Pelage.—Aristiforms on middorsal region: Whitish basally and gradually blackening toward tip which is extended as long, thin filament; total length, 22 to 24 mm; maximum width, 0.5 mm. Setiforms on middorsal region: Gray basally, gradually blackening toward tip but interrupted by Ochraceous-Tawny, subapical zone 3.3 mm long; total length, 19 to 21 mm; maximum width, 0.06 mm. Setiforms on outer thighs: Whitish basally, gradually blackening toward tip but interrupted by Ochraceous-Tawny, subapical zone 3 mm long; total length 14 to 16 mm; maximum width, 0.04 mm.
Skull.—Large and strong; nasals pointed posteriorly; supraorbital ridges moderately developed and extended caudad across anterior third of parietals; zygomatic arches strong; postorbital process of zygoma involving mostly squamosal; incisive foramen elongate and narrow (5 to 6.5 × 2.3 mm) with margins more or less parallel and raised to form ridges which extend posteriorly to within 3 mm of plane of premolars; vomerine sheath complete, with maxillary part thin and extended caudad as medial crest; mesopterygoid fossa extending forward as far as posterior faces of second molars or slightly short thereof; bullae large and inflated.
Teeth.—Molariform teeth large, P4-M3 averaging more than 9 mm in length. Upper molariform teeth: P4 and M3 with three counterfolds; M1 and M2 with four counterfolds each. In lower teeth, p4 with four counterfolds and each molar with three counterfolds.
Comparisons.—Differences from P. g. steerei are given in the account of that subspecies.
Remarks.—Specimens from the type locality were not available. Specimens from Fazenda Paraiso, Faro, were relied upon as representative of the subspecies. These agree with the type according to [Thomas] (1912:89). However, the skin of the type was changed in color by preservative ([Thomas], 1905:587) and the best skin he saw was from Faro (1912:89).
[Thomas] (1920:277) applied the name goeldii also to specimens from Manacaparú, a place a short distance above Manaus on the Solimões (Amazon) River and from Acajutuba, near Manaus, on the Negro River. In referring to these specimens (2 from Manacaparú and 2 from Acajutuba) [Thomas] (loc. cit.) said "Five molar laminae are frequently, if not invariably, present among these specimens." He did not, however, mention whether or not the number of laminae was constant in both M2 and M3. One specimen from Acajutuba, in the collection of Museu Nacional (MN no. 1973 [M]), actually has five laminae in M3, but the specimens in the American Museum from Faro agree absolutely with [Thomas'] original description of goeldii.
[Osgood] (1944:199) doubted that goeldii was a valid species. Evidence that [Osgood's] doubt was unjustified is furnished by the fact that [Thomas] (1912:89) pointed out that his specimen from Faro agrees with the type. Likewise, my two specimens from Faro agree with the type insofar as it has been described. [Thomas] (1912:89) mentioned two additional skulls from the type locality which, he stated, agree with the type which was received from the Museu Goeldi, Pará.
Specimens examined.—Total number, 4, from Brazil as follows: Pará, Faro, Faro, Fazenda Paraiso, 2 (AMNH); Amazonas, Manaus, Manaus, 1 skull (AMNH); Amazonas, Manaus, Acajutuba, 1 (MN).