Zhelestidae

Ekdale, Eric G., Archibald, J. David & Averianov, Alexander O., 2004, Petrosal bones of placental mammals from the Late Cretaceous of Uzbekistan, Acta Palaeontologica Polonica 49 (1), pp. 161-176 : 165-168

“ Zhelestidae ” (type I)

The majority of the “zhelestid” material that has been recovered from the Late Cretaceous consists of teeth. Isolated petrosal bones were assigned to “ Zhelestidae ” based on size and morphology (see above). Although aspects of the “zhelestid” petrosal were used in earlier phylogenetic analyses (Archibald et al. 2001),a detailed description of the struc − ture of the “zhelestid” petrosal has not been published.

Cerebellar surface.—The cerebellar or dorsal (endocranial) surface of the petrosal is marked by two large,distinct openings ( Fig. 2A ). The anteromedial of the two openings,the ovoid in − ternal auditory meatus,is clearly divided by the crista transversa into the foramen acusticum inferius medially and the foramen acusticum superius laterally. The subarcuate fossa is the second distinct opening. In “zhelestids”,this depression is circular in shape and quite deep.

Medial to the subarcuate fossa is the aquaeductus vestibuli. This small opening is directed posterodorsally,and it is followed by a small sulcus traveling towards the caudal border of the petrosal bone. The opening for the aquaeductus is best preserved in ZIN C. 85511,and it appears as a small and circular hole,unlike the slit−like opening observed in other eutherians including archaic ungulates,but rather like the aquaeductus vestibuli in Prokennalestes (PSS−MAE 136; Wible et al. 2001).

Two additional conspicuous openings are found lateral to the internal auditory meatus and anterolateral to the subarcuate fossa in “ Zhelestidae ”. The anterior of the two openings is an anteriorly directed circular opening known as the fenestra semilunaris. It can be observed in ZIN C. 82587 (as well as in URBAC 99−41,ZIN C. 855111,and 85512 to a lesser extent) that a very small foramen exists on the anterior wall of the fenestra semilunaris. This foramen appears to be the endocranial opening of the aquaeductus Fallopii. The aquaeductus Fallopii of “ Zhelestidae ” is a small,anteriorly directed canal that ultimately opens into the tympanic cavity via the hiatus Fallopii (see Fig. 2C ).

The second of the two openings found on the lateral edge of the cerebellar surface of the petrosal is immediately posterior to the fenestra semilunaris ( Fig. 2A ). It is significantly larger than the fenestra semilunaris,and it opens postero − dorsally. This structure is positioned similarly to and appears very much like the endocranial opening of the prootic canal in the Late Cretaceous eutherian Prokennalestes (PSS−MAE 136; Wible et al. 2001). The canal is termed the prootic canal in “ Zhelestidae ” following Wible and Hopson (1995) because the canal most likely transmitted the prootic sinus in life (see vascular reconstructions below for further discussion of this matter). Endocranially,a very broad and distinct sulcus for the prootic sinus can be seen running into the opening of the canal at its posteromedial aspect. The prootic sinus was the anterior distributor of the transverse sinus in “ Zhelestidae ”,and it is found anterolateral to the subarcuate fossa,as it is in eu − therians including Prokennalestes ,rather than posterolateral to the fossa,as it is in metatherians such as Deltatheridium and Didelphis ( Rougier et al 1998; Wible et al. 2001).

A sulcus for the inferior petrosal sinus can be seen traversing the medial edge of the cerebellar side of the petrosal bone ( Fig. 2A ). The sulcus runs posteriorly from the rostral apex of the petrosal to a position posterior to the internal auditory meatus. Immediately posterior to the posterior terminus of the sulcus for the inferior petrosal sinus is a small,cir − cular,medially oriented opening for the cochlear canali − cularis.

In URBAC 99−41,a sulcus can be seen skirting the posteromedial and posterior borders of the subarcuate fossa. This sulcus is not continuous with the sulcus for the inferior petrosal sinus,and it courses towards,but does not join with, the sulcus for the prootic sinus. At its medial aspect,it comes in contact with the posterior lacerate foramen. This is most likely the sulcus for the sigmoid sinus.

Squamosal surface.—The tegmen tympani,a sheet of bone that separates the middle ear cavity from the cranial cavity at the anterolateral aspect of the petrosal bone ( Fig. 2C ),is not inflated in “ Zhelestidae ” as it is in many ungulates,but the “zhelestid” petrosal does possess a distinct flange of bone at the posterolateral corner of the bone ( Fig. 2B,C ). Given the proximity of a similar flange in Prokennalestes ( Wible et al. 2001) ,this pro − cess is termed the lateral flange for “zhelestids”.

A large opening leading to a canal is found at the anterior aspect of the lateral flange. This structure is best preserved in URBAC 99−41,but it can be seen in the other “zhelestid” specimens to a lesser extent. Just inside this opening,on the medial wall of the canal,is a small foramen leading to a me − dially oriented duct. This short and horizontal duct,which opens into the tympanic cavity (see Figs. 2C , 4A,B ),is the ventral aspect of the prootic canal,and it is homologous to the prootic canal in metatherians,including the American opossum, Didelphis virginiana . Dorsal to this horizontal duct is the dorsal aspect of the prootic canal,which opens on the cerebellar surface of the petrosal ( Fig. 2A ). The dorsal aspect of the prootic canal appears homologous to the prootic canal of earlier mammals,including Prokennalestes ( Wible et al. 2001) . The canal extending posteroventrally from the dorsal and ventral aspects of the prootic canal (eventually opening on the squamosal surface of the petrosal bone) is termed the?postglenoid vein canal ( Fig. 2 ) after the vessel that occupied it (see venous reconstructions below).

A deep and distinct sulcus can be seen traveling posteriorly from the opening of the?postglenoid vein canal on the squamosal surface of the petrosal ( Fig. 2B,C ). When the squamosal and petrosal bones were articulated,this sulcus formed the medial wall of a canal known as the posttemporal canal. The petrosal contribution to the posttemporal canal can be seen curving medially along the lateral surface of the petrosal,and onto the lambdoidal (posterior) surface of the bone, ending near the posterior lacerate foramen.

Just posterior to the squamosal opening of the?postglenoid vein canal is a second groove running dorsally perpendicular to the posttemporal canal ( Fig. 2B ). It would form a short bony tube where the petrosal and squamosal bones articulate. The location of this groove is similar to the ascending canal for the superior ramus of the stapedial artery seen in Prokennalestes and other eutherian mammals (Rougier at al. 1992; Wible et al. 2001),and this sulcus is most likely the petrosal contribution of the ascending canal for the superior ramus of the stapedial artery in “ Zhelestidae ” as well. The ventralmost end of this sulcus is confluent with the posttemporal canal traveling posteriorly from the?postglenoid vein canal.

Tympanic surface.—The tympanic (ventral) surface of the petrosal is marked anteriorly by the semiovoid bulge of the promontorium ( Fig. 2C ). The mammalian promontorium houses the cochlea,which includes the organ of hearing. In “ Zhelestidae ”,the promontorium is round and rather bul − bous. It is almond shaped,and the rostral apex appears to be pointed. Lateral to the promontorium is the tegmen tympani, which roofs the middle ear cavity.

In one poorly preserved “zhelestid” specimen (ZIN C. 85514; identified as “ Zhelestidae ” based on the presence of the prootic canal),the ventral aspect of the promontorium has been broken away,conveniently revealing the internal struc − ture of the petrosal ( Fig. 2D ). It is clear that the cochlea of “ Zhelestidae ” was coiled at least 270°,and although this is the extent of the coiling that is visible,the cochlear canal is broken along its path in this specimen,and it most likely completed nearly one full 360° turn.

The promontorium is penetrated by two distinct openings along its posterolateral and posteromedial aspects ( Fig. 2C ). On the posteromedial aspect of the promontorium is the fenestra cochleae,which is also known as the round window given its round shape in most mammals. In some “zhelestid” specimens,especially in URBAC 00−26,it appears as though there is a sulcus traveling up the medial edge of the promontorium from the fenestra cochleae. It is unlikely that this is an actual sulcus,however. Rather,the petrosal expands along its medial border creating a flattened area of bone. This “groove” appears to be associated with the flattening of this petrosal expansion.

In several of the specimens,including URBAC 99−73, ZIN C. 82587 and 85512,there is an anteriorly directed pit just posterodorsal to the fenestra cochleae ( Fig. 2E ). This excavation most likely accommodated the cavum tympani,an air−filled sac that filled the middle ear space in life. The pit is poorly developed (but nonetheless present) in ZIN C. 82587, but it is very well developed in both URBAC 99−73 and ZIN C. 85512. This excavation for the cavum tympani is obscured from view by the posterior wall of the petrosal in ZIN C. 85512,but the lambdoidal surface of URBAC 99−73 is bro − ken giving a clear view of both the fenestra cochleae and the excavation for the cavum tympani. In this specimen,it ap − pears that this excavation is in fact a foramen,but this is al − most certainly caused by damage to the bone. Whether the excavation is a pit or a foramen is indeterminable in ZIN C. 85512,but it is clearly an anteriorly directed pit,and not a foramen, in ZIN C. 82587.

The second opening on the promontorium is the fenestra vestibuli or oval window given its oval shape in most eutherian mammals. It is found on the posterolateral aspect of the promontorium ( Fig. 2C ),and it accommodates the footplate of the stapes. It does not appear that a sulcus for the stapedial branch of the internal carotid artery is present in “ Zhelestidae ”,as neither side of the fenestra vestibuli is notched in any of the specimens. The fenestra vestibuli is oval in most of the “zhelestid” specimens (especially URBAC 02−14),but it appears to be rounded in URBAC 99−73,and even more so in ZIN C. 85511. The average stapedial ratio for “ Zhelestidae ” is 1.6 (the range being 1.2–2.0 and the standard deviation is 0.3). Lateral to the fenestra vestibuli,and ventral to the tegmen tympani,is an area known as the epitympanic recess ( Fig. 2C ). In life,the auditory ossicles articulate within this region of the middle ear.

A distinct and broad sulcus for the inferior ramus of the stapedial artery can be seen running anteriorly just lateral to the promontorium. There is no indication of a transpromontorial sulcus for the internal carotid artery in any of the “zhelestid” petrosals.

Anterolateral to the fenestra vestibuli is the tympanic aperture of the facial canal ( Fig. 2C ). The facial canal itself runs ventrolaterally from its cerebellar entrance (the foramen acusticum superius) to the fenestra semilunaris before opening in a posterior direction on the tympanic surface of the petrosal. The position of the hiatus Fallopii is not clear in all of the “zhelestid” petrosals,and there may be some variation. In some specimens,including URBAC 99−41, ZIN C. 85511,and 82587,the hiatus can clearly be seen opening anterior to the facial canal on the tympanic surface of the petrosal,rather than the lateral edge. A faint sulcus can be seen traveling anteriorly,especially in ZIN C.

85511. The bony sheet flooring the facial canal is damaged in URBAC 99−02,00−26,and ZIN C. 85512 obscuring the position of the hiatus. In URBAC 99−73,it almost appears as if the hiatus Fallopii is incomplete within the petrosal (only a notch that opens laterally is preserved). The foramen may be completed in a fully articulated skull (as is most likely the case with Prokennalestes ; Wible et al. 2001),or this may simply be the result of damage and not the true nature of the structure.

Near the tympanic aperture of the facial canal is the tympanic opening of the prootic canal. The exact position of the tympanic opening of the prootic canal in relation to the tympanic aperture of the facial canal may be variable in “ Zhelestidae ”. In most of the petrosal specimens preserving this opening (URBAC 00−26 and ZIN C. 85512 are too damaged to determine the position of the opening),including URBAC 99−41,99−73,and ZIN C. 82587,the prootic canal opens an − terior and ventral to the lateralmost aspect of the tympanic aperture of the facial canal. The canal appears to open posterior to the facial canal in URBAC 99−02 and ZIN C. 85511. URBAC 99−02 is damaged along its lateral edge exposing the paths of the prootic and?postglenoid vein canals,how − ever,so an exact determination of the position of the prootic canal opening cannot be made,but in ZIN C. 85511,there is a distinct opening just posterior to the lateral aspect of the facial canal. This may not be the opening to the prootic canal, however,as the same specimen possesses a small foramen anterior to the tympanic aperture of the facial canal in the same location as the tympanic opening of the prootic canal in the other “zhelestid” specimens. The purpose of this foramen posterolateral to the facial canal,if a real foramen at all,is unclear.

Figures

Fig. 2. Stereophotographs of the petrosal of “Zhelestidae”. A. Cerebellar view of URBAC 99−41. B. Squamosal view of URBAC 99−41. C. Tympanic view of URBAC 99−41. D. Tympanic view of ZIN C. 85514 with internal structures exposed. E. Lambdoidal view of URBAC 99−73. Scale bar 1 mm. Anterior towards the top in A–D; tympanic towards top in E.

Fig. 4. Soft tissue reconstructions. A. Tympanic view of “Zhelestidae”. B. Squamosal view of “Zhelestidae”. C. Tympanic view of Prokennalestes (modified from Wible et al. 2001). D. Squamosal view of Prokennalestes (modified from Wible et al. 2001). E. Tympanic view of Didelphis virginiana (modified from Wible 1990). F. Squamosal view of Didelphis virginiana (modified from Wible 1990). G. Tympanic view of Kulbeckia kulbecke. H. Squamosal view of Kulbeckia kulbecke. Anterior towards the top in A, C, E, and G. Anterior towards the right in B, D, F, and H.