Part 1

PROCEEDINGS

OF THE

WASHINGTON ACADEMY OF SCIENCES

VOL. III, PP. 111-138. MARCH 26, 1901

MAMMALS COLLECTED BY DR. W. L. ABBOTT ON THE NATUNA ISLANDS.

BY GERRIT S. MILLER, JR.

About three months during the spring and summer of 1900 were spent by Dr. W. L. Abbott in exploring the Natuna Islands in the South China Sea.[1] Specimens were collected at the following localities: Pulo Midei, or Low Island (May 23-26), Pulo Seraia (May 29), Sirhassen Island (June 1-10), Pulo Subi (June 12-13), Pulo Lingung (June 17-19), Bunguran, or Great Natuna Island (June 24-July 31) and Pulo Laut, or North Natuna Island (August 5-13). About 265 mammals were obtained, all of which have been presented to the United States National Museum. This paper contains an account of these, and is published here by permission of the Secretary of the Smithsonian Institution.

Two extensive collections of mammals had been made on the Natuna Islands previous to Dr. Abbott's visit, the first by Mr. A. Everett during September and October, 1893, the second by Mr. Ernest Hose during July, August, September and October, 1894. These have formed, either wholly or in part, the basis of several papers,[2] which constitute the literature relating to the mammals of the islands.[3] Twenty-eight land mammals have been recorded as actually represented by specimens, though several others are mentioned which the collectors ascertained to occur. Dr. Abbott secured forty-four species, but failed to obtain seven[4] of those previously taken. The total number of mammals collected on the islands thus becomes fifty-one. This increase is due, in part to the recognition of a larger number of insular forms than has been admitted by previous writers, but also to a considerable extent to the actual addition of species not hitherto taken. Species new in the latter sense are distinguished in the present paper by absence of reference to previous records.

In regard to the faunal relationships of the Natunas, whether predominantly Bornean or Peninsular, about which much has been written,[5] it may be said that this collection, together with much of the other work recently done by Dr. Abbott, tends to show that there is greater general uniformity in the mammalian fauna of Borneo, the Malay Peninsula, and the intervening islands than has been hitherto supposed. It seems unprofitable therefore to offer conjectures as to the probability of greater nearness of the Natuna mammals as a whole to those of Borneo or to those of the Malay Peninsula.

MANIS JAVANICA Desmarest.

1895. _Manis javanica_ THOMAS and HARTERT, Novitates Zoologicæ, II, p. 492. December, 1895 (Bunguran).

An adult male was taken on Bunguran, June 24, 1900. Total length 914; head and body 508; tail 406.

TRAGULUS BUNGURANENSIS sp. nov.

_Type._--Adult male (skin and skull) No. 104604 U. S. National Museum. Collected on Bunguran Island, North Natunas, July 9, 1900. Original number, 547.

_Characters._--Color pattern essentially as in _Tragulus nigricans_ Thomas, from Balabac. Size equal to that of _T. canescens_ from the Malay Peninsula, therefore much greater than in the Balabac animal.

_Color._--Back uniform ochraceous, fading to buff on sides, the hairs everywhere gray at base. Both back and sides everywhere darkened by black hair tips, but these never sufficiently abundant to produce a dark shading in excess of the ochraceous. The relative proportion of the dark wash to the light under color is precisely the same as in _Tragulus canescens_ and _T. napu_ (from Linga Island) but the black is less conspicuous than in the Bornean form of _T. napu_. Legs, except white area on inner side, like back but slightly brighter and less shaded with black. Entire dorsal and lateral surface of neck clear black to base of hairs, a few ochraceous specks visible on close scrutiny, particularly at sides near throat markings. On shoulders this black area fades abruptly into color of back; on head it passes forward between ears and eyes nearly to muzzle. Cheek, region between eye and ear, and line extending forward over eye to muzzle and separating black median stripe from naked loral space, ochraceous, essentially like that of legs. Throat markings as in _Tragulus nigricans_, but white stripes apparently even more restricted. Region occupied by posterior white stripes black, continuous with that of neck, but distinctly speckled with ochraceous. Region occupied by anterior stripes ochraceous, continuous with that of cheeks and somewhat less pure and more speckled with black. White stripes as follows: (_a_) One on each side of naked chin area. These are about 50 mm. in length and never more than 10 mm. in breadth, but occasionally so narrow as to break up into two or more spots. They are separated from naked chin patch by an ochraceous stripe slightly broader than the white. Chin area narrowly and discontinuously bordered with white, especially in front. (_b_) Two posterior lateral stripes varying from 50 mm. to 80 mm. in length, and never more than 12 mm. wide. They are strongly convergent anteriorly, and sometimes nearly joined together in front by a median spot. These white stripes are always separated from the anterior stripes by an ochraceous median area varying from 10 mm. to 25 mm. in width. (_c_) A median stripe lying between the posterior lateral stripes. Posteriorly this stripe is as wide as the lateral stripes, but it quickly narrows and sometimes disappears at middle of latter, though usually represented again by the median spot already referred to. In none of the specimens is this stripe broad and continuous anteriorly to level of front of lateral stripes as in Nehring's figure of the throat markings of _T. nigricans_.[6] Collar narrow, ochraceous grizzled with black. It is seldom more than 25 mm. in width; therefore much narrower than indicated by Nehring's figure. Behind the collar is a whitish gray median area continuous laterally with narrow light stripe down inner side of fore legs. This light area is sometimes divided by a dark median line joining collar with buff of belly. Belly and chest buff, essentially like that of sides, with which it forms no contrast in color. As on the sides the buff is clouded by black hair tips, but the hairs are scarcely if at all gray at base. On chest the dark hair tips tend to form a median stripe, which is sometimes sharply defined and continuous with the ochraceous line occasionally dividing white of breast. A clear whitish area slightly larger and better defined than that of breast occupies region between hind legs. It is continuous with white stripe down inner side of hind legs. This stripe is usually divided on thigh by encroachment of the surrounding ochraceous. Tail silky white below and at tip, essentially like back above.

_Skull._--The skull of _Tragulus bunguranensis_ fully equals that of _T. canescens_ in size, and distinctly exceeds that of the Bornean form of _T. napu_. It is much larger than that of _T. nigricans_, which proves to be a medium sized species like _T. rufulus_. In general form the skull agrees so closely with that of _Tragulus canescens_ that it is only to be distinguished by its slightly greater relative breadth and smaller, less inflated audital bullæ. As compared with the skull of _Tragulus nigricans_,[7] that of _T. bunguranensis_ is much larger (distance from back of occiput to front of canine 103 instead of 92, zygomatic breadth 53 instead of 45), and the braincase is more conspicuously ridged for muscular attachment. That part of the braincase immediately above posterior root of zygoma is more conspicuously inflated. Otherwise I can detect no salient differences in the skulls of the two animals.

_Teeth._--The teeth are uniformly larger than those of _Tragulus nigricans_, but in form they present no characters of importance. As compared with _T. canescens_ the premolars both above and below are conspicuously more robust, a character in which the Bunguran animal agrees with the Bornean form of _Tragulus napu_.

_Measurements._--External measurements of type: total length 647; head and body 571; tail vertebræ 76; hind foot 146; hind foot without hoofs 128. Average and extremes of five adults from the type locality: total length 643 (628-673); head and body 566 (558-584); tail vertebræ 77 (70-89); hind foot 142 (140-146); hind foot without hoofs 126 (124-128).

Cranial measurements of type: greatest length 114; basal length 107; basilar length 100; occipito-nasal length 106; length of nasals 32; diastema 13 (9);[8] zygomatic breadth 52 (46); least interorbital breadth 33 (28); greatest breadth of braincase above base of zygomata 38 (33); mandible 91 (78); maxillary toothrow (alveoli) 38 (34); mandibular toothrow (alveoli) 44 (39); anterior upper premolar 7 × 3.8 (6.4 × 3); middle lower premolar 7.2 × 3 (5.8 × 2.4).

_Weight._--Weight of type 3.8 kg.; of two other males 3.6 kg. each. Two adult females weigh respectively 3.6 kg. and 4.2 kg.

_Specimens examined._--Six, all from the type locality.

_Remarks._--_Tragulus bunguranensis_ is so distinct from the other known species as to require no detailed comparisons.

TRAGULUS sp.

Two specimens from Sirhassen Island are too immature for determination. Apparently they represent a member of the _napu_ group, allied to that occurring in Borneo. The throat markings show no approach to those of _Tragulus bunguranensis_.

TRAGULUS JAVANICUS (Gmelin).

1894. _Tragulus javanicus_ THOMAS and HARTERT, Novitates Zoologicæ, I, p. 660. September, 1864 (Bunguran).

1895. _Tragulus javanicus_ THOMAS and HARTERT, Novitates Zoologicæ, II, p. 492. December, 1895 (part, specimens from Bunguran).

Six specimens from Bunguran.

TRAGULUS PALLIDUS sp. nov.

1895. _Tragulus javanicus_ THOMAS and HARTERT, Novitates Zoologicæ, II, p. 492. December, 1895 (part, specimen from Pulo Laut).

_Type._--Adult female (skin and skull) No. 104616 U. S. National Museum. Collected on Pulo Laut, North Natuna Islands, August 11, 1900. Original number 625.

_Characters._--Smaller than _Tragulus javanicus_ from Borneo or Bunguran and very pale in color. Black clouding of upper parts inconspicuous, but dark nape band well defined.

_Color._--Back and sides light ochraceous-buff everywhere clouded by the blackish hair-tips, but these never in excess, except perhaps along middle of back and across lumbar region. Flanks, shoulders, neck, outer surface of legs and narrow line dividing color of sides from that of belly pale ochraceous. Nape band clear black, sharply defined from color of sides but quickly fading into that of shoulders. Top of head dull dark brown. A faint pale stripe over and in front of eye. Throat markings normal, the dark bands like neck. Collar very narrow. Under parts and inner surface of legs white. A faint yellowish shade along middle of belly. Tail white beneath and at the tip, ochraceous faintly shaded with brown above.

_Skull._--The skull of the type, though fully adult and with all the teeth distinctly worn, is smaller than in Bunguran specimens so young that the posterior molars are still below the rim of the alveoli. In form, however, it shows no marked peculiarities, though in general it appears to be somewhat broader in proportion to its length than that of the Bunguran animal.

_Teeth._--Teeth as in specimens of _Tragulus javanicus_ from Bunguran except that the premolars, both above and below, are shorter and broader, a difference which may prove to be an individual peculiarity only.

_Measurements._--External measurements of type: Total length 539; head and body 444; tail vertebræ 95; hind foot 107; hind foot without hoofs 95.

Cranial measurements of type: Greatest length 90 (94[9]); basal length 83 (87); basilar length 78 (82); occipito-nasal length 83 (89); length of nasals 25 (29.6); diastema 9.2 (9.8); zygomatic breadth 41.4 (40); least interorbital breadth 26.4 (25); breadth of braincase over roots of zygomata 29.4 (28.4); mandible 72 (75); maxillary toothrow (alveoli) 31.6 (34); first upper premolar 6.4 × 2.8 (7 × 2.6); mandibular toothrow (alveoli) 35.8 (38).

_Specimens examined._--One, the type.

_Remarks._--This is a pallid form of _Tragulus javanicus_, a species which apparently shows very little tendency to become differentiated into local races. The characters of the Pulo Laut animal were pointed out by Thomas and Hartert in 1895.

SUS NATUNENSIS sp. nov.

1894. _Sus_ sp. THOMAS and HARTERT, Novitates Zoologicæ, I, p. 660. September, 1894 (Bunguran).

1895. _Sus_ sp. THOMAS and HARTERT, Novitates Zoologicæ, II, p. 492. December, 1895 (Bunguran).

_Type._--Adult female (skin and skull) No. 104856 U. S. National Museum. Collected on Pulo Laut, North Natuna Islands, August 6, 1900. Original number 609.

_Characters._--Externally much like the Tenasserim form of _Sus cristatus_, but smaller; body brownish in marked contrast with black legs and face; skull conspicuously shorter and broader.

_Fur._--The fur throughout consists of bristles with no admixture of softer hairs. The bristles are everywhere less stiff than in the Tenasserim pig, but the difference is most noticeable in the mane, which, though well developed (about 80 mm. in length), is composed of bristles very slightly coarser than those of the surrounding parts, and of not more than half the diameter of the corresponding hairs in females of _S. cristatus_. Muzzle, chest, belly and ears nearly bare.

_Color._--General color black, clear and unmixed with brown on legs, throat, and face, but elsewhere heavily overlaid with brownish buff, particularly on back and sides. The brownish wash ceases abruptly just in front of ears, leaving the face and cheeks clear black. A conspicuous dull buff streak 100 mm. long and about half as wide at middle extends back from angle of mouth to level of posterior canthus of eye. It is sharply outlined above by black of cheeks, and below by that of chin. A faint buffy mark beneath eye. Tail like back.

_Skull._--The skull while much shorter than that of _Sus cristatus_ from Tenasserim is actually broader. As a result the width across postorbital processes is contained only about three times in occipito-nasal length, as opposed to nearly four times in the related species. Similarly the zygomatic breadth slightly exceeds one half of the basilar length, while in _Sus cristatus_ it is less than half. Width of palate between middle molars almost exactly one sixth distance from posterior edge of palate to front of premaxillaries (measured along median line). In _Sus cristatus_ the palatal width is contained nearly seven times in the same distance. Dorsal profile of skull slightly concave near base of nasals. Zygomata heavier and deeper than in _Sus cristatus_. Audital bullæ noticeably smaller and less inflated than in the Tenasserim pig. Mandible shorter and much more robust than that of _Sus cristatus_, the outward bulge of the ramus a little behind middle of toothrow greatly accentuated.

_Teeth._--As the teeth of the two specimens of _Sus natunensis_ are much worn, while those of the only skulls of _Sus cristatus_ at hand are not fully grown, it is impossible to make any accurate comparisons. The smaller size of the Natuna pig's teeth is, however, evident for the length of the entire upper toothrow does not equal that of _S. cristatus_ without the posterior molar. The crown of the middle upper molar appears to be more nearly square in outline than that of the Tenasserim pig, but in the very different condition of the specimens it would be unsafe to assume that this character is constant.

_Measurements._--External measurements of type; total length 1294; head and body 1117; tail vertebræ 177; height at shoulder 558; hind foot 220 (170); ear from meatus 100; width of ear 75.

Cranial measurement of type: greatest length 295 (332[10]); occipito-nasal length 282 (316); basal length 245 (275); basilar length 235 (263); length of nasals 135 (157); width of both nasals together posteriorly 34 (33); median length of bony palate 168 (183); width of bony palate at middle of second molar 30 (29); breadth between tips of postorbital processes 87 (87); least interorbital breadth 64 (65); zygomatic breadth 130 (133); occipital breadth 58 (62); occipital depth 100 (103); least depth of rostrum between canine and incisor 33 (39); mandible 225 (232); depth of mandible through coronoid process 104 (110); depth of ramus at front of first molar 40 (41); maxillary toothrow to front of canine (alveoli) 113 (131[11]); mandibular toothrow to front of canine (alveoli) 120 (138); crown of first upper molar 12 × 13 (18 × 16); crown of second upper molar 18 × 18 (22 × 16).

_Weight._--Weight of type, 40 kg.; weight of adult female from Pulo Lingung, 35 kg.

_Specimens examined._--Two, one from Pulo Laut, the other from Pulo Lingung.

_Remarks._--While the two specimens agree in all essential characters they differ in numerous minor details. The skin from Pulo Lingung is somewhat darker than the type, but the difference is due to the shade of the brown wash, not to any extension of the black. The skull of this specimen is more rounded posteriorly than that of the type, and the rostrum is shorter. Both specimens show conclusively that their relationships are with the _Sus cristatus_ of the Malay Peninsula and not with the _S. longirostris_ of Borneo, a case which finds an exact parallel in the giant squirrels.

MUS INTEGER sp. nov.

_Type._--Adult male (skin and skull) No. 104837 U. S. National Museum. Collected on Sirhassen Island, South Natunas, June 7, 1900. Original number 455.

_Characters._--A large robust species with coarse but not spinous fur. Relationships with _Mus validus_ Miller, from Trong, Lower Siam, and _Mus mülleri_ Jentink from Sumatra. Differs from the former in smaller size and in the absence of the anterior outer tubercle of the last upper molar, and from the latter in larger size, and yellowish brown (not white) underparts.

_Color._--Back and sides a fine grizzle of black and dull ochraceous (the exact shade intermediate between the ochraceous and ochraceous-buff of Ridgway), the two colors nearly equally mixed on back, but the ochraceous in excess on sides. Underparts and inner surface of legs buff. An ill defined drab-gray median line from throat to pubic region. Head darker and more glossy than back, the cheeks distinctly washed with gray. Lips and chin drab-gray. Feet an indefinite brown, darker on metapodials. Ears essentially naked, dark brown. Tail dark brown throughout. Underfur gray (Ridgway, pl. II, No. 8), becoming paler on under parts where it fades irregularly into the general buff.

_Fur._--The fur is exactly as in _Mus validus_, that is the grooved bristles are so slender that their true nature is not apparent without use of lens. On middle of back the mass of the fur is about 17 mm. in length, the long terete hairs scattered through it reaching about 30 mm. On rump the fur is longer but not conspicuously so, and there is no noticeable increase in length or abundance of the terete black hairs.

_Tail, feet and mammæ._--Tail slightly more coarsely scaled than in _Mus validus_; 9 rings to the centimeter at middle. Hairs scarcely noticeable except toward tip, where they somewhat exceed the breadth of the rings.

Feet heavy and robust. Thumb short, with a flat blunt nail. Soles and palms naked, the former with six well developed tubercles, the latter with five.

Mammæ, p. 2--2, i. 2--2 = 8.

_Skull._--In general appearance the skull of _Mus integer_ resembles that of _Mus validus_.[12] It is shorter (greatest length about 51 instead of 55) and the rostrum is relatively broader and deeper. Audital bullæ similar in form to those of _Mus validus_, but the surface less irregular. Region between anterior bases of zygomata broader than in _Mus validus_ so that the arches are more nearly parallel.

_Teeth._--The teeth are relatively as well as actually smaller than in _Mus validus_ and the enamel pattern is normal, that is, the posterior upper molar consists of two transverse folds, and an anterior internal tubercle. There is no trace of the supplementary outer tubercles of the corresponding tooth of _Mus validus_.

_Measurements._--External measurements of type: total length 463; head and body 235[13] tail vertebræ 228;[13] hind foot 48 (45); ear from meatus 19; ear from crown 15; width of ear 15. In adult male topotype: total length 462; head and body 234;[13] tail vertebræ 228;[13] hind foot 46 (44); ear from meatus 21; ear from crown 16; width of ear 16.

Cranial measurements of type: greatest length 52 (55);[14] basal length 45 (48.6); basilar length 41.6 (45.6); palatal length 23 (26); least width of palate between anterior molars 5 (5); diastema 14 (14.6);[15] length of incisive foramen 8 (9); combined breadth of incisive foramina 3 (3.6); length of nasals 21 (22.6); combined breadth of nasals 6 (6.2); zygomatic breadth 25 (28); interorbital breadth 8 (8); mastoid breadth 19 (19); breadth of braincase above roots of zygomata 18.8 (20); depth of braincase at anterior border of basi-occipital 12.8 (15); frontopalatal depth at posterior extremity of nasals 12.8 (13.4); least depth of rostrum immediately behind incisors 10 (10); maxillary toothrow (alveoli) 9.6 (11); width of front upper molar 3 (3); mandible 30 (31); mandibular toothrow (alveoli) 9 (10).

_Specimens examined._--Four, three from the type locality, and one from Pulo Lingung.

_Remarks._--This rat is probably a near relative of the Bornean _Mus mülleri_ of Thomas.[16] The specimen from Pulo Lingung does not differ appreciably from the others.

MUS SABANUS Thomas.

1887. _Mus sabanus_ THOMAS, Ann. and Mag. Nat. Hist., 5th ser., XX, p. 270. October, 1887 (Mt. Kina Balu, Borneo).

1894. _Mus sabanus_ THOMAS and HARTERT, Novitates Zoologicæ, I, p. 658. September, 1894 (Bunguran).

Thirteen skins and one extra skull, all from Bunguran. There is little probability that this rat is the same as the true _Mus sabanus_ of Borneo.

MUS RAJAH Thomas.

1894. _Mus hellwaldi_ THOMAS and HARTERT, Novitates Zoologicæ, I, p. 658. September, 1894 (Bunguran).

1894. _Mus rajah_ THOMAS, Ann. and Mag. Nat. Hist., 6th ser., XIV, p. 451. December, 1894 (Mount Batu Song, Borneo).

1895. _Mus rajah_ THOMAS, Novitates Zoologicæ, II, p. 26. February, 1895 (Revised determination of Bunguran specimens).

Six specimens (one in alcohol) from Bunguran, two from Pulo Lingung, one from Pulo Laut, four (one in alcohol) from Sirhassen, and one (in alcohol) from Pulo Midei. It is doubtful whether these series are referable to one species or whether any of them are the true Bornean _Mus rajah_. The material is not wholly satisfactory, and I have been unable to examine specimens from Borneo.

MUS NEGLECTUS Jentink.

1894. _Mus rattus_ var. THOMAS and HARTERT, Novitates Zoologicæ, I, p. 658. September, 1894 (Bunguran).

1895. _Mus neglectus_ THOMAS and HARTERT, Novitates Zoologicæ, II, p. 492. December, 1895 (Bunguran).

Five specimens from Pulo Lingung, one from Pulo Midei, and nine from Sirhassen. In the absence of Bornean material, I follow Thomas and Hartert in referring the Natuna rats of the '_alexandrinus_' type to _Mus neglectus_.

SCIUROPTERUS EVERETTI Thomas.

1894. _Sciuropterus phayrei_ THOMAS and HARTERT, Novitates Zoologicæ, I, p. 660. September, 1894 (Bunguran).

1895. _Sciuropterus everetti_ THOMAS, Novitates Zoologicæ, II, p. 27. February, 1895 (Revised determination of Bunguran specimens).

1895. _Sciuropterus everetti_ THOMAS and HARTERT, Novitates Zoologicæ, II, p. 490. December, 1895 (Bunguran).

Two specimens, both from Bunguran; an immature male taken July 4, and an adult female taken July 21, 1900.

PETAURISTA NITIDULA Thomas.

1894. _Pteromys nitidus_ THOMAS and HARTERT, Novitates Zoologicæ, I, p. 660. September, 1894 (Bunguran).