Uromys neobritannicus Tate & Archbold, 1935

Uromys neobritannicus Tate & Archbold, 1935 View in CoL

Type material. HOLOTYPE, AMNH 99881 View Materials , skull only of an adult male collected on New Britain, Bismarck Archipelago by J.W. Eyerdam.

Revised diagnosis. Uromys (u.) neobritannicus differs from all other species of Uromys as follows: i) hypertrophy of the postorbital processes; ii) uniformly black tail (otherwise in subgenus only in U. anak and U. boeadii n.sp., from which it also differs by much larger postorbital processes, square posterior ends of nasals which do not extend as far posteriorly, reddish body colour with deep yellow venter, and more numerous tail scales [16 per cmD; iii) unlike U. anak , but like U. boeadii n.sp., there are no swollen vertical ridges below postorbital processes, despite the large size of the processes; iv) one of the largest species, equal in size to U. anak .

Discussion. Uromys neobritannicus is thus far known only from six individuals, all collected on New Britain. The holotype is a skull only, localised only as far as New Britain. The second specimen (AMNH 119467) is a male skin and skull that was collected at Talasea (on the Mundo Willaumez Peninsula) by Fr Schumm ( Tate, 1951). The third known specimen is held in the Museum of Victoria. It is a subadult female (no. 6890) collected on the Mundo Willaumez Peninsula on the north coast of New Britain by Clive Champion. Three additional specimens (one in spirits (AM M20690 View Materials ), two skins and skulls (AM M20689 View Materials , AM M21118)), from the Fullebom area East New Britain, are in the Australian Museum.

Recent work on New Ireland by one ofus (TFF) failed to find it there despite extensive fieldwork and analysis of extensive archaeological material ( Flannery & White, 1991), and· it seems unlikely that the species inhabits any of the smaller islands off New Britain.

Relationships. Uromys neobrittanicus is clearly the sister species to U. anak , from which it differs in a few skull characters and in colour; of these it is likely that the condition seen in the present species is the plesiomorphic one, with the exception of the striking postorbital processes (a gross enlargement of the small tubercles of U. anak ), and the more numerous tail scales Uromys boeadii n.sp. Figs 8 View Fig , 9 View Fig , Table 2 View Table 2

.

Type material. HOLOTYPE, AMNH 222242 View Materials , young adult male skin and skull, collected by Philip Temple on 22 March 1963, 25 km north-east of Biak, Pulau Biak, Geelvinck Bay , Irian Jaya, Indonesia.

Etymology.ForBp.Boeadi,Indonesia'smosteminent mammalian taxonomist, and one ofthat country's leading biologists.

Diagnosis. This new species is phenetically closest to the U. anaklU. neobrittanicus group, agreeing with the members of this group in its uniformly black tail, the deep rostrum, the tendency towards development of distinct postorbital processes, and the posteriorly extended nasal bones. It differs from them in the following ways: i) size much smaller; ii) postorbital processes poorly developed; iii) posterior ends of nasal bones only just reaching level of anterior orbital margins; iv) tail scales longer than broad, and arranged in clear rings. It differs additionally from U. anak as follows: i) lack of prominent inflation of frontotemporal sutures; ii) flat interorbital area. It differs from U. neobrittanicus as follows: i) less tail scales (7-9 per cm oflength); ii) deep blackbrown colour with sharply marked median pectoral streak.

Description. The holotype (and only known) skin and skull are in good condition. The overall colour is TODO TODO TODO TODO TODO TODO TODO TODO TODO foreparts and with a lighter sheen on the hindparts. The pelage consists largely of long, dark guard hairs, somewhat paler towards the base, with a few light-coloured guard hairs and crimped grey-white underhairs mixed in. The venter is sparsely furred, with the lighter skin tone showing through; the ventral hairs are mostly brown, but there is a sharply marked creamy-white streak, 92 mm long, in the midline of the chest, along the approximate extent of the sternum, beginning as a thin streak at about the level of the axillae, broadening suddenly 22 mm further back, and narrowing gradually behind ( Fig. 8 View Fig ). Hands and feet are lighter brown owing to the sparseness of the hair covering; vibrissae are long and black. The tail scales are somewhat longer than broad, flattened, and arranged in rings.

The skull is deep, especially the rostrum, as in U. anak and U. neobrittanicus , but unlike these the nasals are relatively short, and the postorbital processes are only slightly developed. The parietal crests are well developed, extending forward to form clear supraorbital ridges. The premaxilla extends back on either side nearly to the zygomatic plate. The zygomatic arches sweep downward to just above the level of the posterior molar alveoli, then curve up again to their posterior roots above the external auditory meati. The incisive foramina are slit-like. The palate ends squarely somewhat posterior to the third molars; the mesopterygoid fossa is broad, parallel-sided. The bullae are small ( Fig. 9 View Fig ).

Discussion. Consideration must always be given to the problem of what level of taxonomic recognition should be awarded to entirely allopatric taxa. In the present case, the evidence is overwhelming that, cladistically, U. boeadii is the plesiomorphic sister-taxon to the U. anaklneobrittanicus clade, or to all Uromys (Uromys) ; consequently, if these are to be maintained as separate species, U. boeadii must be given species rank as well.

While most of its features are evidently plesiomorphic, the restriction of the white zone to a pectoral streak and the complete absence of any white in the inguinal region would appear to be autapomorphic states of this new species, perhaps its only ones.