PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, NY 10024 Number 3307, 16 pp., 4 ®gures, 1 table 13 December, 2000

Morphology of the Auditory Region in Paramys copei and Other from North America

JOHN H. WAHLERT1

ABSTRACT The morphology of the external surface of the petrosal and its junction with the basioccipital in Paramys copei is described. Comparison with an outgroup of the Vincelestes and Recent Didelphis and Solenodon reveals that the auditory region retains many primitive features. The Eocene paramyines from North America show only slight differences. Sciuravus is set apart by the lack of a ventral petrosal sinus canal between the petrosal and basioccipital and by the facial nerve and stapedial artery sharing a common foramen in the petrosal. Eu- ropean theridomyids, too, are not as primitive as Paramys but share with it and Sciuravus a ridge on the promontorium that separates the transpromontorial continuation of the internal carotid artery from the origin of the tensor tympani muscle. Twelve characters of the auditory region are analyzed and summarized in a data matrix for use in future studies. Relative prim- itiveness of the auditory region in all of these Eocene rodents suggests that derived characters in later taxa may represent intraordinal relationships but only homoplasy with nonro- dents.

INTRODUCTION in publications on cranial anatomy. Thus, a Rodent auditory regions are suf®ciently wealth of additional characters exists that can known for some characters to be used in phy- be described and evaluated for polarity (prim- logenetic analyses. However, many visible itive or derived). Paramys (family Ischyro- structural details are not routinely described myidae) is a key taxon in this category. It is

College, City University of New York; member of doctoral faculty, CUNY Ph.D. program in biology: ecology, evolutionary biology, and behavior subprogram. 2 AMERICAN MUSEUM NOVITATES NO. 3307 one of the oldest fossil rodents in which suf- 1962, ®gs. 13 and 14; Wahlert, 1974, ®g. 2. ®cient detail has been preserved, and, in Similarly, independent versions of the audi- North American paleontology, the genus has tory region have been published: Wood, been placed near the base of the rodent family 1962, ®g. 14a; Wahlert, 1974, ®g. 4; Parent, tree as a kind of structural ancestor or ex- 1980, pl. 2, ®gs. 2, 4; Lavocat and Parent, ample of primitiveness (Gregory, 1951: ®g. 1985, ®g. 1c. However, I found that there 19.40; Wood, 1962: ®g. 90). One expects it was still matrix obscuring morphology in the to be almost wholly primitive for rodents. auditory region and even a matrix-covered Other contenders for structural primitiveness petrosal displaced posterolaterally on AMNH among rodents are the North American genus 4755. To sit at a microscope and carefully Sciuravus, which McKenna and Bell (1997: use the sharpened tip of a thin beading nee- 186) put in a new and different suborder from dle to push away the ®ne covering matrix the Ischyromyidae, and the European theri- gives one an intimate feel for the shapes and domyids (Lavocat and Parent, 1985: 338). Li edges as they are revealed. Identi®cation of et al. (1989) described the auditory region of these features and comparisons with other Cocomys, a rodent from the early Eocene of living and extinct were made from China; future restudy of this specimen should the literature mentioned below and from the yield important comparative details. ontogenetic studies by MacPhee (1981) and My primary purpose in this paper is to de- Wible (1984). scribe and illustrate the auditory region in Among the ®rst major comparative studies Paramys copei, AMNH 4755 (holotype) and on auditory regions in mammals was that of 4756. These two skulls preserve the ®nest de- van Kampen (1905), who built primarily on tail that I have seen in any fossil rodent, and the ®ne German work in anatomy and de- their early Eocene age, approximately 51 mya velopment. Van der Klaauw (1931) made (million years ago), makes them important by fossil mammals a part of his monograph; he antiquity. The rodent branch of mammalian was able to take advantage of the descrip- phylogeny may have been separate already in tions and collections of fossils from the Asia in the early Paleocene (Heomys Li, American West that had accumulated since 1977); undisputed rodents are known from the the late 1800s. Recent publications provide late Paleocene: Alagomyidae in Asia and character analyses and detailed information North America and Ischyromyidae in North about early mammals (Wible, 1990; Rougier America. In North America, the appearance et al., 1992; Wible and Hopson, 1993; Wible of ischyromyid rodents de®nes the beginning et al., 1995; Rougier et al., 1996; McKenna of the Clarkforkian Age (Woodbur- et al., 2000) and insectivores (McDowell, ne, 1987: 61) about 56 mya (McKenna and 1958; Novacek, 1986). The pioneering com- Bell, 1997: ®g. 1). The early Eocene saw si- parative studies on rodent auditory regions multaneous diversi®cation of rodents in Asia, are relatively recent (Oaks, 1968; Parent, North America and Europe. My second pur- 1980; Lavocat and Parent, 1985). Many stud- pose is to compare the auditory region of Par- ies of living rodents have focused on taxa amys copei with an outgroup of thoroughly with in¯ated bullae (Webster, 1961, 1962, described and illustrated mammals including 1975; Pye, 1965; Webster and Webster, 1971, the early mammal Vincelestes, the marsupial 1975, 1977, 1984; Lay, 1972, 1993; Wahlert, Didelphis, and the insectivore Solenodon, and et al., 1993). Segall (1971) illustrated gliding with other Eocene rodents. My goals are to and non-gliding sciurids. Information is also describe character polarities as a starting point available about speci®c extinct taxa (Lavo- for comparison of rodent auditory regions and cat, 1967; Meng, 1990; Wahlert, 1974, 1977, to assess the relative primitiveness of Para- 1978, 1983; Carrasco and Wahlert, 1999). mys auditory morphology. The two specimens of Paramys copei were SPECIMENS EXAMINED collected by Jacob L. Wortman in 1880 for Edward D. Cope. Images of the skulls have Specimens are listed according to the been independently illustrated and published: classi®cation of McKenna and Bell (1997); Cope, 1884, pl. 24a, ®gs. 1, 1a, 2, 2a; Wood, generic assignments of paramyine species are 2000 WAHLERT: AUDITORY REGION IN PARAMYS COPEI 3 according to Korth (1984, 1985, and 1994). Suborder Sciuravida, Family Sciuravidae Abbreviations: AMNH, American Museum Sciuravus nitidus, AMNH 12531 and 12551, of Natural History; USNM, United States USNM 17683 and 22477, Blacks Fork Mem- National Museum of Natural History. ber, middle Eocene, Bridger Formation, Class Mammalia, Legion Cladotheria, Suble- Bridger Basin, Wyoming gion Zatheria, Family Vincelestidae Vincelestes neuquenianus La Amarga Forma- AUDITORY REGION OF tion, early Cretaceous, Southern NeuqueÂn PARAMYS COPEI Province, Argentina (information from Rougier et al., 1992) The late early Eocene species Paramys copei (AMNH 4755, 4756, and 103390) pre- Sublegion Zatheria, Supercohort Theria, Cohort serves the most ancient, complete example Marsupialia, Magnorder Ameridelphia, Order of auditory region morphology that is avail- Didelphimorphia, Family Didelphidae, Subfam- able for rodents. I describe it below as a stan- ily Didelphinae dard, and then note the differences of other Didelphis virginiana, AMNH 2070, Recent, no Eocene rodent taxa. The region is illustrated data. in slightly oblique ventral view in ®gures 1, Sublegion Zatheria, Supercohort Theria, Cohort 2, and 3. Placentalia, Magnorder Epitheria, Superorder In ventral view the junction of the petrosal Preptotheria, Grandorder Lipotyphla, Order with the basioccipital is oblique, and the ba- Soricomorpha sioccipital is widest posteriorly. The anterior Solenodon paradoxus, AMNH 28270, Recent, part of the basioccipital and adjacent basi- Haiti sphenoid appears swollen and may be pneu- Sublegion Zatheria, Supercohort Theria, Grandor- matized in the anterior portion. The posterior der Anagalida, Mirorder Simplicidentata, Order part is a ¯ange that ¯oors the ventral petrosal Rodentia. All specimens are of Eocene age. sinus. The posterior lacerate foramen is a len- ticular gap medial to the posterior part of the Suborder , Family Ischyromyidae, auditory chamber and situated between the Subfamily Paramyinae: petrosal and the occipital. A broad ascending Tribe Paramyini keel of the petrosal partly divides the anterior Paramys copei, AMNH 4755 (holotype) and part of the gap, which transmitted the sig- 4756; Lost Cabin Member, late early Eocene, moid sinus, from the posterior, for nerves IX, Wind River Formation, Wind River Basin, X, and XI. A prominent dimple in the petro- Wyoming; AMNH 103390, late early Eo- sal just anterior to the keel is the opening of cene, San Jose Formation, San Juan Basin, the cochlear canaliculus. A broad, covered New Mexico passage between the basioccipital and the pe- Paramys delicatus, AMNH 12506 and 13090, trosal may mark the course of the ventral (in- Blacks Fork Member, middle Eocene, Bridg- ferior) petrosal sinus. It opens just anterior to er Formation, Bridger Basin, Wyoming Notoparamys costilloi (type species) (ϭ Lep- the posterior lacerate foramen. The sigmoid totomus costilloi), AMNH 55110 and 55111 sinus and ventral petrosal sinus joined here (holotype), Lost Cabin Member equivalent to form the internal jugular vein. A deep, (Wind River Formation), late early Eocene, horizontal ¯ange of the basioccipital sepa- Huerfano Formation, Huerfano Basin, south- rates the ventral petrosal sinus from the brain eastern Colorado (AMNH 4755, left side). Tribe Manitshini The petrosal ends anteriorly at the piriform fenestra; an anterior prong from the petrosal Pseudotomus hians (type species), AMNH meets a posterior spine from the basisphe- 5025 (holotype), Blacks Fork Member, mid- noid, and together they divide off the medial dle Eocene, Bridger Formation, Bridger Ba- sin, Wyoming part of the fenestra as an anterior carotid fo- Pseudotomus petersoni (ϭ Ischyrotomus peter- ramen. The alisphenoid and possibly a nar- soni) AMNH 2018 (holotype), Myton Mem- row part of the squamosal form the rest of ber, late Eocene, Uinta Formation, Uinta Ba- the anterior edge of the fenestra. The dorso- sin, Utah lateral part of the petrosal abuts the squa- 4 AMERICAN MUSEUM NOVITATES NO. 3307

Fig. 1. Paramys copei (AMNH 4755) auditory region diagram with major features labeled; slightly oblique view of right auditory region; anterior is to the right. Abbreviations: 2ff secondary facial fo- ramen; acf anterior carotid foramen; as alisphenoid; bo basioccipital; bp bullar process of petrosal; bs basisphenoid; cf cochlear fossula; chc cochlear canaliculus; etr epitympanic recess; ® fossa incudis; fo foramen ovale; fv fenestra vestibuli; hF hiatus Fallopii; hy hypoglossal foramen; m mastoid; me mastoid eminence; ost fossa for origin of stapedius muscle; pir piriform fenestra; plf posterior lacerate foramen; pr promontorium; sq squamosal; stc stapedial artery channel; stf stapedial foramen; sty stylomastoid foramen; vpc ventral petrosal sinus canal at posterior lacerate foramen.

Fig. 2. Paramys copei right auditory region Fig. 3. Paramys copei right auditory region slightly oblique view. with arteries and nerves restored. Abbreviations: acc accessory nerve; cc common carotid artery; ec external carotid artery; fa facial nerve; gl glos- mosal, and the mastoid, the posterior part of sopharyngeal nerve; gp greater petrosal nerve; hy the petrosal, ascends and narrows to a point hypoglossal nerve; ic internal carotid artery; icp between the squamosal and occipital at the transpromontorial continuation of internal carotid back of the skull. artery; st stapedial artery; vg vagus nerve. The promontorium, which houses the co- chlea, bulges anterolaterally. The fenestra vestibuli faces laterally; it is oval, about two tongue-shaped ¯ange of the petrosal. Pro- times as long as wide, and the long axis is cesses, canals, and basins permit one to re- nearly horizontal. The cochlear fossula, construct something of the nature of soft tis- which contains the fenestra cochleae, faces sues. posteriorly. Dorsolateral to the promonto- The ventral surface of the promontorium rium, the roof of the tympanic cavity is bears evidence of the course of the internal domed as a fossa and bordered laterally by a carotid artery and its stapedial branch. The 2000 WAHLERT: AUDITORY REGION IN PARAMYS COPEI 5 internal carotid artery would have entered a posterior carotid foramen between the bulla and petrosal; it ran laterally, posterior to a change of surface angle that may mark the extent of bullar overlap, and then within a shallow groove that lies anteroventral to the cochlear fossula; ®nally it turned anteriorly, and the stapedial artery branched off. A deep gutterlike anterior sulcus on the promonto- rium may mark the transpromontorial course of the internal carotid artery. The sulcus is bounded laterally by a stout ridge; the ridge runs anteriorly from the middle of the lateral side of the promontorium. The origin of a moderately large tensor tympani muscle probably extended on the lateral surface of this ridge to the crest. A low swelling forms the medial side of the sulcus, and there is a hole anterior to the sulcus between the petro- Fig. 4. Sciuravus nitidus (AMNH 12531) sal and the pterygoid region that is inter- ventral view diagram of auditory regions and en- preted as the anterior carotid foramen. Evi- docranial cast (with mechanical dot pattern). An- dence of the transpromontorial continuation terior is toward the top of the page. Abbreviations: of the internal carotid artery is stronger in acf anterior carotid foramen; fo foramen ovale; f/ Sciuravus nitidus (USNM 17683) (®g. 4); a s foramen shared by facial nerve and stapedial shallow groove in a similarly situated sulcus artery; hy hypoglossal foramen; ic internal carotid joins the transverse groove of the internal ca- artery channel; icp transpromontorial internal ca- rotid (Wahlert, 1974: ®g. 8). rotid artery channel; ost fossa for origin of sta- The stapedial artery branched from the in- pedius muscle; otc orbitotemporal canal (con- ternal carotid and passed through the inter- tained supraorbital branch of stapedial artery); plf posterior lacerate foramen; ri ramus inferior of crural foramen of the stapes (missing). The stapedial artery (cast); rs ramus superior of sta- artery crossed below the facial nerve, ran an- pedial artery (cast); sf sphenofrontal foramen; st teriorly, above the medially projecting ¯ange stapedial artery (cast); sty stylomastoid foramen. from the lateral tongue of the petrosal, and then entered the cranial cavity at the lateral side of the piriform fenestra. Temporal fo- tended ventromedially onto the ridge along ramina, which are present in the squamosal- the internal carotid sulcus. The bone that parietal suture and in the parietal, probably roofs the fossa alongside the promontorium transmitted temporal branches from the su- may have been formed as the epitympanic perior ramus of the stapedial artery. In Sci- wing of the petrosal (MacPhee, 1981: 52), a uravus nitidus (AMNH 12531) a stapedial separate ossi®cation from the tonguelike lat- groove on the promontorium and an endo- eral ¯ange. Schrenk (1989) found the fossa cranial cast of the stapedial artery channel and lateral ¯ange to be parts of a single struc- are visible (®g. 4). The artery divided into ture in Ctenodactylus. The ¯ange continues inferior and superior rami shortly after en- anteriorly from the crista parotica as does the tering the cranium; the inferior ramus di- embryonic tegmen tympani (Beer, 1985: verged anteriorly and the superior ramus dor- 391). The considerable breadth and domed soanteriorly. shape of the fossa and lateral ¯ange suggest In ventral view the petrosal is very wide dorsolateral in¯ation of the auditory cham- alongside the promontorium, and it is sub- ber. divided into two parts, a shallow median fos- Dorsolateral to the anterior end of the fe- sa that widens anteriorly and a complex lat- nestra vestibuli, the facial nerve emerged eral ¯ange. The origin of the tensor tympani from the fallopian canal into the middle ear muscle occupied part of the fossa and ex- through the posteriorly facing secondary fa- 6 AMERICAN MUSEUM NOVITATES NO. 3307 cial foramen at the posterior end of the fossa. triangle is at the paroccipital process, which The nerve ran posteriorly in the facial sulcus is formed by the occipital bone and but- with the crista parotica on its lateral side; tressed by the mastoid at its base. Novacek then it turned ventrally, where it was accom- (1986: ®g. 20) labeled a similar feature in modated in a broad groove on the medial Leptictis as the sulcus for origin of the di- side of the mastoid eminence. It emerged gastric muscle. There are no mastoid cavities. from a probably incomplete or primitive sty- The posterolateral extension of the mastoid lomastoid foramen. A small canal for the ma- contains a large chamber that housed the jor (greater) petrosal nerve diverges from the para¯occulus of the cerebellum; the smooth fallopian canal and passes through the roof in®lling of this chamber is exposed in of the fossa; it ends at an anteriorly directed AMNH 4755. The tympanohyal is presumed foramen, the hiatus Fallopii, and is continued to be fused to the crista parotica. by a groove to the anterior edge of the pe- The auditory chamber in all rodents is en- trosal at the piriform fenestra. The canal is closed by the petrosal and ectotympanic; exposed in AMNH 4756 by damage; it is there is no participation by the squamosal or widest near the facial nerve as would be ex- other elements in its wall. Traces of the au- pected, since it marks the branching of the ditory bullae can be found in a few fossil major petrosal nerve from the geniculate specimens; bullae are not preserved in Par- ganglion. amys and most of its Bridgerian (middle Eo- The nearly vertical outer surface of the lat- cene) relatives. The anteromedial part of the eral ¯ange of the petrosal was probably ex- promontorium bears a broad facet in Para- posed between the squamosal bone and the mys. The lateral margin of this facet is a low tympanic bulla (missing). The ¯ange forms swell along the edge of the sulcus for the an anteroposterior arch lateral to the fenestra anterior continuation of the internal carotid vestibuli. Its ventrally facing surface contains artery; its posterior margin is formed by a a shallow, anteroposteriorly elongated basin, change in surface angle that is just anterior the epitympanic recess. Just posterior to the to the channel for the internal carotid chan- recess the fossa incudis is indicated by a nel. The facet approximately matches the slight pocket in the anterior surface of the area where the dorsomedial curve of the bul- mastoid eminence. la overlaps the promontorium in Reithropar- The stapedius muscle arose posterolateral amys and Sciuravus. At the lateral side of the to the promontorium in a basin within the petrosal, the ¯ange derived from the tegmen ring of the lateral semicircular canal. The an- tympani bears a ventral facet or seat at its terior boundary of the basin is formed by the anterior end, presumably for articulation of ridge enclosing the lateral semicircular canal; the anterior crus of the ectotympanic. the ridge extends from above the posterior end of the fenestra vestibuli to the mastoid ANALYSIS OF EAR CHARACTERS IN eminence. The canal passes posteromedially PARAMYINES AND OUTGROUPS around the depression for muscle origin and then through the robust bridge from mastoid A summary of character states is presented to the promontorium. If the bulla in para- in table 1. These data are presented as a start- myines was similar to that in Sciuravus ing point for future studies; they do not re- (Dawson, 1961: pl. 3; Wahlert, 1974: ®g. 7) veal any pattern of relationships within the its bony posterior crus did not entirely cover rodents and, when processed with PAUP, this area, although the membrane that en- give a simple comb structure to a clade of closed the chamber may have extended much the paramyines and Sciuravus. Some char- farther posteriorly. acters are included simply because they are In ventral view there is a triangular de- important if other groups of mammals or lat- pression in the mastoid posterior to the sta- er rodents are included in a comparative pedius basin. The base of the triangle extends study. Data for Vincelestes are are taken from from the vertically projecting mastoid emi- Rougier, et al. (1992). nence medially to the posterior end of the Character 1: Anterior part of basioccip- posterior lacerate foramen. The apex of the ital and adjacent basisphenoid in ventral 2000 WAHLERT: AUDITORY REGION IN PARAMYS COPEI 7

TABLE 1 ridges abutting the petrosal, and the junction Character States in Outgroup Mammals and Para- of basioccipital and basisphenoid is primarily myine and Sciuavid Rodents ¯at; Sciuravus also lacks the ventral petrosal sinus between these bones and the petrosal. Character 2: Ventral (inferior) petrosal sinus is transmitted by a canal (enclosed passage) in the petrosal (0); is transmitted by a partly shielded passage between the ba- sioccipital and the petrosal, and a short ba- sioccipital ¯ange may be present (1); is transmitted by a canal that deeply grooves the side of the basioccipital, and the sinus is separated from the brain by a deep, horizon- tal ¯ange of the basioccipital (2); is trans- view are not swollen laterally and may meet mitted endocranially, and no canal is present in a ¯at surface (0); appear swollen laterally between the basioccipital and the petrosal and may project ventrally (1). I do not know (3). The venous sinus passes through a pe- the morphology of this area in Vincelestes. trosal canal in Vincelestes (Rougier et al., In Didelphis the middle of the basioccipital 1996). Wible (1990: ®g. 4) illustrated the swells anteriorly to meet a similarly swollen ventral petrosal sinus in Didelphis as enter- basisphenoid, but there is no lateral swelling; ing a canal anterior to the posterior lacerate a similar increase in thickness can also be foramen. In the adult specimen that I have observed internally. In Solenodon there is no on hand, the canal deeply indents the lateral lateral swelling, and the bones meet in a ¯at side of the basioccipital and enters the cra- surface. nium, as shown by Wible. McDowell (1958: The lateral swellings present in paramy- 139) stated: ``As in other lipotyphlan insec- ines are very different from the condition in tivores, the ventral petrosal sinus of Neso- Didelphis. In Paramys copei lateral swellings phontes and Solenodon is exposed ventrally, of the basioccipital begin at the anterior end owing to the failure of the petrosal and ba- of the posterior lacerate foramen as a ¯oor sioccipital to form a suture beneath the sinus under the ventral petrosal sinuses; the swell- immediately anterior to the jugular (posterior ings expand anteriorly so there is only a me- lacerate) foramen. This exposure extends for- dial trough separating the pair at the anterior ward from the jugular foramen, with which end of the bone; this is continued anteriorly it is continuous, approximately to the level on the basisphenoid, and the relief ¯attens of the basioccipital-basisphenoid suture.'' out at the level of the foramen ovale. The The medial edge of the petrosal forms a par- character is present in P. delicatus, but the tial roof above the sinus canal. swelling is not as broad. The junction of ba- In Paramys copei and P. delicatus there is sioccipital and basisphenoid in Paramys ap- a wedge-shaped gap between the basioccip- pears in side view as a broadly obtuse, down- ital and petrosal for the ventral petrosal sinus. ward projecting angulation of the medial el- It is widest at the posterior lacerate foramen ements of the basicranium. and narrows anteriorly. The channel between Swellings in Pseudotomus hians and P. the two bones is separated from the braincase petersoni begin posteriorly with a low boss by a broad, long dorsal ¯ange of the basi- somewhat anterior and medial to the anterior occipital. This same character is present in end of the posterior lacerate foramen; the Pseudotomus and Notoparamys. Petrosals are swelling continues only as far as the basi- missing from Pseudotomus hians, and the occipital-basisphenoid suture. The basicran- channel incises on the lateral face of the ba- ium is not angled in side view. Dorsoventral sioccipital; it bends sharply dorsal at about ¯attening of specimens of Notoparamys cos- the anterior end of the petrosal. The sinus tilloi makes it impossible to evaluate the must have entered the cranium at the fora- character. Sciuravus has a very slight, if any, men that could also have been for passage of anterior increase in breadth of the lateral the transpromontorial continuation of the in- 8 AMERICAN MUSEUM NOVITATES NO. 3307 ternal carotid artery (anterior carotid fora- monotremes, marsupials and some insecti- men). In Sciuravus the petrosal is closely ap- vores and bats) or helping partly to marginate pressed to the basioccipital, and there is no these openings at the middle lacerate foram- room for the ventral petrosal sinus between ina (in more advanced eutherians)'' (ibid.: them. 56). Separate nameable foramina in this re- Character 3: Piriform fenestra (ϭ mid- gion have been partitioned off from the pir- dle lacerate foramen of some authors) is iform fenestra by changes in the extent of small and medial (0); is large, and prongs of bones. the petrosal and pterygoid separate the an- In Solenodon, the hiatus Fallopii is in the terior carotid foramen from the lateral part anterior edge of the petrosal at the fenestra, of the fenestra (1). In Vincelestes, an aperture whereas in paramyines it is in the middle of identi®ed as the ventral opening of the ca- the fossa for the muscle origin. This suggests vum epiptericum may be the same, at least that the petrosal does not extend as far an- in part, as the piriform fenestra. In Didelphis teriorly in Solenodon as it does in rodents; it is a small opening at the tip of the petrosal, the piriform fenestra in rodents may be cre- and the carotid foramen is anterior to it. It is ated by failure of the alisphenoid and adja- large in Solenodon, paramyines, and Sciuravus. cent basisphenoid to grow posteriorly rather than of the petrosal to extend anteriorly. An- I use the term, piriform fenestra, instead other explanation is that in rodents the an- of middle lacerate foramen as in my previous terior part of the fossa for muscle origin does publications, and I have used anterior carotid not descend far enough (that is, it slopes too foramen as the name for the place where the steeply) to enclose the entire course of the transpromontorial continuation of the inter- nerve. nal carotid artery enters the cranium. The pir- Character 4: Internal carotid artery en- iform (also pyriform ϭ pear shaped) fenestra ters the middle ear, runs transversely, gives is a name coined by McDowell (1958: 128) off the stapedial artery, and then turns an- for the ``large vacuity in the roof of the mid- teriorly across the promontorium (0); does dle ear (corresponding to the area of origin not enter the ear (1). Rougier et al. (1992: of the tensor tympani) of the Soricidae and ®g. 3) reconstructed the internal carotid ar- Solenodontidae.'' He restricted the use of fo- tery in Vincelestes. It ran laterally, gave off ramen lacerum medium to the entrance of the the stapedial artery, and then turned anteri- carotid artery into the skull and pointed out, orly in a channel on the anterior slope of the however, that Story (1951) applied the term promontorium to the carotid foramen. in carnivores to the foramen for the large The internal carotid artery does not enter vein from the ventral petrosal sinus that the ear in adult Didelphis. This is apparently drains to the pharyngeal and pterygoid plex- a derived characteristic of marsupials. Con- uses (McDowell, 1958: 124). MacPhee cerning the morphology of the two isolated (1981: 58) broadened the meaning of piri- petrosals from late Cretaceous didelphoids, form fenestra ``to include the large gap, seen Wible concluded (1984: 310): ``In particular, in all fetal and a few adult mammals, that no vascular grooves appear on the ventral lies between the auditory capsule and its dor- surface of the promontorium for either the sal outgrowths (tegmen tympani, epitympan- internal carotid or stapedial arteries. This ic wing of petrosal) on the one hand and the suggests that the medial position of the in- epitympanic wings of the sphenoid and squa- ternal carotid and the loss of the stapedial mosal on the other.'' Moore (1981: 41) said artery are ancient traits among Marsupialia. that ``posterior to the ala is the middle Also, a large groove along the medial border lacerate foramen which lies close to the of the petrosal's dorsal surface probably con- entrance of the internal carotid artery into tained an intracranial ventral petrosal sinus.'' the cranial cavity .... Theextension of the In conclusion about the living marsupials, basisphenoid posterior to the pituitary results Wible stated (1984: 312): ``The internal ca- in this bone totally enclosing the openings rotid artery follows an extrabullar pathway for the internal carotid arteries (as occurs in medial to the tympanic cavity and promon- typical amphibians and reptiles as well as in torium of the petrosal.... Theproximal sta- 2000 WAHLERT: AUDITORY REGION IN PARAMYS COPEI 9 pedial and its three branches, the posterior, ridge instead of a low one (2). The anterior superior, and inferior rami, are absent.'' The slope of the promontorium is smooth in Vin- internal carotid artery enters the ear of So- celestes, Didelphis, and Solenodon. A prom- lenodon; it turns and continues anterome- inent ridge is present in paramyines (no ev- sially across the sloping anterior surface of idence is retained in Pseudotomus). It sepa- the promontorium to the anterior border of rates the fossa for origin of the tensor tym- the tympanic cavity. Its lateral branch, the pani muscle from a gutter that ascends stapedial artery, passes between the crura of anteriorly on the medial part of the promon- the stapes and divides at the anterior edge of torium. The ridge may have one or both of the petrosal into superior and inferior rami; two functions: to delimit the channel for the the ramus superior passes through the pos- anterior extension of the internal carotid ar- terolateral part of the piriform fenestra to be- tery, and to enlarge the site of origin of the come the middle meningeal artery within the tensor tympani muscle. Sciuravus (USNM cranium (McDowell, 1958: 140). 17683 and 22477, and Parent, 1980: pl. 3, Paramyines and Sciuravus have a channel ®g. 3) has a slight ridge. that runs transversely across the promonto- Character 6: Stapedial artery is exposed rium to the fenestra vestibuli; it indicates the in the middle ear along its entire course (0); presence of the stapedial artery. In Sciuravus enters a foramen anteroventral to the sec- a channel runs anteromedially from it on the ondary facial foramen (1); enters an aper- anterior slope of the promontorium and rep- ture together with the facial nerve and im- resents the transpromontorial continuation of mediately passes laterally through a foramen the internal carotid artery to the anterior ca- into the cranium (2); is absent (3). The sta- rotid foramen. All paramyines have a large, pedial artery is absent in Didelphis. The sta- gutterlike channel in this position that is pedial artery is exposed in the middle ear of bounded on the medial side by the abutment Vincelestes and Solenodon. In Vincelestes, of the petrosal against the basioccipital and Rougier et al. (1992: ®g 3) reconstructed the laterally by a crest that may have been part artery as splitting into a ramus superior that of the origin of the tensor tympani muscle. I exits dorsally through the petrosal and an in- have interpreted this morphology as evidence ferior branch that continues anteriorly in a for the anterior continuation of the internal channel. In Solenodon it branches into su- carotid artery within the ear. However, in perior and inferior rami at the anterior end of paramyines the internal carotid does not the petrosal (McDowell, 1958: ®g. 6). mark the promontorium adjacent to the pos- In paramyines the stapedial artery is partly terior lacerate foramen, and the stapedial or entirely enclosed in a channel or canal that channel is only clear close to the fenestra begins a short distance anterolateral to the vestibuli; the arteries may not have touched intercrural foramen in the stapes. In Paramys the medial part of the promontorium. On the copei it is partly covered by a medially pro- right side of one specimen of Paramys deli- jecting ¯ange from the lateral tongue of the catus (AMNH 12506) I imagine that I see a petrosal. It enters a foramen that is antero- faint, anteriorly running channel that ascends lateral to the secondary facial foramen in to the large anterior carotid foramen at the Paramys delicatus, Pseudotomus petersoni, anteromedial end of the petrosal; it could be and possibly Notoparamys costilloi (crushing a channel for the internal carotid artery, and makes character assessment uncertain). A there seems to be adequate room for both an specimen of Sciuravus nitidus (AMNH artery and the ventral petrosal sinus to pass 12531) shows a complete picture of the ar- through the foramen. However, this vague terial pattern. The basicranium and auditory feature is not conclusive evidence of the ar- regions are relatively intact, but the bone of tery. the cranium is gone, and a clear endocranial Character 5: Anterior part of the pro- cast remains. The stapedial artery passed montorium is smooth (0); has a low ridge through the intercrural foramen in the stapes that separates the course of the internal ca- (missing), turned anterolaterally into a fora- rotid artery from the fossa for origin of the men that also transmitted the facial nerve, tensor tympani muscle (1); has a prominent and continued in an anterolateral direction 10 AMERICAN MUSEUM NOVITATES NO. 3307 through the petrosal into the braincase. A squamosal, the jaw joint is far anterior and cast of the course of the artery and its not directly supported by the petrosal, and branches is visible on the left side of the en- the epitympanic recess is entirely within the docranial cast (®g. 4). The inferior ramus di- petrosal [the squamosal is grooved ventrally verged anteriorly just inside the cranium. The to receive the dorsal part of petrosal] (2). superior ramus continued laterally and then The petrosal participates in the support of the turned anteriorly; the course of the supraor- glenoid fossa in Vincelestes, and it appears bital branch in the orbitotemporal canal is to be exposed in lateral view below the squa- marked by a cast that makes a broad arch mosal (Rougier et al., 1992: ®gs. 1, 2); the that descends to the sphenofrontal foramen epitympanic recess is within the petrosal. in a tiny bit of bone preserved in the back The petrosal is hidden in lateral view by the of the orbit; the artery was presumably ac- squamosal in Didelphis and Solenodon, and companied by a vein. the squamosal makes the lateral side of the Character 7: Petrosal is narrow in width epitympanic recess in both; I regard this as and length alongside the promontorium (0); the primitive condition of the outgroup mam- is wide alongside the promontorium and the mals. In paramyines and Sciuravus the dor- fossa in which the tensor tympani muscle solateral part of the petrosal (the lateral arose, extends, and broadens anteriorly (1). ¯ange) is exposed below the squamosal, and The fossa in which the tensor tympani mus- the epitympanic recess is entirely within the cle arose is narrow and long in Vincelestes, petrosal; this appears to be a different and narrow and short in Didelphis, and narrow derived condition compared to that in Vin- and tapered anteriorly alongside the promon- celestes, in which the petrosal directly sup- torium in Solenodon. In paramyines and Sci- ports the jaw joint. uravus it is wide and broadens anteriorly as Character 10: Epitympanic recess is a a shallow broad channel. The muscle origin shallow channel (0); is deepened and en- leaves no distinct marks, so its actual extent larged (1). Vincelestes and Didelphis have in the fossa is not known with certainty. only shallow epitympanic recesses. The re- Character 8: Hiatus Fallopii is present cess is slightly enlarged dorsally in Soleno- in the medial part of the fossa in which the don. The epitympanic recess is no more than tensor tympani muscle arose (0); is present a shallow groove in the lateral ¯ange of the at the front edge of the petrosal fossa for the petrosal in all of the paramyines and Sciura- tensor tympani muscle (1). In Vincelestes the vus; a slight pocket in the posterior end hiatus Fallopii is in the fossa alongside the marks the fossa incudis. promontorium and not close to the anterior Character 11: Alisphenoid bone makes end of the bone. The hiatus Fallopii is pre- no part of the bulla (0); makes a part of the sent at the anterior edge of the petrosal in bulla (1). The alisphenoid forms no part of Didelphis and Solenodon. The hiatus Fallopii a bulla in Vincelestes. A posteriorly concave is in the medial part of the fossa for the ten- process of the alisphenoid forms the anterior sor tympani muscle in Paramys copei, P. de- part of the bulla in Didelphis. The bulla in licatus, Pseudotomus petersoni, Notopara- Solenodon is formed by a somewhat in¯ated mys costilloi, and Sciuravus nitidus (AMNH but incomplete tympanic ring, and there is no 12551, right side); the region is not pre- contribution from the alisphenoid; the case is served in Pseudotomus hians. the same in paramyines and Sciuravus. Character 9: Petrosal exposure in lateral Character 12: Shallow basin posterior view: the petrosal is visible, supports the jaw to the crista interfenestralis (between the joint, and contains the epitympanic recess fenestra vestibuli and the fenestra cochle- (0); the dorsolateral edge of the petrosal is ae) is simple and encircled by the lateral hidden by a downward overlap of the squa- semicircular canal (0); is enlarged by a shal- mosal, the jaw joint is anterior and not di- low depression that extends posteromedially rectly supported by the petrosal, and the from the basin ventral to the rim made by squamosal makes the lateral side of epitym- the lateral semicircular canal (1). Some part panic recess (1); the dorsolateral part of the of the basin is the fossa in which the stape- petrosal is exposed between the bulla and dius muscle arises. In Vincelestes the basin 2000 WAHLERT: AUDITORY REGION IN PARAMYS COPEI 11 is immediately adjacent to the promonto- stapedial artery is exposed for its entire rium. The space posterior to the promonto- course in the middle ear, and it is not hidden rium in Didelphis is short and deep; a slight by a ¯ange as in Paramys. The promonto- depression in the posterior wall may mark rium lacks the anterior crest that in Paramys the muscle origin. The basin is broad and separates the channel for the promontory shallow in Solenodon; a slight depression in branch of the stapedial artery. the posterior side may mark the muscle ori- The most striking differences are in pro- gin. Paramys copei has a simple, deep basin. portions and possibly stem from a radical Paramys delicatus and Sciuravus nitidus change in stress-bearing function of the au- (AMNH 12551) have a shallow posterome- ditory region. In Vincelestes the glenoid fos- dial extension that is ventral to the semicir- sa is immediately adjacent to the ear region, cular canal. In Pseudotomus the extension which therefore must be structurally inte- appears to have been present, though all grated in the masticatory system. Mono- specimens are damaged. The region is tremes retain the same arrangement, but in crushed in Notoparamys. therian mammals, the jaw articulation is an- terior to the auditory region (Rowe, 1988). DISCUSSION In Paramys and all rodents and lagomorphs, the glenoid fossa is very far anterior, and the The data matrix (table 1), when analyzed shape of the auditory region depends much by PAUP, yields 56 trees of equal length (20 more on the function of hearing, though the steps). The consensus tree shows the para- mastoid has an important role in attachment myines and Sciuravus as a clade with no in- of neck muscles. The entire auditory region ternal resolution. of Paramys is broad. The fossa alongside the Comparison of the petrosals of Paramys promontorium that houses the origin of the copei and the early Cretaceous mammal Vin- tensor tympani muscle widens anteriorly, and celestes neuquenianus (Rougier et al., 1992) there is a lateral ¯ange with a ventral facet is instructive in that there are many similar- for supporting the anterodorsal part of the ities in the correspondence of parts, but there bulla. In addition, the mastoid extends pos- are also structural differences and important teriorly and makes an almost boxlike poster- differences in proportions. The shared prim- oventral corner of the skull. The lack of itive character states are: a low promonto- structural constraint from the masticatory rium with grooves for both the internal ca- system on the auditory region can be seen in rotid artery, which courses anteriorly across the enlargement and elaboration of the bulla, the promontorium, and its lateral branch, the the epitympanic recess, and the mastoid re- stapedial artery; stapedial artery and facial gion in many different living and extinct ro- nerve not sharing a common foramen; and dents. epitympanic recess narrow and shallow and The insectivore Solenodon paradoxus within the petrosal. (McDowell, 1958) retains some primitive Vincelestes has features that are more features seen in Vincelestes: the promonto- primitive than the corresponding structures in rium is low and has a groove running anter- Paramys. A large prootic canal, which en- omedially across the promontory for the in- closes a part of the lateral head vein, is pre- ternal carotid artery; the stapedial artery is sent in the petrosal lateral to the secondary exposed lateral to its passage through the sta- facial foramen; the canal has been lost in pla- pes; and the epitympanic recess is narrow cental mammals, except it is present in the and shallow. But, like therian mammals, the basal eutherian Prokennalestes (Wible et al., prootic canal is lost, and a part of the venous 1997); the prootic sinus and lateral head vein drainage is via the postglenoid foramen, and disappear during early development in all eu- the jaw articulation is anterior to the otic re- therians, and venous drainage from the trans- gion. Solenodon has a large piriform fenestra verse sinus is via a new vessel, the capsu- that is covered with connective tissue and is loparietal emissary vein, that exits from the the site of origin of the tensor tympani mus- skull at the postglenoid foramen (Wible, cle (ibid.: 138±139). Paramys, the rest of the 1990: 192; Rougier et al., 1992: 202). The ischyromyids, and Sciuravus also have a 12 AMERICAN MUSEUM NOVITATES NO. 3307 large piriform fenestra. In Reithroparamys nus was endocranial; the auditory bullae ob- (Meng, 1990: ®g. 1; Wahlert, 1974: ®g. 5), scure the character in Lavocat's (1967) illus- Sciuravus, theridomyids (Lavocat and Par- trations of theridomyids. ent, 1985, ®g. 1a), and many later rodents, Paramyines and Sciuravus share derived the tympanic bulla covers at least the poste- character states that are not present in the rior part of the fenestra. Novacek (1985: 77) outgroup. These include the presence of a pointed out that rodents and lagomorphs ridge on the anterior surface of the promon- share a uniquely enlarged piriform fenestra. torium that separates the channel of the trans- In rodents this feature is anterior to the bulla promontorial continuation of the internal ca- and may represent reduction in ossi®cation rotid artery from the fossa for origin of the of the posteromedial end of the alisphenoid tensor tympani muscle. This fossa alongside and perhaps of the adjoining basisphenoid the promontorium is wide and long with the that is in addition to reduction of the anterior hiatus Fallopii opening in its middle rather part of the petrosal. I suspect that this reduc- than in its anterior edge. The piriform fenes- tion in ossi®cation is possible in rodents and tra is apparently formed by a decreased pos- lagomorphs because the jaw articulation is is terior extent of the ventromedial part of the so far anterior. McDowell (1958: 137) attri- alisphenoid and perhaps of the adjoining ba- buted the fenestra in Solenodon to absence sisphenoid rather than decreased anterior ex- of the anterior part of the petrosal. tent of the petrosal as in Solenodon. Wood (1962: ®g. 3) and Wahlert (1974) Paramys copei and the other fossil rodents identi®ed a carotid canal between the basi- examined also show an important derived fea- occipital and petrosal of Paramys delicatus, ture of both rodents and lagomorphs: the but it probably housed the ventral petrosal squamosal does not participate in the wall of sinus instead; a similar gap is visible in Par- the epitympanic recess. Van der Klaauw amys copei and in other paramyine genera. (1931: 78±79) stated: ``In the rodents we do Wible (1984: 17) stated: ``Wible (1983 [p. not ®nd the squamosal in the lateral wall of 286]) points out that the structure of the me- the epitympanic recess, but the petrotympanic. dial channels in the fossil basicrania is not This small development of the caudal portion always the same, but may be an open groove of the squamosal is a characteristic feature in on the basioccipital-petrosal suture, a patent rodents and causes the total absence of the basicapsular fenestra (an open basioccipital- squamosal in the walls of the tympanic cavity petrosal suture), or a closed canal between and its accessory cavities. As the tympanic the basioccipital and petrosal. Of these three and the petrosal are fused, it often can hardly types of medial-channel morphology, only be said which of the two forms the lateral wall the ®rst (the open groove) is ever associated of the epitympanic recess. Also the bony lat- with an internal carotid artery in extant eu- eral wall can be totally absent. An incisura therians, whereas venous sinuses are known tympanica may be absent or present.'' He de- to occupy channels of all three types.'' Mc- scribed lagomorph genera in the midst of the Dowell (1958: 139) described the condition rodents and presumably considered them as in insectivores: ``As in other lipotyphlan in- part of the same group. The feature can be sectivores, the ventral petrosal sinus of Ne- seen in Novacek's ®gure 4 (1985), which con- sophontes and Solenodon is exposed ventral- veniently groups ®gures (after Kampen, 1905) ly, owing to the failure of the petrosal and of transverse sections of the mid-tympanic re- basioccipital to form a suture beneath the si- gion in several mammals. nus immediately anterior to the jugular (pos- Lavocat and Parent (1985: 338) stated, terior lacerate) foramen. This exposure ex- ``The most primitive rodent auditory region tends forward from the jugular foramen, with presently known by us is that of the Theri- which it is continuous, approximately to the domorpha.'' They (ibid.: 336) proposed that level of the basioccipital-basisphenoid su- the paramyines, in contrast to Sciuravus and ture.'' The anterior extent of the exposure is the Theridomorpha, have a derived character not so great in paramyines; the petrosal and that bars them from ancestry: ``In Sciuravus basioccipital seem to be in continuous con- and Theridomorpha, the auditory region is tact in Sciuravus, and the ventral petrosal si- located at the very posterior part of the skull, 2000 WAHLERT: AUDITORY REGION IN PARAMYS COPEI 13 and the stapedius muscle, according to the creates a deep channel for the transpromon- shape of its insertion, was extending out of torial continuation of the internal carotid ar- the tympanic cavity onto the surface of the tery (Lavocat, 1967: pls. 1, 2; Parent, 1980: posterior region of the skull, in the same pl. 3, ®g. 4; Lavocat and Parent, 1985: ®g. manner as in many Recent , 1a). Lavocat and Parent did not note the de- some archaic Insectivora (Echinosorex, Eri- rived features in theridomyids that are lack- naceus, Tenrec), or Tupaia. In contrast, Par- ing in Paramys and Sciuravus but that can amyinae and Franimys show a very long hor- be seen in Lavocat's (1967) ®gures. First, a izontal mastoid exposed on the ventral part partition anterior to the foramen vestibuli ex- of the skull. This mastoid separates the tym- tends from the promontorium to the medial panic cavity from the posterior part of the side of the lateral petrosal ¯ange; the parti- skull, while the pit of insertion for the sta- tion marks the head of the fossa for origin of pedius muscle, being more vertical, results in the tensor tympani. Such a partition was il- the fact that this muscle was certainly re- lustrated by Carrasco and Wahlert (1999) in stricted to the inside of the bulla.'' Cricetops dormitor, but it is lacking in par- It is possible to assess the condition of ex- amyines and Sciuravus. Second, Lavocat posure of the stapedius muscle origin in Sci- (1967: 492) stated that the epitympanic re- uravus, because two specimens (USNM cess ``constitue une fosse ovale assez profon- 22477 and AMNH 12531) preserve the bul- de,'' and Lavocat and Parent (1985: 334) la, which is an expanded U with the base said the recess is ``dorsally excavated and en- anteromedial. Since the bulla is small and not larged.'' This morphology, shown in Lovo- in¯ated, the area around the stylomastoid fo- cat's (1967) ®gures of adelomyine therido- ramen is uncovered, and presumably the or- myids, is derived relative to the morphology igin of the stapedius muscle was uncovered of Paramys and Sciuravus. However, Lavo- also. However, auditory bullae have not been cat (ibid.: ®gs. 1 and 2) presented a stereo preserved in Paramys, and the character can- photograph of the ear region of a therido- not be properly assessed. Large bullae are myid of unknown species. In it one can see preserved in Reithroparamys, and here the a shallow epitympanic recess as in Paramys region is covered ventrally (Meng, 1990), al- and Sciuravus. though the posterior ends of the bullae are The relationship of theridomyids to other missing. Paramys specimens, with well-pre- rodents is unclear. Vianey-Liaud (1985: 293± served auditory regions, are not as dorsoven- 294) summarized the data supporting descent trally ¯attened (distorted) as are Sciuravus from either the primitive Asiatic ctenodac- specimens. The character described by Lav- tyloids or the European ischyromyid stock. ocat and Parent may be one of posterior com- Martin (1999) showed that all of the early pleteness of specimens, extent of the audi- theridomyids have primitive pauciserial tory bulla, and postdepositional distortion enamel as do the early ctenodactyloids and rather than of morphology. It is not clear ischyromyoids; they could be descended whether they supposed the muscle origin to from either stock. He de®ned and described have reached into the shallower posterome- pseudo-multiserial enamel in Oligocene ther- dial extension from the basin. This would idomyids that is different from multiserial have changed the length and direction of the enamel in later ctenodactyloids and their de- muscle. scendents and thus is not evidence of rela- The question of which rodent is the most tionship. The morphology of the auditory re- primitive is not simple. The theridomyid and gion of theridomyids is similar to that of par- adelomyine specimens are from the Phos- amyines and Sciuravus, but the bridge from phorites of Quercy, which range in age from the promontorium to the lateral petrosal the late Eocene Robiacian into the Oligo- ¯ange, the enlarged epitympanic recess in cene. They are younger than the examined some taxa, and the relatively large bulla are specimens of paramyines and Sciuravus. derived relative to the North American spec- Paramys, the Theridomorpha, and to a lesser imens. Knowledge of the ear region in early degree Sciuravus, share a strong ridge lead- ctenodactyloids may help solve the problem ing anteriorly from the promontorium that of theridomyids origins. 14 AMERICAN MUSEUM NOVITATES NO. 3307

In some features Sciuravus is more prim- specimens and provided a congenial place to itive than paramyines. Sciuravus lacks the carry out the research. Chester Tarka and lateral swelling of the basioccipital, and the Lorraine Meeker gave me helpful guidance ridge on the promontorium alongside the in- in preparation of the illustrations. ternal carotid artery is not as pronounced. However, Sciuravus possesses two characters REFERENCES that are derived relative to paramyines. There is no space between the basioccipital and pe- Beer, G. R. de 1985. The development of the vertebrate trosal for a horizontal ventral petrosal sinus. skull. Chicago: Univ. Chicago Press The distal part of the stapedial artery is not 554 pp., 143 pls. [Reprint, originally exposed; the artery and the facial nerve enter published 1971. Oxford: Clarendon a common foramen and immediately diverge. Press] The morphological similarity of the audi- Carrasco, M. A., and J. H. Wahlert tory region in Paramys copei to the outgroup 1999. The cranial anatomy of Cricetops taxa reveals its primitiveness. The few de- dormitor, an Oligocene fossil rodent rived characters mostly distinguish the genus from Mongolia. Am. Mus. Novitates as a rodent. In future comparative anatomical 3275: 14 pp. studies of rodent auditory regions, Paramys Cope, E. D. 1884. The Vertebrata of the Tertiary forma- copei may be used as a standard of primi- tions of the West, Book I. Rept. U.S. tiveness. The fact that Paramys, Sciuravus, Geol. Surv. Terr. 3 (F. V. Hayden, U.S. and theridomyids are all relatively primitive Geol.-in-charge). Washington, D.C.: makes derived features shared among later Gov. Print. Off. 1009 pp., 75 pls. rodents strong evidence of intraordinal rela- Dawson, M. R. tionship. Derived character states shared with 1961. The skull of Sciuravus nitidus, a middle nonrodents are thus independently derived Eocene rodent. Postilla 53: 1±13. and examples of homoplasy. Gregory, W. K. 1951. Evolution emerging. New York: Mac- millan, v. 1: 736 pp., v. 2: 1013 pp. ACKNOWLEDGMENTS Kampen, P. N. van 1905. Die Tympanalgegend des SaÈugetier- This work was originally prepared in con- schaÈdels. Sonderabdruck Morphol. sultation with Emily Oaks and was presented Jahrb. 34(3): 321±722. in preliminary form as a joint poster at the Klaauw, C. J. van der annual meeting of the Society of Vertebrate 1931. The auditory bulla in some fossil mam- Paleontology in New York City (Wahlert and mals. Bull. Am. Mus. Nat. Hist. 62: Oaks, 1996). I am especially grateful for the 352 pp. input of Dr. Oaks. Ross D. E. MacPhee and Korth, W. W. John R. Wible examined the specimens with 1984. Earliest Tertiary evolution and radia- me and shared their observations and deep tion of rodents in North America. Bull. Carnegie Mus. Nat. Hist. 24: 1±71. knowledge of morphology. Meng Jin read 1985. The rodents Pseudotomus and Quad- the ®nal draft of this manuscript and helped ratomus and the content of the tribe me improve clarity and eliminate errors. I ap- Manitshini (Paramyinae, Ischyromyi- preciate the information about marsupial cra- dae). J. Vertebr. Paleontol. 5: 139±152. nia given me by Darrin Lunde and Christo- 1994. The Tertiary record of rodents in North pher Norris. Marcelo Weksler translated America. New York: Plenum Press. Spanish for me. Malcolm McKenna and Su- 319 pp. san Bell were invaluable resources, as usual. Lavocat, R. Special thanks go to Christina Seiffert of the 1967. Observations sur la reÂgion auditive des Forschungsinstitut und Naturmuseum Senck- Rongeurs TheÂridomorphes. Colloq. Int. Cent. Natl. Rech. Sci. Probl. Actuels enberg for her interest in this work and en- PaleÂontol. 163: 491±501. couragement to publish it. The Department Lavocat, R., and J.-P. Parent of Mammalogy and the Department of Ver- 1985. Phylogenetic analysis of middle ear tebrate Paleontology at the American Muse- features in fossil and living rodents. In um of Natural History gave me access to W.P. Luckett and J.-L. Hertenberger 2000 WAHLERT: AUDITORY REGION IN PARAMYS COPEI 15

(eds.), Evolutionary relationships 1986. The skull of leptictid insectivorans and among rodents, a multidisciplinary higher-level classi®cation of eutherian analysis: 333±354. New York: Plenum mammals. Bull. Am. Mus. Nat. Hist. Press. 183: 1±112. Lay, D. M. Oaks, E. C. J. 1972. The anatomy, physiology, functional 1968. Structure and function of in¯ated middle signi®cance and evolution of special- ears of rodents. Ph.D. diss., Yale Univ., ized hearing organs of gerbilline ro- 1967. Ann Arbor, MI: Univ. Micro®lms, dents. J. Morphol. 138: 41±120. Inc. 411 pp., 99 ®gs., 13 tables. 1993. Anatomy of the heteromyid ear. In Parent, J.-P. H.H. Genoways and J.H. Brown (eds.), 1980. Recherches sur l'oreille moyenne des Biology of the Heteromyidae. Spec. rongeurs actuels et fossiles. MeÂm. Trav. Publ. Am. Soc. Mammal. 10: 270±290. Ecole Practique Hautes Etudes, Inst. Li, C.-k. Montpellier 11: 286 pp., 16 pls. 1977. Paleocene eurymyloids (Anagalida, Pye, A. Mammalia) of Qianshan, Anhui. Ver- 1965. tebr. PalAsiat. 15: 103±118, ®gs 1±2. The auditory apparatus of the Hetero- Li, C.-k., J.-j. Zheng, and S.-y. Ting myidae. J. Anat. 99: 161±174. 1989. The skull of Cocomys lingchaensis, an Rougier, G. W., J. R. Wible, and J. A. Hopson early Eocene ctenodactyloid rodent of 1992. Reconstruction of the cranial vessels in Asia. In C.C. Black and M.R. Dawson the early Cretaceous mammal Vinceles- (eds.), Papers on fossil rodents in honor tes neuquenianus: implications for the of Albert Elmer Wood. Nat. Hist. Mus. evolution of the mammalian cranial Los Angeles Cty. Sci. Ser. 33: 179±192. vascular system. J. Vertebr. Paleontol. MacPhee, R. D. E. 12: 188±216. 1981. Auditory regions of Primates and eu- 1996. Basicranial anatomy of Priacodon frui- therian insectivores. Contrib. Primatol. taensis (Triconodontidae, Mammalia) 18: 82 pp. from the late of Colorado, and Martin, T. a reappraisal of mammaliaform inter- 1999. Evolution of incisor enamel micro- relationships. Am. Mus. Novitates structure in Theridomyidae (Rodentia). 3183: 38 pp. J. Vertebr. Paleontol. 19: 550±565. Rowe, T. McDowell, S. B., Jr. 1988. De®nition, diagnosis, and origin of 1958. The Greater Antillean insectivores. Bull. Mammalia. J. Vertebr. Paleontol. 8: Am. Mus. Nat. Hist. 115: 113±214. 241±264. McKenna, M. C., and S. K. Bell Schrenk, F. 1997. Classi®cation of mammals above the 1989. Zur SchaÈdelentwicklung von Ctenodac- species level. New York: Columbia tylus gundi (Rothmann 1776) (Mam- Univ. Press. 631 pp. malia: Rodentia). Cour. Forschungsinst. McKenna, M. C., Z. Kielan-Jaworowska, and J. Senckenb. 108: 1±241. Meng Segall, W. 2000. Earliest eutherian mammal skull from 1971. The auditory region (ossicles, sinuses) the late Cretaceous (Coniacian) of Uz- in gliding mammals and selected rep- bekistan. Acta Palaeontol. Pol. 45: 1±54. Meng, J. resentatives of non-gliding genera. 1990. The auditory region of Reithroparamys Fieldiana Zool. 58: 27±59. delicatissimus (Mammalia, Rodentia) Story, E. H. and its systematic implications. Am. 1951. The carotid arteries in the Procyonidae. Mus. Novitates 2972: 35 pp. Fieldiana Zool. 32: 477±557, ®gs 82±98. Moore, W. J. Vianey-Liaud, M. 1981. The mammalian skull. Cambridge: 1985. Possible evolutionary relationships Cambridge Univ. Press. 369 pp. among Eocene and Lower Oligocene Novacek, M. J. rodents of Asia, Europe and North 1985. Cranial evidence for rodent af®nities. In America. In W.P. Luckett and J.-L. Har- W.P. Luckett and J.-L. Hartenberger tenberger (eds.), Evolutionary relation- (eds), Evolutionary relationships among ships among rodents: a multidisciplin- rodents: a multidisciplinary analysis: ary analysis: 277±309. New York: Ple- 59±81. New York: Plenum Press. num Press. 16 AMERICAN MUSEUM NOVITATES NO. 3307

Wahlert, J. H. 1984. The specialized auditory system of 1974. The cranial foramina of protrogomor- kangaroo rats. Contrib. Sens. Physiol. phous rodents; an anatomical and phy- 8: 161±196. logenetic study. Bull. Mus. Comp. Wible, J. R. Zool. 146: 363±410. 1983. The internal carotid artery in early eu- 1977. Cranial foramina and relationships of therians. Acta Palaeontol. Pol. 28: 281± Eutypomys (Rodentia, Eutypomyidae). 293. Am. Mus. Novitates 2626: 8 pp. 1984. The ontogeny and phylogeny of the 1978. Cranial foramina and relationships of mammalian cranial arterial pattern. the Eomyoidea (Rodentia, Geomorpha). Ph.D. diss., Duke Univ. Ann Arbor, Skull and upper teeth of Kansasimys. MI: Univ. Micro®lms Int. 705 pp. Am. Mus. Novitates 2645: 16 pp. 1990. Petrosals of late Cretaceous marsupials 1983. Relationships of the Florentiamyidae from North America and a cladistic (Rodentia, Geomyoidea) based on cra- analysis of the petrosal in therian mam- nial and dental morphology. Am. Mus. mals. J. Vertebr. Paleontol. 10: 183±205. Novitates 2769: 23 pp. Wible, J. R., and J. A. Hopson Wahlert, J. H., and E. C. Oaks 1993. Basicranial evidence for early mammal 1996. Primitive morphology of the middle ear phylogeny. In F.S. Szalay, M.J. Nova- in some rodents. Abstr. Pap. J. Vertbr. Pa- cek, and M.C. McKenna (eds.), Mam- leontol. 16(suppl. to no. 3): 70A±71A. mal phylogeny: Mesozoic differentia- Wahlert, J. H., S. L. Sawitzke, and M. E. Holden tion, multituberculates, monotremes, 1993. Cranial anatomy and relationships of early therians, and marsupials: 45±62. dormice (Rodentia, Myoxidae). Am. New York: Springer. Mus. Novitates 3061: 32 pp. Wible, J. R., G. W. Rougier, M. J. Novacek, M. Webster, D. B. C. McKenna, and D. Dashzeveg 1961. The ear apparatus of the kangaroo rat, 1995. Dipodomys. Am. J. Anat. 108: 123±148. A mammalian petrosal form the early 1962. A function of the enlarged middle-ear Cretaceous of Mongolia: implications cavities of the kangaroo rat, Dipodo- for the evolution of the ear region and mys. Physiol. Zool. 35: 248±255. mammaliamorph interrelationships. 1975. Auditory systems of Heteromyidae: Am. Mus. Novitates 3149: 19 pp. postnatal development of the ear of Di- Wible, J. R., G. W. Rougier, M. C. McKenna, and podomys merriami. J. Morphol. 146: M. J. Novacek 377±393. 1997. Earliest eutherian ear region: A petrosal Webster, D. B., and M. Webster of ?Prokennalestes from the early Cre- 1971. Adaptive value of hearing and vision in taceous of Khoobur, Mongolia. [ab- kangaroo rat predator avoidance. Brain stract] J. Vertebr. Paleontol. 17(3): 84A. Behav. Evol. 4: 310±322. Wood, A. E. 1975. Auditory systems of Heteromyidae: 1962. The early Tertiary rodents of the Fam- functional morphology and evolution ily Paramyidae. Trans. Am. Philos. of the middle ear. J. Morphol. 146: Soc., n. ser. 52(1): 1±261. 343±376. Woodburne, M. O. (ed.) 1977. Auditory systems of Heteromyidae: co- 1987. Cenozoic mammals of North America: chlear diversity. J. Morphol. 152: 153± geochronology and biostratigraphy. 170. Berkeley: Univ. California Press. 336 pp.

Recent issues of the Novitates may be purchased from the Museum. Lists of back issues of the Novitates and Bulletin published during the last ®ve years are available at World Wide Web site http://nimidi.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).