Periotic Morphology in the Trichosurin Possums

Copyright q American Museum of Natural History 2003 ISSN 0003-0082

P U B L I S H E D B Y T H E A M E R I C A N M U S E U M O F N AT U R A L H I S T O RY

CENTRAL PARK WEST AT 79TH STREET, NEW YORK, NY 10024

Number 3414, 14 pp., 5 figures, 4 tables October 29, 2003

Periotic Morphology in the Trichosurin PossumsStrigocuscus celebensis and Wyulda

squamicaudata (Diprotodontia, Phalangeridae) anda Revised Diagnosis of the Tribe Trichosurini

KIRSTEN CROSBY1 AND CHRISTOPHER A. NORRIS2

ABSTRACT

The morphology of the periotic in the extant trichosurin possums Strigocuscus celebensisand Wyulda squamicaudata is described and compared with that of various species of extantand extinct phalangerid marsupials. The periotic morphologies of S. celebensis and W. squam-icaudata show a number of similarities with those of brushtailed possums of the genus Tri-chosurus, strengthening the case for a monophyletic tribe Trichosurini consisting of these threegenera. Comparisons with periotics of possums from the Tertiary Riversleigh deposits ofQueensland that were previously assigned on the basis of craniodental anatomy to Trichosurusand Strigocuscus suggest that, despite some similarities, the three extant trichosurins may bemore closely related to each other than to the Riversleigh taxa. A revised morphologicaldiagnosis of the Trichosurini is provided, incorporating these data.

INTRODUCTION

The lesser Sulawesi cuscus Strigocuscuscelebensis and the scaly-tailed possum Wyul-da squamicaudata are small to medium-size

1 Graduate Student, School of Biological Sciences, University of New South Wales, Sydney, Australia, e-mail:[email protected].

2 Curatorial Associate, Division of Vertebrate Zoology (Mammalogy), American Museum of Natural History,e-mail: [email protected]. Present address: Division of Paleontology, American Museum of Natural History.

members of the marsupial family Phalanger-idae. Both species bear a superficial resem-blance to members of the most diverse andabundant group of phalangerids, the true cus-cuses of the genus Phalanger. In the case of

2 NO. 3414AMERICAN MUSEUM NOVITATES

S. celebensis this similarity led to inclusionof the species in Phalanger (e.g., Tate, 1945),a situation which prevailed as the acceptedtaxonomy until comparatively recently. Bycontrast, Wyulda has always been recognizedas a distinct taxon, with craniodental char-acters that indicate a close affinity to thebrushtailed possums of the genus Trichosu-rus.

The division of the Phalangeridae into twogroups, the monogeneric cuscuses and thetwo genera of possums, remained unchal-lenged until the phylogenetic revision ofFlannery et al. (1987), who proposed thatcuscuses were paraphyletic with respect tothe possums. The Phalangeridae were divid-ed into two subfamilies, the Ailuropinae,containing a single species Ailurops ursinus,and the Phalangerinae, containing all the re-maining species of phalangerid. The Phal-angerinae were divided into a further twogroups: the tribes Phalangerini and Tricho-surini. Flannery et al. (1987) resurrected thegeneric name Strigocuscus from synonymywithin Phalanger and placed the genus with-in the Trichosurini, along with Wyulda andTrichosurus. As originally defined by Flan-nery et al., Strigocuscus contained four spe-cies formerly included within Phalanger: S.celebensis, S. gymnotis, S. ornatus, and S.rothschildi. A fifth taxon, S. mimicus, wasraised from subspecific status within Phalan-ger orientalis. Subsequent studies have re-moved all species but S. celebensis fromwithin the group and added a new species,S. pelengensis (George, 1987; Springer et al.,1990; Norris, 1992; Flannery, 1994; Hamil-ton and Springer, 1999), previously a sub-species of S. celebensis (Tate, 1945). The ge-neric placement of this taxon is still underdebate.

A study by Norris (1994) described theperiotic bones of various phalangerid speciesand categorized three periotic morphologieswhich more or less corresponded to the di-visions originally proposed by Flannery et al.(1987), together with the narrower definitionof Strigocuscus made by later authors. Norrisonly studied one trichosurin periotic in detail,that of Trichosurus vulpecula. No loose peri-otic of either S. celebensis or W. squamicau-data was available for examination.

The fact that periotic morphology in phal-

angerids corresponded so closely to the phy-logenetic divisions created by analysis of abroad character set of morphological andmolecular characters suggested that a studyof this region in specimens of fossil phalan-gerids from Tertiary deposits at Riversleigh,Australia, might provide insights into the ear-ly evolution and diversification of the group.In superficial morphology, the Riversleighspecies appear to be trichosurin. However,they also possess a number of cranial char-acters that are plesiomorphic for phalanger-ids as a whole. Flannery et al. (1987) hintedat this in their list of synapomorphies for thetrichosurins, where they noted that the Riv-ersleigh species Trichosurus dicksoni exhib-ited the plesiomorphic state for some char-acters. A comparison of the periotic of thisspecies and of Strigocuscus reidi, also fromRiversleigh, showed a very different mor-phology than that of T. vulpecula, as dis-cussed by Norris (1994). In fact, examinationof these, and other, extinct taxa from River-sleigh suggest that a much broader range ofperiotic morphologies may be present withinthe group than is apparent from its extantmembers. Since other craniodental charactersof these fossil species suggest that the ma-jority of them have trichosurin affinities(K.C., in prep.), a broader knowledge of theperiotic anatomy of the living trichosurins isrequired. To this end, periotics of Strigocus-cus celebensis and Wyulda squamicaudatawere dissected from skulls housed in the De-partment of Mammalogy at the AmericanMuseum of Natural History (AMNH) andstudied in detail.

MATERIALS AND METHODS

The right periotic was dissected from theskull of AMNH 196502, an adult femalespecimen of Strigocuscus celebensis. Theskull of AMNH 196919, an adult femalespecimen of Wyulda squamicaudata, wasprepared from the alcohol-preserved carcassof this animal by dermestid beetle action, andthe left periotic was dissected. Detailed de-scriptions of the two bones were made, andthe morphology was compared with speci-mens of a variety of extant and extinct phal-angerid taxa (table 1).

For the most part, the terminology em-

2003 3CROSBY AND NORRIS: PHALANGERID PERIOTICS

TABLE 1Material Examined

ployed in this paper follows MacIntyre(1972), with modifications from Wible(1990) and Norris (1994). There are certainstructures on the periotic, however, whosehomology needed to be further evaluated be-fore they could give solid phylogenetic in-formation. These were the rostral tympanicprocess, the ectotympanic process, and thecrista promontorii medioventralis. Problemswith these structures include whether theyare homologous with those seen in othergroups, as well as their relative positions.

MacPhee (1981) defined the rostral tym-panic process (RTPP) as being a ‘‘tympanicprocess that arises (apparently always perios-teally) from all or part of the ventral surfaceof the pars cochlearis.’’ Wible (1990), in astudy of periotics of Cretaceous marsupials,described a crest that ‘‘arises from the pro-montorial surface in front of the fenestra co-chleae and runs towards the midline alongthe posterior edge of the medial expansion.’’He noted that this crest is in the same posi-tion as the RTPP in modern marsupials andthat the morphology of this crest in his Pe-trosal Type A (probably a species of Pedi-omys) resembles the RTPP of phalangerids.Norris (1994) described the RTPP in Tricho-surus vulpecula as the medial border of thepars cochlearis, a ‘‘sharp ridge running an-tero-ventrally to fuse with the margin of thetrigeminal fossa.’’ In direct contrast to this,however, Marshall et al. (1995: p. 75) con-

tended that phalangerids have a smooth pro-montorium with no RTPP.

The confusion arises because the RTTPidentified by Wible in Didelphis is only theapex of a much larger structure; this extendsboth rostrally and medially from the summitof the mound formed by the underlying co-chlea, which we term the cupula cochleae. Inphalangerids, the cupula intrudes into thisstructure, and the ventrally extended processseen in Didelphis is absent. As a result, thereappear to be two structures on the ventralsurface of the pars cochlearis: a crest runninganteriorly from the cupula cochleae onto thepromontorium (the RTTP of Norris, 1994),and a variably developed, ventrally orientedprocess. The latter structure was labeled byNorris (1994) as the ectotympanic process(ECTP). In fact, this process is merely thecaudo-lateral extension of the RTTP (fig. 1).Thus, in this paper, we have subsumed theECTP sensu Norris (1994) into the RTPP.

The final structure that is in question is thecrista promontorii medioventralis (CPM).MacIntyre (1972) defined this as a ridge run-ning anteriorly from the margin of the pos-terior lacerate foramen that forms the bound-ary between the cerebellar and tympanic fac-es. In MacIntyre’s figure this structure isshown as lying lateral to the sulcus of theinferior petrosal sinus (SIPS). In the perioticsof Wyulda and phalangerins, however, theventral edge of the cerebellar face can extend


Fig. 1. Tympanic view of a generalized phal-angerid petrosal, with a line showing the crest ofthe rostral tympanic process (RTTP). The cupulacochleae (CUC) intrudes into the RTTP, splittingit into two portions, which Norris (1994) identi-fied as the rostral and ectotympanic processes(‘‘RTTP’’ and ‘‘ECTP’’). In fact, both structuresare part of the RTTP.

→

Fig. 2. Periotic of AMNH 196502, Strigocuscus celebensis. A. Tympanic face. B. Cerebellar face.Abbreviations: AV, aqueductus vestibuli; CC, crus commune; CP, crista petrosa; CPM, crista promontoriimedioventralis; CTP, caudal tympanic process; ER, epitympanic recess; FC, fenestra cochleae; FI, fossaincudis; FNC, facial nerve canal; FNF, facial nerve foramen; FV, fenestra vestibuli; FVN, foramen ofthe vestibulocochlear nerve; IAM, internal auditory meatus; PLF, posterior lacerate foramen; PP, pustularprocess; PR, promontorium; RMTP, recessus mesotympanicus; SAF, subarcuate fossa; SFF, secondaryfacial foramen; SIPS, sulcus of the inferior petrosal sinus; SPS, sulcus of the prootic sinus. TF, trigeminalnerve fossa. Approximately 103 life size.

down medial to the SIPS, forming a ventrallydirected flange. In these cases, the lowestpoint on the pars cochlearis, and the apparentmargin between the cerebellar and tympanicfaces, tends to be the ventral edge of thisflange. Despite this, however, we continue toapply the term CPM to the crest that runslateral to the SIPS, as this is the structure thatmost closely resembles the original definitionof MacIntyre.

Terminology of the nerves and blood ves-sels surrounding the periotic follows Archer(1976). In this paper, the term ‘‘trichosurin’’refers only to the modern species. River-sleigh specimens are discussed separately, astheir exact phylogenetic position has not yetbeen determined (this includes the two de-

scribed Riversleigh species, Trichosurusdicksoni and Strigocuscus reidi).

DESCRIPTIONS OF PERIOTICMORPHOLOGY

Strigocuscus celebensis

The overall morphology of the periotic inS. celebensis is remarkably similar to that ofTrichosurus vulpecula (Norris, 1994), partic-ularly in its elongate promontorium, whichaccounts for almost half the total length ofthe pars cochlearis, and small internal audi-tory meatus (fig. 2). However there are anumber of distinctive features not sharedwith Trichosurus (table 2). The caudal mar-gin of the subarcuate fossa is greatly thick-ened, reducing the overall size of the open-ing. At the anterior margin of the fossa, anoverhang in the crista petrosa produces asmall, pocketed area. As in Trichosurus, thecrista petrosa is expanded to form part of theseptum separating the middle and posteriorcranial fossae and is notched anterodorsallyfor the passage of the prootic sinus. How-ever, unlike in Trichosurus, this expansiondoes not result in the development of a large,platelike lamella or lead to the crista petrosabecoming contiguous with the tegmen tym-pani. In Strigocuscus, the crista petrosa is acomparatively small, robust structure, whichis clearly separated from the tegmen tympa-ni. It is comparatively low at the point whereit meets the margin of the subarcuate fossa,and it is grooved along its lateral margin bythe sulcus of the prootic sinus. In Trichosu-rus this sulcus is deflected laterally by theexpansion of the crista petrosa, but in Stri-gocuscus it follows a more or less straightdorsal-ventral path.

Dorsally, the lamella that is often seenaround the posterodorsal margin of the peri-



TABLE 2Comparison of the Periotics of Trichosurins

otic in phalangerids (Norris, 1994) is absent,with the exception of a small flap just dorsalto the crus commune. On the left periotic ofAMNH 195602, which remained in situ, thisflap carried a small circular foramen, whichis assumed to be the opening of the prooticsinus. There is a faint ridge between the cruscommune and the prefacial commissure,which is sharply depressed. A small aque-ductus vestibule is present just ventral to thecrus commune. The internal auditory meatusis dominated by the large foramen of the ves-tibulocochlear nerve, with the small foramenof the facial nerve lying just dorsolateral toit.

From a cerebellar aspect, the promonto-rium is separated from the remainder of thepars cochlearis by a faint discontinuity,caused by the SIPS crossing the surface ofthe periotic just rostral to the internal audi-tory meatus. On the tympanic face, a small,pustular process arises from the ventral sur-face of the promontorium just caudal to thesulcus of the trigeminal nerve; this structurewas not seen in any other phalangerid taxastudied. As in Trichosurus, the promonto-rium itself is long and slender when viewedfrom both the tympanic and cerebellar as-pects. Unlike Trichosurus, however, it retainsthis slender appearance in its squamosal as-pect. This is the result of differences in themorphology of the tympanic surface of the

pars cochlearis in the two taxa. In Strigocus-cus, the apex of the cupula cochleae is verysmall and indistinct (unlike that of Tricho-surus, which is prominent and well defined),and the rostral tympanic process is small.The squamosal surface of the promontoriumis depressed by a narrow trigeminal fossa.The large opening of the hiatus Fallopii liesjust posterodorsal to the fossa; just anteriorto the fossa, the medial margin of the pro-montorium is depressed by the SIPS.

Ventrally, the pars cochlearis is borderedby a long CPM. This structure is not nearlyas large and ventrally extensive as that seenin Trichosurus, where the CPM completelyfills the gap between the basioccipital and thetympanic wing of the alisphenoid (Norris,1994). This is linked, in part, to the devel-opment of the ventral portion of the rostraltympanic process in the two taxa. In Tricho-surus this is a large, well-defined structure,which is greatly expanded ventrally. In Stri-gocuscus, by contrast, the process, thoughlarge, is only poorly developed; it is roundedand shows little or no ventral expansion. Thelambdoidal face of the pars cochlearis is rel-atively smooth and lacks the deep sulci seenin a number of extinct and living phalangeridtaxa. The acqueductus cochleae is small andhas only a very shallow sulcus leading to it.

Further differences between the perioticsof Strigocuscus and Trichosurus can be seen


Fig. 3. Periotic of AMNH 196920, Wyulda squamicaudata. A. Tympanic face. B. Cerebellar face.Abbreviations: DL, dorsal lamella; MCF, exposure of periotic in middle cranial fossa. Other abbrevia-tions as for figure 2. Approximately 103 life size.


→

Fig. 4. Periotics of living phalangerids. A, B, Trichosurus vulpecula; C, D, Phalanger (Spilocuscus)maculatus; E, F, Phalanger gymnotis. A, C, E, tympanic face; B, D, F, cerebellar face. Abbreviations:hF, hiatus Fallopii; TNS, sulcus of the trigeminal nerve. Other abbreviations as for figure 2. Approxi-mately 53 life size.

in the roof of the tympanic sinus. The fenes-tra cochleae and fenestra vestibulae are muchsmaller in Strigocuscus. The recessus meso-tympanicus is curved rather than straight andis bordered by a caudal tympanic process thatis only poorly defined. The epitympanic re-cess is shallow and seems to lack the prom-inent bordering septa seen in Trichosurus:this is reflected in the lack of continuity be-tween the crista petrosa and tegmen tympaniin Strigocuscus.

Wyulda squamicaudata

As in Trichosurus, the crista petrosa ofWyulda is expanded to form a flattened, an-teriorly directed, platelike lamella (fig. 3).The prefacial commissure is not especiallydepressed and instead forms more of a rightangle between the crista petrosa and the pro-montorium. The subarcuate fossa is large andlacks the thickening of the caudal marginseen in Strigocuscus, and the internal audi-tory meatus is relatively large. The moststriking feature, however, is the overall mor-phology of the pars cochlearis, which al-though almost as long as those of the othertwo trichosurin taxa, is much broader, espe-cially in its tympanic and squamosal aspects.The ventral edge of the cerebellar face ex-tends downward as a ventrally directed la-mella, bordering the inferior petrosal sinusmedially and giving the cerebellar surface ofthe periotic a flattened appearance. This la-mella ends rostrally at the point where theinferior petrosal sinus emerges onto, andcrosses, the promontorium. The depth of thelamella exaggerates the degree of disconti-nuity between the promontorium and the re-mainder of the pars cochlearis. The shallowsulcus of the inferior petrosal sinus is rela-tively close to the tip of the promontoriumand stops part way across the cerebellar face;this is in contrast to the situation seen in theRiversleigh specimens, where the sulcus isdeep and runs the entire width of the cere-

bellar surface of the promontorium (K.C., inprep.).

The breadth of the pars cochlearis is par-ticularly striking when the periotic of Wyuldais viewed from its tympanic aspect. The ros-tral tympanic process is robust, well defined,and ventrally oriented, as in Trichosurus.The apex of the cupula cochleae is better de-fined than in Strigocuscus, and the rostraltympanic process is larger. The fenestra co-chleae is more ventrally oriented than in ei-ther Trichosurus or Strigocuscus. The epi-tympanic recess is well defined, with a dis-tinct lamina connecting the crista petrosa tothe tegmen tympani. The recessus mesotym-panicus is curved, and the caudal tympanicprocess is small and poorly defined.

From the squamosal aspect, the muchgreater depth and bulk of the pars cochlearisin Wyulda is clearly evident, and it can beseen that this is a function of the size of therostral tympanic process. The discontinuitybetween the promontorium and the cerebellarface of the periotic is also very apparent, to-gether with the sulcus of the inferior petrosalsinus that is responsible for the discontinuity.As in Strigocuscus, the acqueductus cochleaeis small and indistinct. The platelike lamellaof the crista petrosa dominates the lambdoi-dal aspect of the specimen. Expansion of thisstructure leads to a marked lateral displace-ment of the sulcus of the prootic sinus, sim-ilar to that which is seen in Trichosurus(Norris, 1994).

DISCUSSION

In a number of respects, the periotics ofStrigocuscus celebensis and Wyulda squam-icaudata are very similar in morphology tothose seen in species of Trichosurus. Theseform a tight morphotype when compared tothe periotics of species of Ailurops, Phalan-ger, and Spilocuscus (fig. 4). The morpho-type is distinguished primarily by consider-able elongation of the promontorium (around



→

Fig. 5. Periotics of extinct phalangerids. A, B, Trichosurus dicksoni. C, D, Strigocuscus reidi. E,F, DSS 23338. A, C, E, tympanic face; B, D, F, cerebellar face. Abbreviations: SMN, stylomastoidnotch. Other abbreviations as for figure 2. Approximately 53 life size.

45% of the total length of the pars cochlearis,compared with 30% in Ailurops and thephalangerins). Further features, most notablythe extensive expansion of the crista petrosaand the crista promontorii medioventralis,support a closer relationship between Tricho-surus and Wyulda than between either ofthese taxa and Strigocuscus. All three showa more simply built promontorium than thatof the phalangerins. In the latter group, adeeply grooved sulcus of the inferior petrosalsinus separates the rostral tympanic processand crista promontorii medioventralis, whilethe sulcus of the facial nerve cuts deeply intothe pars cochlearis anteroventral to the hiatusFallopii. These two features greatly increasethe structural complexity of the promonto-rium in phalangerins and cause a marked dis-continuity between the promontorium andthe rest of the pars cochlearis from both cer-ebellar and tympanic aspects, which is absentin the trichosurins. Finally, while the antero-posterior proportions of both the subarcuatefossa and the internal auditory meatus arefairly consistent across all phalangerid taxa,the dorsoventral axis for the two openingsshows consistent differences when the tricho-surins are compared to the phalangerins. Thesubarcuate fossa accounts for around 70% ofthe dorsoventral axis of the periotic, com-pared with 55% in the phalangerins, and ap-pears much larger. By contrast, the equiva-lent axis for the internal auditory meatus intrichosurins accounts for only 50% of theheight of the pars cochlearis, compared to75% in phalangerins, and appears muchsmaller. These results appear to solidify therelationship between Trichosurus, Wyulda,and S. celebensis, showing that the latter spe-cies is definitely a trichosurin, rather than aphalangerin or the sister group of a tricho-surin 1 phalangerin clade.

One aim of this study was to see whatlight, if any, the periotics of the modern tri-chosurin genera could throw on the radiationof phalangerid marsupials in the Tertiary.Two phalangerid species were described by

Flannery and Archer (1987) from the Mio-cene of Riversleigh and assigned to the mod-ern genera Strigocuscus and Trichosurus. Apreliminary look at their associated perioticssuggests that these species are more closelyrelated to each other than to either of themodern genera. The distinctive morphologyof Miocene phalangerids will be looked at ina separate paper (K.C. in prep.), but a generalcomparison of the various specimens of peri-otic from Tertiary phalangerid species withthose of modern trichosurins follows.

Periotic morphology of Riversleigh speci-mens shows more similarities to trichosurinsthan to phalangerins. Like trichosurins, Riv-ersleigh specimens have an elongate promon-torium, small internal auditory meatus, and alarge epitympanic recess (fig. 5). However,there are also some striking differences. TheRiversleigh specimens lack the expandedcrista petrosa that is seen in the trichosurins,and although the crista promontorii medio-ventralis is expanded ventrally, the spacebounded by the edge of the crista, the rostraltympanic process, and the cupula cochleae isconcave, rather than smoothly sloping as inthe trichosurins. This morphology of the parscochlearis is also seen in a number of otherdiprotodontian groups, including pseudoch-eirids and potoroids and, although absentfrom the ailuropine petrosal (Norris, 1994),it may be plesiomorphic within phalangerids.All Riversleigh phalangerid taxa show a dis-tinctive discontinuity across the cerebellarface of the periotic formed by the sulcus ofthe inferior petrosal sinus. This is also seenin Wyulda. Unlike Wyulda, the Riversleighperiotics are all delicate rather than robust.The promontorium is also quite curved ros-trally when compared to the modern tricho-surins.

Other features of the periotic in the tricho-surins are harder to apportion to individualclades and seem likely to be autapomorphicfor the taxa concerned. Most notable here isthe shape of the pars cochlearis: slender andelongate in Strigocuscus, broad and stocky in



TABLE 3Synapomorphies of the Trichosurini (sensu Flannery et al., 1987)

Wyulda, or with a prominent, ventrally di-rected ‘‘blade’’ in Trichosurus.

Although a broader examination of cran-iodental morphology in the Trichosurini isoutside the scope of this particular study (butis the subject of ongoing work by K.C.), itis worth making a brief overview of the sit-uation in order to put the anatomy of theperiotic into a wider context. Flannery et al.(1987) listed a set of five synapomorphiesdefining the tribe Trichosurini and a furthereight defining the clade Trichosurus 1 Wyul-da (table 3). Subsequent investigation by anumber of authors (Springer et al., 1990;Norris, 1992, 1994; Hamilton and Springer,1999), drawing on both molecular and mor-phological evidence, has led to a number ofthese characters being discarded as cases of

homoplasy between the trichosurins and cer-tain members of the phalangerin radiation,but have also produced new characters, mostnotably in the otic region. A revised list ofdiagnostic characters for the Trichosurini isgiven in table 4. When combined with theresults of our study of periotic anatomy, wethink that this character set provides a suffi-cient basis for support of a monophyletic Tri-chosurini including S. celebensis.

One outstanding issue concerning the Tri-chosurini is the phylogenetic position ofPhalanger pelengensis, which is still uncer-tain. This species was originally considereda subspecies of Strigocuscus celebensis(Tate, 1945). Flannery et al.. (1987) suggest-ed that the species might be a plesiomorphicsister taxon to both Phalanger and Spilocus-


TABLE 4Revised Diagnostic Characters of the Tribe Trichosaurini

cus, but left the species as ‘‘Phalanger’’.George (1987) also kept P. pelengensis inPhalanger rather than following S. celeben-sis to Strigocuscus. Groves (1987) main-tained the close relationship with S. celeben-sis, but kept both in Phalanger. Flannery(1994, 1995) also maintained this relation-ship, but used the genus Strigocuscus. How-ever, preliminary study of skull morphologyby the authors suggests that, despite the out-ward similarity of these species, the skull ofPhalanger pelengensis is much more similarto phalangerins than to trichosurins (includ-ing S. celebensis): the periotic of P. pelen-gensis lacks the elongate promontorium andexpanded crista petrosa seen in the tricho-surin possums; there is no I3-C1 diastema;the ventral surface of the orbits shows onlya limited overhang of the cheek tooth row;and the molars are slightly crenulate. It ispossible that the taxonomic affinities of thisspecies may lie with the endemic cuscuses ofthe neighboring Moluccan archipelago, P. al-exandrae, P. ornatus, and P. rothschildi: this

hypothesis is still under study, and will beexplored in detail in a future paper (Crosbyand Norris, in prep.).

ACKNOWLEDGMENTS

Travel by K.C. to England and the UnitedStates was funded by a Sylvester-BradleyAward through the Palaeontological Associ-ation, an American Museum of Natural His-tory collections study grant, a travel grantfrom the University of New South WalesSchool of Biological Sciences, and by fund-ing from the Riversleigh Project. Thanks aredue to Matthew Wills and Malgosia Atkinsonat the Oxford University Museum of Zoolo-gy, Paula Jenkins and Daphne Hills at theNatural History Museum, London, and BobRandall of the American Museum of NaturalHistory for their help when K.C. was on herstudy trip. The skull of Wyulda was extractedby Darrin Lunde and prepared by Neil Dun-can, both of the AMNH Department ofMammalogy. Eric Stiner (AMNH Mammal-


ogy) provided invaluable technical supportfor the imaging of the various periotics stud-ied.

REFERENCES

Archer, M. 1976. The basicranial anatomy of mar-supicarnivores (Marsupialia), interrelationshipsof carnivorous marsupials, and affinities of theinsectivorous marsupial peramelids. ZoologicalJournal of the Linnean Society 59: 217–322.

Bensley, B.A. 1903. On the evolution of the Aus-tralian Marsupialia: with remarks on the rela-tionships of the other marsupials in general.Transactions of the Linnean Society of London9 (2nd ser.): 83–217.

Flannery, T.F. 1994. Possums of the world. Amonograph of the Phalangeroidea. Chatswood,Australia: GEO Productions.

Flannery, T.F. 1995. Mammals of the south-westPacific and Moluccan Islands. Ithaca, NY: Cor-nell University Press.

Flannery, T.F., and M. Archer 1997. Strigocuscusreidi and Trichosurus dicksoni, two new fossilphalangerids (Marsupialia: Phalangeridae) fromthe Miocene of northern Queensland. In M. Ar-cher (editor), Possums and opossums: studiesin evolution: 527–536. Sydney: Surrey Beatty& Sons and the Royal Zoological Society ofNew South Wales.

Flannery, T.F., M. Archer, and J. Maynes 1987.The phylogenetic relationships of living phal-angerids (Phalangeroidea: Marsupialia) with asuggested new taxonomy. In M. Archer (edi-tor), Possums and opossums: studies in evolu-tion: 477–506. Sydney: Surrey Beatty & Sonsand the Royal Zoological Society of New SouthWales.

George, G.G. 1987. Characterization of the livingspecies of cuscus (Marsupialia, Phalangeridae).In M. Archer (editor), Possums and opossums:studies in evolution: 507–526.Sydney: SurreyBeatty & Sons and the Royal Zoological So-ciety of New South Wales.

Groves, C. 1987. On the cuscuses (Marsupialia:

Phalangeridae) of the Phalanger orientalisgroup from Indonesian Territory. In M. Archer(editor), Possums and opossums: studies inevolution: 569–579.Sydney: Surrey Beatty &Sons and the Royal Zoological Society of NewSouth Wales.

Hamilton, A.T., and M. Springer 1999. DNA se-quence evidence for placement of the groundcuscus, Phalanger gymnotis, in the tribe Phal-angerini (Marsupialia: Phalangeridae). Journalof Mammalian Evolution, 6: 1–17.

MacIntyre, G. T 1972. The trisulcate petrosal pat-tern of mammals. In T. Dobzhansky, M.K.Hecht, and W.C. Steere (editors), Evolutionarybiology 6: 275–303. New York: Plenum Press.

MacPhee, R.D.E. 1981. Auditory regions of pri-mates and eutherian insectivores: morphology,ontogeny, and character analysis. Contributionsto Primatology 18: 1–282.

Marshall, L., C. de Muizon, and D. Sigogneau-Russell 1995. Pucadelphys andinus (Marsupi-alia, Mammalia) from the early Paleocene ofBolivia. Memoires du Museum Naturald’Histoire Naturelle 165: 1–164.

Norris, C.A. 1992. The phylogeny and biogeog-raphy of the marsupial family Phalangeridae.D. Phil. thesis, University of Oxford.

Norris, C.A. 1994. The periotic bones of possumsand cuscuses: cuscus polyphyly and the divi-sion of the marsupial family Phalangeridae.Zoological Journal of the Linnean Society 111:73–98.

Springer, M.S., J.A.W. Kirsch, K.P. Aplin, and T.F.Flannery 1990. DNA hybridization, cladistics,and the phylogeny of phalangerid marsupials.Journal of Molecular Evolution 30: 298–311.

Tate, G.H.H. 1945. Results of the Archbold Ex-peditions. No. 52. The marsupial genus Pha-langer. American Museum Novitates, 1283: 1–41.

Wible, J.R. 1990. Petrosals of Late Cretaceousmarsupials from North America, and a cladisticanalysis of the petrosal in therian mammals.Journal of Vertebrate Paleontology 10: 183–205.

Recent issues of the Novitates may be purchased from the Museum. Lists of back issues of theNovitates and Bulletin published during the last five years are available at World Wide Web sitehttp://library.amnh.org. Or address mail orders to: American Museum of Natural History Library,Central Park West at 79th St., New York, NY 10024. TEL: (212) 769-5545. FAX: (212) 769-5009. E-MAIL: [email protected]

a This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

Documents

Periotic Morphology in the Trichosurin Possums