Morphological:

Institutional:

Chronological and Operational:

Merychippus insignis:

The following characterization is based on contrasting the Echo Quarry sample of Merychippus insignis with Parahippus leonensis and C. goorisi. As indicated in figure 13 of Hulbert and MacFadden (1991), C. goorisi [“M.” goorisi] is more derived than M. insignis and both are more derived than Parahippus leonensis. I utilize information from the lower dentition as derived from the Echo Quarry sample here allocated to M. insignis. As discussed below, this allocation is consistent with the morphology found in lower dentitions and mandibles associated with crania from Trinity River Pit 1, Texas, and Deep Creek, Nebraska.

As figured below, Merychippus insignis is a mesodont equid that differs from Parahippus leonensis in: larger cranial size, longer muzzle, nasal notch more deeply retracted (to about midway between C1 and P2, vs. about above C1), POB ca. 17 mm wide (vs. about 7 mm or less), DPOF distinctly expressed with dorsal, posterior, ventral, and (palpable) anterior borders (vs. less distinct); the DPOF is (slightly) pocketed posteriorly (vs. unpocketed) and is relatively short (ca. 64 mm vs. ca. 72 mm) in comparison to the cheek tooth row length (ca. 116 mm vs. 92 mm); the anterior end of the DPOF is located above P3 (rather than above P2); the posterior portion of the DPOF is short (ca. 15 mm) compared to the anterior part (ca. 50 mm), in contrast to the two parts being about equally long in P. leonensis; the vertical ridge that separates the two parts of the DPOF is thus located above M1 rather than above P4; the IOF is located above the P4/M1 boundary versus above P3; the bisector of the rear of the IOF lies 13–17 mm above an anterior projection of the facial crest (rather than being only about 7 mm above it); the lacrimal is dorsoventrally broad and extends into the rear of the DPOF for a distance of about 8 mm, versus about 20; upper cheek teeth are mesodont, ca. 24–27 mm tall at the mesostyle in the unworn condition, versus ca. 16 mm or less; the protocone is uniformly connected to the protoloph in P2 except in earliest wear, but remains isolated until latest wear in other cheek teeth (versus being confluent with the protoloph in all teeth after midwear); fossette borders are moderately complex in the upper half of the cheek tooth crown (vs. simpler), with as many as four but usually two plications on the opposing faces of the pre- and postfossettes (vs. rarely more than two, except in very early wear), and at least one pli commonly present on the distal borders of these fossettes (vs. rarely present); protocones are uniformly (and persistently; late wear) spurred (as in P. leonensis, but in early wear only); those of the premolars are nearly circular, whereas those of the molars are more elongate (vs. uniformly circular in mature wear); the hypoconal groove commonly has a single plication (vs. smooth or lost early in P. leonensis); dp1 is rarely preserved in the lower cheek tooth dentition (vs. commonly present); metaconids and metastylids are subequal in size and mostly have rounded outlines and are separated by deeper but U-shaped linguaflexids (vs. smaller, and with shallower linguaflexids); protostylids are developed on p3–m2 in later wear and are usually little more than an angulate bend in the enamel at the pertinent part of the tooth (as in P. leonensis, but only developed faintly and in later wear).

The suite of characters present in Merychippus insignis includes a number of plesiomorphic ones relative to more derived taxa, such as C. goorisi, for example, POB still relatively short; DPOF still relatively moderately developed; lacrimal dorsoventrally tall; lacrimal still penetrates DPOF; dP1 larger (table 16.1); protocone connected to protoloph in P2; protocones with persistent spur; metaconids/metastylids subequal in size with premolar ectoflexids early penetrating the isthmus. The DPOF/IOF character states described above are distinct in M. insignis relative to both C. goorisi and P. leonensis.

Merychippine:

This concept is utilized for mesodont to hypsodont Neogene equids that resemble Merychippus insignis wherein the DPOF has well-defined posterior, dorsal, ventral, and at least palpably (but not sharply as in C. goorisi) distinctive anterior borders; the IOF is located above the P4/M1 boundary, about aligned with the lower boundary of the orbit, and occurs distinctly below the well-defined ventral boundary of the DPOF (figs. 16.7–16.10). Well-defined is taken to mean that the fossa border forms a definite shoulder, so that the IOF opens well lateral to, and below, the medial surface of the DPOF, in contrast to, for example, P. leonensis (fig. 16.7) where the IOF is much farther forward and the DPOF much less well defined, and in contrast to C. goorisi where the IOF is both anterior in position and located very close to the much more strongly defined anterior portion of the ventral border of the DPOF. A malar fossa is absent. None of the other taxa referred by Hulbert and MacFadden (1991) as “Merychippus,” including “M. goorisi” has this combination of features where relevant cranial material is known (“M.” primus, “M.” intermontanus, “M.” carrizoensis, “M.” tertius, “M.” nr. sejunctus, “M.” coloradense, “M. goorisi”). “M.” gunteri is represented only by dental remains and cannot be directly compared in this context, but Hulbert and MacFadden (1991: fig. 10) show “M.” gunteri as a sister taxon to Parahippus leonensis and below all other “Merychippus” on their cladogram. Aside from the above, two other samples representing species-rank taxa are referred to herein as merychippine: a sample from the early Barstovian Trinity River Pit 1 fauna, of Texas, and that from a Deep Creek, Nebraska, site in the Valentine Formation (late Barstovian). In addition to having the relevant cranial features, this material demonstrates the association of cranial and upper dental features with those of the mandible and lower dentition.

Devil's Gulch Member, Valentine Formation, Nebraska:

AMNH 126899, nearly complete ♂ cranium (fig. 16.11A), laterally crushed and skewed, lacking left zygomatic arch, both paroccipital processes, with R&L I1–3, canines, dP1, P2–M3, in relatively early wear (M3 having breached enamel only on protocone and protoloph). AMNH 126900, partial ♂ cranium, dorsoventrally crushed, lacking virtually all of the right side and neurocranial portion of both sides, with L canine, LdP1, P2–M3 in late wear (protocone of P3 and M2 showing incipiently confluent protocone with the protoloph) and associated left mandible lacking symphyseal region, but having the mandibular and ascending rami, with p2–m3 in late wear stage, and no evidence of dp1 ever having been present (fig. 16.11C, D).

Comments: This material is briefly described to set out the reasons for considering the Deep Creek taxon to be included as a merychippine and to list the characters of its lower dentition that are relevant to corroborating the allocation of Echo Quarry specimens to Merychippus insignis. Further descriptions will be given in a revision of Merychippus s.s. now in progress.

AMNH 126900, with the cheek teeth in adult wear (fig. 16.11C, D) shows (table 16.2) that the P2 mesostyle is 15.6 mm tall. The p2 protoconid is about 12.3 mm tall (table 16.3). The hypoconulid of m3 is still distinct from the hypolophid (fig. 16.11D). Merychippine affinity is shown by ovoid and still spurred protocones on P4–M3 (fig. 16.11B), with the premolar protocones being more nearly circular than those of the molars, the presence of single pli caballins, the remnants of a reasonably complex fossette plication pattern, relatively large dP1 (ca. 14 mm long; fig. 16.11B, C), a remnant DPOF comparable to, but medially deeper (#40; table 16.4 vs. table 16.5) than that of the Echo Quarry sample of M. insignis, and the IOF being located above the P4/M1 boundary (see cranium of AMNH 126899; fig. 16.11A). In addition to the medially deeper (but still only slightly pocketed) DPOF, this specimen differs from the Echo Quarry sample of M. insignis in having the P2 protocone confluent with the protoloph [vs. isolated from it at virtually all ontogenetic wear stages in M. insignis], but the merychippine characters of the Deep Creek material are emphasized here.

The merychippine traits are better expressed by AMNH 126899. In addition to the features described above, the better preserved DPOF (fig. 16.11A;) more clearly resembles that of the Echo Quarry sample of M. insignis (figs. 16.7, 16.9). In spite of the wider POB (table 16.4), the lacrimal penetrates the DPOF (fig. 16.11A), which is proportionately smaller than in M. insignis (compare tables 16.4, 16.5). The upper cheek tooth dentition of AMNH 126899 is in an earlier stage of wear (fig. 16.11A, B) than the previous specimen (fig. 16.11C). The P2 mesostyle in AMNH 126899 is 25 mm tall, M3 shows only incipient wear, and confirms the presence of spurred protocone, the moderately complex fossette plication pattern, the double pli caballins on the premolars, the isolated protocone of P2, and the relatively large dP1 (11.2 mm long = 58% length of P2; table 16.1; note that the length of dP1 is about 53% that of P2 in merychippines, but much less, ca. 45% in Cormohipparion goorisi). In addition to its overall larger size and isolated P2 protocone, AMNH 126899 differs from the Echo Quarry sample of M. insignis in that the nasal notch is retracted to a position above dP1, the POB is wider, and the DPOF proportionately smaller. I emphasize the merychippine similarities of the Deep Creek, Echo Quarry, and Trinity River samples, rather than these differences.

The lower dentition of the merychippine from the Valentine Formation is represented by the associated lower ramus of AMNH 126900 (fig. 16.11D). Although incomplete, this specimen is relatively shallow below p2 (31.0 mm) and has been restored as being relatively shallow below the boundary between m2 and m3 (ca. 47 mm; table 16.6B). Whether or not this restoration is absolutely correct, enough of the actual bone is preserved to warrant the interpretation that this was a relatively slender-jawed form, comparable to, even though absolutely larger than, Merychippus insignis (table 16.7).

The lower cheek teeth are in an adult wear stage, with the occlusal pattern well worn on p2–m1, the ectoflexid still nearly touching the labial enamel of the metaconid/metastylid on m2–3, and the hypoconulid still separate from the hypolophid of m3. Protostylids are present on m1 and appear to be faintly represented on p3–4, as well (although not shown on fig. 16.11D). Note that, as in M. insignis (fig. 16.13) such protostylids are poorly expressed and do not form a major transverse element of the anterior border of the tooth. It is important that, even at this stage of wear, the ectoflexids of p3–4 reach well lingual to, or beyond (p4) the postflexid (fig. 16.11D) such that a well-developed isthmus does not seem to have been present in the premolars. Although in a relatively late stage of wear (V or later), p4 of AMNH 126900 preserves the deep lingual penetration of the ectoflexid typical of M. insignis.

Discussion: These specimens are referred to as merychippines, and it is likely that the Deep Creek form is specifically distinct from Merychippus insignis. Features in which the Deep Creek taxon appears to differ from M. insignis include: larger overall size, proportionately somewhat smaller dP1, P2 protocone remaining separate from the protoloph in late wear (e.g., fig. 16.11C), more posteriorly retracted nasal notch, longer POB, concomitantly shorter DPOF, and other features summarized in table 16.4.

Trinity River Pit 1, Fleming Formation, Texas:

Fossils from this quarry are of early Barstovian age (e.g., Tedford et al., 1987). The sample includes equid specimens having a cranial and upper cheek tooth morphology similar to that just described for the Devil's Gulch sample. Further, these Trinity River Pit 1 specimens are similar to and different from M. insignis in ways comparable to the Devil's Gulch material described above, and also are associated with mandibular rami and lower dentitions in which p2–4 show early penetration of the metaconid/metastylid isthmus by the ectoflexid (fig. 16.13), in contrast to the condition in Cormohipparion goorisi (fig. 16.14). Based on these comparisons, those Texas specimens having the morphology just described are referred to Merychippus s.s. Furthermore, the Texas Merychippus s.s. samples under discussion also have a mandible that is overall more slender and of relatively more uniform depth below the cheek teeth (tables 16.6, 16.7) than in C. goorisi (table 16.8; fig. 16.15B). Thus, the three samples, the Texas material allocated as merychippine, the Devil's Gulch material, and specimens herein allocated to M. insignis, all are similar to each other in the relatively shallow mandible depth below the cheek teeth and in that the ectoflexid in the lower premolars penetrates deeply between the metaconid and metastylid, in contrast to C. goorisi wherein the premolar isthmuses are strongly developed and retained ontogenetically longer (compare figs. 16.13, 16.14, tables 16.6–16.8, 16.9). In that it is possible that more than one specific-rank taxon is represented by the Trinity River Pit 1 Merychippus s.s. material, this sample is not treated as a hypodigm. The material is also under study by R.L. Evander, so the comments made here are brief.

Material: F:AM 109883, partial cranium, RdP1, P2–M3; F:AM 109894, partial cranium, RdP1, P2–M3; F:AM 109892, R maxillary fragment, dP1, P2–M3; F:AM 109893, partial cranium, RP2–M3; F:AM 109891, associated cranium, LI2–3, R&L dP1, P2–M3, R&L rami, i1–3, p2–m3; F:AM 109884, partial cranium, RP2–M3; F:AM 109885, R maxillary fragment, P2–M3; F:AM 109874; partial cranium, RdP1, P2–4, M1 erupting; F:AM 107837, partial cranium, RP2–M3; F:AM 109882, RP2–M3; F:AM 109876, RP2–M3; F:AM 113054, R ramal fragment, p2–m3; F:AM 113052, R ramal fragment, p4–m3; F:AM 113053, R ramal fragment, p2–m2; F:AM 113056, L ramal fragment, p2–m3; F:AM 113055, R partial symphysis, ramal fragment, p2–m3; F:AM 113061, L ramal fragment, p2–m3; F:AM 113062, L partial symphysis, ramal fragment, p2–m3; F:AM 113059, L ramal fragment, p2–m3; F:AM 113064, L ramus, angular process and ascending ramus, p3–m3; F:AM 113057, R ramal fragment, p2–m3; F:AM 113060, R partial symphysis, ramal fragment, p2–m3; F:AM 113065, R&L rami, lacking symphysis, R&L p3–m3; F:AM 113070, L partial symphysis, ramus, p2–m3; F:AM 113071, L ramal fragment, p2–m3; F:AM 113066, partial symphysis, R&L rami, R&L i1, Ri2–3, c1, p2–m3, angular process; F:AM 113067, R ramus, p2–m3, angular process, partial ascending ramus; F:AM 113052, R ramal fragment, p4–m3. F:AM 113075, R ramus with symphyseal region, p2–m3.

Features indicating affinity with Merychippus s.s. rather than with Cormohipparion goorisi (found in the same quarry) include the following.

Crown height. The least worn P2 of the Trinity River Pit 1 merychippine is 16.2 mm tall at the mesostyle (table 16.2); the hypocone is still isolated from the metaloph (i.e., in very early wear). It becomes confluent with the metaloph at 14.5 mm in this sample. In M. insignis, the hypocone is confluent with the metaloph at 21 mm in P2. In C. goorisi, this occurs at 24 mm. Unworn P2 mesostyle heights for these taxa are 25 mm (M. insignis) and 26 mm (C. goorisi).

M3 in the merychippine sample is unworn at ca. 20 mm. Comparable figures for M. insignis are 24 mm; for C. goorisi, 26 mm. Thus, the Texas merychippine is somewhat lower crowned than in M. insignis from Nebraska, and much lower crowned than contemporary C. goorisi.

Protocone. The protocone of P2 is virtually circular, with a persistent spur in the Trinity River Pit 1 merychippine sample, and otherwise identical to that seen in M. insignis. The protocone in P3–4 in the merychippine sample is similar to the condition seen in M. insignis.

Fossette borders. P4–M2 fossette borders are less complex in the Trinity River Pit 1 merychippine sample, and the pli prefossette loop is distinctly smaller and less prominent than in C. goorisi, and generally similar to M. insignis.

DP1 length. This ranges from 10.5 to 15.0 mm (x = 12.5; N = 5) and is intermediate in the Trinity River Pit 1 merychippine sample between C. goorisi (smaller) and M. insignis (table 16.1).

Segregated according to these characters, the Trinity River Pit 1 merychippine sample shows other features consistent with its referral to Merychippus s.s. The lacrimal reaches into the rear of the DPOF (e.g., F:AM 109883, 109894, 109893, 109891, 109884, 109874, 107837). The IOF is located above the posterior part of P4 (or dP4) or the border between P4 and M1 (or dP4 and M1; e.g., F:AM 109883, 109894, 109892, 109893, 109891, 109884, 109885, 107837). IOF location does not show significant ontogenetic variation. When visible, the DPOF is either not pocketed or is only slightly pocketed (e.g., F:AM 109883, 109884, 109891, 109892, 109893, 109894).

Table 16.9 shows a segregation of lower cheek teeth from the Trinity River Pit 1 sample, with specimens compared at similar wear stages. Those originally referred by MacFadden and Skinner (1981) to C. goorisi are distinct from the others, of which F:AM 109891 is associated with a cranium and upper cheek tooth dentition similar to Merychippus insignis. The Texas merychippine lower cheek teeth differ from those of C. goorisi in maintaining a deep penetration of the premolar isthmuses by their ectoflexids and being ontogenetically precocious in developing premolar and molar protostylids.

In summary, dentally and cranially these particular Trinity River Pit 1 specimens resemble M. insignis and other merychippines (e.g., the Deep Creek, Nebraska, material described above), and appear to have a number of cranial characters in common with that taxon. The Texas merychippine differs, at least, from M. insignis and from the Deep Creek, Nebraska, form in being still lower crowned.

The Texas specimens also are distinctly lower crowned than those assigned here to C. goorisi, have a simpler coronal pattern, and more ovate and distinctly spurred protocones (even into very late wear). In their overall similarity to the dental morphology of M. insignis, these Texas specimens are of merychippine type. Similarly, a suite of mandibular rami and lower cheek tooth dentitions differ from those of C. goorisi and resemble the merychippine from the Devil's Gulch Member of the Valentine Formation, as discussed above. These Trinity River Pit 1 specimens display strong penetration of premolar isthmuses by the ectoflexids and an ontogenetically earlier development of premolar and molar protostylids in contrast to C. goorisi. Associated crania and mandibular rami link the upper and lower cheek tooth features and additionally show that dP1 is relatively long (table 16.1), that the facial region is of Merychippus s.s. type (lacrimal penetrates the rear of the DPOF, IOF is located above the P4/M1 boundary and below a well-defined ventral border of the DPOF) and that the mandibular ramus is of relatively even depth from M3–P2.

The above analysis of associated cranial, dental, and mandibular data in the Deep Creek, Nebraska, and Trinity River Pit 1, Texas, samples led to a determination that not only are they referable as merychippines, but also which specimens from Echo Quarry, Nebraska, were referable to M. insignis as based originally on the cranium and upper dentition. The appraisal also indicates which of the Texas materials can be confidently allocated to C. goorisi. Most of the referrals of material to that species by MacFadden and Skinner (1981) are corroborated herein.

Type Specimen and Locality:

ANSP 11276, maxillary fragment with RdP2–3 from the Fort Randall Formation, Bijou Hills, South Dakota (Skinner and Taylor, 1967: 18).

Distribution and Age:

Late Barstovian, Bijou Hills, South Dakota, and early Barstovian, northwestern Nebraska.

Referred Material:

From Echo Quarry, Olcott Formation, Sioux County, Nebraska (Skinner et al., 1977). F:AM 87006, RdP2–4; F:AM 87007, RdP2–4, unerupted P2–M2; F:AM 87009, RdP2–4; F:AM 87010, LdP2–4; F:AM 87015, partial cranium, RdP3, R&L dP4, unerupted RP4, R&L M1–2; F:AM 87000, juvenile ♂ cranium with LdP1, R&L dP2–4, M1; F:AM 87001, ♂ cranium with RdP1, R&L P2–M3; F:AM 87002, ♂ cranium with RP2–M3; F:AM 87003, ♂ cranium with R&L dP1, P2–M3; F:AM 87004, ♂ cranium with R&L dP1, P2–M3; F:AM 87005, elderly ♂ cranium with R&L dP1, P2–M3; F:AM 87008, R maxillary with P2–M3; F:AM 87049, L maxilla with P3–M2; F:AM 87048, L maxilla with P3–M3; F:AM 87050, LP4–M2, M3 in crypt; F:AM 87058, RP4–M3; F:AM 109983, unerupted RP3–4, RM1; F:AM 87056, RP2–M2, M3 in crypt; F:AM 87014, partial cranium with R&L P3–M3; F:AM 87052, LP4–M3; F:AM 109984, RM1–3; F:AM 87042, palate with R&L P2–M3; F:AM 87013, palate with R&L dP1, P2–P3, RP4–M3; F:AM 87041, fragmentary cranium with LP2, M1–3, RP3–M3; F:AM 87022, RP3–M3; F:AM 87035, LM1–3; F:AM 87036, LM1–3; F:AM 87039, LM1–3; F:AM 87037, LM1–3; F:AM 87019, RP3–M2; F:AM 87060, LP4–M3; F:AM 87046, LdP1, P2–M3; F:AM 87062, LM1–M3; F:AM 87043, LdP1, P2–M3; F:AM 87047, LP2–M3; F:AM 87057, elderly RdP1, P2–M3; F:AM 110374, RM1–3; F:AM 110387, LP3–M1; F:AM 110389, RP2–4; F:AM 110379, RM1–3; F:AM 87059, RP2–M1; F:AM 87064, LP3–M3; F:AM 87061, LdP1, P2–4; F:AM 87051, LP2–M2; F:AM 87055, palate with R&L dP1–P3, LP4; F:AM 87045, palate with R&L dP1–P2, RP3–M1; F:AM 87030, LdP1, P2, P4–M1; F:AM 87012, palate with R&L P2–M2; F:AM 87016, RP3–M3; F:AM 87023, RP2–M1; F:AM 87029, LP3–M2; F:AM 87028, LP3–M1; F:AM 87032, LdP1, P2–M1; F:AM 87040, LP2–4; F:AM 87034, LP2–M2; F:AM 111771, L ramal fragment, p2–m2; F:AM 87074, R ramal fragment, p2–m3; F:AM 111781, L&R symphysis, L ramal fragment, p2–m1; F:AM 111791, R ramal fragment, p3–m2; F:AM 111793, ♂, R&L symphysis, Li3, ramal fragment, p2–m2; F:AM 112061, R ramal fragment, p3–m3; F:AM 87072, L symphysis, ramal fragment, p2–m3; F:AM 112062, L ramal fragment, p3–m3; F:AM 111794, symphysis, R&L c1, Li3 erupting, Lp2–m2; F:AM 87076, symphysis, erupting R&L i3, R ramal fragment, p2–m2; F:AM 111774, R ramal fragment, p2–m2; F:AM 111780, L ramal fragment, p3–m2; F:AM 111792, L ramal fragment, p3–m2; F:AM 87075, L ramal fragment, p3–m3; F:AM 112070, symphysis, R&L ramal fragment, c1, p2–m3; F:AM 87078, R&L symphysis, angular, condylar and coronoid processes, R&L p2–m3; F:AM 112064, L ramal fragment, p2–m3; F:AM 112066, L ramal fragment, p2–m3; F:AM 112063, L ramal fragment, p2–m3; F:AM 87080, L symphysis, ramal fragment, p2–m3; F:AM 111757, R symphysis, ramal fragment, p2–m3; F:AM 111760, R symphysis, ramal fragment, p2–m3; F:AM 111761, R ramal fragment, p3–m3; F:AM 111759, R ramal fragment, p3–m3; F:AM 112067, L ramal fragment, p2–m3; F:AM 87085, R&L symphysis, ramal fragment, p2–m3; F:AM 111750, L symphysis, ramal fragment, p2–m3; F:AM 87088, ancient R symphysis, ramal fragment, p2–m3; F:AM 111754, R ramal fragment, p2–m3.

Specimens Probably Referable:

These specimens are in a very old or very early ontogenetic stage of wear, and may be confused with other taxa, but likely pertain to the Echo Quarry hypodigm of Merychippus insignis. F:AM 87017, RP2–M3; F:AM 110373, LP3–M3; F:AM 87063, LP3–M3; F:AM 71218, RM1–3; F:AM 110376, RM1–3; F:AM 87027, LP4–M3; F:AM 110378, RP4–M3; F:AM 110376, RM1–3; F:AM 110377, RM1–3; F:AM 110375, RP3–M3; F:AM 87038, LM1–3; F:AM 110380, very ancient LdP1, P2–M3; F:AM LP4–M3; F:AM 87043, LP2–M3; F:AM 87028, LP3–M1; F:AM 87054, LP2–4; F:AM 110386, RP4–M2; F:AM 110381, LdP1, P2–M1; F:AM 110388, LP4–M2; F:AM 110385 (very early wear), RP2–4; F:AM 110383 (very early wear), RP4–M2; F:AM 110384, RdP1, P2–4; F:AM 87021, RP3–M2; F:AM 110382 (very early wear), RdP1, P2–4.

Specimens Probably not Referable:

These specimens, currently catalogued as Merychippus insignis from Echo Quarry, probably do not pertain to this species. F:AM 71190, RdP1, P2–M3. M1 in early but nearly adult wear has a mesostyle height of 30 mm and thus originally much taller than typical of unworn M1 in M. insignis. Also, for this stage of wear, the protocone tends to be more elongate, the hypocone is too slender, fossette plications are too deep and numerous. F:AM 71216, RP4–M2. P4 is ca. 17 mm tall (mesostyle), but is nearly worn out with a very much simpler enamel pattern than typical of M. insignis at this crown height; protocone is confluent with the protoloph in M1 and P4, M2 has very simple fossette patterns, but retains an isolated protocone. F:AM 71227, LdP1, P2–M2 has a 17 mm tall (mesostyle) P2, with the protocone confluent with the protoloph and the protoloph connected by a crochet to the metaloph (never occurs in M. insignis; in P3–M1 the protocone is confluent with the protoloph at 18, 17, and 16 mm, respectively (much sooner than in M. insignis), and the isolated M2 protocone lacks a spur.

Original Characterizations: (some additions signified by * based on the present study):

Leidy (1857) gave no diagnosis, but characterized the (immature) dentition of M. insignis as resembling those of ruminants, including those of deer and oreodonts. Based on the referred material from Echo Quarry, Skinner and Taylor (1967: 29–35) characterized, but did not diagnose, Merychippus insignis and (with few and minor exceptions) did not otherwise compare it with other equid species. The following is paraphrased and somewhat condensed from Skinner and Taylor (1967: 29–35), with emphasis on features determined to be useful in phyletic analysis by Hulbert and MacFadden (1991). In this context M. insignis is a mesodont horse in which the upper cheek teeth consist of a large and functional double-rooted dP1; dP2 and dP3 have prominent mesostyles, weak metastyles, an arcuate protoloph with a single plication between (but not connecting) the paracone and protocone, an incipient anterocrochet present but not uniting with the metaloph, protocone round to tear-shaped, crescentic metaloph with incipient to distinct crochet, and connecting to the anterior base of the metacone and to the hypocone, respectively, distinct hypostyle uniting anteriorly with the metaloph and bearing one or two plis hypostyle in the hypoconal groove. DP4 was not described by Skinner and Taylor (1967), but based on their figure 5 and personal inspection, generally resembles dP3.

The permanent cheek teeth were characterized as being mesodont, moderately curved, and covered by cement, except for the centers of fossettes. The long parastyle of P2 is centrally placed; those of P3–4 are large and overlap the metastyles of the next anterior tooth. Molar parastyles are neither so large nor as overlapping as in the premolars. Premolar mesostyles are stronger than those of the molars. Except for an anterior rib on P2, ribs are weak on all other cheek teeth. On P2 the protoloph connects to the ectoloph posterior to the parastyle, whereas on remaining cheek teeth it connects to the parastyle. The pli protoloph is weakly developed to absent in all adult teeth. The protoloph is mostly separate from the metaloph on P2, but is usually connected to the metaloph by a crochet on the other teeth. Protocones, elongate and isolated in early wear, connect with the protoloph and become rounded in later wear, and are oriented with their long axis about parallel to the anteroposterior axis of the tooth. Hypocones show changes in shape with wear comparable to the protocones and have a similar orientation. In early wear the fossette borders are moderately complex, the pli caballin is usually double in premolars, single in molars. One or two plications from the hypostyle extend into the hypoconal groove.

The cranium of M. insignis (Skinner and Taylor, 1967: 33–35) has a moderate-sized muzzle, the nasal notch (* not specifically mentioned by Skinner and Taylor, 1967) is relatively short (incised to a position above and about midway between C1 and P2); the buccinator fossa is separate from the DPOF; the POB is about 17 mm (*x̄ = 17.2, table 16.5, #32) wide and the lacrimal extends anterior to it; the DPOF is ovate, located high on the face, and ends anteriorly above P3. The upper border of the DPOF is sharp, especially at midlength, and the posterior border is sharp and rounded. In contrast, the ventral and anterior borders are shallow to indistinct, although *they are visibly and palpably distinguishable from the adjacent facial surfaces. The DPOF is faintly pocketed, if at all, and a malar fossa is absent. *The DPOF is bipartite, with a shallow ridge located just anterior to the tip of the lacrimal separating an anterior from a posterior portion (fig. 16.7C). *The infraorbital foramen lies low on the face, with the bisector of its posterior border being about 10–12 mm below the DPOF, 17–20 mm above an anterior projection of the facial crest, and above the rear half of P4. The orbit has a distinctly formed anteromedial fossa for the lacrimal sac, with a small lacrimal duct found slightly externally and dorsally that penetrates into the nasal cavity. The frontal bones are flat, not domed, and the lacrimal is prominent, effectively subrectangular in shape in facial expression with a variably tapered or notched anterior tip contained within the rear portion of the DPOF. In ventral view, the muzzle is U-shaped, and palatine fissures are prominent. The posterior nares are widest opposite M2 in adult skulls. The occipital condyles have shallow and wide condyloid fossae.

Additional Description and Comparison:

A fuller appreciation of the significance of various cranial parameters measured as recommended by the 1981 Hipparion Conference (Eisenmann et al., 1988) will follow the measurement of remaining Cormohipparion samples. For the present, the following comments can be advanced both as basic data (table 16.5) and in comparison with the cranium of Cormohipparion goorisi. Using parameter #4 (combined length of the basioccipital and basisphenoid) as a basilar length indicative of basic cranial and thus body size (e.g., Radinsky, 1984), cranial dimensions of M. insignis may be compared with those of C. goorisi. Parameter #4 is about 20% longer in M. insignis than in C. goorisi. Other parameters that are comparably different in the two taxa (with C. goorisi being 16–20% smaller; log—0.05 to 0.1, fig. 16.16) include: diastemal width (#14), trans-glenoid width (#19), DPOF height (#35), and height of rear center of DPOF above the alveolus (#38). Thus, in all the above, M. insignis is actually larger than but proportionally similar to C. goorisi. This suite is here referred to as Cranial Group 1 to facilitate comparison.

Cranial Group 2 includes parameters in which M. insignis is dimensionally about the same as C. goorisi. In terms of Group 1, these features actually are about 20% proportionally larger in C. goorisi. These parameters include, postpalatal length (#3), total cranial length (#6), length of premolar series (#7), length of molar series (#8), length of combined premolar and molar series (#9), width of muzzle at I3 (#15), inion height (#22), incision of nasal notch (#30), and facial length from orbit to nasal notch (#31).

Cranial parameters in which C. goorisi is distinctly larger than M. insignis would be proportionally much (27–46%) larger in C. goorisi. This suite is referred here to Cranial Group 3. The parameters include: muzzle length (#1), palatal length (#2), length of preorbital bar (#32), height of the IOF above the alveolus (#37), and the distance between the anterior end of the lacrimal and the rear of the DPOF. In M. insignis, the anterior end of the lacrimal actually penetrates beyond the rear of the DPOF by an average of about 8 mm, so the value is a negative figure (#39, table 16.5). In C. goorisi, the rear of the DPOF lies about 10 mm anterior to the lacrimal tip.

In summary, it appears that M. insignis is a relatively small hipparionine (sensu Hulbert and MacFadden, 1991), with a basilar length (#4) of about 86 mm; the muzzle is short (ca. 67 mm, #1) and relatively wide (ca. 36 mm, #15), the premolar series is slightly longer (ca. 64 mm, #7) than the molar series (ca. 55 mm, #8) and the total cheek tooth length is rather short (ca. 116 mm, #9). The palate is relatively wide (ca. 48 mm, #13, the frontals are flat and relatively wide (ca. 92 mm, #18). The preorbital region of the face is relatively long (ca. 132 mm, #31), and the DPOF relatively long (ca. 64 mm, #33) and medially shallow (9–12 mm as measured in this study), relatively well-defined dorsally, well-defined posteriorly, and weakly defined ventrally and anteriorly; it is faintly, if at all, pocketed posteriorly; the overall rectangularly shaped lacrimal has a pointed anterior tip that extends on average about 8 mm into the rear of the DPOF (#39); the DPOF is relatively high (ca. 37 mm, #35) and located well above (ca. 31 mm, #36) the facial crest; the POB is of moderate size (ca. 17 mm, #32). The IOF is located above the rear of P4, or the boundary between P4 and M1 or, in juvenile specimens, comparably with respect to dP4 and M1, and is virtually below the longitudinal midpoint of the DPOF, irrespective of ontogenetic age.

Description of the Upper Cheek Tooth Dentition:

The following remarks are derived from the data obtained for each tooth in the cheek tooth series, and summarized in tables 16.10–16.16. Table 16.10 forms the crown height framework upon which the other data shown in tables 16.11–16.16 were constructed. These record the following data for each tooth position: height (mesostyle) at which the protoloph connects to the metaloph, interval at which the prefossette is displayed (compare the incomplete fossettes in P3 of fig. 16.7D versus P3 in fig. 16.10F) (or lost), comparable data for the postfossette; for opening of the hypocone (isolated in M1, fig. 16.7D, vs. open and connected to metaloph in fig. 16.10F), presence and complexity of the hypoconal groove, isolation versus opening of the protocone to the protoloph, and presence and complexity of the pli caballin, of the pli protoloph, of the pli prefossette, of the pli postfossette and of the pli hypostyle.

A detailed description of the upper cheek teeth is presented in Appendix 1. The salient morphology of the adult upper cheek tooth dentition of Merychippus insignis is summarized as follows.

The dP1 is large (table 16.1), double rooted, and persistent into final stages of wear (e.g., F:AM 87032). P2–M3 are mesodont (unworn crown heights range from 24 to 27 mm; tables 16.11–16.16), moderately curved (Skinner and Taylor, 1967), with generally ovate protocones (figs. 16.7, 16.12) that have a strong spur persisting virtually to the end of wear. A stable (= adult) wear pattern is established after about 24% wear in P2, 11% in P3, 15% in P4, 22% in M1, 22% in M2, 12% in M3, and the hypocone connects to the metaloph in early wear. Within the interval of maximum complexity of the occlusal pattern, the pli protoloph is present in 50% (P2), 85% (P3), 53% (P4), 13% (M1), 39% (M2), and 29% (M3) of the specimens.

Although the protocone and hypocone are subequal in size (Skinner and Taylor, 1967), the protocone is always somewhat larger. The protocone is usually subovate in outline, and (as indicated here) is nearly equidimensional in the P2 and P3, with the width/length ratio of the means of the dimensions being 0.90 (P2) and 0.81 (P3). In P4 (0.78), M1 (0.75), and M2 (0.61), the protocone tends to be more elongate (table 16.17). Because of relatively early wear stages represented for many M3s, they were not considered in this analysis.

This report reveals that, whereas the protocone remains distinct from the protoloph until latest wear in all cheek teeth, the hypocone early connects to the metaloph. The protocone remains closed until very late wear (12 mm or below, or at levels consistent with 50% or greater wear). This is best developed in P4, M2, and M3. In P2, some specimens display an open protocone from the 17-mm level (32% wear; table 16.11). For P3 and M1 (tables 16.12, 16.14) comparable figures are 18 mm (33%) and 16 mm (41%). In all teeth, the protocone retains a spur until it becomes confluent with the protoloph. The protoloph consistently connects via a crochet to the metaloph in all but P2, where the two lophs are uniformly separate until very latest wear.

Within the interval of greatest pattern complexity (tables 16.11–16.16), the pli caballin is present in about 50% of the specimens in P2, 85% in P3, 69% in P4, 97% in M1, 92% in M2, and 32% in M3. When present, the pli caballin is double more frequently in the premolars than in the molars (22% in P2, 57% in P3, and 69% in P4 versus 8% in M1, 4% in M2, and never double in M3).

The pli hypostyle is present in about 94% of the specimens in P2, and is composed of more than a single pli in 29% when present. Comparable figures for the other cheek teeth are: P3, 70% and 5%; P4, 80% and 17%; M1, 84% and 13%; M2, 71% and 0%; and M3, 53% and 44%.

Fossette borders are moderately complex, and opposing margins of the pre- and postfossettes may bear as many as 5–6 plications at 19%–33% wear in P4–M2. The pli protoloph, pli prefossette, pli postfossette, and pli hypostyle are typically present to, and beyond, 50% wear, except in M1, which generally lacks a pli protoloph. Although less frequent and less numerous than on opposing faces of the pre- and postfossettes, P2 and P3 show extra plications on the anterior and posterior faces of those fossettes in 15–70% of cases where such are present. P4, M1, and M3 show no additional crenulations of the pli protoloph, whereas M2 has an extra plication in 10% of such cases. An additional crenulation of the pli hypostyle occurs in about 15% of such cases in P4 and M1, in no cases for M2, and in about 20% in M3. This indicates that there are virtually no crenulations on the anterior border of the prefossette, except for the pli protoloph, and none on the posterior border of the postfossette except for the pli hypostyle in P4–M3. Table 16.18 shows average plication complexity for P2–M3 in the interval of greatest complexity. Virtually all taxonomically important details of the tooth have been lost by the time the teeth are worn to 10–13 mm above the base of the crown.

Description of the Lower Cheek Tooth Dentition:

The morphology of the material discussed here from Echo Quarry (Olcott Formation), Nebraska, is consistent with that discussed above for Merychippus s.s., and is of the proper size to successfully occlude with comparably aged (ontogenetically) upper dentitions associated with the crania of M. insignis. The description is presented in an ontogenetically developmental sequence, couched, in part, on table 16.19 where the specimens are listed in progressively increasing wear stages. The wear stages used here are generally comparable to those used for the Trinity River Pit 1 specimens. Because of specific differences, however, some morphological details may differ between the two samples.

In wear stage I (e.g., F:AM 111771), the p2 metaconid is single, slightly later bifid into the presumptive metaconid/metastylid (fig. 16.13). A spur from the anterolingual portion of the postflexid extends posterolingually. For p3, the postflexid spur (fig. 16.13) is directed toward an opposing structure from the entoconid (pre-entoconid spur) and, as throughout, the labial margins of the protoconid and hypoconid are rounded and, for all premolars, the ectoflexid does not penetrate the isthmus between the metaconid and metastylid. The p4 is generally similar to p3, but in a more advanced stage of wear. In about equivalently worn m1, the metaconid/metastylid are already distinct, and longitudinally drawn out, whereas in m2, the ectoflexid penetrates the metaconid/metastylid isthmus, and the pre-entocristid extends to the prehypocristid (fig. 16.13).

In stage II (e.g., F:AM 111781, 111793), p2 shows a better developed metaconid and metastylid; a posterolingual spur is present on the paraconid, comparable to that from the prehypocristid and the pre-entoconid spur still is present. Although separate and connected, the metastylid and metaconid are slightly penetrated by the ectoflexid (fig. 16.13). In p3, short-lived spurs occur on the posterolingual border of the paralophid; the hypoconulid is distinct; the ectoflexid does not penetrate the metaconid/metastylid isthmus. The p4 resembles p3 but lacks paralophid spurs. In m1, preflexids and postflexids are nearly absent, and ectoflexid penetrates deeply lingually. The m2 is much like m1 but has a shorter pre-entoconid spur.

In stage III (e.g., F:AM 112070), all premolars show deep penetration into the metaconid/metastylid isthmus by the ectoflexid; above-mentioned spurs are absent; the postflexid is flattened and open (p2) to isolated p3; in m1 the preflexid is essentially lost and the postflexid is flat. The hypoconulid of m3 is still not breached.

In stage IV (e.g., F:AM 87087), the premolar morphology is generally like that of stage III but the p2 postflexid is reduced, and virtually absent on p3–4. Protostylids are present on p3–4, m2, and likely on m1. The hypoconulid of m3 is worn, but dentine not yet exposed.

In stage V (e.g., F:AM 111761), the ectoflexid now is at the labial edge of the postflexid in p2 and still is near the labial enamel of the metaconid/metastylid in p3–4; m3 hypoconulid is breached. In older stages, the dental pattern progressively deteriorates to the extent that the morphological characterization of Merychippus insignis is not based on such specimens.

Based on the materials evaluated here, the lower cheek tooth dentition of Merychippus insignis is characterized as being relatively small; premolar and molar metaconids/metastylids (when developed in stage II and later) are subequal in size and have rounded enamel borders; the protoconid and hypoconid are rounded in outline; premolar ectoflexids penetrate deeply lingually and only in latest stages of wear terminate at or labial to the pre- or postflexids whereas molars show deeply penetrating ectoflexids at least to wear stage III or IV; premolar and molar protostylids appear at about wear stage III and IV and rarely are more than a visible sharp bend in the enamel pattern at the anterolabial corner of the tooth.

As shown in table 16.7, the mandibular ramus of M. insignis is gracile and diminishes only slightly from below the anterior end of m3 to the anterior end of p2. Based on less exhaustive comparisons with other, approximately contemporaneous, taxa, the lower mandible and cheek tooth dentition of M. insignis appears to be conservative in the combined character suite of: smaller size, less prominent protostylids, shorter symphysis, rounded protoconid and hypoconid, subequal metaconid/metastylid, rounded outline of metaconid and metastylid with broad, shallow, U-shaped linguaflexid; deep lingual penetration of premolar ectoflexids.

Summary:

Based on the information detailed above, Merychippus insignis can be characterized on the Echo Quarry sample as: relatively small size with a short muzzle; nasal notch retracted to a position about midway between C1 and P2; a POB that is of intermediate width (e.g., 17 mm); the DPOF faintly but distinctly expressed with definite dorsal, posterior, and ventral borders; the DPOF but slightly pocketed posteriorly and relatively long (ca. 64 mm) in comparison to the cheek tooth row length (ca. 116 mm); the IOF located above the P4/M1 boundary, with the bisector of the rear of the IOF lying about 13–17 mm above the anterior projection of the facial crest; the lacrimal dorsoventrally broad and penetrating the rear of the DPOF for a distance of about 8 mm; dP1 persistently large (ca. 13 mm long), two-rooted, and present into old age; upper cheek teeth mesodont, ca. 25–27 mm tall at the mesostyle in the unworn condition; the protocone uniformly connected to the protoloph in P2 except in earliest wear, but remaining isolated until latest wear in other cheek teeth; fossette borders moderately complex in the upper half of the cheek tooth crown, with as many as four to six but usually two plications on the opposing faces of the pre- and postfossettes; protocones virtually uniformly spurred; those of the premolars nearly circular, whereas those of the molars are more elongate; the hypoconal groove commonly with single plication; dp1 rarely preserved (generally suppressed); metaconids and metastylids subequal in size and mostly with rounded outlines and separated by shallow but U-shaped linguaflexids; premolar ectoflexids early penetrate the metaconid/metastylid isthmuses and maintain a position at or lingual to the pre- or postflexid into late wear; protostylids developed on p3–m2 in later wear and are usually little more than an angulate bend in the enamel at the pertinent part of the tooth; mandibular rami have short symphyses and are relatively uniform in depth from the anterior end of m3 to p2.

Type Specimen and Locality:

“F:AM 73490, well preserved skull with alveoli for incisors, right canine and P1, and right and left P2–M3, from Trinity River Pit 1, Fleming Formation, San Jacinto County, Texas Gulf Coastal Plain.” (MacFadden and Skinner, 1981: 620).

Distribution and Age:

Early Barstovian of the Gulf Coastal Plain of Texas and Florida (ca. 15 Ma; Hulbert and MacFadden, 1991: 36), including material from the Sweetwater Branch site, Arcadia Formation, Polk County, Florida not mentioned by MacFadden and Skinner (1981). The present discussion is limited to the holotype and topotypic material. Comments are not presented as to the affinity of the Sweetwater Branch specimens.

Hypodigm (* = original hypodigm of MacFadden and Skinner, 1981: 621):

*F:AM 73940, holotype ♂ cranium with RC1, RdP1, R&L P2–M3; *F:AM 73941, juvenile cranium with R&L dP2–4; *F:AM 73942, ♀ laterally skewed cranium with R&L dP1 (alveoli), P2–M3; *F:AM 73943, ♀ facial region, with R&L dP1, P2–M3; *F:AM 73944, R&L dP2–4, M1–2, M3 barely erupting; *F:AM 73952, ♀ partial facial region, with R&L I2–3, C1, dP1, P2–M3, LP4–M3; *F:AM 73945, R maxillary fragment with dP2–dP4; *F:AM 73946, R maxillary fragment with dP2–4; *F:AM 73947, R maxillary fragment with dP2–4, erupting M1; F:AM 109875, R maxillary fragment with P2 and P3 in crypt, RdP3–4, M1; F:AM 109886, maxillary fragment with LP3–M3; F:AM 109887, RM1–3; F:AM 109889, LP3–M2; F:AM 109890, RP4–M3; F:AM 73948, R ramus with symphyseal region, Ri2, Lc1, Rp2–m3; F:AM 73949, L ramus with p2–m3; F:AM 113058, L ramus with symphyseal region, p2–m3; F:AM 113063, R&L rami with Rp2–m1, Lp3–m3; F:AM 113068, R ramus with symphyseal region, Rp2–m3; F:AM 113069, L ramus with p2–m3. F:AM 107873, adult skull fragment with RP2–M3, LC1, LP2–4 is here removed from the hypodigm of C. goorisi on the basis of IOF being located above the anterior half of M1 in this ontogenetically very old specimen.

Original Diagnosis (MacFadden and Skinner, 1981: 621):

“Small hipparion. Cheek teeth mesodont. Protocone rounded with anterior spur. Cheek teeth covered with cement. Skull small relative to hipparions such as species of North American Hipparion s.s., Neohipparion, and the other species of Cormohipparion. Nasal notch shallow and it extends posteriorly to a position that lies dorsal to the canine. Prominent nasomaxillary fossa lying over P3–M1 with well-developed and continuous anterior and posterior rims. Very deep posterior pocket of fossa that extends almost to the anterior margin of the orbit. Fossa positioned far forward of orbit resulting in a wide preorbital bar. The anterior margin of the lacrimal bone usually does not touch the posterior rim of the fossa [see below]. In the upper cheek teeth, protocones rounded with anterior spur (particularly during early wear stages), fossette borders moderately plicated, pli caballin consists of single or double loops and deep hypoconal groove. In the lower cheek teeth, ectoparastylids [= protostylids of this report] absent or rudimentary, deep ectoflexids, pli caballinid rudimentary or absent, metaconids and metastylids small, rounded, and moderately separated, and moderately developed enamel plications”.

Other Original Comments:

These are excerpted from MacFadden and Skinner (1981) as being useful in broader comparisons. I have verified all features subsequently, with particular additions indicated here by *. In addition to being a mesodont horse (noting that mesodont has not been defined except only vaguely, and that unworn upper cheek teeth of this species range from 26 to 32 mm in height*), C. goorisi is additionally characterized as having a shallow nasal notch that is about midway between C1 and P2; the posterior tip of the premaxilla overlies the rear of the buccinator fossa; the postcanine diastema is short; the buccinator fossa is moderately developed (and is separated from the DPOF*); the infraorbital foramen lies over P3 or P4 at or *very near the anteroventral margin of the *very well-defined nasomaxillary fossa (DPOF of subsequent usage, and as used in this report); the malar crest is moderately inflated; the malar fossa is absent. The DPOF is teardrop shaped, has well-defined and continuous *dorsal, ventral, and anterior and posterior rims, and is deeply pocketed. The DPOF is anteroposteriorly oriented*. The DPOF is located well forward of the orbit so that the preorbital bar is wide. The anterior edge of the lacrimal bone either “barely touches” (e.g., F:AM 73941) or “does not touch” (e.g., F:AM 73940) the rear of the DPOF (MacFadden and Skinner, 1981: 622). These statements are open to other interpretation. Woodburne (1996a) reappraised the morphology of these skulls in conjunction with F:AM 73942 and concluded that the anterior tip of the lacrimal extended at least a short distance into the rear of the DPOF as shown in figure 16.12. Of special note, however, is the fact that the jugal makes a rather extensive contact with the lacrimal and reduces the length of the maxillo-lacrimal suture, a feature unique to C. goorisi.

Additional Description:

The following is developed from tables 16.5 and 16.20, based on cranial parameters recommended by the 1981 Hipparion Conference (Eisenmann et al., 1988), and indicates some cranial features in which C. goorisi differs from M. insignis. Using parameter #4 (combined length of basioccipital and basisphenoid) as a basilar length indicative of cranial and thus body size (e.g., Radinsky, 1984), cranial dimensions of C. goorisi may be compared with those of M. insignis. Parameter #4 is about 20% shorter in C. goorisi than in M. insignis. Other parameters that are comparably different in the two taxa (with C. goorisi being smaller) include: diastemal width (#14), frontal width (#18), trans-glenoid width (#19), occipital height (#22), muzzle height (#25), DPOF length (#33), length from rear of DPOF to IOF (infraorbital foramen; #34), DPOF height (#35), height of DPOF above maxillary (facial) crest (#36), and height of rear center of DPOF above the alveolus (#38). Thus in all the above, C. goorisi is actually smaller than, but proportional to, the dimensions found in M. insignis Cranial Group 1.

Cranial parameters in which C. goorisi is dimensionally about the same as M. insignis would actually be proportionally smaller in the merychippine form. These pertain to Cranial Group 2: length of premolar series (#7), length of molar series (#8), length of combined premolar and molar series (#9), width of muzzle at I3 (#15), facial length from orbit to nasal notch (#31).

Cranial parameters in which C. goorisi is distinctly larger than M. insignis would be proportionally much smaller in the merychippine species. These pertain to Cranial Group 3: muzzle length (#1), palatal width (#13), length of preorbital bar (#32), height of the IOF above the alveolus (#37). In M. insignis, the anterior end of the lacrimal penetrates beyond the rear of the DPOF by an average of about 8 mm, so the value is a negative figure (#39, table 16.5). In that the DPOF is pocketed in C. goorisi, the anterior tip of the lacrimal touches the rear of the DPOF, giving a value of 0 for #39 in table 16.20).

In summary, it appears that C. goorisi is a relatively small hipparionine (sensu Hulbert and MacFadden, 1991), with a basilar length (#4) of about 72 mm, versus 86 in M. insignis; the muzzle is short (ca. 77 mm, about 10 mm longer than in M. insignis; #1) and relatively wide (ca. 36 mm, vs. 36 in M. insignis), premolar series is slightly longer (ca. 64 mm, #7) than the molar series (ca. 55 mm, #8) and the total cheek tooth length is rather short (ca. 116 mm, #9 in both species). The palate is relatively wide in both (ca. 46–48 mm, #13), the frontals are flat and relatively wide (ca. 101, vs. 92 mm, #18). The preorbital region of the face is relatively long (ca. 128, vs. 132 mm, #31), the DPOF relatively shorter (ca. 50 vs. 64 mm, #33), medially moderately deeper (23 mm vs. 9–12 mm; column 40, table 16.20 vs. table 16.5), well defined dorsally, posteriorly, ventrally and anteriorly versus mostly less well defined overall, except posteriorly; DPOF is strongly (= nearly to orbit) versus faintly, if at all, pocketed posteriorly; the overall triangularly versus rectangularly shaped lacrimal reaches the rear margin of DPOF (#39); the DPOF is relatively shallower dorsally (ca. 29 vs. 37 mm, #35) and located less dorsally above (ca. 22 vs. 31 mm, #36) the facial crest; the POB is wide (ca. 24 vs. 17 mm, #32). The IOF is located above P3 or the P3–P4 boundary (or dP3–dP4 boundary), versus above P4 or the P4–M1 (or dP4–M1 boundary), and is virtually below the anterior portion versus the longitudinal midpoint of the DPOF, irrespective of ontogenetic age.

The Upper Cheek Tooth Dentition:

The following remarks are derived from the data obtained for each tooth in the cheek tooth series, and summarized in tables 16.21–16.27. These show that the upper dentition of C. goorisi is sparsely represented, there being no examples of the lower half of each tooth. Estimated unworn crown heights are: P2, 26 mm; P3, 27 mm; P4, 34 mm; M1, 34 mm, M2, 30 mm; M3 26 mm. Due to the small sample size, the detailed analysis accomplished for M. insignis is not attempted here. The following general remarks apply for C. goorisi.

The upper cheek tooth dentition of Cormohipparion goorisi is characterized by having incisors with cement-filled cups (infundibula; MacFadden and Skinner, 1981); dP1 is of moderate size, although smaller than in M. insignis (mean = 10.9 mm; table 16.1), two-rooted, and persistent until late ontogenetic age; P2–M3 are mesodont; unworn crown height from P2–M3 ranges about 26–34 mm; P2 protoloph connects to the metaloph (table 16.22) at about 30% wear (contra M. insignis; isolated until very late wear; table 16.11).

Other salient comparisons with M. insignis include: In P2 the pli prefossette is present after about 19% wear, with the posterior border of the prefossette scoring from 2.67 (total sample; N = 3) in C. goorisi. This is three times the mean complexity as compared to M. insignis (0.88; table 16.18). For P3 (table 16.23), the pli prefossette is present after 15% wear, commonly bifid, and associated with 2–4 additional plis. The posterior border of the prefossette scores at 4.25, considerably more complex than in M. insignis (1.32; table 16.18). Comparable statements for P4 (table 16.24) are: present after about 20% wear, mostly bifid, the fossette border scoring 5.00 (more than twice as complex as in M. insignis, 1.81, table 16.18); for M1 (table 16.25), pli prefossette is present, after about 32% wear, and associated with 1–5 plis, posterior fossette border scores at 3.25, again more complex than in M. insignis (2.74, table 16.18); for M2 (table 16.26), pli prefossette is present after ca. 20% wear, associated with 1–3 plis, posterior fossette border scores 3.40, about twice the complexity of M. insignis (1.81, table 16.18); for M3 (table 16.27), pli prefossette is present (sometimes bifid) after ca. 4% wear, is associated with 3–4 additional plis, with the posterior fossette border scoring as 3.60, almost three times more complex than in M. insignis (1.00, table 16.18).

The pli postfossette in P2 is present, after 19% wear) and commonly associated with additional plis, with the anterior border of the postfossette scoring at 2.00. This is more complex than comparable figures for M. insignis (1.21; table 16.18). For P3, the pli postfossette is present after 15% wear, the anterior fossette border scoring 2.00 (about comparable to M. insignis, 1.81); for P4, the pli postfossette is present (except in the unworn state) and associated with 1–4 additional plis; anterior fossette border scores 3.43, about 35% more complex than seen in M. insignis (2.21); for M1 the pli postfossette is present after ca. 32% wear, associated with 1–5 additional plis, its anterior fossette border scoring 3.42, distinctly more complex than in M. insignis (2.32); for M2, pli postfossette is present after about 20% wear, associated with 1–3 additional plis, with its anterior fossette border scoring 2.88, more complex than in M. insignis (1.87); for M3, the pli postfossette is mostly present, associated with 1–3 additional plis, the anterior fossette border scoring 3.00, three times more complex than in M. insignis (0.93).

The pli hypostyle is present in about 79% of all specimens in adult wear and is rarely composed of more than a single pli.

The pli caballin is present in about 50% of the specimens in P2, and in 90–100% of the specimens in the other cheek teeth with adult wear. It is doubled more frequently in the premolars than in the molars (50% in P2, 20% in P3, 42% in P4 versus 29% in M1, 13% in M2, and never double in M3, as represented here).

The protocone remains closed until about 45–56% wear for those teeth (P3, M1) in which it is shown to be open eventually to the protoloph, and except for earliest wear (especially in P2) never has an anterolabial spur, in distinct contrast to M. insignis. P2 protocones average more nearly circular than those of P3–M3. The prominent pli prefossette loop projects well beyond the posterior border of the postfossette, and is commonly bifid at its terminal tip, especially on P3–M3.

Virtually all taxonomically important details of the tooth are still present into later wear where this can be shown (P4, 59%; M1, 68%; M2, 57%; all minimum figures).

The premolars and molars are relatively tall for a mesodont horse, with unworn crown heights being about 26 (P2), 27 (P3), 34 (P4), 34 (M1), 30 (M2), and 26 mm (M3). For M. insignis, comparable unworn crown heights are 25, 26, 27, 24 (?27), 28, and 24 mm.

The fossette borders are relatively complex, especially in P4–M2, with up to 6, 6, and 4 plications on the posterior border of the prefossette, respectively, and 5, 6, and 4 on the anterior border of the postfossette. The pli prefossette is commonly bifid or trifid. The enamel loop that outlines the pli prefossette is very large and conspicuous, sometimes forming an isolated enamel lake, and is commonly bifid at its posterolingual tip. The posterior border of the prefossette is always significantly more complex than in M. insignis, as is the anterior border of the postfossette (possible exception of P2). The protocone is subovate in P2, but more elongate in P3–M2, and the associated spur is neither as prominent nor as persistent with wear as in M. insignis.

The Lower Cheek Tooth Dentition:

The lower incisors have cement-filled cups, and dp1 is absent (MacFadden and Skinner, 1981). The cheek tooth dentition (table 16.28) is described as an ontogenetic sequence in table 16.9 and figure 16.14. Based on other Trinity River Pit 1 specimens, the earliest wear stage includes specimens in which the metaconid/metastylid of p2 is still single, the postflexid is still open lingually, and m3 is in very early wear. No specimens of C. goorisi are represented at this stage.

In stage II, (e.g., F:AM 73948), the hypoconid of m3 is breached, and the hypoconulid is slightly worn; the metaconid and metastylid of p2 are beginning to be separate, and the p2 postflexid is nearly closed lingually. A preentocristid extends anterolabially toward a posterolingually directed spur from the anterior part of the prehypocristid. The ectoflexid does not penetrate the metaconid/metastylid isthmus. In p3 and p4, the ectoflexid still is labial to the isthmus.

In stage III (e.g., F:AM 73949), the metaconid is distinct from the metastylid in p2, a preentocristid and remnant opposing spur from the anterior part of the prehypocristid still is present, and a spur within the preflexid extends toward the metaconid. In p3, the postflexid is slightly open, the hypoconid of m3 is worn, the hypoconulid breached. In p3 and p4 the ectoflexid still is labial of the metaconid/metastylid isthmus, and protostylids are present in p3–m1.

Stage IV is not represented. In stage V, the postflexid is still open in p3–4 (p2 not represented; e.g., F:AM 113058), and the ectoflexid still is well labial of the postflexid (not penetrating the isthmus) in p3.

In stage VI, p2 pre- and postflexids are open; in p3 the postflexid is an isolated loop; premolar ectoflexids still reach only to about the level of the postflexids; protostylids are present on p3–m1.

In stage VII (e.g., F:AM 113069), premolar ectoflexids are still about at or slightly labial to the postflexids; protostylids are present from p3–m2.

As shown in table 16.9, the premolar ectoflexids in C. goorisi never penetrate the isthmus of the metaconid/metastylid to the degree seen in a Trinity River Pit 1 Merychippus s.s., and the development of protostylids appears to be retarded by about one wear stage relative to that merychippine form. Otherwise the two taxa are similar in overall appearance, with the rounded shape and subequal size of the metaconids and metastylids, the shallow and U-shaped linguaflexid, and the rounded labial borders of the protoconid and hypoconid.

As shown in tables 16.7 and 16.8 and figure 16.15, the mandible of C. goorisi is attenuated anteriorly from m3 to p2 more than in M. insignis.

Summary:

Based on the information detailed above, Cormohipparion goorisi can be characterized on the basis of the Trinity River Pit 1 sample as: relatively small size with a short muzzle; nasal notch retracted to a position about midway between C1 and P2; relatively wide POB (e.g., 24 mm) and DPOF distinctly expressed with strongly defined anterior, dorsal, posterior, and ventral borders; the DPOF is strongly pocketed posteriorly (virtually to a position opposite the anterior edge of the orbit), but its facial expression is relatively short (ca. 48 mm) in comparison to the cheek tooth row length (ca. 116 mm); the IOF is located above the P3/P4 boundary, at or very close to the anteroventral rim of the DPOF; the lacrimal is dorsoventrally narrow and reaches the rear of the DPOF; dP1 is of moderate size (ca. 10 mm long, but shorter than in M. insignis), two-rooted, and present into old age; upper cheek teeth are mesodont (but taller than in M. insignis, ca. 26–34 tall at the mesostyle in the unworn condition); the protocone is isolated from the protoloph in P2 except in late wear, as in the other cheek teeth; fossette borders are relatively complex in the upper half of the cheek tooth crown, with as many as six but usually four plications on the opposing faces of the pre- and postfossettes; the pli prefossette loop is usually conspicuously large and bifid posterolingually; protocones are virtually uniformly not spurred; those of P2 are nearly circular, whereas those of the other premolars and molars are more elongate; the hypoconal groove commonly lacks plications; dp1 is not preserved in the lower cheek tooth dentition; metaconids and metastylids are subequal in size and mostly have rounded outlines and are separated by shallow but U-shaped linguaflexids; premolar ectoflexids virtually never penetrate the metaconid/metastylid isthmuses and maintain a position at or labial to the pre- or postflexid throughout wear; protostylids are developed on p3–m1 in later wear and are usually little more than an angulate bend in the enamel at the pertinent part of the tooth; mandibular rami have short symphyses and become markedly shallower in depth from the anterior end of m3 to p2.

In comparison with Merychippus insignis, Cormohipparion goorisi differs in all of the above-mentioned traits. If the relative position for these taxa is that shown in Hulbert and MacFadden (1991; C. goorisi is more derived), the changes involved in comparing M. insignis and C. goorisi appear to involve modifications that result in a taxon with a somewhat shorter basicranium, having undergone negative allometry with respect to characters allocated to Cranial Group 1 (above), undergone positive allometry with respect to characters allocated to Cranial Group 2, and especially with respect to those in Cranial Group 3, developed a more hypsodont and complex cheek tooth dentition, a lower jaw that is more attenuated anteriorly, and a lower cheek tooth dentition in which the premolar ectoflexids penetrate the metaconid/metastylid isthmus, and develop protostylids on p3–m3 by wear stage III, rather than at wear stage IV in M. insignis.

Ancestry of Cormohipparion goorisi:

Based on the foregoing, C. goorisi is strongly different both cranially and dentally from Merychippus insignis. If each species can be said to stand at the base of its respective clade, then it seems clear that the generic-rank taxa that they each represent are phyletically separate, and that neither was the ancestor of the other. Further, the posterior location of the IOF in conjunction with its plesiomorphic mesodont cheek-tooth dentition separates M. insignis and other merychippine species from all other mesodont to hypsodont equids of late Hemingfordian to Barstovian age. Based on specimens surveyed here, and others under review, the merychippine fossil record is limited to the Barstovian.