VOLUME 13 PART 5 MEMOIRS OF THE QUEENSLAND MUSEUM BRISBANE VOLUME 13 PART 5 MEMOIRS OF THE QUEENSLAND MUSEUM ISSUED BY THE AUTHORITY OF THE MINISTER FOR EDUCATION, THE HON. J. C. A. PIZZEY. 199 THE GENERA PROPLEOPUS AND HYPSIPRYMNODON AND THEIR POSITION IN THE MACROPODIDAE Jack T. Woops Queensland Museum Any assessment of the systematic position of the extinct Propleopus oscillans (De Vis) depends wholly on comparison with similar parts of the living Hypsiprymnodon moschatus Ramsay and related forms. Palaeontological evidence of its affinities is lacking. Propleopus Longman was included with Hypsiprymnodon Ramsay and Burramys Broom in the subfamily Hypsiprymnodontinae of the Macropodidae by Tate (1948). However Ride (1956) has shown conclusively that Burramys parvus Broom, known only from fossils of presumed Pleistocene age from Wombeyan Caves, New South Wales, is not a macropodid. He placed it in the subfamily Phalangerinae of the Phalangeridae. The type of P. oscillans is in the collections of the Queensland Museum. De Vis’ original description (1888) is erroneous in some features, and many characters are only compared with those in other genera, which themselves have never been adequately described or figured. A redescription and new figure of the type are provided below. It is surprising to find that, while H. moschatus has been much discussed in the literature since its original description by Ramsay (1876), no comprehensive account of its dentition has been provided. Indeed the interpretation of Carlsson (1915) of the deciduous teeth will be shown to be erroneous, and these errors have been perpetuated by later authors. Fortunately there is a good series of skulls, including those of juveniles, in the collections of the Queensland Museum, and a description of the dentition of H. moschatus is given. The evidence of the dentition and certain other characters in determining the systematic position of Hypsiprymnodon is also discussed. PROPLEOPUS OSCILLANS (De Vis) (Figure 1) Triclis oscillans De Vis, 1888, Proc. Linn. Soc. N.S.W., 3, p. 8, pl. 1. Propleopus oscillans (De Vis). Longman, 1924, Mem. Qd. Mus., 8, p. 20. MatTertaL.—Type, F. 3302, incomplete left ramus, with I,, P;—M,, King Creek, Darling Downs, 8. E. Queensland. F. 681, left I,, Darling Downs, Q.; F. 3287, cast of incomplete right ramus, with M,—M,, Wellington Caves, New South Wales (cast of Univ. Calif. Mus. Paleont. No. 45171). 200 MEMOIRS OF THE QUEENSLAND MUSEUM Redescription of Type Mandible robust ; ventral margin arcuate, with greatest depth below M,. Symphysis ligamentary ; elongate, extending to point opposite anterior edge of P,;; subalveolaris fossa shallow ; then prominent postsymphysial swelling (for root of I,). Diastema long; diastemal crest rounded behind alveolus of I,, becoming sharper and swinging posterolaterally toward nearly obliterated alveolus of P,. Mental foramen large, slightly anteroventral to P,. Ramus incomplete posteriorly, but masseteric canal confluent with inferior dental canal and extending to below anterior molars ; postalveolar shelf short, triangular, flanked by prominent postalveolar ridge. I, rather short, but deeply rooted ; slightly curved in lateral view, ascending to extensive subhorizontal surface of wear, gently rounded at tip ; rounded-cuneate in transverse section, much higher than wide; enamelled ventrolaterally, with enamel extending half-way Jaterally, less so mesially ; mesial facet of wear towards tip. I, missing, but open alveolus small, subhorizontal. P, larger than molars, suboval at base of crown; trenchant, with cutting edge oblique, making an angle of about 20° with molar axis; high, ascending anteriorly ; surface of wear labial, more conspicuous posteriorly. Lingual and labial faces each with seven nearly vertical, opposed. ridges; penultimate anterior pair rising to cuspule slightly above cutting edge; posterior pair short, weaker ; anteriorly with median ridge descending, swinging lingually, weakening. M,< M.< M, >M,; molar row straight, converging slightly anteriorly with axis of mandible, displaying pronounced anterolabial roll. Molars brachyodont, subrectangular, more lophoid anteriorly. M, broader posteriorly ; protoconid low, rounded, with short forelink Figure 1.—Propleopus oscillans (De Vis). Occlusal and lateral views of left ramus ; F.3302, type, natural size. THE GENERA PROPLEOPUS AND HYPSIPRYMNODON 201 descending slightly and swinging mesiad to anterior cingulum, and similar posterior projection contributing to low midlink ; metaconid high, somewhat trihedral, with labial spur descending to protoconid, forming short protolophid, and with curved lingual face extending from anterior cingulum, to tiny stylid in central valley ; anterior cingulum, descending labially, recurving beyond forelink ; hypoconid low, rather cresentic, separated from smaller protoconid by deep labial cleft, posteriorly with hindlink passing to short posterior cingulum ; entoconid rather similar to, but lower than metaconid, largely contributing to hypolophid. M, with anterior moiety subequal in width to, but longer than, talonid ; protoconid and hypoconid low, rounded ; metaconid higher than entoconid, both trihedral; hypolophid medianly low, divided. M, more distinctly quadrituberculate, wider anteriorly ; with few grooves nearly obliterated in floor of lingual part of central valley ; hypoconid subequal to entcconid. M, displaying greater reduction of talonid, with entoconid inferior to hypoconid, posterior cingulum short ; cusps more distinctly separated, especially those of talonid. Measurements (mm.) —— ¥.3302 (Type) F.3287 ¥F.681 I, in dorsal view 22-9 x 6-9 I, in section : o T 5 or 9-8 x 6-9 11-7 x 7:6 Length of on ‘0 43 Tip of I, to posterior of postsymphysial eciites _ 76 Diastema (I,-P;) oo bey : 5 ..| 39 Fy sh ma sa oa at pe xs 13-9 x 9-7 M, 9-5 x 8-7 9-3 x 9-2 M, wn . a si M3 > M!; molar rows parallel in juveniles, appreciably divergeat anteriorly later, markedly anisognathous with lower molars, very feebly convex labially ; displaying some anterolingual roll ; lingual cusps wearing at greater rate. Molars brachyodont, quadrituberculate ; trirooted, with single posterior root. M1 with anterior and posterior moieties equal; paracone high, subconical, with smooth curved labial surface between forelink, descending to small parastyle, posterior link descending to meet anterior link of metacone and delimit central valley ; protocone lower than paracone, cresentic, anteriorly passing to short cingulum, posteriorly separated from hypocone by deep lingual cleft, closed at base of crown; metacone rather similar to paracone, but with summit of cusp more distinctly linguad to and transversely linked with crest of smooth labial surface, hindlink descending to wider posterior cingulum ; hypocone slightly higher than protocone, but lower than metacone, cresentic, posteriorly passing to cingulum ; occlusal basins finely ridged ; base of crown expanded. M2? of similar form to M', but anterior cingulum wider, parastyle weaker ; anterior moiety appreciably wider than posterior at base of crown ; metacone equal to hypocone ; incipient loph formation of descending spurs from opposite cusps. M? of lower crown height and greater posterior contraction ; metacone slightly reduced; weak loph development, more from labial cusps. M*‘ erupted after greater hiatus than those between appearance of anterior molars, reduced, especially posteriorly, subtriangular in outline ; metacone vestigial ; hypocone small; small parastyle present. T, very long, ascending at low angle, approximated ventrally towards tips, with development of mesial facet of wear ; root compressed oval in section ; crown subquadrantal at base, tapering and blade-like distally, slightly upcurved and incurved towards tip, with three longitudinal flanges—dorsolabial, ventromesial, and dorsomesial towards base; surface of wear with upper incisors subhorizonal, near planar (by attrition with I?—*), with slight rounding of tips (against I’). I, minute, tubular, anteriorly directed, adpressed to ramus and base of I,, not functional and lost at varying stages. 204 MEMOIRS OF THE QUEENSLAND MUSEUM THE GENERA PROPLEOPUS AND HYPSIPRYMNODON 205 P, emplaced after DP, ; generally similar to P?, but without rudimentry lingual cingulum. DP, submolariform ; compressed, especially anteriorly ; metaconid high, in continuation with secant edge of P,, with forelink descending to prominent parastylid, sharp hindlink. carrying metastylid ; protoconid absent, but an inconspicuous labial link descends steeply from metaconid to join cresentic hypoconid ; entoconid compressed, anteriorly separated by deep notch from metastylid, posteriorly joined by curved cingulum to lower hypoconid. P, generally similar to P®, but without lingual cingulum ; erupted with secant edge diverging from molar axis by about 25°, increasing later to about 30°. M,< M, >M, > M,; molar rows slightly convergent anteriorly in juveniles, parallel in adults, displaying some anterolabial roll. Molars descending in height posteriorly ; birooted ; crowns with enamel more extensive labially, but cusps there subject to greater wear. M, contracted anteriorly ; metaconid compressed, oblique, almost collinear with, but lower than, secant edge of P,, short forelink descending steeply via tiny parastylid to high narrow anterior cingulum, labial spur descending to join low arcuate protoconid ; entoconid trihedral, with smooth steep lingual surface, labial spur descending to meet corresponding shorter process of hypoconid in incipient lophid development, posteriorly passing to low curved cingulum; hypoconid lower, cresentic, separated from protoconid by prominent labial cleft, closed at base of crown; occlusal basin initially finely ridged. M, with anterior moiety much longer and slightly wider than posterior ; metaconid trihedral, slightly in advance of lower cresentic protoconid, joined by wide curved anterior cingulum; incipient lophid development. M, smaller, but generally similar to M, ; hypoconid equal to entoconid. M, erupted after longer hiatus than those between appearance of anterior molars ; posteriorly reduced. Figures 2. 4.—Hypsiprymnodon moschatus Ramsay. 2a, occlusal view of cranium, showing left P?-M*; 2b, occlusal view of mandible, showing right P,-M,; 2c, lateral view of right ramus; J.6824, x 2. APF, anterior palatine foramen PE, petrosal AS, alisphenoid BO, basioccipital BS, basisphenoid PL, palatine PEF, postero-external palatine foramen PGF, postglenoid foramen CF, condylar foramen ECF, entocarotid foramen EO, exoccipital FO, foramen ovale JF, jugular foramen JU, jugal MA, maxilla MS, mastoid NA, nasal OC, occipital condyle PM, premaxilla POP, paroccipital process PPV, posterior palatal vacuity PT, pterygoid SO, supraoccipital SPF, sphenopalatine foramen SQ, squamosal TY, tympanic VO, vomer 3a, occlusal view of right DP* ; 3b, occlusal view of right DP, ; J.10231, x 2. 4a, occlusal view of right M4; 4b, occlusal view of right M,; J.10233, x 2. MEMOIRS OF THE QUEENSLAND MUSEUM 206 O-EX SE TEX GG OEX GZ TEx ae TEX SG SEX OG TEX 9G TEX G6 STIX SG SLX ERS O-GX SSG LEx oe’ GT XS-S GTX SE SLX o~ GTX SG XG-G XG GEX GE GEXEG GEXO-G SSX LZ 6-6 X 9-6 GG X16 FEXLG FEXLG TEX 9-6 GEXDG SEX LG GEX LE Lex 26 GEXBE GEXLSG SEX LE SEX LG GEXSG FEXS-G FEXSS FES OE FEXO-E GEXO-E SEX OE CEXO-€ GEXO-E GX GG SEX OG FGEX TE Pex Le PEXOS FGX O06 FEX 0-8 FGXO-E SGX TE SGX LE 66 X08 £6 X08 G6 X OE SEX LE FEXSG FEXOE SEX OG 66X08 PExo~a EEX64E EEX 6G GEX6E SEX OS GEXOS GEXOG SEX O-' TSG X OE LEX O06 BERLE VEX LE SEXO" SEX GE DEX 8-E LG@X Ss LEX Ls EX LE GEXOE G-3EX9-§ LEXSE BEX Ss VEX LE LEX LSE DEX LE SEX B8 LEX LE LEX L8 GEX Ls PEXSBE ST XES LT X86 ATX ESE LTX SE 9-1 X6-Z LIX GZ 6TX6T OLX GL BIX ST SIX S81 OT X6-T OTST GTX TLS | &SEXOS OGX EES LEX LS LéEx Te GexX GGx TEx s-~ GLS GE GTX G-E TEX GG LEX FG 0-6 X 0-6 OGX OG LEX G6 LEX T-@ ral FEXHG FEXLG FEXBS FEX LZ FSX LZ FEXO-G FEX LG SEXSG GEX OG PEXLs FEX OG PEXLs PEXLS PUX LZ F-ZX OG SEX Se FEXSG SEX Le GEXLS 96 X BSG LEXBS VG X OG LEXGG LEX6-G LEXOG VEX GG LEXOE G-EXS-6 PEXSS LEX6-% L6XO6-G LEX 86 VEXSG 9SXS-G DEX SG DEXOSE VEX GCG FEXGG VEX EG FBX 0-8 FEXO-€ X6-Z X63 (TIM) syUsToIMsvayy LEXSE LG XBS 9-6 X O08 EX O06 LEX OE L@XO-8 SEXO-E 6X OE GEXLE’ SEX LG LEXOG SEX O86 LEXO8 LEX O86 PEXS-G LEXS6G EX OE 9G X OS GE XO-8 LEX OS 9X 6G DEX 6G FEX OS F-EXG-S GEXBS SGX SSG TA GEXG6-' BEX SE 6-6 X OF 0-6 X OF 0-8 X O-F 0-6 X O-F 6-2 X O-F SEX 6s SEX BE O-EX LE OE XS-E BEX LEX LE OE XL BEXS-E BEX LS 8-3 X6-E LEXS8-S BSEXB-8 LOXS-6 ec 06X04 I I TéEX9G | LEX OG 1 TéEx9-e | TEx Oe! OZXOE | LEeXOe I OGX9G |OGXELTA OZXLE | 0GX LET GEXLE | OEXOGI ed zd 99 89 co 19 6¢ 69 SP GF 98 (worur -WOTY URI) qysue'T LESCOL es6g £ 9ec0T LC cecore Geko £ FEZOTL Secor le crre 9689' Faso e eseOl L Tecor'e Ose0Tl ‘ON “SOI THE GENERA PROPLEOPUS AND HYPSIPRYMNODON 207 Notes on the Skull Cranium, small, with maximum length of about 68mm.; elongate, with markedly tapering rostrum. Maxillae dorsally in considerable contact with frontals; laterally with infraorbital foramen, often compound, with main (anterior) opening above anterior margin of P? ; inferior process of anterior zygoma root weak ; palate posteriorly with large vacuities, separated by narrow median septum. Palatines ventrally reduced to slender bar behind vacuities, forming two-thirds of median septum, and narrowly flanking posterior third of vacuities; dorsally in extensive contact in midline in roof of narial passage, obscuring the presphenoid. Jugals laterally excavated, for superficial layer of masseter ; zygomatic arches rather flattened, converging anteriorly, Roof of braincase gently arched ; no sagittal crest, but old individuals have weak temporal ridges posteriorly confluent on well-developed interparietal. Squamosal anteriorly in narrow contact with frontal ; subsquamosal foramen immediately above external auditory meatus; and prominent foramen just within meatus opening anteriorly to sinus in root of arch ; no postglenoid process of squamosal, Tympanic tubular, but incomplete dorsally, and not smoothly uniting with squamosal in roof of meatus. Alisphenoidal bullae weakly inflated, in considerable contact with basioccipital, almost obscuring petrosal in ventral view; alisphenoid anteroventrally with small subvertical process extending to palatine and bounding pterygoid fossa, Paroccipital processes feebly developed. Supraoccipital inflated medially for vermis. Mandible with ventral margin arcuate, greatest depth below M,. Symphysis ligamentary, elongate, extending to point opposite anterior of P,; postsymphysial swelling prominent. Diastemal crest conspicuous in lateral view. Masseteric canal deeply invading body of ramus, confluent with inferior dental canal; anterior margin of coronoid process thickened, ascending at about 50°, Angle broadly inflected, produced somewhat posteriorly towards margin. Condyle relatively low ; articulating surface transverse, slightly convex longitudinally, THE SYSTEMATIC POSITION OF PROPLEOPUS AND HYPSIPRYMNODON In the original description De Vis (1888) compared the type of T'riclis oscillans ; with mandibles of species of Hypsiprymnodon, Hypsiprymnus, and Phalangista. He concluded that its affinities with H. moschatus were paramount, but suggested that “the Pleopodidae, embracing thus an ultimate stage of differentiation in Hypsiprymnodon and a more generalized type in T'riclis, are a continuation of a stock whence had arisen the Phalangistidae on the one hand, the Hypsiprymnidae on the other’. However, the pattern of tooth replacement and the development of the masseteric canal definitely establish the position of Propleopus and Hypsiprymnodon within the Macropodidae, in the current interpretation of this family ; and there is no close relationship with any phalangerid genus. The structure of the mandible of P. oscillans is closest to those of H, moschatus and species of Bettongia, In the relative shortness and heaviness of I,, the resemblance is to species of Bettongia. In the presence of a small, functionless I,, the length of the diastema, and the position of the mental foramen, the resemblance is clearly to H. moschatus. In the number of grooves of Ps, its shape and size relative to the molars, the condition is closer to H. moschatus, but the degree of anterolabial deflection of its secant edge is intermediate between that of H. moschatus and B. penicillata. The 208 MEMOIRS OF THE QUEENSLAND MUSEUM molar rows are straight as in H. moschatus, but slightly convergent anteriorly as in species of Bettongia. Molar gradation and cusp development are more comparable with the condition in B. penicillata and B. gaimardi, except that in P. oscillans the lophids are not so well developed and M, is less reduced posteriorly ; there is notable similarity in the production of the anterior cingulum labially beyond the forelink in lower molars and in the degree of anterior development of M,. The great extent of the subhorizontal surface of wear of I, in P. oscillans is a feature which cannot be matched in any related genus. It indicates comparable elongation of the crowns of I? and FI. While Propleopus displays an intermediate condition in the development of several features of the mandible between that of Hypsiprymnodon and Bettongia, if size be disregarded, there is no uniform gradation and the separation of all three as distinct genera must be maintained. Individuals of P. oscillans far exceeded in size any of their living relatives ; judging from the mandibular remains they must have been comparable in bulk to the living red and grey kangaroos. The species affords vet another instance.of the climax of gigantism of terrestrial vertebrates on the Austratian continent in the Pleistocene. One consequence of the structural relationship of Propleopus with both Hypsiprymnodon and Bettongia is that the validity of the division of the rat-kangaroos into the subfamilies Hypsiprymnodontinae and Potoroinae becomes more questionable than ever. Tate (1948) maintained these as distinct subfamilies in his review of the classification of the Macropodidae, and associated Potoroops, Potorous, Caloprymnus, Bettongia, and Aepyprymnus in the Potoroinae. Previously Simpson (1945) had included both Hypsiprymnodon and Propleopus in the Potoroinae, without discussing the systematics of the group. On the other hand, Pearson (1946) had proposed the separation of the rat-kangaroos, inclucing Hypsiprymnodon, as a distinct family, the Potoroidae, from the Macropodidae. He recognized the basic similarity in the structure of the female urogenital system in genera of his Potoroidae, and regarded its development in this group as more specialized than in the Macropodidae, Later Pearson (1950) maintained the Hypsiprymnodontinae and the Potoroinae as sub- families of the Potoroidae. . Pearson’s separation of the kangaroos into two families is based essentially on characters not utilized in the discrimination of other families of phalangeroid marsupials. It is contended that the presence of the basic body form including foot-structure ; gait; skull structure, including the development of the masseteric canal; dental formula, and characteristic pattern of tooth replacement necessitates their grouping in the one family—the Macropodidae. The findings of Pearson however strongly support the division of the Macropodidae into two subfamilies, as in the classification of Simpson (1945). THE GENERA PROPLEOPUS AND HYPSIPRYMNODON 209 From the description of the dentiton it can be seen that the tooth change in H.moschatus conforms to the general pattern of the Macropodidae in that P: resemble Pp’ in Shape and function, are lost penecontemporaneously with the displacement of DP; by P’ at a stage approaching adult body size, DP. being more comparable in form with the true molars than the true premolars. The teeth mentioned are sometimes designated F, Fr and DP* respectively by other authors. The views of Tate (1948) and Ride (1956) that H. moschatus is atypical in this regard are based. on the work of Carlsson (1915). As was originally suggested to me in conversation by Mr. George Mack, Director of the Queensland Museum, Carlsson misinterpreted the anterior cheek teeth in the juvenile skull she examined. ‘ Die beiden linken ausfallende Pramolar ” figured (1915, pl. 2, fig. 12) are in reality DP® and M! and the corresponding lower teeth (fig. 14) are DP, and M,. Apparently P* had not yet erupted in the very young specimen studied (with skull length 24 mm.). The brief coexistence of P? and P* in the tooth row of H. moschatus (fig. 2a) is due to the fact that while the crown of P* is much longer than that of DP®, it exhibits considerable axial deflection from the line of the cheek teeth on eruption. E are usually longer than DP* in the Macropodidae and mechanically displace a and DP’ if the former have not been lost already. Ride (1956) has already discussed in detail the resemblance between a (Pi of Ride) of H. moschatus and the phalangerid Burramys parvus. It is agreed that the occurrence of this type of sectorial in diverse mammalian groups renders specious any argument that its presence in these two marsupials is due to inheritance from a common ancestor rather than convergence. It is commonly recognized that the presence of quadrituberculate crowns, which show, however, varying degrees of weak lophoid development, readily distin- guishes the molars of the rat-kangaroos from the bilophodont structures of the Macropodinae. It is reasonable to assume that the Macropodidae were derived from omnivorous phalangerid ancestors with quadrituberculate molars. In the Macropodinae radiation has been towards browsing and grazing forms with bilophodont molars. In the rat-kangaroos, the trend, as reflected in the dentition, is incomplete, least of all in Hypsiprymnodon., The molar gradient, however, shows that Hypsip- rymnodon is not an unmodified stem macropodid, but should be associated with the Potoroinae. It may be argued that the presence of I, in Propleopus and Hypsiprymnodon contributes a reason for the separation of these genera from the Potoroinae. In both cases the tooth shows only vestigial development, and in Hypsiprymnodon, at least, it may be interpreted as a relict structure in an animal whose diet and dentition show but little specialization beyond those of its ancestors. Such a structure is 210 MEMOIRS OF THE QUEENSLAND MUSEUM useful in tracing phylogeny, but on the other hand its presence or absence need not be critical in any scheme of classification purporting to be consistent with that phylogeny. Pearson (1950) has pointed out that the separation of the frontal and squamosal by the junction of the parietal seems to represent the primitive condition in the marsupial cranium. In addition he showed that this relation is maintained in the Macropodinae, but not in the rat-kangaroos, where Hypsiprymnodon shows divergence in the narrow contact of the frontal and squamosal, and the others greater deviation in the much wider junction of these bones. As Abbie (1939) has demonstrated, the development of the masseteric canal in H. moschatus, in its confluence with the inferior dental canal, is comparable with the condition in the other rat-kangaroos. In the Macropodinae the masseteric canal, though extensive, merely opens into the inferior dental canal through a variously enlarged masseteric foramen, Comparison with the various developments of the posterior part of the mandible of phalangerids suggests that the structure of the masseteric canal in the Macropinae is likely to be nearer the condition in the ancestral macropodids than that of the rat-kangaroos. The lack of pronounced reduction of the fore-limb, the lack of pronounced elongation of the pes, especially the fourth digit, and the presence of the reduced hallux are three external characters which readily distinguish H. moschatus in appearance from its relatives, In determinining their taxonomic significance, these structures can be considered collectively, in that they all relate to locomotion and less directly to food-gathering activity. H. moschatus is probably the only ground- dwelling phalangeroid marsupial with a strict rain-forest habitat. It is feasible that its early phalangerid ancestors were arboreal animals with a generally similar diet in a similar habitat, Other than limited divergence in the structure of the pes (and hind-limb as a whole), towards digitigrade modificaton, would have no adaptive significance in this habitat. It would appear that the trend towards specialized saltatory locomotion progressed outside this environment in forest and grassland areas, with concomitant specialization for herbivorous diets. Further, as the trend in limb development has apparently operated in both macropodid subfamilies, the retention of the primitive unspecialized condition in H. moschatus neither confirms nor denies its association with either group, The naked tail in H. moschatus does not appear to be prehensile. Its adaptive significance may be that it is less likely to encourage parasites from the rain-forest floor. Pearson (1946, 1950) has discussed the female urogenital system of the rat- kangaroos in detail. He demonstrated that they exhibit a marked specialization in several characters beyond the generally primitive condition in the Macropodinae, THE GENERA PROPLEOPUS AND HY PSIPRYMNODON 211 and furthermore that the structure in H. moschatus conforms to the pattern of the group, with the exception that the fused median vaginal cul-de-sac is short and not produced posteriorly to reach the posterior vaginal sinus in parous adults. He used this as evidence in support of his view that Hypsiprymnodon is the most primitive genus of his Potoroidae. CONCLUSION The accumulated evidence indicates that in most of its characters which can be termed specialized, Hypsiprymnodon is comparable with the rat-kangaroos. Their development exceeds that of relative structures in the Macropodinae and shows that Hypsiprymnodon should not be considered as the surviving ancestral form for all Macropodidae, as maintained by Raven and Gregory (1946). It is not disputed that in characters relating to diet and locomotion Hypsip- rymnodon need show little specialization beyond the condition of hypothetical ancestral forms. Furthermore these generalized characters allow Hypsiprymnodon to be well adapted to its way of life in the rain-forest habitat. It is probable that this environment has been more stable over a longer period of time than other habitats, with less selection pressure than in the open forests and grasslands. The classification of Hypsiprymnodon as a less specialized member of the Potoroinae is the more satisfactory one in terms of an analysis of the gross morphology and what can be inferred of the phylogeny. The genus may not have evolved far from the ancestral of stock of the group, but the derivation of any particular genus from Hypsiprymnodon, sach as Bettongia in the scheme of Bensley (1903), is beyond the available evidence. Propleopus should also be included in the Potoroinae. The available characters of P. oscillans, waich by comparison with those of the living forms are few in number, indicate that it stands closer to Hypsiprymnodon than any living genus does, but no direct line of relationship can be inferred. The classification of the Macropodidae reverts to that of Simpson (1945). Palorchestes, separated in the subfamily Palorchestinae by Tate (1948), has been shown by Woods (1958) to belong to the Diprotodontidae. Sthenurus, with which Palorchestes was once associated in subfamily Sthenurinae by Raven and Gregory (1946), appears to be merely a macropodine form, specialized for browsing on coarse herbage. Both Sthenurus and the closely related genus, Procoptodon, need revision. B 212 MEMOIRS OF THE QUEENSLAND MUSEUM LITERATURE CITED Abbie, A. A., 1939. A Masticatory Adaptation peculiar to some Diprotodont Marsupials. Proc. zool, Soc. Lond., ser. B, 109, pp. 261-279. Bensley, A., 1903. On the Evolution of the Australian Marsupialia; with Remarks on the Relationships of the Marsupials in General. Trans. Linn. Soc. Lond. (Zool.), ser. 2, 9, pp. 83-217, pls. 5-7. Carlsson, A., 1915. Zur Morphologie des Hypsiprymnodon moschatus. K. svenska Vetensk. Akad. Handl., 52, No. 6, pp. 1-18, pls. 1-3. De Vis, C, W., 1888. On an Extinet Genus of the Marsupials Allied to Hypsiprymnodon. Proc. Linn. Soc. N.S.W., 3, pp. 5-8, pl. 1. Etheridge, R., in Jack, R. L., and Etheridge, R., 1892. The Geology and Palaeontology of Queensland and New Guinea. Brisbane, Government Printer, and London, Dulau & Co. XX XII+767 pp., 46 pl. Longman, H. A., 1924. Some Queensland Fossil Vertebrates. Mem. Qd. Mus., 8, pp. 16-28. Pearson, J., 1946. The Affinities of the Rat-kangaroos (Marsupialia) as revealed by a Comparative study of the Female Urogenital System. Pap. roy. Soc. Tasm. (1945) pp. 13-25. ——.-- 1950. The Relationships of the Potoroidae to the Macropodidae (Marsupialia). Pap. roy. Soc. Tasm. (1949) pp. 211-229. Ramsay, E. P., 1876. Deseription of a new genus and Species of Rat Kangaroo, allied to the genus Hypsiprymnus, proposed to be called Hypsiprymnodon moschatus. Proc. Linn. Soc. N.S.W., 1, pp. 33-35. Raven, H. C., and Gregory, W. K., 1946. Adaptive Branching of the Kangaroo Family in Relation to.Habitat. Amer. Mus. Nov., No. 1309, pp. 1-33. Ride, W. D. L., 1956. The Affinities of Burramys parvus Broom a Fossil Phalangeroid Marsupial. Proc. zool. Soc. Lond., 127, pp. 413-429, pls. 1-2. Simpson, G. G., 1945. The Principles of Classification and a Classification of Mammals. Bull. Amer. Mus. nat. Hist., 85, pp. 1-350. Tate, G. H. H., 1948. Results of the Archbold Expeditions, No. 59. Studies on the Anatomy and Phylogeny of the Macropodidae (Marsupialia). Bull. Amer. Mus. nat. Hist., 91, pp. 239-351. © 1952. Results of the Archbold Expeditions, No. 66. Mammals of Cape York Peninsula with Notes on the Occurrence of Rain Forest in Queensland. Bull. Amer. Mus. nat. Hist., 98, pp. 565-616. Tedford, R. H., 1955. Report on the Extinct Mammalian Remains at Lake Menindee, New South Wales. Rec. 8. Aust. Mus., 11, pp. 299-305. _ Woods, J. T., 1958. The Extinct Marsupial Genus Palorchestes Owen. Mem. Qd. Mus., 13, pp. | 177-193. S. G. Retp, Government Printer, Brisbane CONTENTS * VOLUME 13 PART 5 (Issued 5th September, 1960) PAGE The Genera Propleopus and Hypsiprymnodon and Their Position in the Macropodidae .. Jack T. Woods .. _ _ _ .. 199