ON THE TAYASSUID AFFINITIES OF XENOHYUSGINSBURG , 1980 , AND THE DESCRIPTION OF NEW FOSSILS FROM SPAIN *

In 1980, Ginsburg created the new genus Xenobyus for a large suiform artiodactyl which he considered to belong to the family Suidae. Among the distinguishing characters of the genus, Ginsburg cited the enlarged, strongly curved and inflated central upper incisors possessing a distal accessory cusplet, the shortened muzzle, the enlarged lower fllSt and second incisors, and the c10se packing of the entire tootb row (i.e. reduced to absent diastemata). These and other characters, such as the vertically implanted mandibular canines, tbe sympbyseal morpbology, and the shape of the third upper incisor, suggest that Xenobyus belongs instead to the family Tayassuidae, subfamily Doliochoerinae. Wbereas Ginsburg suggested that Xenobyus had an unknown ancestry and that it represented an immigrant into Europe at about the beginning of wne MN2b, it now seems more likely tbat it represents a late doliochoere of the sort represented by DoJiocboerus quercyi which occurs in late Stampian deposits (Ginsburg, 1974). In many ways it is merely an enlarged version of tbis species, but tbere are sufficient morphological difIerences to warrant retention of tbe genus Xenobyus. Viewed witbin a tayassuid framework, Xenobyus is not seen to be unusual and not to run counter to evolutionary trends in tbe Suidae, the latter view being a necessity following its identifiation as a suid. Sorne fundamental similarities between tbe molars of Xenobyus and Kenyapotamus could provide a link between late doliochoeres and early hippopotamids, an hypothesis already mentioned by Pickford (1983). Newly discovered fossils from Loranca, Cuenca, Spain belong to two different species of Xenobyus, X venitor and an undescribed larger species.

• Este trabajo ha sido posible gracias a las subvenciones de la Junta de Comunidades de Castilla-La Mancha para las excavaciones del yacimiento de Loranca del Campo (Cuenca).This reassignment of the genus naturaUy affec18 the discussions made by Ginsburg regarding the palaeoethology and origíns of Xenohyus.

New fossils from Spain
Collections of fossils from Loranca, Cuenca Province, during the past four years contain six specimens of Xenohyus.Three of these fall comfortably within the range of variation of X. venitoT but three fossils faH well ouside this range.These latter specimens are appreciably larger than their homologues in X. venitoT (Table 1) and may well represent a separate species.Morphologically, however, they are typical of the French species.
These tbree fossils are associated with a very rich fauna containing other large mammals as well as micromammals whose age is Ramblian, zone Z, more or less equivalent to MN3a of Mein's Biozonation (Ginsburg, et al, 1987).
The three smaller specimens were found in surface scatters of fossils at abaut 100 metres distance from where the series of larger teeth were found in situ.
Their age could be Ramblian, but an age more recent canoot be excluded because relatively clase to the area where they were found also exis18 a Middle Aragonian locality.Thus, the fossils found on the surface could have two different origins.

Description
Right upper central incisor lacking the root and a portion of the crown on the mesial edge: This tooth is moderatly worn, but shows the typical inflated and strongly curved external surface, the 10w lingual ciñgulum and the distal accessory cusp.The course of the central lingual ridge is shown by a curve in the outline of the wear facet near the lingual cingulum.
Left upper first or second molar: This unwom tooth crown lacks the roo18 and a small portion of the distolingual enamel near the cervix.The four principal cusps are bunodont and well separated from each other by grooves.There are anterior, median and posterior accessory cusps, those on the anterior and posterior par18 of the crown being incorporated into the anterior and posterior cingula.The median accessory cusplet is clase to the hypocone.The buceal cingulum is large.
Left lower third molar: This unworn rootless crown was found close to the upper molar described immediately aboye, and considering their compatible stage of wear, the two teeth might represent a single individual.There are four principal cusps arranged in two transverse pairs, and a fifth talonid cusp at the rearo There is a distinct median accessory cusp clase to the entoconid.The bypoconulid lies between the hypoconid and the distal talonid cusp and is bordered lingually and labially by cingula.Anteriorly there is a large cingulum which extends onto tbe front part of the labial surface of the protoconid.
These tbree fossils clasely resemble their homologues in X. venitoT, both in their morphology and in size.

Description
Rigbt upper central incisor: This tooth lacks mast of the root and the distal accessory cusplet is damaged.The crown is lightly wom, and is in every res- • pea except for size, similar to tbose of X. venitor.
The wear facet is principally al tbe apex and down tbe distal marginal ridge towards tbe distal accessory cusplet, as in material from France.
Left upper third premolar, unworn, lacking the roots: This specimen is c10sely similar, except for its larger size, to {bose from france.The only significan{ observable difference resides in tbe fact that the crest running from the apex of tbe principal cusp to the rear cingulum has two beaels on it.TItese beaels are very small, and with light wear would be unobservable.
TIte tooth lacks an external cingulum, ahhougb tbere is a slight swelling in the labial surface of the crown wbere a cingulum would normally appear.The distolingual cusplet and cingula are comparable in sbape and position to those oC X venitor.
Wom right third molar in mandible fragment: This wom tooth has httle detailed morphology remaining, but its general shape, and {he presence oC a cingulum on lhe front portion of tbe protoconid indicates tbat it belongs to the same group as tbe upper central incisor and the third premolar described aboye.
The talonid is simple, and has labial and lingual cingula leading forwarels Crom it towarels the bypoconid and entonocid.Between these two cingula lies the bypoconulid.Tbe median accessory cusplet is c10sely attached to the entoconid, and is difficult to disting uish in this worn and cracked specimen.1959).The main difference is one oC size.
The upper second incisor illustrated by Ginsburg shows the typical morphology oC Doliochoerus and also retains the sire proportions relative to the central incisors (i.e. the second incisor is considerable smaller than the central incisor).
Upper premolars: The third premolar oC Xenohyus closely resembles its homologue in D. quercyi apart Crom its larger size.Ginsburg reports that most individuals oC X. venitor possess labial cingula, but tbis Ceature seems to be variable, the spanish specimens having only an incomplete labial cingulum, as in a specimen from France.
The upper Courth premolars oC X. venitor are closely similar to those oC D. quercyi except Cor their larger sire and more marked labial cingulum.In the position oC the three main cusps (two labial, one lingual) and the degree oC development oC the anterior and posterior cingula, and in crown height, the upper fourth premolars oC the two genera are difficult to distinguish Crom each other.
Upper molars: The rather bunodont, simple upper molars oC X. venitor, resemble those oC D. quercyi in a number oC Ceatures including the proportions oC the principal cusps, the positions and size relations of the anterior, median and posterior accessory cusplets and the strength of the anterior and posterior cingula.The only major difference in the available specimens, is that the labial cingulum in the upper molars oC Xenobyus are more inflated and more complete than they are in Doliochoerus.Furthermore, the height of the cingula in relation to total crown height is similar in the two genera, and it should be pointed out that in this morphology the doliochoeres resemble the primitive hippopotamid Kenyapotamus described by Pickford (1983).
Lower incisors: The two central incisors of Xenohyus are enlarged, not only relative to the cheek teeth but also in relation to the third incisor.They have strongly developed central lingual ridges, and are bordered mesia11y and dista11y by marginal ridges.Their procumbent orientation in the symphysis and their position close to the canine, as was noted by Ginsburg, is another feature by which X. venitor resem-bIes D. quercyi.
The third lower incisor oC Xenobyus is a sma11 tooth with a mesiodistally elongated crown, in which the distal part projects over the sloping root, precisely as in Doliocboerus.
Lower Canine: The position oC the canine close to the incisors in Cront and the premolars behind, together with its vertical orientation within the jaw is a Curther indication oC tayassuid rather than suid affinities Cor Xenohyus.seaux, 1959), their gradual increase in sire and complexity of crown morphology from mesial to distal is also typical of doliochoeres.Lower Molars: Although at fírst glance the lower molars look like those of suids, there is nothing in their morphology to exclude them from being tayassuids.If anything, the clase contact between the entoconid and the.median accesory cusplet is more like the situation in Old World peccaries in general (Pickfard, 1978) than it is in suids, but it must be admitted that in their molar morphology there has been a certain amount of convergent evolution between peccaries and pigs, as epitomised by the striking similarity between the molars of Listriodon and Schizocboerus.
Mandible: Ginsburg (1980) already noted the shortness of the snout in Xenobyus.Comparison oC the symphyseal region with Doliochoerus reveals the fo11owing overa11 similarities: The symphyseal section, the shape, orientation and extent of the superior surface of the symphysis, its extension to the year as far as the second premolar, the shape of the geniohyoid fossae and the shape of the ventral surface, a11 recall Doliocboerus rather than suids.Furthermore, the absence of symphyseal splaying in the camine area is also a tayassuid character.
There are mental Coramina below the second and fourth premolars about half way down the depth of the mandible.In tbis Xenobyus is similar to D. quercyi.
Maxilla: Although the maxilla of Xenohyus is poorly known, the available specimen from Laugnac accords we11 with the tayassuid hypothesis.The zygoma are placed forwards over the second molar and the anterior part of the third molar, whereas in many suids, the zygoma are located further to the rear on the face of the maxilla.It is difficult to determine the distal extension of the palate in the sample of Xenohyus, but it seems reasonable to suggest, on the basis of specimen Lg 120, that there was a distal extension of the palate as in Schizochoerus as shown by Pickford (1978).

Discussion
It appears to us that it is likely that Xenohyus is an Old World doliochoerine tayassuid.Furthermore, it is probable that it evolved in Europe from an Oligocene doliochoere such as DoJiochoerus quercyi or something similar.The latter is of late Stampian distribution according to Ginsburg (1974) while Xenohyus is known from lower Miocene strata of France (MN2b and MN3b) and Spain (Ramblian, zone Z), although the possibility that it can also be found in Middle Aragonian levels cannot be excluded until the relationships of the different Loranca sites is firm1y established.
It was previously thought that Xenohyus became extinct before MN4, but the similarities between the molars of Xenohyus and Kenyapotamus may provide evidence that the lineage persisted after migrating to Africa in Faunal Set III (Pickford, 1981) where it gave rise to the hippopotamids.Clearly, however, a better sample of fossils would be needed in order to verify too possibility.
The detailed relationships of Xenohyus with respect to other Old World tayassuids such as Schizochoerus, Taucanamo and DoJiochoerus, must await the discovery of substantailly better fossils of Xenohyus.Den-taUy, the closest relationships are undoubtedly to DoJiochoerus.
•• Institut de Paléootologie.8, rue BufIoo, 75005, París.••• Museo Nacional de Ciencias Natl\fales.J. Gutiérrez Abascal, 2. 28006 Madrid.GinsbuIg placed the genus within the family Suidae, all his comparisons and commen18 of the genus being made with other suids.However, a re-examination of the material, including new specimens from Spain, leads us to propose a different hypothesis regarding the familial af'finities of the genus.All the distinctive characters mentioned by Ginsburg for defming the new genus, are, in our opinion, typically doliochoerine tayassuid features, which relate the genus to the Old World peccaries rather than to the pigs.
The canine in situ in the mandibu-Xenobyus are closely compatible with those of Doliocboerus except for their larger sire.Their position in a closed series close to the canine, the strongly developed «trigonid» in the fourth premolar (Descha-