On the Mammals collected by Mr, A, Robert at Chapada *, Matto Grosso (Percy Sladen Expedition CentraI Brazil).

PROCEEDINGS OF THE GENERAL MEETINGS FOR SCIENTIFIC BUSINESS OF THE ZOOLOGICAL SOCIETY. OF LONDON. 1903, vol. II. (l\IAY-DEOEl\fBER.) PRINTED FOR THE SOCIETY: AND SOLD A'l' TH:)3JIR HOUSE IN HANOVER-SQUARE. LONDON: MESSRS. LONGlVIA NS, GREEN, AND qo., PATERNOSTRR RffW. 232 MR. OLDFIELD THOMAS O:N THE :MAMMALS 2. On the lVIammals collected by Mr, A, Robert at Uhapada *, l\Iatto Grosso (Percy Sladen Expedition to Oentra I Brazil). By 0LDFlELD THOl\IAS, F .H.S. [Receiyed ,July 23, 1903.] (Plate XXVII, t) By the generosity of Mrs. Percy Sladen, Mr. Alphonse Robert, who had already done such good work in Siio Paulo and Parana, was enabled to make, during the latter half of 1902, a collecting expedition for the benefit of the National Museum to Matto Grosso, Central Brazil, a region in respect to which the Museum *Chapada heing a word of Portuguese Qrigin (signifying a plateau) bas its accent on the second ;i.; Cuyaua and other names of Guarani derivation on lhe final a. t l:'ur explanation of the l'latc, �re p. 2'H, 1903.] OF THE PERCY SLADEN EXPEDITIO!{, 233 collections had hitherto been extremely deficient. Carried out as it was with all his accustomed energy and courage, Mr. Robert's expedition was completely successful in spite of the many difficul ties in his way, and of his being unaccompanierl by any European helper. The collection, of which the present paper gives an account, is an astonishing one for him to have been able to obtain and prepare with his own hands, in so bad a climate, especially as so many of the mammals are of considerable size. Collections of mice are much more easily made than of dogs, peccaries, coati mondis, monkeys, &c. ; and the long series of the different species here enumerated speaks volumes for Mr. Robert's working qualities. The specimens are all prepared in modern fashion, with flesh measurements and separate skulls, and there are besides a number of skeletons, so that, bearing in mind the inaccessibility of the locality, the collection may be looked upon as one of the most valuable that the National collection has received for many years. The thanks of all zoologists are therefore due both to Mrs. Sladen for her generous help, and to Mr. Robert for the admirable way in which he carried out his instructions. Mr. Robert started from Santos on the 15th April, 1902, and after passing through the usual difficulties and delays of quaran tine &c. incidental to entering Argentina, proceeded via Buenos Ayres, Asuncion, and Cuyaba, Matto Grosso, to Santa Anna de Chapada, a village situated at an altitude of about 800 m., on the Serra do Chapada, some thirty miles N.E. of Cuyaba, arriving there on the 17th June, and staying there collecting until the 29th November, when he came down the river again with the results of his labours. Mr. Robert spent most of his time collecting mammals, bnt he also obtained a considerable number of birds, some reptiles ;ii, mollusca t, insects, &c., all of which have been presented to the National Museum by Mrs. Sladen. The present is the third considerable collection of mammals that has been made in the district of Cuyaba. The first was that formed by Natterer in 1825-1829, and described in conjunction with his other Brazilian mammals by .. Wagner and Pelzeln. The i:;econd was that of the "American Naturalist Exploring Expedi tion" of 1882, under the leadership of Mr. H. H. Smith, and was made at the very village, Sta. Anna de Chapada, where Mr. Robert worked. The mammals were described by Prof. E. D. Cope t in combination with those obtained at Sao Joao, Rio Gran<le do Sul, where the expedition stayed before passing on to Chapada. No other important series from single localities have ever been made nearer to the present region than those from the Rio Jordfio (Robert), Lagoa Santa (Lund and Reinhardt), Paraguay (Azara, Rengger, and Foster), and those obtained along the eastem side of the Andean chain by P. 0. Simons. * C'f. Boulcn:;cr, s11p1•1i,, p. 6� ; Smith, sup1·a1 p. 70. t Am. �at. niii. p. l'.!8 (1S8?). 234 1\IR. OLDFIELD THOMAS OX THE l\IAl\IMALS [Nov. 3, Ouyaba is on the river level, and the country towards Ohapada is all at the same low altitude until the abrupt rise of the plateau edge is reached, comparatively close to Santa Anna. Mr. Robert has discovered a considerable number of new and interesting species, of which the most notable is the Wild Dog (Pl. XXVII. ), which I have named in honour of the late Mr. Percy Sladen. 1. OEBUS AZAR.iE Rengg. 0.1022, 1041, 1083, 1085, 1093, 1188, 1143, 1171, 1174, 1196, 1197, 1198, 1208. �. 1045, 1082, 1086, 1094, 1104, 1105, 1152, 1175, 1176. In the absence of any certainty that Spi."'{'s name Cebits libidi nosiis (type locality, Rio San Francisco) really belongs to this monkey, I use that of Rengger, which antedates Geoffroy's C. elegans by many ye::trs. This fine series emphasizes the difficulty of working out members of the present genus, for there is considerable variation both in the general colour, in the relative distribution of black. and yellow on the head, and in the length and development of the temporal tufts. 2. AoTus AZAR.iE Humb. llycti'pitlieciis azcirce auct. Q. 1150. 28 September, 1902. 3. OALLITHRIX �* l\IELANURA Geoff. d'. 1130. �. 1132. 4. HrsTroTus VELATUS Geoff. �. 1186. 5. MICRONYCTERIS l\lEGALOTIS Gray �. 1014, 1015. G. HE}IIDERl\IA PERSPICILLATUl\I Linn. o. 1063. �. 1012, 1013. 7. LoNCHOGLOSSA CAUDIFERA Geoff. d'. 1000, 1001, 1002. �. 999, 1111. 8. ARTIBEUS PLANIROSTRIS Spix. d'. 1089. *For the use of Callith1·i,i: instead of Hanale see Ann �fanN H (7) xii P• 455 (1903). r ' • ~ o• • • · • 1903.] OF THE rERCY SLA.DEN EXrEDITIO�. 9. V AMPYROPS LINEATUS Geoff. cJ. 1091, 1092, 1099, llOO, 1101, 1200. � -1090, 1096, 1097, 1098. Quite similar to topotypes from Paraguay. 10. FELIS WIED II Schinz. Flat skin, native made. 11. CANIS CANCRIVORUS Desm. cJ. 1139, 1184. �. 1170. 12. CANIS SLADENI, sp. n. (Plate XXVII.) cJ. 1080. l1 August, 1902. Type. (B.lVI. No. 3.7.7.40.) Allied to C. 'Vetulus, but larger, with longer skull, greyer in colour and with black feet. An additional upper molar present on each side in the type. Size rather larger than in C. vetulus. Fur close and uniform, not intermixed with long black piles; long hairs of back about 33-35 mm. in length. General colour above clear grey suffused with the buffy of the underfur, the net result being nearest to Ridgway's "smoke-grey." Individually the longer hairs are light for their basal and black for their terminal halves, with a white subterminal ring about a quarter of an inch from their tip. Underfur brownish plumbeous bamlly, buffy clay-colour terminally. Under surface little paler, dull brownish, lighter in the axillre and groins. Muzzle and chin blackish. Crown like back, but the light ends to the underfur almost obsolete. Back of ears brown, grizzled with clay-colour; longer hairs of inner surface creamy; a large postauricular patch yellowish clay-colour. Fore arms and thighs like body; hands and feet, both above and below, blackish brown. Tail brownish grey, broadly washed with black at the end and over the caudal gland. Skull, as compared with that of C. vetulus, markedly larger throughout; interorbital region longer and narrower, the part behind the postorbital processes specially lengthened; frontals smooth and swollen ; palatal foramina very long ; palate ending opposite to middle of m.2; mesopterygoid fossa long, the distance from the hinder edge of the palate to the concavity below the hamular processes 22 mm.; front edge of prresphenoid 6 mm. behind the palatal edge. Penis-bone of the usual canine canoe shape, but much less strongly keeled than in the one example of C. vetulits available. Teeth of the general type of those of C. vetiil'lls; p.4 short and rounded; a supplementary and probably abnormal m.3 present on each side, longer on the left than on the right. Dimensions of the type, measured in the flesh :- Head and body 650 mm.; tail 310; hind foot (s. u.) 130; ear 76. 236 MR. OLDFIELD TH0:\:IAS OX THE 1'IA:MlHALS [Nov. 3, Skull-greatest length 122 mm. ; basal length 112 ; zygomatic breadth 63·5; nasals, length mesially 38; interorbital breadth 19; tip to tip of postorbital processes 26; intertemporal breadth 17·8; breadth of brain-case 43·5; palatal foramina (singly) 8 X 2:5; palate length 56; length of os penis 47. Teeth-p.4, length on outer border 8·2; m.1 8 X 9·5; m.2 6·4 X 8·6; combined length of m.1 and m.214·5. P.,, length 64; m., 10, m.� 6·7, m.3 4·4. This Dog differs from C. vetulus, to which it is no doubt allied, in its darker and greyer general colour, its blackish feet, and larger skull. The possession by the only specimen of a third pair of upper molars is probably an individual abnormality, but occur ring here in one of the most primitive types of Canidre it presents a very interesting and suggestive case of atavism. For this reason alone further examples of this animal will be welcomed with interest, to see how frequently the extra teeth are present. I have named this distinct animal in honour of the late Mr. Percy Slaclen, whose valuable zoological work is known to all naturalists, whose death was lamented by a wide circle of friends, and whose widow has chosen the most useful and suitable way of perpetuating his memory by bearing the expense of Mr. Robert's expedition to Central Brazil. 13. OANIS VETULUS Lund. o. 1108. �. 1189. Quite alike externally, these two specimens differ surprisingly in the size of their molars, the female having the larger. Further specimens will be needed before this difference can be explained, but it may be noted that the larger-toothed specimen agrees closely with typical C. vetulus, whilst that with the smaller teeth has an equal resemblance to the wrongly localized type of Gray's "Lycalopex fulvicaudus, var. 1. chiloensis." * 14. SPEOTHOS VENATICUS Lund. �. 1127. The Bnsh-dog is seldom captured by collectors, and this excellent specimen is a valuable accession to the Museum. 15. GALERA BARBARA L. o. 1064, 1117. �. 1065. 16. NASUA NASUA L. o. 1043, 1054, 1056, 1057, 1069, 1087, 1164. �. 1055, 1070, 1071, 1129, 1136 1137 1141 1142, 1156, 1157, 1163. ' ' ' Among this fine series of eighteen Coatis it might be said that *P, Z. �- 18li8, p. 511 ; Cat. Carn. B. �I. p. HlS (l,S68). 1903.] OF THE PERCY SLADEX E.XPEDITIOX. 237 no two can be found that are exactly alike, the set ranging in colour from uniform red to nearly uniform black, and varying as much in details as in general tone. The skulls also vary very ccmsiderably both in size and in the development of their bullre, the extremes appearing at first sight to be quite different species, but being connected together by every gradation. 17. ScrnRus LANGSDORFFI Brandt. 0.1088, 1154, 1177, 1182, 1195. �. 1081, 1124, 1155. A valuable series 11epresenting the true S. langsdorffi, which name has been applied by various authors to many different members of the group. 18. RHIPIDO:\IYS ROBERTI, sp. n. o. 1023, 1140. A Rhipidomys intermediate in size of skull between the large species of the R. rnacrurus-latimanus group, and the small ones allied to R. phceotis; most nearly related to the latter. Fur of medium length; hairs of back about 7-8 mm. in length. General colour above dull fulvous or "clay-colour," finely lined with black. Sides clearer itnd more buffy, bordered below by a clear ochraceous-buff line edging the pure sharply-defined white of the belly. Face like back, rather less heavily lined. Cheeks buffy, like sides. Eyes without darker rings. Ears of medium length, brown, not strongly contrasting with the general colour. Ouoor side of a.1·ms and l8gs like flanks, inner pure white on the fore limb down to the wrists, but on the hind limb only halfway down to lower leg, the fulvous passing round the ankles; hands and feet dull whitish. Tail fairly long, rather less hairy than usual in this genus, its basal inch furry all round and coloured like the body; the remainder uniformly brown. Skull shaped most like that of R. benevoleus, although much larger. Supraorbital ridges fairly well developed, evenly divergent backwards. Outer plate of anteorbital foramen scarcely developed anteriorly. Palatal foramina of medium length, narrow an<l pointed anteriorly, broadly open posteriorly; not reaching back to the level of m.1 Interpterygoid fossa of normal shape, with rounded anterior border. Dimensions of the type, measured in flesh :- Head and body 110 mm.; tail 145; hind foot (s. u.) 25, (c. u.) 26·7; ear 16. Skull-greatest length 32 mm.; basilar length 24·5 ; greatest breadth 16 ; length of nasals 11 ; interorbital breadth 5·6 ; breadth of brain-case 14 ; length of palate 13·6 ; diastema 8 ; palatal foramina 5 x 3· 1 ; length of upper molar series 4·8. Type. Adult cJ. B.lVI. No. 3.7.7.67. Original number 1023. Collected 6 July, 1902. Although Mr. Robert's trunk measurement of the type of this species is slightly less than that giYen by M1·. Simons for R. phreot1's, 238 MR. OLDFIELD TH01\IAS ON THE MAMMALS [Nov. 3, its skull is very markedly larger, and its teeth even more so. That is, however, the species with which it seems most nearly allied. The presence of Rhipidomys robe1·ti, like that of Xeacomys, aives evidence of the affinity of the Cuyaba region of Matto Grosso �ith the eastern slopes of the Andes, from. which all the previous members of this group have hitherto come. In naming this distinct species after Mr. Robert I wish to make very special acknowledgment of the admirable work he has been doing i.n South America during the last three years. His collec tions, perfectly prepared in spite of all the difficulties incidental to a damp tropical climate and uncivilised surroundings, have revolutionised the material available in the British Museum for the study of the Mammals of Southern Brazil. Numbers of species, previously known to us only by the original descriptions or by a few unmeasured, discoloured, or otherwise imperfect specimens, are now represented by series of perfect skins with all the details incidental to good modern work. And again, the present collection, formed in a still more difficult country, links up in the centre his own Eastern Brazilian material with that obtained along the edges of the Andean chain by the late Mr. Perry 0. Simons. 19. NECTOi\'.IYS SQUAMIPES MATTENSIS, subsp. n. 0.1005, 1019, 1025, 1037, 1050, 1203. �. 1031, 1033, 1048, 1109, 1131. External characters as in true sqitamipes and in garkppi; the latter also appears to grade into squarnipes. Skull rather shorter and more rounded, less slender and elongated than in true squccmipes. Nasals as usual narrow and elongate. Supraorbital edges with a marked raised bead, more developed than in squarnipes, much more than in garleppi. Inter parietal small, narrow antero-posteriorly, its anterior edge generally directly transverse. Palatal foramina long, widely open, with a slight angular constriction at the junction of their anterior and middle thirds. Opening of posterior nares wide, the edge of the palate squarely transverse, the median pterygoid fossa broader anteriorly than posteriorly, its breadth in front greater than the length of m.1 Lateral pterygoid fossre propor tionally narrow and sharply pointed anteriorly. (In N. squamipes the median fossa is narrow anteriorly, broadening backwards, with a rounded anterior border.) Dimensions of the type, measured in the flesh:- Head and body 200 mm.; tail 200; hind foot (s. u.) 47 (extremes 44-47), (c. n.) 50; ear 24. Skull-greatest length 43 mm.; basilar length 34·6; zygomatic breadth 22·2 ; interorbital breadth 7; interparietal 3·4 x 11 ·5 ; palate length 20·4 ; palatal foramina 7·5 x 3·2 · breadth of posterior palatal fossa anteriorly 3·8; length of 'upper molar series 6·3. 1903.] OF THE PERCY SLADEN EXPEDITION. 239 1'ype. Adult male. B.M. No. 3.7.7.71. Original number 1050. Collected 30 July, 1902. The difference in the structure of the pterygoid fossre, perfectly uniform throughout the series, seems to make it necessary to give a special name to the Matto Grosso Water-rat; but as lower down the same river, in Paraguay, specimens agreeing with the true N. squamipes occur, I only distinguish it as a subspecies. It may be noted as a curiosity that a very large proportion of the skuHs of Nectomys, even when quite old, have their median interparietal suture open, a characteristic very rare in mammals generally. 20. ORYZOMYS LA.TICEPS Lund. 0.1003, 1007, 1008, 1020, 1052, 1053, 1062. � -1004, 1006, 1010, 1016, 1026, 1049. These specimens appear to be certainly identical with those obtained by Mr. Robert on the Rio Jordao *, S. Minas Geraes, not far from the typical locality of the species. 21. NEACOMYS SPtN0SUS AM<ENUS, subsp. n. 0.1061, 1077, 1188. �. 1036, 1051. Size averaging slightly larger than in the true Peruvian 1¥. spinosus, and tail rather longer. General characters very much as in that animal, in spite of the great distance between the localities. Fur shorter and thinner than in spinosus, especially on the under surface. Colour above bright ochraceous, liberally lined with black along the back, but the dorsal area nevertheless not nearly so dark as in spinosus. Sides clearer fulvous, the yellowish line along their lower edge not so sharply defined as in spinosus. Belly and inner sides of limbs white, the hairs white to their roots; line of demarcation not very sharply defined. Face rather greyer than body. Ears uniformly brown. Arms and legs like sides, or rather duller; hands and feet dull whitish. Skull as usual varying a good deal in size and shape, but in a general way that of subsp. amceniis is a little larger than that of spinosu,s, with more rounded brain-case, less developed supraorbital ridges, and slightly larger palatal foramina, but none of these points can be regarded as really distinctive. Dimensions of the type, measured in the flesh :- Head and body 90 mm. (extremes 80-90); tail 102 (102-112); hind foot (s. u.) 23 (22-24), (c. u.) 24·5; ear 16 (14-16). Skull-greatest length 24·6mm.; basilar length 18·1; zygomatic breadth 12·7; interorbital breadth 4·3; palatal foramina 4; length of upper molar series 3·1. 1'ype. Old male. B.M. No. 3.7.7.84. Original number 1077. Collected 11 August, 1902. The occurrence of a J.Yeciconiys in this region was quite *C'f. Ann. l\fog. N'. H. (7) viii. p. 530 (HlOl).