5’. Left femur of a Turtle (Chelone = which wergned 150 Ibs., nat. size. Day see tith la che ee J Lradeben MONOGRAPH ON Pee Oss th th Pris OF THE LONDON CLAY, AND OF THE BRACKLESHAM AND OTHER TERTIARY BEDS. eeAU Relea ale Paces 5—50; Puares [—XIfand IIa. CROCODILIA (Crocoptius, &c.). BY PROFESSOR OWEN, D.C.L., F.R.S., F.L.S., F.G.S., &e. sued in the Volume for the Year 1849. LONDON: PRINTED FOR THE PALZONTOGRAPHICAL SOCIETY. 1850. PART II. OrDER—CROCODILIA. CROCODILES, ALLIGATORS, GAVIALS. Or the numerous and various kinds of Reptiles, the fossil remains of which have been discovered in the tertiary and secondary strata of Great Britain, many are found to have their nearest representatives, amongst the actual members of the class, in the present order; and here more particularly in the long and narrow-snouted genus called, through a corrupt latinization of its native name, Gavialis, which is now represented by the Gavial or, more properly, Garrhial, of the river Ganges. In the interpretation of the fossil remains of Reptiles, no skeleton has more frequently to be referred to than that of the Gavial or Crocodile, or has thrown more light on the nature of those singularly modified forms of the class which have long since passed away. It is accordingly requisite for the paleontologist who would describe the fossil remains of reptiles, to make himself, in the first place, thoroughly conversant with the osteology of the recent Crocodila. This knowledge can be gained only by assiduous study of the skeletons themselves, with the aid of the best descriptions, or the guide of a competent teacher. But to enable the reader to follow or comprehend the description of the fossil Saurians, some elementary account of the Crocodilian skeleton is at least necessary, accompanied with illustrations of the parts which, in the sequel, will have to be frequently referred to under special or technical names. In Tab. XI of the present part of this Monogragh is given a reduced or miniature side view of the skeleton of a Gavial which was twenty-five feet in length—dimensions which are rarely found to be surpassed in the present day. Beneath it is a restoration of the skeleton of the Teleosaur, or extinct Gavial of the Triassic or Oolitic period, showing how closely the general type of conformation has been adhered to, the modi- fications of the more ancient form of Crocodile evidently adapting it for moving with greater speed and facility through the water, and indicating it to have been more strictly aquatic, and probably marine. The particular nature of these modifications will be explained when I come to describe the Crocodiles of the secondary strata. I propose at present to give a pre- liminary sketch of the osteology of the recent Crocodilia. A glance at a natural or well-articulated skeleton of one of these reptiles, such as 6 FOSSIL REPTILIA OF THE LONDON CLAY. is figured in T. XI, will show that it consists mainly of a series of segments, more or less alike. From the back of the head to the end of the tail, the chief part of each segment consists of a cylindrical portion or ‘body,’ differmg only m its proportions, and diminishing as it recedes from the trunk. Every segment sends a plate of bone upwards from its upper or dorsal surface, which plate or ‘ spine’ is supported by an arch of bone, except in the diminishing segments at the end of the tail. Other plates of bone, of more variable forms and dimensions, project from each side of the segments of the trunk and basal part of the tail. In a less proportion, but still in a great number of the segments, an arch of bone is formed below, or on the ventral side of the cylindrical body ; but this lower arch is more variable in its proportions and mode of composition than the upper arch: it is open or incomplete in the neck. Under all these variations, however, there is plainly manifested a fundamental unity of plan in the composition of the different segments, which have accordingly received the common appellation of ‘ vertebra.’ For the convenience of description, the vertebree are divided, though somewhat arbitrarily, imto groups bearing special or specific names. Those next the head, with the inferior arch incomplete below, are called ‘ cervical vertebree;’ they are usually nine in number: those that follow with the inferior arch closed below, or which have the laterally projecting parts slender and freely moveable, are called ‘dorsal vertebre ;’ the other vertebree of the trunk that have no lateral moveable appendages, are called ‘lumbar vertebrae; the last vertebree of the trunk, always two in number in the Crocodilia, the inferior arches of which coalesce to support and be supported by the hind limbs, are the ‘ sacral vertebree ;’ the segments of the tail are the ‘ coccygeal,’ or ‘caudal vertebree, whether they possess or not an inferior arch, or whatever other modifications they may offer. These names, ‘ cervical,’ ‘dorsal,’ ‘lumbar,’ ‘ sacral,’ ‘coccygeal,’ were ori- ginally applied to corresponding segments or vertebree in the human skeleton, from the study of which the nomenclature of osteology takes its date: 1t may well be supposed, therefore, that a classification and designation of vertebree based upon knowledge limited to their characters in a single example of the vertebrated series, and that example one in which the common type has been most departed from, to adapt it to the peculiar attitude and powers of the human species, would fall far short of what is required to express the general ideas derived from a comparison of all the leading modifications of the vertebrate skeleton; and accordingly the anatomist who passes from a previous acquaintance with human osteology only, to the study of those of the lower Vertebrata, finds that he has to rectify, in the first place, the erroneous notions which anthropotomy has taught him of the nature of the primary segment of his own and other vertebrated skeletons, and to acquire true ideas, with the concomitant nomenclature, of the essential constituents or anatomical elements of such segment. In human anatomy, for example, the costal elements are only recognised when they CROCODILIA. 7 retain throughout life that distinctness, or moveable union with the rest of their segment, which they manifest at their first appearance ; and they are then classified as distinct bones from the rest of their segment, to which the term ‘vertebra’ is restricted, and which is equally regarded as a single bone; as, e. g., in the dorsal region of the skeleton. In the cervical region the whole segment is called ‘ vertebra,’ and is recognised as the equivalent bone to a dorsal vertebra, although it includes the costal elements, because these have coalesced with the rest of their segment, which anchylosis is misinterpreted as a mere modification of a transverse process; and the ‘cervical vertebra’ is distinguished by having that process ‘perforated, and not entire as in the other vertebre. But, in the Crocodile, the embryonic condition of the cervical ribs in Man is retained throughout life; and, therefore, if we were to be guided by the characters laid down by the recognised authorities in anthropotomy for the classification of its vertebra, we should seek in vain for any vertebree with “ transverse processes perforated for the transmission of vertebral arteries,’ whilst we should find all the vertebre from the head to the loins, ‘‘ with articular surfaces, either on their sides or their transverse processes, where they join with ribs,” and should accordingly have to reckon these as “ dorsal vertebra.” These and many similar instances which might be adduced, have compelled me to premise a few brief explanations of the principles and nomenclature by which I shall describe the fossil remains of the Repézlia, and illustrate their nature by reference to the skeletons of their existing representatives, in the present and succeeding Monographs. The primary segment of the skeleton of all Vertebrata isa natural group of bones, which may be severally recognised and defined under all the modifications to which such segment may have been subjected in subservient adaptation to the habits and exigencies of a particular species. A view of such a segment, as it exists in the thorax of the crocodile, the tortoise, and the bird, is given at p. 5, Part I, of the present Monograph. The part marked ¢ is the ‘centrum,’ or body of the vertebral segment; it is always developed originally as a separate element, and retains its character of individuality in the tortoise and crocodile. The bony arch above the centrum was formed originally by two distinct side-plates,—the ‘ neurapophyses,’ x, which coalesce with one another at their summits and thence develope a median plate or process of bone called the ‘neural spine’ ms. Other bony processes which shoot out from the neurapophyses are more variable, and will be afterwards noticed. The arch so formed coalesces with the centrum in the bird, and constitutes an apparently single bone, to which, in anthropotomy, the name ‘ vertebra’ would be restricted. But it would be as reasonable to confine it to the central element (c) in the tortoise and crocodile ; for the parts of the inferior arch are not less essentially parts of the same natural segment, than the neurapophyses which have formed the upper arch. The next pair of elements, 8 FOSSIL REPTILIA OF THE LONDON CLAY. then, which we have to notice, is marked in figs. 4, 5, and 6, p/, signifying ‘ pleura- pophysis, the name of these elements. In the segments figured they retain their primitive distinctness, and acquire unusual length, in order to aid in encompassing the dilated canal or eavity for the heart and lungs: so modified, these elements are commonly called ‘ribs,’ or ‘ vertebral ribs.’ The elements more constantly employed to protect the vascular or ‘ heemal’ axis, in other words, to form the inferior or heemal canal, are those marked / in figs. 4 and 6: they are the ‘ hemapophyses,’ which are usually articulated, like the neurapophyses, with the centrum, but are displaced by the great centres of the vascular system in the thorax, where they have got the special name of ‘ sternal ribs,’ and also that of ‘ costal cartilages,’ or ‘cartilages of the ribs,’ when they do not become ossified. The hemal arch in the thorax is usually completed by a median element (4s), called a ‘hzemal spine, but which itself becomes vastly expanded in the bird (fig. 4) ; it is, nevertheless, the part in the hemal arch which repeats below, or answers to the part (zs) m the upper arch. In the segments of the trunk and tail, the element (zs) retains its normal size and form as a ‘neural spine; but where the central axis of the nervous system becomes unusually developed, as in the head, e. g., analogously to the development of the vascular centres in the chest, the neural canal is correspondingly expanded, and the cavity acquires a special name, and is called ‘ cranium,’ just as the analogously expanded heemal canal is called ‘thorax.’ Into the formation of the wall of the cranium other vertebral elements enter besides the neurapophyses, those e. g. which are numbered 8 and 12 in fig. 9, p. 17, of the present Part; the neural spine (7 and 11 in the same figure) retains its primitive distinctness, is expanded horizontally, and, like the ‘ sternum’ in the thorax of the bird (As, fig. 4, p. 5, Part I), it receives a special name (7), e. g. of ‘ parietal’, and (9) of ‘ frontal’ in fig. 9. The elements a a (figs. 4 and 6, Part I) form a symmetrical pair of bones or cartilages, attached at one end to the heemal arch, and projecting out- wards and backwards. These are the ‘ prosartemata, or appendages ; they are, of all the elements of the vertebral segment, those that are least constant in regard to their presence, and, when present, are subject to the greatest amount of development and metamorphosis: they become, e. g., the pterygoid appendages in the nasal segment of the fish’s skull, the opercular bones in the frontal segment of the fish, the branchiostegal rays in the parietal segment, the pectoral fins in the occipital segment, and they are deve- loped into the fore limbs and hind limbs, the arms, wings, and legs of other Vertebrata.* As the nervous and vascular centres become reduced in size, the bony canals or arches protecting them are simplified and contracted, and the vertebra assumes a symmetrical character. In the Crocodile, the hemal arch, in the tail,e. g., is formed by the hemapophyses, which ascend and articulate directly with the centrum; the pleurapophyses are shortened, directed outwards, and become anchylosed to * The facts and arguments in support of this conclusion, are detailed in my works ‘On the Nature of Limbs,’ and ‘On the Archetype of the Vertebrate Skeleton,’ 8vo (Van Voorst). CROCODILIA. 9 form ‘transverse processes ;’ but such a vertebra, when analysed as it is developed, resolves itself very nearly into the ideal type Hig. 7. given in the subjomed diagrammatic cut (fig. 7) ; 2 is the neural axis, called ‘myelon,’ or ml 5 ‘spinal marrow ;’ /is the heemal axis, the chief VAS iE sea te trunk of which is called ‘aorta,’ and ‘ caudal ii Wiis mes artery. The names of the vertebral elements which, being usually developed from distinct parapophysis<” centres, are called ‘ autogenous,’ are printed in Vv [. "+ hamapophysis. \_-——---neural spine. ay Roman type; the /¢a/ics denote the ‘ exogenous’ eygapophysis.” parts, more properly called ‘ processes,’ which | aa hemal spine. shoot out from the preceding elements. Ideal typical vertebra. On comparing this form of the primary segment with that figured in Cut 4, p. 5, Part I, it will be seen that they differ by altered proportions with some change of position of certain elements; but every modification resulting in the various forms of the parts of the skeleton figured in T. XI, has its seat in one or other of the segmental or ‘vertebral’ elements above defined ; and the same principle I believe that I have established with regard to the internal skeleton in all vertebrate animals. With this preliminary explanation, the nature and relations to the typical vertebra of the parts of the Crocodilian vertebre, figured in T. IV, V, IX, will be, itis hoped, readily appreciated. In T. IX, in which are figured, through opportunities kindly afforded by the Marchioness of Hastings, and Searles Wood, Hsq., F.G.S., some of the most perfectly-preserved fossil reptilian vertebrae which have hitherto been discovered, the elements and processes are indicated by the initial letter of their names. Figs. 1 and 2 give a side view and a back view of a cervical vertebra, apparently the fourth, of the Crocodilus Hastingsie, from the Eocene deposits at Hordwell; cis the centrum, ~ the neural canal formed by the neurapophyses, which have coalesced superiorly with each other, and with the neural spine (ws). Inferiorly they articulate by a suture (which is shown by the wavy line on each side of the process d in fig. 1) with the centrum; p/ is the pleurapophysis, which articulates by two parts, the lower one called the ‘head’ to the process from the centrum, the upper one called the ‘ tubercle’ to the process from the neurapophysis ; beyond the union of the head and tubercle, the pleurapophysis projects freely outwards and downwards, but instead of being elongated in that direction, it becomes expanded in the direction of the axis of the body, i. e. forwards and backwards, and so acquires a shape which has given rise to the name ‘hatchet bone’ or ‘ hatchet-shaped process,’ * applied to this element in the Plesiosaurus. * «lo compensate for the weakness that would have attended this great elongation of the neck, the Plesiosaurus had an addition of a series of hatchet-shaped processes on each side of the lower part of the cervical vertebra.” (Buckland, Bridgewater Treatise, vol. i, p. 206, and vol. ii, p. 30, 1836.) Cuvier recognised in these lateral bones, “en forme de hache,” the homologues of the “ petites cétes cervicales” of the Crocodile. (Ossemens Fossiles, 4to, tom. v, pt. ii, p. 479, 1824.) And Conybeare had 2 10 FOSSIL REPTILIA OF THE LONDON CLAY. The purport of this modification is the same in the Crocodilia as that which seems to be more called for in the Plesiosaurus, viz. to augment the strength of the cervical region of the skeleton ; and this is so effectually done by the overlapping of the hatchet- shaped ribs of this region in the Crocodila, as shown in T. XI, that the flexibility of the neck is much restricted, although the joint of the head allows that part to be bent from side to side at nearly right angles with the neck. When, however, the head is held firmly forwards by its powerful muscles, the imbricated vertebree of the neck transmit with great effect the impulse which the strong and long tail gives to the rest of the body in the act of swimming. In T. IX, fig. 3 the cervical vertebra is represented minus its pleurapophyses, and it answers accordingly to that portion of the natural segment to which the term ‘ vertebra’ is usually restricted in the dorsal region of the trunk. The exogenous processes shown in this view of the vertebra are, p, the ‘parapophysis’ or inferior transverse process, developed from the centrum; d, the ‘diapophysis’ or upper transverse process de- veloped, as in most cases it is, from the neurapophysis; z, 2’, are the ‘ zygapophyses’ or ‘oblique processes, which, from their function in articulating together contiguous vertebree, are also called ‘articular processes.’ In most of the cervical, and in some of the dorsal, vertebree of the Crocodile, an exogenous process is developed from the under surface of the centrum, called ‘hypapophysis ;’ it is indicated by the letters Ay in fig. 2, T. IX. In some species it is double,* and beneath the atlas it becomes ‘ auto- genous’ or is developed as a separate element, ca, ex, of the subjoined Cut, fig. 8, in which condition the part is found beneath the centrums of two or three of the anterior cervical vertebree in the Ichthyosaurus.t+ The first and second vertebree of the neck are peculiarly modified in most air-breath- ing Vertebrata, and have accordingly received the special names, the one of ‘atlas,’ the other of ‘ epistropheus’ or ‘ axis.’ In Comparative Anatomy these become arbitrary terms, the properties being soon lost which suggested those names to the human anatomist; the ‘atlas e. g. has no power of rotation upon the ‘axis’ in the Crocodile, and it is only in the upright skeleton of man that the Atlas and Axis vertebre 4rge globular head is sustained upon the shoulder-like processes of the Crocodile. of the ‘ atlas.’ In the Crocodile, these vertebree are concealed by the peculiarly prolonged angle of the lower jaw in the side view of the skeleton in T. XI, previously extended the same homology to the “particularly prominent wing-like appendages to the transverse processes in many of the long-necked quadrupeds, and the long styloid processes of the cervical vertebra of birds.’ (See his admirable Memoir of June 14th, 1822, in the Geol. Trans., 2d series, vol. i, p. 384.) * In Crocodilus basifissus, e. g., see the Quarterly Journal of the Geological Society, November 1849, p. 38], pl. x, fig. 2. + This teresting discovery was communicated by its author, Sir Philip de M. Grey Egerton, Bart., F.G.S., to the Geological Society of London, in 1836, and is published in the fifth volume of the second series of their Transactions, p. 187, pl. 14. CROCODILIA. 11 and a separate view of them is, therefore, given in figure 8. The pleurapophyses are retained in both segments, as in all the other vertebre of the trunk. That of the atlas, fig. 8, p/a, is a simple slender style, articulated by the head only, to the mdependently developed inferior part of the centrum, or ‘hypapophysis’ (ca, ev). The neurapophyses (wa) of the atlas retain their primitive distinctness ; each rests in part upon the proper body of the atlas (ca), in part upon the hypapophysis. The neural spine (zs, @) is also here an independent part, and rests upon the upper extremities of the neurapophyses. It is broad and flat, and prepares us for the further metamorphosis of the corresponding element in the cranial vertebree. The centrum of the atlas (ca), called the odontoid process of the epistropheus in Human Anatomy, here supports the abnormally-advanced rib of that vertebra, which in some Crocodilia is articulated by a bifurcate extremity, like the ribs of the succeeding cervical vertebree ; but it is not expanded or hatchet-shaped at the free extremity. The proper centrum of the axis vertebra (cv) is the only one in the cervical series which does not support a rib ; it articulates by suture with its neurapophyses (zz), and is characterised by having its anterior surface flat, and its posterior one convex. With the exception of the two sacral vertebre, the bodies of which have one articular surface flat and the other concave, and of the first caudal vertebra, the body of which has both articular surfaces convex, the bodies of all the vertebree beyond the axis have the anterior articular surface concave, and the posterior one convex, and articulate with one another by ball-and-socket joints. This type of vertebra, which I have termed ‘ procce- lian, * characterises all the existing genera and species of the family Crocodilia, with all the extinct species of the tertiary periods, and also two extinct species of the Greensand formation in New Jersey. Here, so far as our present knowledge extends, the type was lost, and other dispositions of the articular surfaces of the centrum occur in the vertebrae of the Crocodilia of the older secondary formations. The only known Crocodilian genus of the periods antecedent to the Chalk and Greensand deposits with vertebre articulated together by ball-and-socket joints, have the position of the cup and the ball the reverse of that in the modern Crocodiles, and the genus, thus cha- racterised by vertebre of the ‘opisthoccelian’ type, has accordingly been termed Streptospondylus, signifying ‘vertebre reversed.’ The aspects of the zygapophyses are, however, more constant ; the anterior ones, T. IX, fig. 3 z, look obliquely inwards ; the posterior ones, ib. 2’, obliquely outwards. In looking, therefore, upon the cut surface of a vertical longitudimal section of a Crocodilian vertebra, the smooth, flattened inner surface of the anterior zygapophysis is turned towards the observer, and the convex outer surface of the posterior zygapophysis. Thus the anterior and posterior extremity of the vertebra being determined by observation of the aspect and direction of the zygapophyses, it is at once seen whether the body has the * TIpos, before ; cotAos, concave. + Quarterly Journal of the Geological Society, November 1849. 12 FOSSIL REPTILIA OF THE LONDON CLAY. proccelian structure, as im the true Crocodiles, T. IV, T. IX, or the opisthoccelian structure, as in the Streptospondylus. But the most prevalent type of vertebra amongst the Crocodilia of the secondary periods was that in which both articular surfaces of the centrum were concave, but in a less degree than in the single concave surface of the vertebra united by ball andsocket. A section of a vertebra of this ‘ amphiccelian’ type, such as existed in the 7é/eosawrus and Steneosaurus, will be figured in a subsequent Monograph. In the Jchthyosaurus, the concave surfaces are usually remarkable for their depth, the vertebrae resembling in this respect those of fishes. Some of the most gigantic of the Crocodilia of the secondary strata had one end of the vertebral centrum flattened, and the other (hinder) end concave; this ‘ platyccelian’ type we find in the dorsal and caudal vertebre of the gigantic Cetiosaurus. With a few exceptions, all the modern Reptiles of the order Lacertilia have the same proceelian type of vertebra as the modern Crocodilia, and the same structure prevailed as far back as the period of the MWosasawrus, and in some smaller members of the Lacertilian order in the Cretaceous and Wealden epochs. Resuming the special description of the osteology of the modern Crocodila, we find the proccelian type of centrum established in the third cervical, which is shorter but broader than the second; a parapophysis is developed from the side of the centrum, and a diapophysis from the base of the neural arch; the pleurapophysis is shorter, its fixed extremity is bifid, articulating to the two above-named processes; its free ex- tremity expands, and its anterior angle is directed forwards to abut against the inner surface of the extremity of the rib of both the axis and atlas, whilst its posterior pro- longation overlaps the rib of the fourth vertebra. The same general characters and imbricated coadaptation of the ribs characterise the succeeding cervical vertebree to the seventh inclusive, the hypapophysis (Ay, fig. 2, T. IX) progressively though slightly increasing in size. In the eighth cervical the rib becomes elongated and slender; the anterior angle is almost or quite suppressed, and the posterior one more developed and produced more downwards, so as to form the body of the rib, which terminates, however, in a free point. In the ninth cervical the rib is increased in length, but is still what would be termed a ‘ false’ or ‘floating rib’ in anthropotomy. In the succeeding vertebra the pleurapophysis articulates with a heemapophysis, and the hzemal arch is completed by a heemal spine; and by this completion of the typical segment we distinguish the commencement of the series of dorsal vertebrae. With regard to the so-called ‘ perforation of the transverse process,’ this equally exists in the present vertebra, as in the cervicals, as may be seen by comparing fig. 6, p. 5, Part I of this Monograph, with fig. 2,T. IX, Part II; in both, the foramen is the vacuity intercepted between the bifurcate extremity of the rib and the rest of the vertebra with which that rib articulates; and, on the other hand, the cervical vertebree equally show surfaces for the articulation of ribs. Cuvier, in including the proximal portions of the ribs with the rest of the vertebra, in his figure of a dorsal vertebra of a CROCODILIA. 13 Crocodile,* so far follows nature, and produces a parallel to his figure of a cervical vertebra; but the entire natural vertebra or segment includes the parts delineated in outline in Cut 6, p. 5, Part I. In that figure is shown the semiossified bar // which is interposed between the pleurapophysis p/ and hzmapophysis 4 in the Crocodilia and some existing Lizards. The typical characters of the segment due to the completion of both neural and hemal arches, is continued in some species of Crocodila to the sixteenth, in some (Crocodilus acutus) to the eighteenth vertebra. In the Crocodilus acutus and the Alhgutor lucius, the heemapophysis of the eighth dorsal rib (seventeenth segment from the head) jos that of the antecedent vertebra. The pleurapophyses project freely outwards, and become ‘floating ribs’ in the eighteenth, nineteenth, and twentieth vertebree, in which they become rapidly shorter, and in the last appear as mere appendages to the end of the long and broad diapophyses : but the hemapophyses by no means disappear after the solution of their union with their pleurapophyses ; they are essentially independent elements of the segment, and they are continued, therefore, in pairs along the ventral surface of the abdomen of the Crocodilia, as far as their modified homotypes the pubic bones. They are more or less ossified, and are generally divided into two or three pieces. Another character afforded by the hemal arch is the more important in reference to palzontology, as it affects the centrum and neural arch of the vertebra as well as the pleurapophysis; and thus aids in the determination of the vertebra. The parapophysis progressively ascends upon the side of the centrum in the two anterior dorsal vertebree, and disappears in the third, or, passing upon its neurapophysis, blends with the base of the diapophysis. In this segment, therefore, the proximal end of the rib ceases to be bifurcate, but is simply notched, the curtailed head being applied to the end of the thickened anterior part of the transverse process, and the tubercle abutting against its extremity ; in the five following dorsals the head and tubercle of the rib progressively approximate and blend together, or the head disappears m the tenth dorsal, in which the rib is simply attached to the end of the diapophysis. The hypapophysis ceases to be developed after the third or fourth dorsal vertebra. The zygapophyses become gradually more horizontal, the anterior ones looking more directly upwards, the posterior ones downwards. The ‘lumbar vertebre’ are those in which the diapophyses cease to support moveable pleurapophyses, although they are elongated by the coalesced rudiments of such which are distinct in the young Crocodilia. The development and persistent individuality of more or fewer of these rudimental ribs determimes the number of the dorsal and lumbar vertebre respectively, and exemplifies the purely artificial character of the distinction. The number of vertebree or segments between the skull and the sacrum, in all the Crocodilia I have yet examined, is twenty-four. In the skeleton of * Ossemens Fossiles, 4to, tom. v, pt. 11, pl. iv, fig. 4. 14 FOSSIL REPTILIA OF THE LONDON CLAY. a Gavial I have seen thirteen dorsal and two lumbar; im that of a Crocodilus cata- phractus twelve dorsal and three lumbar; in those of a Crocodilus acutus, and Alligator /ucius, eleven dorsal and four lumbar, and this is the most common number; but in the skeleton of the Crocodile, I believe of the species called Croc. diporcatus, described by Cuvier,* he gives five as the number of the lumbar vertebre. But these varieties in the development or coalescence of the stunted pleurapophysis are of little essential moment; and only serve to show the artificial character of the ‘dorsal’ and ‘ lumbar’ vertebree. The coalescence of the rib with the diapophysis obliterates of course the character of the ‘costal articular surfaces ;) which we have seen to be common to both dorsal and cervical vertebre. The lumbar zygapophyses have their articular surfaces almost horizontal, and the diapophyses, if not longer, have their antero-posterior extent somewhat increased ; they are much depressed, or flattened horizontally. The sacral vertebre are very distinctly marked by the flatness of the coadapted ends of their centrums: there are never more than two such vertebre in the Crocodilia recent or extinct; in the first the anterior surface of the centrum is concave, in the second it is the posterior surface; the zygapophyses are not obliterated m either of these sacral vertebrae, so that the aspects of their articular surface—upwards in the anterior pair, downwards in the posterior pair—determines at once the corresponding extremity of a detached sacral vertebra. The thick and strong transverse processes form another characteristic of these vertebra ; for a long period the suture near their base remains to show how large a proportion is formed by the pleurapophysis. This element articulates more with the centrum than with the diapophysis developed from the neural arch ; + it terminates by a rough, truncate, expanded extremity, which almost or quite joins that of the similarly but more expanded rib of the other sacral vertebrae. Against these extremities is applied a supplementary costal piece, serially homologous with the superadded piece to the proper pleurapophysis in the dorsal vertebra (“, fig. 6, p- 5, Part I), but here interposing itself between the pleurapophyses and heemapophyses of both sacral vertebrae, not of one only. This intermediate pleurapophysial piece is called the ‘ilium ;’ it is short, thick, very broad, and subtriangular, the lower truncated apex forming with the connected extremities of the hemapophysis an articular cavity for the diverging appendage, called the ‘hind leg. The hemapophysis of the anterior sacral vertebra is called ‘pubis: it is moderately long and slender, but expanded and flattened at its lower extremity, which is directed forwards towards that of its fellow, and joined to it through the intermedium of a broad, cartilagmous, hemal spine, completing the hemal canal. The posterior heemapophysis is broader, subdepressed, and subtriangular, expanding as it approaches its fellow to complete the second heemal * Tom. cit., p. 95. It is to be observed that Cuvier begins to count the dorsal vertebra when the rib has changed its hatchet-shape for a styloid shape. + Cuvier, who well describes this structure, remarks, ‘aussi méritent-elles plutét le nom des cétes que celui @’apophyses transyerses.”” (Tom. cit., p. 98.) CROCODILIA. 15 arch; it is termed ‘ischium.’ The great development of all the elements of these hemal arches, and the peculiar and distinctive forms of those that have thereby acquired, from the earliest dawn of anatomical science, special names, relates phy- siologically to the functions of the diverging appendage which is developed into a potent locomotive member. This limb appertains properly, as the proportion contributed by the ischium to the articular socket and the greater breadth of the pleurapophysis show, to the second sacral vertebra; to which the ilium chiefly belongs. The first caudal vertebra, which presents a ball for articulating with a cup on the back part of the last sacral, retains, nevertheless, the typical position of the ball on the back part of the centrum ; it is thus biconvex, and the only vertebra of the series which presents that structure. I have had this vertebra in three different species of extinct Eocene Crocodilia. In the Crocodilus toliapicus,T.1V, fig. 7; in the Croc. champsoides, T. V, fig. 10; and in the Crocodilus Hastingsie, T. IX, fig. 7. The advantage of possessing such definite characters for a particular vertebra is, that the homologous vertebra may be compared in different species, and may yield such distinctive characters as will be hereafter pointed out in those of the three species above cited. The first caudal vertebra, moreover, is distinguished from the rest by having no articular surfaces for the heemapophyses, which in the succeeding caudals form a hzemal arch, like the neurapophyses above, by articulating directly with the centrum. The arch so formed has its base not applied over the middle of a single centrum, but like the neural arch in the back of the tortoise and sacrum of the bird, across the interspace between two centrums. The first hzemal arch of the tail belongs, however, to the second caudal vertebra, but it is displaced a little backwards from its typical position. The detached centrum of a caudal vertebra, besides being more slender and com- pressed, is distinguished from those of the before-described vertebrze by the two articular surfaces at the posterior border of their under surface T. IV, fig.9. The zygapophyses become vertical as far as the sixteenth or seventeenth, beyond which the two posterior zygapophyses coalesce in an oblique plane notched in the middle, which is received into a wider notch at the fore part of the neural arch of the succeeding vertebra. The sutures between the pleurapophyses and diapophyses are maintained during a long period of the animal’s growth, and demonstrate the share which these two elements respectively take in the formation of the transverse process. So constituted, these processes progressively decrease in length to the fifteenth or sixteenth caudal vertebra, and then disappear. The neural spmes progressively decrease in every dimension, save length, which is rather increased as far as the twenty-second or twenty-third vertebra, beyond which they begin again to shorten, and finally subside in the terminal vertebree of the tail. 16 FOSSIL REPTILIA OF THE LONDON CLAY. The caudal hemapophyses coalesce at their lower or distal ends, from which a spinous process is prolonged downwards and backwards; this grows shorter towards the end of the tail, but is compressed and somewhat expanded antero-postenunag The heemal arch so constituted has received the name of ‘ chevron bone.’ A side view of the body of a middle caudal vertebra of the Crocodilus tohapicus is given in T. III, fig. 8, and an under view of the same in fig. 9, showing the two hypapophysial ridges extending from the articular facets for the hemapophyses at one end to the other end of the centrum. The segments of the endo-skeleton composing the skull are more modified than those of the pelvis ; but just as the vertebral pattern is best preserved in the neural arches of the pelvis, which are called collectively ‘sacrum,’ so, also, is it in the same arches of the skull, which are called collectively ‘ cranium.’ The elements of which these cranial arches are composed preserve, moreover, their primitive or normal individuality more completely than in any of the vertebrz of the trunk, except the atlas, and consequently the archetypal character can be more completely demonstrated. In fossil Crocodilia, and many other reptiles, the bones of the head are very hable from this cause to a greater extent of dislocation and separation than happens to the skull of the warm-blooded animal, in which a greater proportion of those primitive bones coalesce with age. It not unfrequently happens that detached bones of the skull of a reptile are found fossil, and the usually much modified form of these vertebral elements renders their determination difficult. In order to diminish this difficulty, especially as the bones of the cranium are least familiar to the paleontologist in their detached state, I have subjoined a side view of them, fig. 9, nearly as they are arranged in the formation of the successive natural segments of the skull. Such figures are the more necessary in the present state of anatomy and paleontology, since the illustrations of the osteology of the crocodile which have hitherto been prefixed to the descriptions of the fossil remains of the Reptilian class, as, e. g., in the great work of Cuvier, include only figures of the bones in question as they are naturally combined together in the entire skull. For the anatomical description and determination of the individual bones, as con- stituent elements of the vertebral segments of the head, I must refer the reader to my work ‘On the Archetype of the Vertebrate Skeleton, pp. 115-25, figs. 18-21, and pl. 2, fig. 3; and here limit myself to an exemplification of the natural arrangement and names of the bones according to the letters and numbers in figure 9. The bones of the head of the Crocodiles, as of all other vertebrate animals, are primarily classified into those of The ENDO-SKELETON, The SPLANCHNO-SKELETON, and The ExXo-SKELETON. CROCODILIA. 17 Disarticulated bones of the Skull of an Alligator, N1 to rv the neural arches; H1 to 1v the hemal arches and appendages. The bones of the exdo-skeleton of the head form naturally four segments, called Occipital vertebra, N 1, H1; Parietal vertebra, N 11, H 11; Frontal vertebra, N 111, H 111; Nasal vertebra, N tv, H tv. These segments are subdivided into the neural arches, called Epencephalic arch (1 basioccipital, 2 exoccipital, 3 superoccipital, 4 connate paroccipital) ; Mesencephalic arch (5 basisphenoid, 6 alisphenoid, 7 parietal, s mastoid) ; Prosencephalic arch (9 presphenoid, 10 orbitosphenoid, 11 frontal, 12 post- frontal) ; Rhinencephalic arch (13 vomer, 14 prefrontal, 15 nasal) : and into the hzmal arches and their appendages, called Maxillary arch (20 palatine, 21 maxillary, 22 premaxillary) and appendages (24 pterygoid, 24’ ectopterygoid, 26 malar, 27 squamosal) ; Mandibular arch (28 tympanic, 29—32 mandible) ; Bigs) 9: 18 FOSSIL REPTILIA OF THE LONDON CLAY. Hyoidean arch (39 epihyal, 40 ceratohyal, 41 basihyal) ; Scapular arch (50 suprascapula, 51 scapula, 52 coracoid) and appendages (53—58 bones of fore-limb). The bones of the splanchno-skeleton, are The petrosal (16) and otosteals (16/) ; The sclerotals (17) which in most retain their primitive histological condition as fibrous membrane. The turbinals (18 and 19) and teeth. The bones of the evo-skeleton, are The lacrymals (73). The superorbitals (present in Alligator sclerops). To the foregoing brief analysis of the constituent parts of the framework of the Crocodilia, which are petrifiable or conservable in a fossil state, and from the study and comparison of which we have to gain our insight into the nature and affinities of the extinct Reptiles, it seems here only requisite to add a few observations on the characteristic mode in which the bones are associated together in certain parts of the skeleton in the present order, and especially in the skull. With regard to the trunk, the Crocodila are distinguished from the Lacertilia and from all other existing orders of Reptiles, by the articulation of the vertebral ribs (pleurapophyses) in the cervical and anterior part of the dorsal segments by a head and tubercle to a parapophysis and diapophysis. As this double joint is associated with a double ventricle of the heart, and as the single articulation of every rib in other Reptiles is associated with a single ventricle of the heart, we may infer a like difference in the structure of the central organ of circulation in the extinct reptiles, manifesting the above-defined modifications in the proximal joints of the ribs. The sacrum consists of two vertebree only, in Crocodilia as in Lacertilia: they are modified in the present order, as before described, p. 14. The skull consists, as above defined, of four segments. The hinder or occipital surface of the skull presents, in the Crocodilia as in the Lacertilia, a single convex occipital condyle, formed principally by the basioccipital, and not showing the trefoil character which it bears in the Chelonia (Part I, T. XV, fig. 4), in which the exoccipitals con- tribute equal shares to its formation. In the Batrachia, the exoccipitals exclusively form _ the joint with the atlas, and there are accordingly two condyles. The occipital region of the crocodilian skull is remarkable for its solidity and complete ossification, and for the great extent of the surface which descends below the condyle. (T. VI, fig. 2.) In the Lacertilia, a wide vacuity is left between the mastoid, exoccipital, and par- occipital: but in the Crocodilia this is reduced to the small depressions or foramina near 3, fig. 2,T. VI. The tympanic pedicles (28) extend outwards and downwards, firmly wedged between the paroccipital, mastoid, and squamosal; in the Lacertians CROCODILIA. 19 these pedicles are suspended vertically from the point of union of the mastoid and paroccipital. The chief foramen in the occipital region is that called ‘foramen magnum’ (between 2 and 2, in fig. 2, T. VI), through which the nervous axis is continued from the skull. On each side of the foramen magnum isa small hole, called ‘ precondyloid foramen,’ for the exit of the hypoglossal nerve. External to this is a larger foramen, marked z in fig. 2, for the transmission of the nervus vagus and a vein. Below this is the “carotid foramen’ c. All these are perforated in the exoccipital. Below the condyle there is usually a foramen, and sometimes two, for the transmission of blood-vessels. Lower down, at the suture between the basioccipital and basisphenoid, is a larger and more constant median foramen, indicated by the dotted line from e f; it is the bony outlet of a median system of eustachian tubes, peculiar to the Crocodila. On each side of the median eustachian foramen, and in the same suture, is a smaller foramen, which is the bony orifice of the ordinary lateral eustachian tube. The membranous continuations of the lateral eustachian tubes unite with the shorter continuation from the median tube, and all three terminate by a common valvular aperture, upon the middle line of the faucial palate, behind the posterior or palatal nostril. The large, bony aperture of this nostril is formed by the pterygoids (24 in fig. 2). The carotid canal, c, opens by a short bony tube into the tympanic cavity, and is described as the ‘eustachian canal’ in the ‘Lecons d’Anatomie comparée’ of Cuvier. The artery crosses the tympanic cavity, and enters a bony canal at its fore part, which conducts to the ‘sella turcica’ in the interior of the cranium. The median eustachian foramen is described by Cuvier as the ‘arterial foramen,* the canal from which divides and terminates in the ‘sella turcica. | By MM. Bronn, Kaup, and De Blainville, the median eustachian foramen is contended to be the bony aperture of the posterior nostrils. { The results of the dissections and injections of recent Crocodiles and Alligators, by which I have been able to rectify the discrepant opinions regarding the carotid, eustachian, and naso-palatal foramina, and which have led to the discovery of a third median eustachian canal, or rather system of canals, between the tympanic cavities and fauces, peculiar to the Crocodilian Reptiles, are given ‘in detail in the ‘ Philosophical Transactions for 1850. The complexity of the superadded system has doubtless chiefly contributed to mislead the justly-esteemed authorities who have believed that they saw in it characters of the carotid canals or of the posterior nasal passages. The eustachian apparatus in the Crocodilia may be briefly described as follows: From the floor of each tympanic cavity two air-passages are continued ; the canal from the fore part of the cavity extends downwards, backwards, and inwards, in the basisphenoid, which * Ossemens Fossiles, tom. v, pt. li, p. 133. + Ib. p. 78. ft Abhandlungen iiber die Gavialartigen Reptilien der Lias-formation, folio, 1841, pp. 12, 16, 44. 20 FOSSIL REPTILIA OF THE LONDON CLAY. unites with its fellow from the opposite tympanum, to form a short median canal, which descends backwards to the suture between the basisphenoid and the basioccipital, where it joins the median canal formed by the union of the two air-passages from the back part of the floor of the tympanum, which traverse the basioccipital. 'The common canal formed by the junction of the two median canals descends along the suture to the median foramen e/, fig. 2, T. VI. The air-passage from the back part of the tym- panum, which traverses the basioccipital, swells out into a rhomboidal sinus im its convergent course towards its fellow, and from this sinus is continued the normal lateral eustachian canal, which, on each side, terminates below in the small aperture, external to the-median eustachian foramen. That part .of the outer surface of the skull which is covered by the common integument is more or less sculptured with wrinkles and pits in the Crocodilia: the modifications of this pattern are shown in T. I, fig. 1, in the nilotic Crocodile, and in T. VI, in the eocene Crocodile from Hordwell. The flat platform of the upper surface of the cranium is perforated by two large apertures, surrounded by the bones numbered 7, 8, 11, 12; these apertures are the upper outlets of the temporal fossz, divided from the lower and lateral outlets by the conjoined prolongations of the mastoid s and postfrontal 12: if ossification were continued thence to the parietal 7, the temporal fossee would be roofed over by bone, as in the Chelones. In old Crocodiles and Alligators there is an approximation to this structure, and the upper temporal apertures are much diminished in size. In the Gavials (T. XI, fig. 1 a) they remain more widely open, and, in the fossil Gavials of the secondary strata, they are still wider, as seen in T. XI, fig.2@; by which the structure of the cranium approaches more nearly to that of the Lacertian reptiles, where the temporal fossa is either not divided into an upper and lateral outlet, or is bridged over by a very slender longitudinal bar from the postfrontal to the mastoid. The lateral outlets of the temporal fossee (T. VI, fig. 1) are divided from the orbits by a bar of bone developed from the postfrontal (12) and malar (26), and against the imner side of the base of which the ectopterygoid abuts; the posterior boundary of the fossa is made by the tympanic (28) and squamosal (27). The orbits, having the postfronto-malar bar (12, 26) behind, are surrounded in the rest of their circumference by the frontal (11), the prefrontal (14), the lachrymal (73), and the malar (26). The supraorbital or palpebral ossicle is rarely preserved in fossil specimens. The facial or rostral part of the skull anterior to the orbit, is of great extent, broad and flat in the Alligators and some Crocodiles, narrower, rounder, and longer in other Crocodiles, always most narrow, cylindrical, and elongated in the Gavials. The anterior or external nostril is single, and is perforated in the middle of the anterior terminal expansion of the upper jaw. ‘This expansion is least marked in the broad- headed species (compare T. VI, fig. 1, with T. III, fig. 1); in existing Crocodiles and Alligators the points of the nasal bones penetrate its hind border, as at 15, fig. 1, CROCODILIA. 21 T. I. In the Gavials (T. XI, fig. 1) the nasals (~) terminate a long way from the nostril. The Crocodilia resemble the Chelonia in the single median nostril.* In the Lacertilia there is a pair of nostrils, one on each side the median plane, which is occupied by a bridge of bone extending from the usually single premaxillary to the nasals. The plane of the single nostril is almost horizontal in all existing and tertiary Crocodilia. On the inferior or palatal surface of the skull (T. VII, fig. 2), the most anterior aperture is the circular prepalatal foramen surrounded by the premaxillaries 22; then follows an extensive smooth, horizontal, bony plate, formed by the premaxillaries (22), the maxillaries (21), and the palatines (20). The postpalatal apertures are always large in the Crocodilia, and are bounded by the palatines (20), maxillaries (21), pterygoids (24), and ectopterygoids (25). The posterior aperture of the nostril is formed wholly by the pterygoids; it is shown in T. VI, fig. 3, between the bones marked 24. Behind it is the median and lateral eustachian foramen already described, as belonging rather to the posterior region of the head. The scapulo-coracoid arch, both elements (fig. 9, 51, 52) of which retain the form of strong and thick vertebral and sternal ribs in the crocodile, is applied in the skeleton of that animal over the anterior thoracic heemal arches (T. XI). Viewed as a more robust heemal arch, it is obviously out of place in reference to the rest of its vertebral segment. If we seek to determine that segment by the mode in which we restore to their centrums the less displaced neural arches of the antecedent vertebree of the cranium or in the sacrum of the bird,t we proceed to examine the vertebre before and behind the- dis- placed arch, with the view to discover the one which needs it, in order to be made typically complete. Finding no centrum and neural arch without its pleurapophyses from the scapula to the pelvis, we give up our search in that direction ; and in the opposite direction we find no vertebra without its ribs until we reach the occiput : there we have centrum and neural arch, with coalesced parapophyses—the elements answering to those included in the arch N1, fig. 9—but without the arch H1; which arch can only be supplied, without destroying the typical completeness of antecedent cranial segments, by a restoration of the bones 50—52, to the place which they naturally occupy in the skeleton of the fish. And since anatomists are generally agreed to regard the bones 50—52 in the crocodile as specially homologous with those so numbered in the fish,{ we must conclude that they are likewise homologous in a higher sense; that in the fish, the scapulo-coracoid arch is in its natural or typical position, whereas in the crocodile it has been displaced for a special purpose. Thus, agreeably with a general principle, we perceive that, as the lower vertebrate * Ina skeleton of the Alligator lucius in the Museum of the Royal College of Surgeons, a slender bar of bone is continued from the nasals to the premaxillary, across the median nasal aperture, as if is in the skull of the same species figured in the ‘Ossemens Fossiles,’ tom. v, pt. ii, pl. i, fig. 8. + See ‘On the Archetype and Homologies of the Vertebrate Skeleton,’ p. 117, p. 159. t Op. cit., fig. 5, p. 17. 22 FOSSIL REPTILIA OF THE LONDON CLAY. 5 animal illustrates the closer adhesion to the archetype by the natural articulation of the scapulo-coracoid arch to the occiput, so the higher vertebrate manifests the superior influence of the antagonising power of adaptive modification by the removal of that arch from its proper segment. The anthropotomist, by his mode of counting and defining the dorsal vertebree and ribs, admits, unconsciously perhaps, the important principle in general homology which is here exemplified, and which, pursued to its legitimate consequences and further applied, demonstrates that the scapula is the modified rib of that centrum and neural arch which he calls the ‘occipital bone,’ and that the change of place which chiefly masks that relation (for a very elementary acquaintance with comparative © anatomy shows how little mere form and proportion affect the homological characters of bones) differs only in extent and not in kind from the modification which makes a minor amount of comparative observation requisite, in order to determine the relation of the shifted dorsal rib to its proper centrum in the human skeleton. With reference, therefore, to the occipital vertebra of the crocodile, if the com- paratively well-developed and permanently distinct ribs of all the cervical vertebre prove the scapular arch to belong to none of those segments, and, if that heemal arch be required to complete the occipital segment, which it actually does complete in fishes, then the same conclusion must apply to the same arch in other animals, and we must regard the occipital vertebra of the tortoise as completed below by its scapulo-coracoid arch and not, as Bojanus supposed, by its hyoidean arch.* Having thus endeavoured to show what the scapular arch of the crocodile is, I proceed to point out the characteristic form of its chief elements. The upper and principal part of the scapula (51, fig. 9) is flattened, and gradually becomes narrower to the part called its neck, which is rounded, bent inwards, and then suddenly expanded to form a rough articular surface for the coracoid, and a portion of a smoother surface for the shoulder-jomt. The contiguous end of the coracoid (52) presents a similar form, having not only the rough surface for its junction with the scapula, but contributing, also, one half of the cavity for the head of the humerus. It is perforated near the interspace between these two surfaces. As it recedes from them, it contracts, then expands and becomes flattened, terminating in a somewhat broader margin than the base of the * Anatome Testudinis Europea, fol., 1819, p. 44. Geoffroy St. Hilaire selected the opercular and sub- opercular bones to form the inverted arch of his seventh (occipital) cranial vertebra, and took no account of the instructive natural connexions and relative position of the hyoidean and scapular arches in fishes. With regard to the scapular arch, he alludes to its articulation with the skull in the lowest of the vertebrate classes as an ‘amalgame inattendue’ (Anatomie Philosophique, p. 481): and elsewhere describes it as a “disposition veritablement trés singuliére, et que le manque absolu du cou et une combinaison des piéces du sternum avec celles de la téte pouvoient seules rendre possible.” —Annales du Muséum, ix, p. 361. A due appreciation of the law of vegetative uniformity or repetition, and of the ratio of its prevalence and power to the grade of organization of the species, was, perhaps, essential in order to discern the true signification of the connexion of the scapular arch in fishes. CROCODILIA. 23 scapula, which margin is morticed into a groove at the anterior border of the broad rhomboidal cartilage continued beyond the ossified part of the manubrium, which forms the key-bone of the scapular arch. The anterior locomotive extremity is the diverging appendage of the arch, under one of its numerous modes and grades of development.* The proximal element of this appendage or that nearest the arch, is called the ‘humerus’ (53, fig. 9): its head is subcompressed and convex; its shaft bent in two directions, with a deltoid crest developed from its upper and fore part; its distal end is transversely extended, and divided anteriorly into two condyles. The shaft of this bone has a medullary cavity, but relatively smaller than in the mammalian humerus. The second segment of the limb consists of two bones: the larger one (54) is called the ‘ulna:’ it articulates with the outer condyle of the humerus by an oval facet, the thick convex border of which swells a little out behind, and forms a kind of rudimental ‘olecranon ;’ the shaft of the ulna is compressed transversely, and curves slightly out- wards; the distal end is much less than the proximal one, and is most produced at the radial side. The radius (55) has an oval head; its shaft is cylindrical; its distal end oblong and subcompressed. The small bones (56) which intervene between these and the row of five longer bones, are called ‘ carpals:’ they are four in number in the Crocodilia. One seems to _be a continuation of the radius, another of the ulna; these two are the principal carpals; they are compressed in the middle and expanded at their two extremities ; that on the radial side of the wrist is the largest. A third small ossicle projects slightly backwards from the proximal end of the ulnar metacarpal: it answers to the bone called ‘ pisiforme’ in the human wrist. The fourth ossicle is interposed between the ulnar carpal and the metacarpals of the three ulnar digits. These five terminal jointed rays of the appendage are counted from the radial to the ulnar side, and have received special names: the first is called ‘ pollex,’ the second ‘index, the third ‘ medius,’ the fourth ‘annularis, and the fifth ‘mimimus.’ The first joint of each digit is called ‘metacarpal;’ the others are termed ‘phalanx.’ In the Crocodilia the pollex has two phalanges, the index three, the medius four, the annularis four, and the minimus three. The terminal phalanges, which are modified to support claws, are called ‘ ungual’ phalanges. As the above-described bones of the scapular extremity are developments of the appendage of the scapular arch, which is the heemal arch of the occipital vertebra, it follows, that, like the branchiostegal rays and opercular bones in fishes, they are essentially bones of the head. The diverging appendage of the pelvic arch being a repetition of the same element * See my Discourse ‘On the Nature of Limbs,’ 8vo, Van Voorst, 1849, pp. 64-70. 24 FOSSIL REPTILIA OF THE LONDON CLAY. as the appendage of the scapular arch modified and developed for a similar office, a close resemblance is maintained in the subdivisions of the framework of both limbs. The first bone of the pelvic limb, called the ‘femur,’ is longer than the humerus, and, like it, presents an enlargement of both extremities, with a double curvature of the intervening shaft, but the directions are the reverse of those of the humerus, as may be seen in T. XI, where the upper or proximal half of the femur is concave, and the distal half convex, anteriorly. The head of the femur is compressed from side to side, not from before backwards as in the humerus; a pyramidal protuberance from the inner surface of its upper fourth represents a ‘trochanter ;’ the distal end is expanded transversely, and divided at its back part into two condyles. The next segment or ‘leg,’ includes, like the corresponding segment of the forelimb called ‘forearm,’ two bones. The largest of these is the ‘tibia,’ and answers to the radius. It presents a large, triangular head to the femur ; it terminates below by an oblique crescent with a convex surface. The ‘ fibula’ is much compressed above; its shaft is slender and cylindrical, its lower end is enlarged and triangular. All these long bones have a narrow medullary cavity. The group of small bones which succeed those of the leg, are the tarsals; they are four in number, and have each a special name. ‘The ‘astragalus’ articulates with the tibia, and supports the first and part of the second toe. It is figured in Cuvier’s ‘Ossemens Fossiles,’ tom. vy, pt. ii, pl. iv, figs. 19.4, B, C, D. The ‘calcaneum’ inter- venes between the fibula and the ossicle supporting the two outer toes; it has a short but strong posterior tuberosity. The ossicle referred to represents the bone called ‘cuboid’ in the human tarsus. A smaller ossicle, wedged between the astragalus and the metatarsals of the second and third toes is the ‘ ectocuneiform.’ Four toes only are normally developed in the hind-foot of the Crocodilia ; the fifth is represented by a stunted rudiment of its metatarsal, which is articulated to the cuboid and to the base of the fourth metatarsal. The four normal metatarsals are much longer than the corresponding metacarpals. That of the first or innermost toe is the shortest and strongest; it supports two phalanges. The other three metatarsals are of nearly equal length, but progressively diminish in thickness from the second to the fourth. The second metatarsal supports three phalanges; the third four; and the fourth also has four phalanges, but does not support a claw. The fifth digit is represented by a rudiment of its metatarsal in the form of a flattened triangular plate of bone, attached to the outer side of the cuboid, and slightly curved at its pointed and prominent end. In the skull of the Crocodile, as of most other vertebrates, there are intercalated a few bones, or ossified parts of special organs, which, as is shown in the classed Table of the bones of the head, do not belong to the vertebral system of bones. CROCODILIA. 25 The bone anterior to the orbit, marked 73 in fig. 9, and in T. I and T. VI is perforated at its orbital border by the duct of the lachrymal gland, whence it is termed the ‘lachrymal bone, and its facial part extends forwards between the bones marked 14, 15, 21, and 26 in the same plates. In many Crocodilia there is a bone at the upper border of the orbit, which extends into the substance of the upper eyelid; it is called ‘ superorbital.’ In the Crocodilus palpebrosus there are two of these ossicles. Both the lachrymal and superorbital bones answer to a series of bones found com- monly in fishes, and called ‘ suborbitals’ and ‘superorbitals.. The lachrymal is the most anterior of the suborbital series, and is the largest in fishes; it is also the most constant in the vertebrate series, and is grooved or perforated by a mucous duct. These ossicles appertain to the dermal or muco-dermal system or ‘ exoskeleton.’ The little slender bone, marked 16’ in fig. 9, has one of its extremities in the form of a long, narrow, elliptic plate, which is applied to the ‘fenestra ovalis’ of the internal ear; from this plate extends a long and slender bony stem, which grows somewhat cartilaginous, expands and bends down, as it approaches the tympanum or ear-drum, to which it is attached. The cartilaginous capsule of the labyrinth or internal ear is partially ossified by sinuous plates of bone connate with the neurapophyses (fig. 10, 2 and 6), between which that organ is lodged; I apply the term ‘ petrosal’ to the principal and most independent of those ossifications of the ear-capsule, to that, e. g., which retains some mobility after it has con- Fig. 10. tracted a partial anchylosis to the exoccipital (2), and which appears upon the inner surface of the cranial walls at the part marked 16 in the subjoined Cut 10, between 2 and 6. It is the only independent bone on that surface of the cranium which, in my opinion, answers to the ‘ petrous portion of the Vertical longitudinal section of the cranium of a Crocodile (Crocodilus acutus). temporal’ in human anatomy, and to which the term ‘roc her can be properly applied, in the language of the French comparative anatomists. Cuvier, however, restricts that name to the ‘alisphenoid’ (6, figs. 9, 10) in the Crocodiles. The ossicles, (16 and 16’, fig. 9), together with the partial ossifications in the sclerotic capsule of the organ of sight, (17, fig. 9)—always more distinct in Chelonia than in Crocodilia— belong to that category of visceral bones to which the term ‘splanchno- skeleton’ has been given; they also are foreign to the true vertebrate system of the skeleton. 4 26 FOSSIL REPTILIA OF THE LONDON CLAY. The teeth—The most readily recognisable character by which the existing Crocodilians are classified and grouped in appropriate genera, are derived from modifications of the dental system. 18—18 | 2222. 0 : 18—18 ~ 2222 the fourth tooth of the lower jaw is received into a cavity of the alveolar surface of the upper jaw, where it is concealed when the mouth is shut. In T. VIII, fig. 2, these pits are shown behind the last premaxillary tooth e, in an Eocene Alligator from Hordwell. In old individuals of the existing species of Alligator, the upper jaw is perforated by the large inferior teeth in question, and the fossze are converted into foramina. In the Crocodiles (genus Crocodilus) the fourth tooth in the lower jaw is received nto a notch excavated in the side of the alveolar border of the upper jaw, as in fig. 1, T. VIII, behind the tooth e, and is visible externally when the mouth is closed, as in T. VI, fig. 1. In most Crocodiles, also, the first tooth in the lower jaw perforates the premaxillary bone when the mouth is closed, as in T. I, between the teeth In the Caimans (genus A//igator) the teeth vary in number from marked @ and 6. In the two preceding genera the alveolar borders of the jaw have an uneven or wavy contour, and the teeth are of an unequal size. In the Gavials, (genus Gavialis) the teeth are nearly equal in size and similar in form in both jaws, and the first as well as the fourth tooth in the lower jaw, passes into a groove in the margin of the upper jaw when the mouth is closed, T. XI. In the Alligators and Crocodiles the teeth are more unequal in size, and less regular in arrangement, and more diversified in form than in the Gavials: witness the strong thick conical laniary teeth at the fore part of the jaw, as shown in T. VII and T. III, fig. 6, as contrasted with the blunt mammillate summits of the posterior teeth, as shown in T. V, fig. 12. The teeth of the Gavial are subequal, most of them are long, slender, pointed, subcompressed from before backwards, with a trenchant edge on the right and left sides, between which a few faint longitudinal ridges traverse the basal part of the enamelled crown. The teeth of both the existing and extinct Crocodilian reptiles consist of a body of compact dentine forming a crown covered by a coat of enamel, and a root invested by a moderately thick layer of cement. The root slightly enlarges, or maintains the same breadth to its base, which is deeply excavated by a conical pulp-cavity extending into the crown, and is commonly either perforated or notched at its concave or inner side. The dentinal tubes in the crown of a fully-developed tooth form short curvatures at their commencement at the surface of the pulp-cavity, and then proceed nearly straight to the periphery of the crown; they very soon bifurcate, the divisions slightly diverging; then contiuing their course with gentle parallel undulations, they CROCODILIA. 27 subdivide near the enamel, and terminate in fine and irregular branches, which anastomose generally by the medium of cells. The dentinal tubes send off from both sides, throughout their progress, minute branches into the intervening substance, and terminate in the dentinal cells. These cells are subhexagonal, about ;1, of an inch in diameter, and are traversed by from ten to fourteen of the dentinal tubes; they are usually arranged in planes parallel with the periphery of the crown, near which they are most conspicuous, and towards which their best defined outline is directed: they combine with the parallel curvatures of the dentinal tubes to form the striz, visible in sections of the teeth by the naked eye, which cause the stratified appearance of the dentine as if it were composed of a succession of superimposed cones. The diameter of the dentinal tube before the first bifurcation is ;,4,,th of an inch, both the trunks and bifurcations of the tubes have interspaces equal to four of their respective diameters. The enamel viewed in a transverse section of the crown presents some delicate strie parallel with its surface, whilst the appearance of fibres vertical to that surface is only to be detected, and these faintly, on the fractured edge. It is a very compact and dense substance; the dark brownish tint is strongly marked in the middle of the enamel when viewed by transmitted light. The cells with which the fine tubes of the basal cement communicate, are oblong, about —-,1,,th of an inch across their long axis, which is transverse to that of the tooth ; the inter-communicating tubes, which radiate from the cells, giving them a stellate figure. I have entered into these particulars of the microscopic texture of the teeth of the Crocodile because it will be seen in the sequel that important modifications of the dental tissues characterise some of the extinct Reptilia. In the black Alligator of Guiana the first fourteen teeth of the lower jaw are implanted in distinct sockets, the remaining posterior teeth are lodged close together in a continuous groove, in which the divisions for sockets are faintly indicated by vertical ridges, as in the jaws of the Ichthyosaurs. A thin compact floor of bone separates this groove, and the sockets anterior to it, from the large cavity of the ramus of the jaw; it is pierced by blood-vessels for the supply of the pulps of the growing teeth and the vascular dentiparous membrane which lines the alveolar cavities. The tooth-germ is developed from the membrane covering the angle between the floor and the inner wall of the socket. It becomes in this situation completely enveloped by its capsule, and an enamel-organ is formed at the inner surface of the capsule before the young tooth penetrates the interior of the pulp-cavity of its predecessor. ; The matrix of the young growing tooth affects, by its pressure, the inner wall of the socket, and forms for itself a shallow recess; at the same time it attacks the side of the base of the contained tooth; then, gaining a more extensive attachment by its basis and increased size, it penetrates the large pulp-cavity of the previously formed 28 FOSSIL REPTILIA OF THE LONDON CLAY. tooth, either by a circular or semicircular perforation. The size of the calcified part of the tooth matrix which has produced the corresponding absorption of the previously formed tooth on the one side, and of the alveolar process on the other, is represented in the second exposed alveolus of the portion of jaw figured in Pl. 75, fig. 4, of my ‘Odontography, the tooth marked « in that figure, having been displaced and turned round to show the effects of the stimulus of the pressure. The size of the perforation in the tooth, and of the depression in the jaw, proves them to have been, in great part, caused by the soft matrix, exciting dissolution and absorbent action, and not by mere mechanical force. The resistance of the wall of the pulp-cavity having been thus overcome, the growing tooth and its matrix recede from the temporary alveolar depression, and sink into the substance of the pulp contained in the cavity of the fully- formed tooth. As the new tooth grows, the pulp of the old one is removed; the old tooth itself is next attacked, and the crown being undermined by the absorption of the inner surface of its base, may be broken off by a slight external force, when the point of the new tooth is exposed. The new tooth disembarrasses itself of the cylindrical base of its predecessor, with which it is sheathed, by maintaining the excitement of the absorbent process so long as the cement of the old fang retains any vital connexion with the periosteum of the socket; but the frail remains of the old cylinder, thus reduced, are sometimes lifted off the socket upon the crown of the new tooth, when they are speedily removed by the action of the jaws. This is, however, the only part of the process which is immediately produced by mechanical force: an attentive observation of the more important pre- vious stages of growth, teaches that the pressure of the growing tooth operates upon the one to be displaced only through the medium of the vital dissolvent and absorbent action which it has excited. Most of the stages in the development and succession of the teeth of the Crocodiles are described by Cuvier* with his wonted clearness and accuracy ; but the mechanical explanation of the expulsion of the old tooth, which Cuvier adopts from M. Tenon, is opposed by the disproportion of the hard part of the new tooth to the vacuity in the walls of the old one, and by the fact that the matter impressing—viz. the uncalcified part of the tooth-matrix—is less dense than the part impressed. No sooner has the young tooth penetrated the interior of the old one, than another germ begins to be developed from the angle between the base of the young tooth and the inner alveolar process, or in the same relative position as that in which its imme- diate predecessor began to rise, and the processes of succession and displacement are carried on, uninterruptedly, throughout the long life of these cold-blooded carnivorous reptiles. From the period of exclusion from the egg, the teeth of the crocodile succeed each other in the vertical direction; none are added from behind forwards, like the true * Op. cit., pp. 90-3. CROCODILIA. 29 molars in Mammalia. It follows, therefore, that the number of the teeth of the cro- codile is as great when it first sees the light as when it has acquired its full size; and, owing to the rapidity of the succession, the cavity at the base of the fully-formed tooth is never consolidated. The fossil jaws of the extinct Crocodilians demonstrate that the same law regulated the succession of the teeth at the ancient epochs when those highly organized reptiles prevailed in greatest numbers, and under the most varied generic and specific modi- fications, as at the present period, when they are reduced to a single family, composed of so few and slightly varied species, as to have constituted in the Systema Nature of Linnzus, a small fraction of the genus Lacerta. CROCODILUS TOLIAPICUS, Owen. Tab. II, fig. 1, and Tab. II 4. Syn. CrocopiLE pe Sueppy (?), Cuvier. Ossemens Fossiles, 4to, tom. v, pt. ii, p. 165. Crocopitus SpENnceRI, Buckland. Bridgewater Treatise, vol. i, p. 251. ‘Crocodile , with a short and broad snout.’ Vol. ii, p. 36, pl. 25’, fig. 1. _— — Owen. Reports of the British Association, 1841, p. 65. In proceeding to the comparison, and preparing for the description of the British fossil Crocodilia, 1 endeavoured, in the first place, to obtaim the bones of the species which now exists in a locality nearest to Great Britain, and also of an individual of that same species which had lived at a remote period; and I have been favoured by the kindness of my esteemed friend Philip Duncan, Esq., Fellow of New College, Oxford, and Conservator of the Ashmolean Museum, with the opportunity of examining the bones of a mummified Crocodile froma sarcophagus at Thebes, in that collection at Oxford. Two views of the skull of this old Egyptian Crocodile are given in T. I. The total length of the skull from the bone marked 2s to the end of 22, is twice the breadth of the back part of the skull. The upper apertures of the temporal fossee are subcircular ; the point of the squamosal (27) projects into the lateral aperture. The breadth of the back part of the sculptured cranial platform (s,§8), is less by one fourth than the breadth of the skull anterior to the orbits. The breadth of the interorbital space is nearly equal to the transverse diameter of the orbit. The pomts of the nasals (15) project into the external nostril. The postpalatal apertures reach as far forwards as the seventh tooth, counting from the hindmost; there are nineteen alveoli on each side of the upper jaw, the five anterior teeth being lodged in the premaxillary, which is perforated by the first tooth of the lower jaw. Geoffroy St. Hilaire has applied the old Egyptian name Zovyoc to the mummified Crocodiles of that country; but there is no good specific character which distinguishes them from the modern Crocodiles of the Nile, to which Cuvier has given the name of Crocodilus vulgaris. Cuvier appears to have first called the attention of paleontologists to the remains 30 FOSSIL REPTILIA OF THE LONDON CLAY. of Crocodilia in the Eocene clay forming the Isle of Sheppy, in the last volume of the second edition of his great work on the ‘ Ossemens Fossiles,’ p. 165, 1824. He there specifies a third cervical vertebra, which was obtained by M. G. A. Deluc, at Sheppy, and of which Cuvier made a drawing at Geneva; he says it much resembles the corresponding vertebra in one of our living Crocodiles, and might have come from an individual about five feet in length. ‘“M. Deluc,” he adds, ‘found very near it a much smaller vertebra, which I recognised as belonging to a monitor or some allied genus.”* ‘ Our knowledge of the Eocene Crocodiles of Sheppy received a remarkable accession at the publication of the highly interesting and instructive ‘ Bridgewater Treatise’ of Dr. Buckland, in which he states that “ true Crocodiles, with a short and broad snout, like that of the Caiman and the Alligator, appear, for the first time, in strata of the tertiary periods, in which the remains of mammalia abound. . . . One of these,” he adds, “found by Mr. Spencer in the London Clay of the Isle of Shoop is engraved PI. 25; fig. 1,” and the name ‘ Crocodilus Spencer’ is appended to that figure. In preparing my ‘ Report on British Fossil Reptiles’ for the British Association in 1841, I examined the original specimen figured by Dr. Buckland, in which unfortunately the end of the snout with the intermaxillaries and an indeterminate proportion of the maxillaries having been broken off and lost, no exact idea could be formed of the pro- portions of the facial or rostral part of the skull. In a larger specimen of the fossil skull of a Crocodile from Sheppy, in the British Museum, T. II 4, the whole of the upper, as well as the lower jaw, were preserved, and as the proportions of the snout agreed with those of some true Crocodiles, and differed in an equal degree with those species from the Gavial; and as, like the Crocodiles and Caimans, it presented the more important distinction of a different composition of that part of the skull, I retained for the specimen in that ‘ Report’ the name of Crocodilus Spencert, proposed by the author of the Bridgewater Treatise for the Sheppy Crocodile, so differmg from the Gavial. The able keeper of the Mineralogical Department of the British Museum, Charles Konig, K.H., F.R.S., to whom I am indebted for every facility m describing and figuring this specimen, has suggested that the name by which Baron Cuvier first indicated the existence of a true Crocodile in the Eocene clay of Sheppy, should have the priority, and I adopt, therefore, the name Crocodilus toliapicus, which he has attached to the specimen in question, and with the more readiness since I have now reason to doubt whether the mutilated cranium, figured in the ‘ Bridgewater Treatise,’ belongs to the same species. * Could this have been a vertebra of the large serpent, which I have subsequently described under the name of Paleophis? I have not as yet met with a single lacertian vertebra from Sheppy. If the collection of M. Deluc be still preserved at Geneva, the vertebra in question might be compared with the figures of the Paleophus toliapicus, ‘Ophidia,’ T. XV. CROCODILIA. 31 The more entire fossil skull in question presents the following dimensions : Feet. Inches. Lines. Total length from the hindmost part of the lower jaw. 2 2 0 Breadth between the articular ends of the tympanics 0 10 0 Do. across the orbits 0 7 6 Do. of the intertemporal space 0 0 9 Do. of the interorbital space 0 1 4 From the articular end of the tympanic to We pit 0 8 6 From the occipital condyle to the orbit . 0 7 0 From the orbit to the external nostril 5 0 14 0 Breadth of the cranium five inches in advance of the orbits 0 3 8 Do. across the external nostril 0 2 8 Depth of the lower jaw at the vacuity between the aaguine and surangular $ : : 0 . 0 3 6 Length of that vacuity . : . 0 3 0 Breadth of the base of one of the larger Caviar teeth, 0 0 8 This remarkably fine fossil skull, which is figured one third of its natural size in T. Il, fig. 1, presents proportions which come nearest to those of the Crocodilus acutus, being longer in proportion to its basal breadth than in the Crocodilus Suchus, in which the diameter between the articular ends of the tympanis (28) is just half the length of the entire skull. The interorbital space in the Crocodilus toliapicus is relatively narrower and flatter than in the Croc. acutus or Croc. Suchus, and the facial part of the skull becomes narrower before the expansion of the upper jaw, at the figure 15, than it does in either of those species. ‘The narrow elongated nasals on which the figure 15 is placed, extend forwards to the external nostril (22), as in the true Crocodiles, and the alveolar border is festooned as is shown in the side view in T. II 4. The teeth perme 20 20 than in the recent species above cited, and resemble in this respect the teeth of the Crocodilus Schlegel of 8. Miiller, which is from Borneo. The extent of the symphysis of the lower jaw is greater in the Crocodilus toliapicus than in the Croc. acutus, and still greater than in the Croc. Suchus; the Sheppy species in this respect more nearly resembles the living species from Borneo above cited. =——+-==84 in number: they are more uniform in size, and more regularly spaced CROCODILUS CHAMPSOIDES, Owen. Tab. III. (Tab. IL fig. 2 ?) Syn. Crocopite DE Suepry (?), Cuvier. Loc. cit. Crococttus SpeNnceERI, or ‘Crocodile with a short and broad snout” (?) Buckland. Bridgewater Treatise, vol. ii, pl. xxv, fig. 1. The fossil skull already described establishes the fact of the existence of a true Crocodile in the London Clay at Sheppy, but not of a species with a short and broad snout ; the present specimen equally demonstrates the presence at the earliest period of the Tertiary geological epoch of Crocodilia with those modifications of the cranial 32 FOSSIL REPTILIA OF THE LONDON CLAY. and dental structure on which the characters of the restricted genus Crocodilus of modern Zoology are founded; but they are associated with a general form of the head which approaches more nearly to the Gavials than does that of the Crocodilus toliapicus, and which are most nearly paralleled amongst the known existing true Crocodiles by the Crocodilus Schlegelit. ‘This Bornean species was, in fact, originally described as a new species of Gavial, but the nasal bones, as in the fossil from Sheppy figured in T. II, 15, extend to the hind border of the external nostril. The fine subject of T. III, forms part of the collection of J. S. Bowerbank, Esq. F.R.S., which is well known for its rich and varied illustrations of the fossils of the Isle of Sheppy. The following are some of its admeasurements : Feet. Inches. Lines. Total length from the occipital condyle to the end of the premaxillaries 5 5 a 1 4 0 Breadth across the hinder angles of the supracranial platform 0 4 0 Do. across the orbits . Bp hee F a : 0 5 0 Do. of the intertemporal space 0 0 4 Do. of the interorbital space 0 1 0 Do. across the external nostril : : 5 3 0 2 0 From the occipital condyle to the orbit 0 3 4 From the orbit to the external nostril 0 10 0 The skull yielding the above dimensions is much smaller than that of the Crocodilus toliapicus, T. IL; but it cannot have belonged to a younger individual of the same species, because, in existing Crocodiles, the part of the skull anterior to the orbits is proportionally shorter in the young than in the old individuals, as may be seen by comparing the figures which Cuvier has given of the skulls of three individuals of different ages of the Crocodilus biporcatus, in figures 4, 18, and 19, of plate 1 of the last volume of the ‘ Ossemens Fossiles,’ 4to, 1824; whereas the part of the skull anterior to the orbits is relatively longer and more slender in the smaller fossil skull now described than in the larger one on which the species Croc. toliapicus is founded. We have, therefore, satisfactory proof that two species of true Crocodile existed during the de- position of the Eocene Clay at the actual mouth of the Thames, and have left their remains in that locality. Their specific distinction is further illustrated by the different forms and propor- tions of particular parts of the skull. The alveolar border is more nearly straight; the transverse expansion of the maxillaries (21) is less, whilst that of the premaxillaries (22) is greater: the interorbital space is broader and more concave. The teeth are more uniform in size, are more regularly spaced, and are wider apart: they are, likewise, upon the whole, larger in proportion to the size of the jaw. Figure 5, T. IIL, shows the crown of a new tooth just emerging from the second socket of the maxillary bone of the natural size ; figure 6 is the fourth tooth of the premaxillary, fully formed; fig. 7 CROCODILIA. 33 is the displaced tooth which is cemented by the matrix to the palatal surface of the premaxillary in fig. 2. The enamelled crown shows the fine raised longitudinal ridges better developed than one usually sees them in modern Crocodiles. There are twenty- one alveoli on each side of the upper jaw. In all the particulars in which the skull under description differs from that of the , Crocodilus toliapicus, it departs further from the nilotic crocodile, and resembles more the Gavial-like Crocodile of Borneo; and as one of the old Egyptian names of the Crocodile, Champsa, has been applied generically to the Gavials by some recent Erpetologists, I have adopted the term ‘ Champsoides’ to signify the resemblance of the present extinct species of Eocene Crocodile to the Gavials. The basioccipital condyle, together with the condyloid processes of the exoccipital, project backwards in the Croc. champsoides farther than in any modern Crocodile; and the supraoccipital 3, fig. 4, T. III, descends nearer to the foramen magnum. The upper jaw is more depressed, and the suborbital part of the maxillary bone is much less inclined to the vertical in the present skull than in the original of Dr. Buckland’s figure of the Crocodilus Spenceri, which in other respects more nearly resembles the Croc. champsoides than the Croc. toliapicus; the difference above specified seems to be greater than can be accounted for by any accidental pressure to which the fossil skull figured in T. III can have been subjected. _ The mutilated skull to which the term Croc. Spenceri was originally applied, is defective, as I have said, in the facial or maxillary portion which is requisite for its unequivocal determination to either of the two species which the more perfect specimens since acquired have proved to have existed at the Eocene tertiary period. The form of the mutilated portion of skull, and the figure of it given in Pl. 25’ of the ‘ Bridgewater Treatise,’ might well appear to indicate a short and broad snouted species of true Crocodile ; but if it be not distinct from the two better represented species above described, I should be more inclined to refer it to that which has the longest and narrowest snout, from the conformity of the characters of the part of the skull which is preserved. A view of the palatal surface of the specimen in question is given in T. II, fig. 2. Crocodilian vertebre referable to the two foregoing species of Sheppy Crocodiles. Not more than two species of Crocodile are indicated by the detached vertebre from Sheppy; but the different proportions of the homologous cervical vertebre, fig.3 and 7, T. V, and of the characteristic biconvex caudal vertebra, fig. 7, T. IV, and fig. 10, T. V, would have determined the fact of there being two distinct species, had their cranial characters, which are so satisfactorily demonstrated in T. II 4 and T. III, remained unknown. I refer, provisionally, the shorter and thicker vertebre to the Crocodilus toliapicus with the shorter and thicker snout, and the longer and thinner vertebree to the Croc. champsoides with the snout of similar proportions. 34 FOSSIL REPTILIA OF THE LONDON CLAY. Vertebre of the CROCODILUS ToLIAPiIcus, Tab. IV and Tab. V, fig. 1, 2, 3, 5, 6. The vertebra, fig. 1, 2, T. V, is the fourth cervical; it differs from that of the Crocodilus acutus, Croc. Suchus, and Croc. biporcatus, in the greater breadth and squareness ‘of the base of the hypapophysis (fig. 2/4), which extends almost to the bases of the parapophyses p; the vertical diameter of the parapophyses is greater in comparison with their antero-posterior extent in the fossil than in the above-cited recent Crocodiles ; the neurapophyses are thicker, and terminate in a more rounded border both before and behind; their bases extend inwards, and meet above the centrum, whilst a narrow groove divides them in the recent Crocodiles above cited; the length of the centrum is greater in proportion to the height and breadth in the fossil vertebra. In other respects the correspondence is very close, and the modern crocodilian characters are closely repeated. Traces of the suture between the centrum and neurapophysis remain, as shown at , 2, fig. 1. The diapophysis ¢, and the upper portion of the neural arch, with the zygapophyses and neural spine, have been broken away; the borders of the articular ends of the centrum have been worn away. The vertebra (fig. 3, T. V) is the sixth cervical: im this specimen the base of the hypapophysis is contracted laterally and extended antero-posteriorly ; the side of the centrum above the parapophysis (7) has become less concave; the vertebra has increased more in thickness than in length; in these changes it corresponds with the modern Crocodiles; it has been mutilated and worn in almost the same manner and degree as the fourth cervical. The vertebra (fig. 1, 2, T. IV) is a seventh cervical of a smaller individual of the Crocodilus toliapicus. ‘The hypapophysis has become more compressed and more extended antero-posteriorly ; the parapophysis has become shortened antero-posteriorly, and increased in vertical diameter. The anterior concave surface of the centrum (fig. 1) is more circular, less extended transversely, than in the corresponding vertebra of the recent Crocodiles compared with the fossil. Fig. 3, 4, T. IV, are two views of the eighth cervical of an individual of about the same size as that to which the fourth and sixth cervicals in T. V belong. Fig. 4, exemplifies the same difference which fig. | presents in regard to the more circular contour of the anterior concave surface of the centrum as compared with recent Crocodiles ; the bases of the neurapophyses are thicker and more rounded anteriorly ; the neural canal is rather more contracted; the base of the hypapophysis more extended in the axis of the vertebra (see fig. 3) than in the recent Crocodiles compared. The parapophyses have now risen, as in those Crocodiles, to the suture of the neurapophysis, and the diapophysis springs out at some distance above that suture. Fig. 6, T. IV, shows the under surface of a dorsal vertebra, in which the hypapophysis ceases to be developed (probably the fourth or fifth). Fig. 5, T. IV, gives the same view of one of the lumbar vertebra, showing the CROCODILIA. 35 elongation of the centrum, and the broad bases of the depressed diapophyses ; there is an indication of two longitudinal risings towards the back part of the under surface of the centrum. Fig. 5 and 6, T. V, give two views of the anterior sacral vertebra of the Crocodilus toliapicus ; it is concave and much expanded transversely at its fore part (fig. 5), flattened and contracted behind. Traces of the suture remain to show the proportion of the anterior articular surface which is formed by the base of the pleurapophysis p ; and fig. 6 shows the extension of that base from the side of the centrum upon the diapophysis or overhanging base of the neurapophysis; the under surface of the centrum of this vertebra has a slight median longitudinal rising. Fig. 7, T. IV, gives a side view of the characteristic, biconvex, anterior caudal vertebra of the Crocodilus toliapicus. Fig. 8, 9, T. IV, give two views of a middle caudal vertebra: in fig. 9 are shown the characteristic hypapophysial ridges extending from the articular surfaces for the heemapophyses at the hind part of that aspect of the centrum: in fig. 8 the processes of the neural arch are restored in outline; a thick and low ridge extends from the middle of the side of the centrum to the base of the transverse process which it strengthens, like an underpropping buttress. Vertebre of the CROCODILUS CHAMPSOIDES. Fig. 7 and 8, T. V, give two views of the third cervical vertebra of the above- named gavial-like Crocodile, which vertebra, besides its longer and more slender propor- tions, differs in the smaller size of its hypapophysis from the corresponding vertebra in any existing species of Crocodile or Gavial: the process in question being in the form of a low crescentic ridge, as shown at figure 8, between the bases of the parapophyses (). Both parapophyses terminate by a convex surface, which appears to have been anatural one. Between the parapophysis (y) and diapophysis (d), fig. 7, the side of the centrum is more deeply excavated than in the Crocodilus toliapicus. The centrum contributes a small part to the base of the diapophysis, as in the third cervical vertebra of modern Crocodiles. The neurapophysis are thinner than in the Croc. toliapicus, and their bases do not join one another above the centrum. The longitudinal ridge extending from the anterior to the posterior zygapophysis is sharply defined. Fig. 4,T. V, is the first dorsal vertebra of the Crocodilus champsoides, in which, as in existing Crocodiles, the parapophysis () has passed almost wholly from the centrum upon the neurapophysis, the diapophysis (¢) having been subject to a corresponding ascent. The base of the compressed hypapophysis extends over the anterior third of the middle line of the under surface of the centrum. There is a remarkable transverse constriction at the base of the posterior ball of the centrum, as if a string had been tied round that part when it was soft, and there is no appearance of this groove having been produced by any erosion of the fossil, or being otherwise than natural. 36 FOSSIL REPTILIA OF THE LONDON CLAY. The same character is repeated, though with less force, in the posterior dorsal vertebra, fig. 9, T. V, and, together with the general proportions of the specimen, supports the reference of that vertebra to the Crocodilus champsoides. 'There is a slight longitudinal depression at the middle of the side of the centrum near the suture with the neurapophysis (7, 7). Fig. 10 isa side view of the first caudal vertebra of the Crocodilus champsoides : besides being longer and more slender than that vertebra is in the Croc. toliapicus, the inferior surface of the centrum is less concave from before backwards. The evidences of Crocodilian reptiles from the deposits at Sheppy less characteristic of particular species than those above described, are abundant. Mr. Bowerbank possesses numerous rolled and fractured vertebre, condyloid extremities, and other portions of long bones; with fragments of jaws and teeth. Mr. J. Whickham Flower, F.G.S., has transmitted to me some fragments of the skull of a Crocodile from Sheppy, cluding the articular end of the tympanic hone, equalling in size that of a Crocodilus biporcatus the skull of which measures two feet eight inches in length. Mr. Leifchild, C.E., possesses a considerable portion of the lower jaw of a Crocodile of at least equal dimensions, also from Sheppy, showing the angle of union of the rami of the lower jaw which corresponds with that in the Crocodilus toliapicus, Pl. 2. In the museum of my esteemed and lamented friend, the late Frederic Dixon, Esq., F.G.S., at Worthing, is preserved a portion of the fossilized skeleton of a Crocodile, from the Eocene clay at Bognor, in Sussex ; it consists of a chain of eight vertebree, including the lumbar, sacral, and the biconvex first caudal, which are represented of their natural size in tab. xv, of Mr. Dixon’s beautiful and valuable work on the ‘ Geology of Sussex.’ A dorso-lateral bony scute adheres to the same mass of clay close to the vertebree, and doubtless belonged to the same individual. The proportions of the vertebree agree with those of the Crocodilus toliapicus. This fine specimen was dis- covered, and presented to Mr. Dixon, by the Rev. John Austin, M.A., Rector of Pulbrough, Sussex. Mr. Dixon had also obtained from the same locality a posterior cervical vertebra of a Crocodile, similar in its general characters to those above mentioned, but belonging to a larger individual. The length of the body of this vertebra is two inches and a half. I have examined some remains of Crocodilia from the London Clay at Hackney; but as these also are not sufficiently perfect or characteristic for decided specific determination, no adequate advantage would be obtained by a particular description, or by figures of them. The chief conclusion arrived at from the study of the Crocodilian fossils from the Island of Sheppy has been the proof, by the specimens selected for depiction in the present work, that at least two species of true Crocodile have left their remains in that locality; that neither of these had a short and broad snout like the Caimans, but that one of them—the Croc. champsoides—much more nearly resembled the Gavial of CROCODILIA. 37 the Ganges in the proportion of that part of the skull; although, in its composition, especially as regards the length and connexions of the nasal bones, it is a true Crocodile. Amongst the existing species of Crocodile the Croc. acutus of the West Indies offers the nearest approach to the Croc. ¢oliapicus, and the Croc. Schlegelii of Borneo most resembles the Croc. champsoides. But there are well-marked characters in both the skull and the vertebrae which specifically distinguish the two fossil Crocodiles of Sheppy from their above-cited nearest existing congeners. Crocopitus HAsTINGsta, Owen. Tab. VI, VIL, VIII, UX, and ai XIL fig. 2 and 5. Reports of the British Association, 1847, p. 65. That Crocodiles with proportions of the jaws assigned to the Eocene species noticed in Dr. Buckland’s ‘ Bridgewater Treatise’ and especially adapted for grappling with strong mammiferous animals, actually existed at that ancient tertiary epoch, and have left their remains in this island, is shown by the singularly perfect fossil skull figured in the above-cited plates. This specimen was discovered by the Marchioness of Hastings, in the Eocene fresh-water deposits of the Hordle Cliffs in Hampshire, which her _Ladyship has described in the volume of ‘ Reports of the British Association’ above cited, (p. 63). When the specimen was originally exposed, it was in the same extremely fragile and crumbling state as the beautiful carapaces of Zrzonyx obtained by Lady Hastings from the same locality, and described and figured in the First Part of this Monograph on the Chelonia ; but thanks to the skill and care with which the noble and accomplished discoverer readjusted and cemented the numerous detached fragments of those specimens, the present unique fossil has been in like manner restored as nearly to its original state as is represented in the plates; and all the requisite characters for deter- mining the nature and affinities of the species, can now be studied with the same facility as in the skulls of existing Crocodiles. If the reader will compare the plates above cited with the section of Cuvier’s ‘Ossemens Fossiles,’ in which the distinctions between the Alligators and Crocodiles are specified,* he will see, (in fig. 1, T. VII) for example, that the fourth tooth or canine of the lower jaw is not received into a circumscribed cavity of the upper jaw, * < Tes tétes des caimans, outre le nombre des dents, et surtout la maniére dont la quatriéme d’en bas est recue, outre les différences qui dependent de la circonscription totale, se distinguent de celles des Crocodiles proprement dits, 1°, parce que le frontal antérieur et le lacrymal descendent beaucoup moins sur le museau ; 2°, en ce que les trous percés a la face supérieure du crane, entre le frontal postérieur, le pariétal et le mastoidien, y sont beaucoup plus petits, souvent méme y disparaissent tout-a-fait, comme dans le caiman & paupiéres osseuses; 3°, en ce que l’on apercoit une partie du vomer dans le palais, entre les intermaxillaires et les maxillaires; 4°, en ce que les palatins avancent plus dans ce méme palais, et s’y élargissent en avant; 5°, en ce que les narines posterieures y sont plus larges que longues, etc.” (tom. vy, pt. ii, p. 105.) 38 FOSSIL REPTILIA OF THE LONDON CLAY. but is applied to a groove upon the side of the upper jaw, and is exposed. Fig. 1, T. VI, shows that the prefrontal (14) and lachrymal (73) bones, instead of descending much less upon the facial part of the skull, extend much more, and advance nearer to the end of the muzzle than in any Alligator, or even than in any actual species of broad- nosed Crocodile. The vacuities left between the postfrontal (12), the parietal (7), and the mastoid (8) (T. VI, fig. 1, and T. UH, fig. 3), are as wide as in the skull of a Crocodilus biporcatus of equal size, and are larger than in the Alligator lucius or All. sclerops. Fig. 2,'T. VII, shows that no part of the vomer is visible between the premaxillaries (22) and maxillaries (21), or elsewhere on the palate. But the palatine expansion of the vomer is nota constant character; it is wanting, for example, in the Alligator lucius of North America. The palatines (20) are not more advanced in the fossil in question than they are in the true Crocodiles, and their anterior portion does not expand to its anterior truncated termination. The posterior nostril, the entire contour of which is shown in the portion of the skull of the same species figured in T. VI, fig. 3, is longer than it is broad. There is but one character in which the fossil skull in question differs from the true Crocodile, and agrees with most species of Alligator; it is in the reception of the two anterior teeth of the lower jaw into cavities of the premaxillaries, shown nm fig. 2, T. VII, which are not perforated; so that there are no foramina anterior to the bony nostril, as in T. I, in the bone marked 22. These foramima are not, however, absent in all Alligators; the skull of the Alligator sclerops, figured by Cuvier (tom. cit., pl. i, fig. 7), shows them, as do all the species of true Crocodile the skulls of which are figured in the same plate. There is one character by which the Crocodilus Hastingsie differs from all known species of both-Crocodile and Alligator: it is that afforded by the broad and short nasal bones (15, fig. 1, T. VI), which do not reach the external nostril; this being formed, as in the Gavials, exclusively by the premaxillaries 22. In the general proportions, however, of the skull of Croc. Hastingsia, especially the great breadth, shortness, and flatness of the obtusely-rounded snout, it resembles that of the Alligators more than that of any known species of true Crocodile, the length from the tympanic condyle to the end of the snout being to the breadth taken at the condyles as 16 to 9. The following are dimensions of the fossil in question : Feet. Inches. Lines. Length of skull from the angle of the lower jaw to the end of 1 the snout : 1 6 6 Do. from the iyeapanie condyle to ditto. 1 4 6 Do. to the orbit 0 5 4 Do. from the ee to the external nostril 0 7 0 Breadth of the skull across the tympanic condyles 0 9 3 Do. the orbits : 0 7 0 Do. the external nostril 0 4 0 Longest diameter of upper temporal aperture 0 1 9 Do. the post-palatal vacuities 0 4 9 0 3 0 0 4 3 Depth of the lower jaw at the posterior vacuity . Depth of the occipital region : : CROCODILIA. 39 The occipital region of the skull (T. VI, fig. 2), in the proportion of its breadth to the depth of the lateral parts formed by the conjoined paroccipitals (4) and mastoids (s), resembles that of the true Crocodiles rather than that of the Alligators, in which that region is proportionally deeper than in the Crocodiles; the vertical extent of the supraoccipital is less, and that of the conjoined parts of the exoccipitals above the foramen magnum is greater; the vertical extent of the descending part of the basioccipital is also greater in proportion to its breadth than in the Alligators. The proportion of the ‘basisphenoid (5) and of the conjoined parts of the pterygoids (24) which appear in this view (fig. 2),is less than in the Alligators, but is greater than in most Crocodiles, thus presenting an intermediate character; but the entire exclusion of any part of the posterior nostril from this view is a character of the Alligators, and is due to the horizontal plane of that aperture in them, and to its position in advance of the posterior border of the pterygoids, from which it is partitioned off usually by a bony ridge. The posterior nostril has the same position and aspect in the Orocodilus Hastingsia, and these characters of the posterior nostril are perhaps more distinctive between Alligator and Crocodile than the shape of the aperture, in which the present fossil differs both from the Alligators and from most of the Crocodiles with which I have compared it. The backward extension of the exoccipitals and of the basioccipital condyle, is such as to bring both parts into view in looking directly upon the middle of the upper surface of the skull, asin T .VI, fig. 1. In this character the fossil resembles the Crocodiles more than the Alhgators, but the projection is greater than in existing Crocodiles, and equals that in the Sheppy Crocodilus champsoides. On the upper surface of the skull a distinctive character has been pointed out by Cuvier in the different proportions of the supra-temporal apertures in the Alligators and Crocodiles. The horizontal platform in which these apertures are perforated, is also square in the Alligators ; the mastoidal angles being less produced outwards and backwards, and the postfrontal angles less rounded off; this difference is shown in the skulls figured in Cuvier’s pl. i, tom. cit. The Croc. Hastingsia, both by the obtuseness of the postfrontal angles, and the acuteness and production of the mastoidal angles, resembles the Crocodiles, as well as by the size of the supra-temporal apertures; these are ovate with the small end turned forwards and a little outwards. Another character may be noticed in the figures of the skulls of the three species of Alligators as compared with those of the three species of Crocodile in Cuvier’s pl. i, viz. the larger proportional size of the orbits in the former, in which the orbit much exceeds in size the lateral temporal aperture. In the Alligator niger, also, I find the orbits enormous, and it is the encroachment of the narrow anterior part of the orbital cavity upon the facial part of the prefrontal and lachrymal, that renders that part of those bones relatively shorter in the Alligators. In the Crocodilus Hastingsi@ the proportions of the lateral temporal apertures (T. VI, fig. 1, 12, 26) and orbital (11, 14, 73) apertures, are like those in the species of Crocodile in which the orbits are smallest. The extent of 40 FOSSIL REPTILIA OF THE LONDON CLAY. the facial part of the prefrontal (14) and lachrymal (73) is greater in the Croc. Hastingsie than in any existing species of true Crocodile. Another characteristic of the present fossil presented by the upper surface of the skull, is the shortness as well as breadth of the nasal bones, and their almost truncate anterior termination at nearly one inch from the external nostril. In all the Alligators’ skulls that I have examined or seen figured, the nasal bones are broadest at their posterior third part, and converge to a point anteriorly, where in the Alligator lucius, e. g., they extend across the nasal aperture. The interorbital space is slightly concave in the Crocodilus Hastingsie ; two broad + and slightly elevated longitudinal tracts are contmued forwards upon the face from the fore part of the orbits ; but they are not developed into ridges, as in the Croc. biporcatus. The maxillaries swell out a little in advance of the middle of the nasals, and then contract to the crocodilian constriction at the suture with the premaxillaries, where the tips of the lower canines appear in the upper view (fig. 1, T. VI), and their whole crown is exposed in the side view (fig. 1, T. VII). The conjoined parts of the premaxillaries send a short pointed projection into the back part of the external nostril. On the under or palatal surface of the skull (T. VII, fig. 2) the maxillo-premaxillary suture runs almost transversely across, as in the Crocodilus rhombifer, figured by Cuvier ~ in pl. iii, fig. 2, of the volume above cited. There is no appearance of the vomer upon the palate. The palatal bones (20), though somewhat broader anteriorly, and more abruptly truncate than in any existing Crocodile that I have seen, are more like those bones in the true Crocodiles than in the Alligators. The portion between the post- palatal vacuities is longer and narrower; the posterior end of the palatines is nar- rower, and the part of the bone anterior to the notch receiving the posterior angle of the palatal plate of the maxillary does not expand in advancing forwards, as it does in the Alligators: in the Aligutor niger this expansion is greater than in the A//. lucius, and the posterior ends of the palatines are also remarkably expanded, and applied to the anterior borders of the pterygoids almost as far as their articulation with the ectopterygoids, the postpalatal vacuities not at all encroaching on the pterygoids, as they are seen to doat 24, T. VII, fig. 2, and also in the figure of the Crocodilus rhombifer above cited, and in other true Crocodiles. The form of the pterygoids (24, T. VII, fig. 2) is peculiar in the Crocodilus Hastingsi@ : they are contracted anteriorly, and send forwards a short truncated process to meet the narrow posterior ends of the palatines (20); and the same character being repeated in another skull of the same species, from Hordle, also in the collection of Lady Hastings, in which this part of the bony palate (T. VL, fig. 3) is more perfect than in the subject of T. VII, fig. 2, it may be regarded with some confidence as specific. In the Crocodilus champsoides of Sheppy it will be seen, by fig. 2, T. II, that the pterygoids (24, 24) are not produced where they join the palatines (20). In the Alligators, the posterior border of the conjoined pterygoids is deeply notched behind the posterior nostrils, the angles of the notch being slightly extended backwards: in most Crocodiles, the sides of the notch are so developed that CROCODILIA. Al it does not sink deeper than the line of the posterior border of the pterygoids ; and this modification is exaggerated in the Crocodilus Hastingsia (T. VI, fig. 3) in which the notch in question is merely the interval between two slender diverging processes from the middle of the back part of the pterygoids, 24. The posterior aperture of the nasal passages is wholly surrounded in the Crocodilus Hastingsie by the horizontal plate of the pterygoids, and has the same position and aspect as in the Alligators ; but its form is heart-shaped, with the apex directed backwards, and the antero-posterior diameter exceeding the transverse one. I have not met with this form of the posterior nostril in any other species of Crocodilian; but it is repeated in two individuals of the Croc. Hastingsie, and may be regarded as a specific character. The ectopterygoid, 25, T. VI, fig. 3, T. II, fig. 2 (@ fig. 2, pl. ii, ‘Ossemens Fossiles,’ t. v, pt. ii) articulates with a larger proportion of the outer surface of the pterygoids (24) in the Crocodiles than in the Alligators: it agrees with the Crocodiles in the extent of this articulation in the Croc. Hastingsia. Boo 22 94 20—20 In the upper jaw the fourth, ninth, and tenth are the largest; and the fifteenth and sixteenth exceed in size those immediately before and behind them. The alveolar border of the jaw increases in depth to form the sockets requisite for firmly lodging these larger teeth, and gives rise to the festooned outline of the jaw, which is found in all Crocodiles and Alligators in proportion as the teeth are unequal in size. The lower jaw presents the same compound structure as that in the Crocodilia, with the general form characteristic of that in the Alligators and in most of the true Crocodiles: the symphysis, e. g. is as short as Crocodilus biporcatus and the Alligator miger, in which it extends as far back as the interval between the fourth and fifth socket. ‘This is the relative position of the back end of the symphysis in a fine and perfect under jaw of the Crocodilus Hastingsie in the collection of the Marchioness of Hastings. Ina portion of the under jaw of apparently the same species of Crocodile, from the same locality, in the collection of Searles Wood, Esq., F. G. 8., the symphysis terminates opposite the interval between the third and fourth tooth. The chief distinction observable between the modern Crocodiles and Alligators in the lower jaw is the greater relative size of the vacuity between the angular (30) and surangular (29) pieces, and the greater relative depth of the ramus at that part, in the Alligators. In these characters the lower jaw of the present species more resembles that of the true Crocodiles; although, as the vacuity in question is somewhat larger, a slight affinity to the Alligator might be inferred from that circumstance. The comparative figures of the hinder third of the mandibular ramus in T. XII, fig. 4, 5, 6, will exemplify the difference in question, and the degree of proximity to the crocodilian and alligatorial characters respectively. With regard to another character deducible from the relation of the backwardly- produced angle of the jaw to the articular surface, the Cracodilus Hastingsie more 6 The number of teeth in this species is A2 FOSSIL REPTILIA OF THE LONDON CLAY. decidedly resembles the Alligator: I allude to the depth of the excavation between the articular cavity (29) and the end of the angular bone (30), and to the lower or higher level of the angle itself: the fossil jaw (fig. 5) resembles the Alligator (fig. 6) in this respect more than the Crocodile (fig. 4). The alveoli are twenty in number in each ramus of the Crocodilus Hastingsie : the third and fourth are large, of equal size, and close together; behind these the eleventh, twelfth, and thirteenth are the largest, and the alveolar ridge is raised to support them : after the seventeenth the summits of the crowns of the teeth become obtuse, and the crowns mammilloid, and divided by a constriction or neck from the fang; they each, however, have a separate socket, as in the Crocodiles, the septa not being incomplete at the hinder termination of the dental series, as in the A//igator niger figured in my ‘Odontography.”* . Fig. 3, T. IL, gives a representation, of the natural size, of the cranial platform of a young Crocodilus Hastingsiea in the collection of Searles Wood, Esq.; the hemi- spheric depressions in the surface of the bone are more regular, distinct, and relatively larger, and the interorbital part of the frontal is narrower, concomitantly with the larger proportional eyeballs and orbits of the young animal. The relatively larger supratemporal apertures form another character of nonage; but there is no ground for deducing a specific distinction from any of the differences observable between this part of the young crocodile’s cranium and the corresponding part of that of the more mature specimen (T. VI). ALLIGATOR HantToNIENSIS, Wood. Tab. VIII, fig. 2. London Journal of Palezeontology and Geology. On reviewing the characters of the skull of the Crocodilus Hastingsie we perceive that they combine to a certain extent those which have been attributed to the genus Crocodilus and the genus Alligator ; in general form it resembles most the latter, but agrees with the former in some of the particulars that have been regarded by Cuvier and other palzeontologists as characteristic of the true Crocodiles. I allude more particularly to the exposed position of the inferior canines when the mouth is shut. Respecting which, however, I am disposed to ask, whether this be truly a distinctive character of importance? One sees that it needs but a slight extension of ossification from the outer border of the groove to convert it intoa pit; yet the character has never been found to fail as discriminative of the several species of existing Crocodiles and Alligators hitherto determined. It constitutes, however, the only difference between the skulls of the Crocodilus Hastingsié in the collection of the Marchioness of Hastings and that fine portion of skull now, by the kindness of Mr. Searles Wood, before me, on which he has founded the species named at the head of the present section. So closely, in fact, * Tom. ui, pl. Ixxy, fig. 3. CROCODILIA. 43 do those specimens from the same rich locality correspond, that any other comparative view than that given in T. VIII appeared superfluous. In both the broad nasal bones terminate at the same distance from the external nostril, which is accordingly formed exclusively by the premaxillaries ; in both, the palate-bones present the same narrow, truncate posterior ends, and the same equal breadth of their anterior portions included between the maxillaries; only these terminate rather more obliquely in Mr. Wood’s specimen, their anterior ends forming together a very obtuse angle directed forwards. But this is comparatively an unimportant difference, and I regard as equally insignifi- cant the slight interruption of the transverse line of the maxillo-premaxillary suture, at the middle part, which will be seen by comparing fig. 2 with fig. 1, in T. VIII. The teeth are the same in number, arrangement, and proportion in the Heine (6 See YaleeseiEi Se FA) = VPs a) ts Ho > \e <} E& le ; . S & {8 ) > le A) eS > s ae Kus E Kus) 2 XS E Rus 2 Kus EGG: Sy ea M ; m Linoseyy 2 Sass m W pt Sasa m tins SY & . m LLSNI NVINOSHLIWS Saiuvyugia LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNINVINOSHLIWS S31y! Zz n ie Ze. g ne 2) z 2) 7 Ly, = = Wy, : = z 3 = = - My, F = Yop = \ = Yili ys 5 ee 2 yt eS z B NWS 2 8 a = Bo oe Bete B = RI ES SMITHSONIAN _ INSTITUTION PNCITOTIESNTENVINOSHIINS Sa lyvuad I1_LIBRARI ES SMITHSONIAN _INSTI ul a “ & ul @ us Z feed se ox = oc = o | < a= < c a Ce a < = fn = cc = ce = cc Ss cm : = fa) 5 a = fea = = Oo —_ = (o} = fo) = ae J) 2 a z= : ~~) Fe y ILSNI NVINOSHLINS S31YVvVYdIT LIBRARIES_ SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3IUY Ss z is S = z fe z ® Gs 2 @ = <5 e 3 as = a Yi tip P 2D = S = 3 2 IY 2 5 2 5 a — a E ey a = D = = = = i = E ae & m 2 us = o Zz a z R | ES SMITHSONIAN _ 3 Soe eae poaluwud ul BRARI ES SMITHSONIAN _ Ae = z 2 = = 2 & = z Sale z a z =a S WSS 3 z ee z Ree 5 ay aed: Rie s ILSNI NVINOSHLIWS” $3 Iyvyuydiq LIBRARI ES SMITHSONIAN _INSTITUTION | NOILNLILSNI NVINOSHLINS S314 = iy: @) S = eee n Z Six Ww i) uw 2 NER Ww ZX = =. ip Jp 3 : ce =i \ASN aivPipp > < = WS S x AN = = this, S ~ cS \\, = 2 NY 5 a Up 3 e 3. Ne Ly aj aT - oa = ae S my RIES” SMITHSONIAN” INSTITUTION NONLALILSNI NVINOSHLINS SSIYVYGI7 LIBRARIES SMITHSONIAN au ; 2 iY, oo o Q 2 wo 2 y o = GLY, = a \ = a E es) ree LE > > NS ee > = i> - VP ne De oN AY iE Fs) = 23) n* ee: — —_ S\S on ox es. ox ASNI_NVINOSHIINS Saiuyugi7_L!BRARIES SMITHSONIAN INSTITUTION NOILMLILSNI_ NVINOSHLINS S314h =~ n “ ® = ; + Se op) re) 2 > 2 SMITHSONIAN SMITHSONIAN NVINOSHLINS SMITHEONIAN NVINOSHLIWS NVINOSHLIW my NVINOSHLINS S3IYVYS!IT LIBRARIES SMITHSONIAN INSTIT RIES SMITHSONIAN INSTITUTION SatuyvudI? LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI _ > Zp) = —_— : z a z i : = = @ & . faa

= 2 2 = E - Gy cass E 2 Ez RIES) SMITHSONIAN INSTITUTION NOILNLILSN! NYINOSHLINS S3IuUVYuaIT SMITHSONIAN _ 2 z ier n Z n uz a) = ~ s Se NES < = < = s R ENN 2 gs = Z = zZ fe} 2 8 GY 2 6 £ rs} 2 . < = zy Li, Ee z = z = aye Baty eM eda ee a 2 z PoNTENVINOSHLINS, S34 1YVYGIT_LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI. NVINOSHLIWS S314 ee Qs (ay Peet Sy creep Ce er aA — => Ree bd ua AW a cn be <} & Me fp fi» = \y < = ro fe FE B WWE ey) 2 Mew) EC Ke 2 apy = mae 2 = a \ WS - A CU ONG, Vo] = 47 f — Kiapys/ & “Sabye \ e m K\ Bo Nise om NnossS 2 m INOS oe ASE na DN: S nw . ss = w p= — aoa RARIES SMITHSONIAN INSTITUTION NOILMIILSNI_NVINOSHLIWS, SaluvugiT [LIBRARIES SMITHSONIAN _ If z z Z ty, < = W: < = < =i Zz =] Mp. = 5 Ay, = z 2 a = a 2 o Fa a i pod SVYGIT_LIBRARIES SMITHSONIAN INSTITUTION NV NOS NS =z us Z wl J z w o WwW Uy, = = Z a = Yi, ° 2 = Ww a a < a < aA < Se WN a | = 3 =! Aa a a = NS Pra S F S = Sg = Bae is -! -! 2 BRARIES_ SMITHSONIAN, INSTITUTION NOILNLILSNI_NVINOSHLIWS S31YVua 11 LIBRARIES SMITHSONIAN _It iS) 2 o = o es 2 a = ee yee es i= 2 = ay eins eatnee 2 YY > 5 D 5 2 2 PM > = GG = = 2 = 2 : yy 2 = n =) on %5 = n = HINLILSNI_ NVINOSHLINS SSINVEEIT_LIBRARIES SMITHSONIAN INSTITUTION | NOILMLILSNI_ NVINOSHLIWS |S 2 z Reo wn oS Zz ek z 2 = SZ = Hig =] = Sj = = SSF ’ oS x= WY = — = Sie = = = z % S = S = = S (a 2 ae a . 8 3 ae . 2 BRARIES SMITHSONIAN INSTITUTION NOILNLILSNI_NVINOSHLIWS Sa1uvedi7_ i. fe | w & Z PAC SS & = & =. fie! = BIS = z. = = : : : 5 £43 5 > Na: =e) i = = b LE La aw SSS — = Ee its ee Anes bes = \s Ea - INSTITUTION NOILMIILSNI_ NVINOSHLINS, S31uvyaIT LIBRARIES SMITHSONIAN _ 1 z ete) ae = 5 z = x = z Si iY z = z z g Na : Ag 5 fefi2 NX: g NE z AW? ie SMA ENN 2 = i = =i WY Se = =. s +S" 5 = , 2 a fas z a 2 7) oS 2 a NLNLILSNI_NVINOSHLINS Sa 1YVYGIT_ LIBRARIES SMITHSONIAN_INSTITUTION Nea NING S ae & w % tl % ul BY w = a we sic) = wc = <= at < eA <= c UES RS a = S cc sn Pee GS OUWSSN o oO —_— 0 } = faa) ry > o z2 55 S = 2 = en = = 2 =D) Fs —_ BRARIES_ SMITHSONIAN, INSTITUTION NOILALILSNI_ SIIYVUIT_LIBRARIES_ SMITHSONIAN | S i S — 5 = ° = — aod o = ao == a = eo = =o = a = 2 - a = = Bo = Fe Z m = m 2 m => ie pl = a = n Sec o = ILNLILSNI_NVINOSHLINS Saluvud tT SMITHSONIAN INSTITUTION | NOILMLILSNI_ NVINOSHLUNS | s <= = as = =z) 6s = <= = z= =| prey ee 4 = = z 4 “typ . ro) = : Ifa 2 g 3B z : é = 2 iy = Z E S = Si ge _ 2 : | JSON oie 2 SMITHSONIAN INSTITUTION NOILNLILSNI_NVINOSHLINS S31YVY@IT_ LIBRARIES SMITHSONIAN _| Ww a aso ee a re PN rial heave epee tan asoe oF RIES RA LB oO oO 00 © oO (o>) 2 11 9 048