The brontothere genus Palaeosyops, known from the Early Eocene of both North America and Asia, has a complex and extensive taxonomic history. Palaeosyops was one of the first brontothere genera to be named, by Joseph Leidy in 1870. Since Leidy's description of the type species P. paludosus, North American fossils now considered to belong to Palaeosyops have been referred to over twenty species, in several genera. In Asia, Palaeosyops is represented by the species P. dayi, originally named as a species of the related genus Eotitanops.
Modern revisions of the fossil material have recognized relatively few species as valid, but reassessments have not reached a consensus. The most recent revisions recognize three or five North American species of Palaeosyops as valid. The distinctions used to separate the species have also been questioned, and some have suggested that the North American fossils may represent only a single species.
Taxonomic issues are compounded by the lectotype specimen of P. paludosus, which is a probably non-diagnostic isolated tooth. P. paludosus and the genus Palaeosyops are generally maintained as valid, but may thus be nomina dubia, in which case the genus as currently understood would have to be treated under the next oldest available name, Limnohyus (currently considered a junior synonym).
P. paludosus, and type specimens Palaeosyops paludosus was described in 1870 by Joseph Leidy, based on four isolated teeth found at Church Buttes in the Bridger Formation of Wyoming. The teeth include three molars and one premolar (the four teeth are now catalogued as the three specimens USNM 758, 759, and 762), and were found by Ferdinand Vandeveer Hayden during his geological explorations of the Great Plains and the Rocky Mountains. Palaeosyops was the first hornless brontothere to be described, and one of the first brontothere genera to be described overall, just 24 years after the first brontothere fossils ("Menodus") were scientifically described. Leidy compared the teeth to Chalicotherium, Palaeotherium, and "Titanotherium", but failed to identify Palaeosyops as close to the (then fragmentarily known) brontotheres, partly because he erroneously believed brontotheres to be artiodactyls. Only in 1873 did Leidy realize that Palaeosyops was a perissodactyl, when he described a skull, teeth, and limb bones and noticed their resemblance to those of Palaeotherium and Tapirus.
The name Palaeosyops means "ancient boar face", derived from Ancient Greek παλαíο (ancient), σύς (boar), and ὤψ (face). Leidy gave no reason for the name. In 1929, Henry Fairfield Osborn interpreted the name as a reference to similarities Leidy noted between the molars of Palaeosyops and those of the anthracothere Hyopotamus (=Bothriodon). The species name paludosus means "marshy", or "dwelling in the marshes". Leidy gave no reason for the species name either; Osborn interpreted it as a possible reference to the Bridger Formation containing ancient lake basins.
In 1929, Osborn pointed out that the isolated teeth representing Leidy's syntypes may not necessarily all belong to the same individual. Osborn thus designated USNM 759, a second lower molar from the sample, as the lectotype of P. paludosus. USNM 759 is described first of the four teeth in Leidy's original description of the material. USNM 759 has generally been accepted as the valid lectotype since. Leidy's original teeth are of questionable diagnostic value. The lectotype being a lower molar is especially problematic, since lower molars in primitive brontotheres differ very little between species. As early as 1872, Othniel Charles Marsh remarked that the P. paludosus type material made it almost impossible to determine what fossils belonged to that species. The problematic type material has led to questions concerning the validity of P. paludosus, and thus also Palaeosyops as a genus, and has led some researchers to propose neotype specimens.
Osborn was the first to propose a neotype for P. paludosus, suggesting the well-preserved lower jaw AMNH 11680, from Millersville in the Bridger Formation. Osborn justified the jaw as a neotype by the "absolute similarity in form and size of the second inferior molar" to USNM 759. In 2000, Gregg F. Gunnell and Vicki L. Yarborough noted that they believe the lectotype tooth to be indeterminate, which would in turn make Palaeosyops a nomen dubium. To preserve the name, Gunnell and Yarborough proposed the UM 98890 as the neotype; this specimen includes a relatively well-preserved skull, jaw, and some postcranial elements. Gunnell and Yarborough noted that UM 98890 was found near Church Buttes, at the same stratigraphic interval as Leidy's original teeth, and that its second lower molar was the same size as USNM 759. Because the original lectotype is still available, neither proposed neotype holds any formal status or nomenclatural significance.
Despite the probably nondiagnostic lectotype, Palaeosyops has been maintained as a valid genus in modern revisions. The validity of Palaeosyops has been supported by the lectotype being from the Blacks Fork Member of the Bridger Formation, which has only reliably yielded fossils of what appears to be a single valid brontothere taxon, for which Palaeosyops paludosus would be the oldest available name. Should Palaeosyops be deemed to be invalid, the next oldest available genus name corresponding to the genus as currently understood would by Limnohyus, named by Marsh in 1872.
Descriptions of other species
Leidy's species Leidy went on to describe three more species of Palaeosyops from the Bridger Formation, all based on fragmentary fossils. In 1871, Leidy named the new species Palaeosyops major, based on a jaw fragment from Grizzly Buttes, near Fort Bridger. Leidy noted that the jaw appeared more robust than jaws he had assigned to P. paludosus, which he used to justify the new taxon. Leidy also noted that the new jaw appeared to have been swollen and altered due to some disease the animal had in life. In 1929, Osborn proposed neotype specimens for P. major, consisting of a skull (AMNH 12182) and a jaw (AMNH 12181). Like Osborn's proposed neotype for P. paludosus, these fossils have no formal status or nomenclatural significance.
In 1872, Leidy named the species Palaeosyops humilis, based on a single upper molar from "Valley of Dry Creek 40 miles from Fort Bridger". The P. humilis tooth is small, and Leidy assumed that it may be a deciduous tooth. Later in 1872, Leidy named the species Palaeosyops junius, based on an isolated molar and an isolated premolar, from the vicinity of Fort Bridger. P. junius was transferred to the genus Mesatirhinus by Osborn in 1929.
Marsh's species Othniel Charles Marsh named the new species Palaeosyops minor in 1871, based on a single isolated molar from Grizzly Buttes. Marsh noted that the tooth was similar to Leidy's Palaeosyops paludosus, which had been found in the same deposits, but that it was less than half the size of the corresponding teeth assigned to that species. Marsh did not figure the P. minor tooth. In 1929, Osborn did not believe Marsh's P. minor to belong to Palaeosyops, or a brontothere at all. Instead, Osborn referred it to Tillodontia, as "Anchippodus minor". In 2010, Mader contended that Marsh may have been correct in his original assessment that the tooth belonged to Palaeosyops, though Mader did not personally examine the tooth.
In 1872, Marsh described a first lower premolar, later identified as another Palaeosyops fossil, from Grizzly Buttes. The tooth was associated with fragmentary skull and postcranial material, but Marsh did not recognize the tooth as belonging to Palaeosyops, or to a perissodactyl at all. Due to the tooth's deceptive resemblance to the premolar of a dog, Marsh described the tooth as belonging to a new species in the genus Canis, Canis montanus. The name Canis montanus is preoccupied by Canis montanus (Pearson, 1836), a synonym of the extant hill fox. In 1899, Oliver Perry Hay erected the replacement name Canis marshii for Marsh's "Canis montanus".
Marsh named two more species in 1872, Palaeosyops laticeps and Limnohyus robustus. P. laticeps was based on a nearly complete skeleton found "near Marsh's Fork, about 15 miles from Fort Bridger". Marsh compared P. laticeps to Leidy's P. paludosus teeth, and noted that the P. laticeps skeleton, which belonged to an adult individual, appeared to be smaller, and that its teeth were much smoother than those of P. paludosus. Limnohyus robustus was based on several fossils, with a fragmentary skull selected as the type specimen. Marsh believed that two genera of primitive brontotheres were represented in the Bridger Formation, one with a hypocone (one of the cusps) on the third molar and one without it, and applied the name Palaeosyops to animals with a hypocone, and his new genus Limnohyus to those that lacked the hypocone. The name Limnohyus means "marshy boar" or "shore boar", derived from Ancient Greek Λíμνη (a marshy lake, or shore) and σύς (boar).
In 1872, Leidy pointed out that the original molars he had referred to P. paludosus included a third molar without a hypocone, and that Marsh could thus not use that trait to define Limnohyus. Limnohyus was thus recognized as a junior synonym of Palaeosyops. In 1890, Marsh used to same logic to instead move P. laticeps to a new genus, Limnohyops, which was distinguished from Palaeosyops by the presence of the hypocone. Limnohyops means "marshy boar face" or "shore boar face", adding the suffix ὤψ (face) to Limnohyus.
Cope's species In 1872 and 1873, Edward Drinker Cope described four new species of Palaeosyops and Limnohyus: P. vallidens, L. laevidens, L. fontinalis, and P. diaconus. P. vallidens was based on several fossils from the Washakie Basin, including a partial lower jaw and isolated teeth. Cope distinguished P. vallidens by its large size, larger than Leidy's P. major, and the presence of strong transverse ridges on molars and premolars. L. laevidens was based on several fossils from the Bridger Basin, including a relatively well-preserved skull. Cope identified L. laevidens as a larger form. Cope considered the molars of L. laevidens to be similar to those of L. robustus, but differing in that the second superior molar had only one outer tubercle, and that the cingula of L. laevidens were much less developed than in L. robustus. L. fontinalis was based on the fragmentary skull of a young individual, found on "on a bluff on Green River, near the mouth of the Big Sandy, Wyoming"; Cope found the characteristics of its last superior molar to be similar to P. paludosus. P. diaconus was based on a partial maxilla from "Henrys Fork of Green River, Wyoming". Cope described P. diaconus as of the same size as P. major, but differing in the proportions of the teeth.
Cope described a fifth species in 1880, Palaeosyops borealis, based on a maxilla fragment from the Wind River Basin in Wyoming. P. borealis was transferred to the new genus Eotitanops by Osborn in 1907. Osborn distinguished this genus from Palaeosyops mainly by its more elongated skull.
Osborn's species In 1889, Osborn named the new species Palaeosyops hyognathus, based on a mandible from the Washakie Basin. In 1929, Osborn transferred this species to the genus Dolichorhinus.