Content uploaded by James A. Hopson
Author content
All content in this area was uploaded by James A. Hopson on Oct 21, 2014
Content may be subject to copyright.
ARTICLE
LATE EARLY JURASSIC MAMMALIAFORMS FROM HUIZACHAL CANYON,
TAMAULIPAS, ME
´
XICO
MARISOL MONTELLANO,
1
JAMES A. HOPSON,
*,2
and JAMES M. CLARK
3
1
Departamento de Paleontología, Instituto de Geología, Ciudad Universitaria, Delegación. Coyoacán, México, D.F. México;
2
Department of Organismal Biology and Anatomy, University of Chicago, 1027 E. 57th Street, Chicago, Illinois 60637,
jhopson@uchicago.edu;
3
Department of Biological Sciences, The George Washington University, Washington, D.C. 20052
ABSTRACT—Previous reports of the fossil vertebrates from the upper Lower Jurassic La Boca Formation of Tama-
ulipas, México, presented descriptions of a tritylodontid therapsid, a burrowing diapsid, crocodyliforms, a pterosaur and
sphenodontians. Mammaliaform material is rare, consisting primarily of partial lower jaws, some of which are extremely
small. Five partial dentaries and one maxilla all have molars of “triconodont” morphology. Two lower jaws clearly show
the internal mandibular trough, suggesting attachment of postdentary bones, another two do not possess a trough, and in
another its presence or absence cannot be determined. This diversity of forms suggests that during the late Early Jurassic
variations on the primitive “triconodont” molar patterns and in lower jaw morphology took place. The mammaliaform
specimens from Tamaulipas appear to represent new taxa, but, because of the incomplete preservation of most of the
material, only three new genera and species are erected: Bocaconodon tamaulipensis, Victoriaconodon inaequalis, and
Huasteconodon wiblei.
RESUMEN
En informes previos de vertebrados fósiles de la tardía jurásica
temprana Formación La Boca, Tamaulipas, México, se describi-
eron un terápsido tritilodóntido, un diápsido cavador, cocodri-
lomorfos, un reptil volador y esfenodontes. El material de
mamíferos (mamaliaformes) es escaso, y consiste principalmente
de fragmentos de mandíbulas, algunas extremadamente
pequeñas. Cinco fragmentos de dentarios y una maxila poseen
molares con morfología “triconodonta.” Dos ejemplares mues-
tran claramente el canal interno sugiriendo que los huesos post-
dentarios estaban presentes; otros dos no poseen el canal y en el
otro su presencia o ausencia no se puede determinar.
Esta diversidad de formas sugiere que durante finales del Ju-
rásico temprano existieron en las variaciones de los patrones
molares “triconodontes” primitivos y en la morfología de la
mandíbula. Los mamaliaformes de Tamaulipas parecen todos
representar nuevos taxa, pero, debido a la conservación tan in-
completa de casi todo el material, sólo tres nuevos géneros y tres
especies fueron establecidos: Bocaconodon tamaulipensis, Victo-
riaconodon inaequalis, y Huestaconodon wiblei.
INTRODUCTION
Mesozoic terrestrial vertebrate faunas from México are poorly
known, being primarily represented by Late Cretaceous assem-
blages in the northern part of the country. These faunas consist
mostly of dinosaur remains. Mesozoic mammals of Campanian
age have been described by Lillegraven (1972, 1976) and Clem-
ens (1980) from the El Gallo Formation in Baja California Norte;
these include the multituberculates Mesodma cf. M. formosa,
?Stygimys sp., a pediomyid marsupial Pediomys sp., and a eu-
therian, Gallolestes pachymandibularis, that is allied with Zhe-
lestidae by Archibald et al. (2001) and Wible et al. (2007). Re-
cently, Aguillón-Martinez et al. (2004) reported a multitubercu-
late tooth from the Late Cretaceous Cerro del Pueblo
Formation, Coahuila, increasing the Mesozoic mammalian re-
cord for Mexico.
In 1982, the skull of a tritylodontid therapsid was discovered in
Jurassic beds in Huizachal Canyon, west of Ciudad Victoria,
Tamaulipas, and described as Bocatherium mexicanum by Clark
and Hopson (1985). Subsequent expeditions to this area between
1985 and 1993 recovered a large sample of small- to medium-
sized vertebrate fossils (Fastovsky et al., 1987; Clark et al., 1991;
Clark et al., 1994). The fauna includes at least three previously
unknown sphenodontians (Reynoso-Rosales, 1992). These are a
Sphenodon-like taxon, Cynosphenodon huizachalensis (see Rey-
noso, 1996), a dwarf form, Zapatodon ejidoensis (see Reynoso
and Clark, 1998), and, a possibly venomous sphenodontid, Sphe-
novipera jimmysjoyi (see Reynoso, 2005). A primitive burrowing
diapsid, Tamaulipasaurus morenoi, was described by Clark and
Hernández-Rivera (1994) and, more recently, the partial skel-
eton of a primitive pterosaur was described as Dimorphodon
weintraubi by Clark et al. (1998). Other elements of the fauna
include crocodyliforms, represented by several partial skulls and
postcranial skeletons, and ornithischian and theropod dinosaurs,
represented by isolated teeth.
The presence of mammaliaforms (sensu Rowe, 1988) was re-
ported by Fastovsky et al. (1987) and briefly described by Clark
et al. (1991, 1994), Fastovsky et al. (1995), and Montellano et al.
(1995), but as yet none has been given a detailed description. The
purpose of this paper is to describe the mammaliaform maxilla
and lower jaws collected in the Jurassic redbeds of Huizachal
Canyon. The partial skull (IGM 6618) briefly described by Clark
et al. (1994) will be described at a later time.
GEOLOGIC AND GEOGRAPHIC SETTING
Huizachal Canyon is one of a series of east-west trending ero-
sional canyons exposing pre-Late Jurassic strata in the Sierra
*
Corresponding author.
Journal of Vertebrate Paleontology 28(4):1130–1143, December 2008
© 2008 by the Society of Vertebrate Paleontology
1130
Madre Oriental, the north-south trending mountain range abut-
ting the Gulf of Mexico coastal plain province in northeastern
Mexico. The canyon, located approximately 15 km southwest of
Ciudad Victoria (Fig. 1), exposes east-dipping bedded red sand-
stones, mudstones, and siltstones, which compose the type sec-
tion of the Huizachal Formation (Imlay et al., 1948). This for-
mation was later raised to group status and divided into the La
Boca Formation and the overlying La Joya Formation (Mixon et
al., 1959). Recent work (Strater, 1993; Fastovsky et al., 2005)
suggests that, at least in Huizachal Canyon, the red silty mud-
stone sequence of the La Boca Formation can be divided into an
upper and a lower unit representing, respectively, epiclastic and
pyroclastic sedimentation.
Recent structural and petrographic studies indicate that the
Huizachal Group was deposited unconformably upon an older,
undescribed, sequence of pyroclastic volcanic rocks (Fastovsky
et al., 2005). These authors suggested that the rocks underwent
extensive late-stage or post-depositional silicification, and the
trace element geochemistry of the volcanic unit shows that these
rocks range from sub-alkaline basalt to rhyolite suggesting sub-
alkalic ocean-continent Andean volcanism. Fastovsky et al.
(2005) proposed that the volcanic and sedimentary rocks of Hu-
izachal Canyon were deposited in a convergent plate margin
setting.
The fossils come from a <10 m thick sequence in the lower part
of the Huizachal Group that is the result of pyroclastic volcanic
deposition (Fastovsky et al., 1995, 2005). U-Pb isotopic data from
zircon in a volcaniclastic rock from the lowest part of the Hu-
izachal Group (La Boca Fm.) yield an age of 189 ± 0.2 Ma
(analytical error) (Fastovsky et al., 2005). The fossil-bearing
sedimentary rocks are above this unit and conformable with it.
The radiometric date places these beds in the Pleinsbachian ma-
rine stage of the late Early Jurassic, which is consistent with their
contained fauna.
Abbreviations—IGM, Instituto de Geología, UNAM; Uni-
versidad Nacional Autóhoma de México; Fm., Formation; P/p,
premolar; M/m, molar; Ma, million years ago. We use the cusp
nomenclature of Crompton and Jenkins (1968) in which cusps
are identified by letters, upper case in the upper dentition and
lower case in the lower dentition.
SYSTEMATIC PALEONTOLOGY
Because the systematic relationships of the La Boca mamma-
liaforms are uncertain, a cladistic analysis of mammaliaforms
with “triconodont”-like molars was performed in order to deter-
mine the likely affinities of the best-preserved specimens (IGM
3493, 6617, 6619, 6855). We used the data matrix of Rougier et al.
(2007), adding two additional characters (see Appendices 1 and
2). We ordered the eight multistate characters (1, 5, 10, 12, 19, 23,
24, 62) that Rougier et al. (2007) ordered in their analysis.
The strict consensus of 10 trees (Fig. 2) closely resembles the
single most parsimonious tree of Rougier et al. (2007, fig. 7),
except that our consensus tree reverses the positions of Dinne-
therium and docodonts (Haldanodon and Docodon), groups Tri-
conodon and Trioracodon as sister taxa within Triconodontidae,
and places Jeholodens as the sister taxon to Gobiconodon and
Repenomamus. The “amphilestid”-like and therian stem taxa
form a pentachotomy consisting of Phascolotherium, Am-
philestes, the gobiconodont clade (including Jeholodens), the
Hakusanodon clade of Rougier et al. (2007), and the therian
stem clade. The Adams consensus tree groups Phascolotherium
and Amphilestes in a trichotomy with the stem therian clade. One
of our ten trees duplicates the distribution of “amphilestid”-like
and therian stem taxa in the single most parsimonious tree of
Rougier et al. (2007), except for the location of Jeholodens in the
gobiconodont clade.
The results of our cladistic analysis were utilized in determin-
ing the taxonomic allocations of La Boca specimens in the fol-
lowing systematic descriptions. Two specimens, IGM 6617
(named Bocaconodon tamaulipensis below) and 6855, fall be-
tween the non-mammalian mammaliaforms Dinnetherium and
docodonts and are thus considered to be Mammaliaformes, Fam-
ily incertae sedis. The other two specimens (IGM 3493 and 6619)
are considered to be mammals because the cladistic analysis
places them within groups usually considered to fall within Mam-
malia. IGM 3493 (named Victoriaconodon inaequalis below) is
nested within the family Triconodontidae and IGM 6619 (named
Huasteconodon wiblei below) is nested within a gobiconodont
clade (Repenomamus and Gobiconodon). Of the two specimens
not included in the cladistic analysis, one (IGM 6622) possesses
a prominent internal mandibular trough, whereas the other
(IGM 6856) lacks a trough. By convention, the first should be
excluded from and the second included in Mammalia. However,
as discussed below, the absence of the mandibular trough should
no longer be used as a diagnostic feature of Mammalia. There-
fore, we classify both of these specimens as Mammaliaformes,
Family incertae sedis.
We accept the node-based definition of mammals put forward
by Rowe (1988), in which Mammalia comprises the most recent
common ancestor of living Monotremata and Theria (Marsupia-
lia + Placentalia) and all of its descendants. Certain Late Triassic
and Jurassic taxa (e.g., Morganucodon, Megazostrodon, Din-
netherium, Docodon), traditionally regarded as mammals (see
Kielan-Jaworowska et al., 2004), are generally agreed to be more
primitive than monotremes and therians and thus lie outside of
FIGURE 1. Localition of Huizachal Canyon in Tamaulipas, México.
MONTELLANO ET AL.—LATE EARLY JURASSIC MAMMALIAFORMS 1131
Mammalia sensu Rowe. We follow Rowe by placing them in the
more inclusive taxon Mammaliaformes, defined by Rowe (1988)
as comprising the last common ancestor of Morganucodontidae
and Mammalia and all its descendants. Rowe’s diagnoses of
Mammaliaformes and Mammalia contain few characters rel-
evant to the specimens described here, the most significant being
the loss of the internal mandibular, or postdentary, trough
(“meckelian sulcus” of Rowe) as diagnostic of Mammalia, a fea-
ture correlated with the freeing of certain postdentary bones
from the mandible and their suspension from the skull as middle
ear bones. Non-mammalian mammaliaforms generally possess
an internal mandibular trough, known to have housed a rod of
postdentary elements, including the homologs of middle ear
bones in mammals. Mammals, by and large, lack the internal
mandibular trough and, presumably, the attachment of putative
middle ear bones to the lower jaw. Thus, it appears that absence
of the internal mandibular trough in certain La Boca lower jaws
should identify them as mammals, whereas its presence in other
jaws should exclude them from Mammalia and cause them to be
allocated to the more inclusive Mammaliaformes.
In recent years the distribution of the internal mandibular
trough among Mesozoic mammaliaforms, including mammals,
has come to cast doubt on the diagnostic value of this feature.
Luo et al. (2001) described Hadrocodium wui, based on tiny skull
and jaws from the Early Jurassic of China, which lacks the in-
ternal mandibular trough (and by inference possessed free
middle ear bones), but cladistically falls outside the Mammalia as
defined by Rowe (1988). More recently, Rich et al. (2005a,b)
described an internal mandibular trough in the Early Cretaceous
monotreme Teinolophos trusleri, interpreting its presence to in-
dicate the attachment of putative ear bones to the lower jaw and
thus the diphyletic origin of the definitive mammalian middle ear
in monotremes and other mammals. This interpretation is sup-
ported by phylogenetic studies of early mammals (Luo et al.,
2002; Kielan-Jaworowska et al., 2004) that group monotremes
with the Australosphenida, Jurassic and Early Cretaceous Gond-
wanan mammals retaining an internal mandibular trough or
other indications of a contact of postdentary bones with the
lower jaw. Even in the absence of an internal mandibular trough,
middle ear elements can still be attached to the mandible via an
ossified Meckel’s cartilage, as seen in the “triconodont” Yanoco-
nodon and several gobiconodontids (Luo et al., 2007). Thus, the
presence or absence of an internal mandibular trough is no
longer sufficient for determining whether middle ear bones were
attached to or separated from the mandible or whether a fossil
taxon should be excluded from or included in Mammalia.
FIGURE 2. Strict consensus of 10 most parsimonious trees of 172 steps found by PAUP version 4.0b10, using the Branch and Bound option. Data
matrix is that of Rougier et al. (2007) to which two additional characters have been added: 63—lingual curvature of rear of upper molar tooth row:
absent (0), present (1); 64—interlocking mechanism between adjacent lower molars: absent (0), cusp d fits between cusps e and b (1), cusp d fits
between cusps e and f (2), cusp d fits embayment or groove of b (3), posterior cusp of preceding tooth overlaps anterior cusp of succeeding tooth
(4). Consistency Index (CI)⳱0.5000, Homoplasy Index (HI)⳱0.5000, Retention Index (RI)⳱0.7604, Rescaled CI⳱0.3802.
JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 28, NO. 4, 20081132
All of the mammaliaform specimens from the La Boca For-
mation of Huizachal Canyon have molars of “triconodont” mor-
phology, i.e., the three main molar cusps are aligned in a single
row parallel to the long axis of the jaw. However, only two
specimens can be attributed to a known family. The term “tri-
conodont” has long been used to denote a variety of basal mam-
maliaform taxa with non-tribosphenic dentitions, but these do
not together form a clade (Rougier et al., 2001; Kielan-
Jaworowska et al., 2004; Rougier et al., 2007). Inasmuch as Rou-
gier et al. (2007) no longer recognize a sister taxon relationship
of gobiconodontids with triconodontids (Rougier et al, 2001) and
now place gobiconodontids in a paraphyletic series with “am-
philestids”, symmetrodonts, and therians, we provisionally re-
strict the formal taxon Triconodonta (and the essentially equiva-
lent Eutriconodonta of Kermack et al., 1973) to include the fam-
ily Triconodontidae only. The family Amphilestidae is also
paraphyletic (Rougier et al., 2007) and its included taxa are
herein referred to informally as “amphilestids”. Late Triassic
and Jurassic taxa of a primitive mammaliaform grade are usually
placed in the Morganucodontidae (e.g., Morganucodon and
Erythrotherium) and Megazostrodontidae (e.g., Megazostrodon,
possibly Dinnetherium), and are grouped within the Order Mor-
ganucodonta by Kielan-Jaworowska et al. (2004). They form a
paraphyletic series in the trees of Rougier et al. (2007) and in this
paper and are referred to hereafter as morganucodont-grade
mammaliaforms or “morganucodonts”.
Mammaliaformes Rowe 1988
Family incertae sedis
Bocaconodon gen. nov.
(Fig. 3)
Type and Only Species—Bocaconodon tamaulipensis.
Etymology—After La Boca Formation, plus Greek conos,
(cone) and odontos, (tooth).
Diagnosis—As for the species, by monotypy.
Bocaconodon tamaulipensis sp. nov.
Holotype—IGM 6617, partial right mandible, measuring 3.7
mm in preserved length, with last two molars and portion of third
tooth from rear. This specimen has not been completely re-
moved from the matrix and is exposed only on its medial side. (It
is described as “Taxon A” and figured as “’amphilestid’ mam-
maliaform” in Clark et al., 1994, and as “’amphilestid’ mammal”
in Fastovsky et al., 1995.)
Locality and Horizon—Jim’s Joy, Huizachal Canyon, Muni-
cipio de Ciudad Victoria, Tamaulipas, Mexico. Lower part of the
La Boca Formation.
Age—Late Early Jurassic (Fastovsky et al., 2005).
Etymology—In reference to the name of the state of Tama-
ulipas where Huizachal Canyon is located.
Diagnosis—Mammaliaform with molars resembling those of
Morganucodon, Megazostrodon, and Erythrotherium in having
three main cusps oriented longitudinally, with cusp a anterior to
the middle of the crown and cusp b much smaller than cusp c.
Bocaconodon differs from these taxa in having a taller, more
recurved cusp a with a concave rather than convex distal margin;
taller, more anteroposteriorly compressed cusps b and c; and a
lingual cingulum that lacks cusps e and g and is generally smooth
and non-cuspidate. It also differs from these taxa in that cusp b
lies well above the cingulum. The cingulum resembles that of
Amphilestes in being concave below cusps b and c and highest
below cusp a. The dentary possesses an internal trough (for post-
dentary bones) and an internal mandibular (meckelian) groove
that lies parallel to the lower border of the jaw rather than con-
verging on the lower border as in Morganucodon and Mega-
zostrodon.
Description—The horizontal ramus is slender, only slightly
deeper than the height of the largest tooth, appearing shallowest
below the last molar. The alveolar margin of the dentary is
straight and the lower border curves smoothly down and forward
from behind the level of the last tooth. A short distance below
the alveolar margin is a shallow longitudinal groove that may be
homologous with the groove for the dental lamina seen in Mor-
ganucodon (Parrington, 1971; Kermack et al., 1973). Below the
molars is a shallow internal mandibular (meckelian) groove that
parallels the lower border of the jaw. The groove widens below
the rear half of the last molar and merges into a large, deep
longitudinal trough. The trough is similar in morphology to that
which in Morganucodon and other Early Jurassic mammalia-
forms houses a slender rod of postdentary jaw bones. At the
anterior end of the trough lies the inferior alveolar foramen,
through which the mandibular nerve entered the lower jaw; the
foramen lies below cusp d of the last molar, a position seen in
some specimens of Morganucodon (Kermack et al., 1973).
Above the trough, on the medial surface of the base of the
coronoid process, is a depressed, roughened area that undoubt-
edly was the site of attachment of a coronoid bone. The lower
jaw of this new taxon was at the same evolutionary grade as that
of Morganucodon in that the postdentary bones, which in living
mammals form free ear ossicles, are still functional lower jaw
elements undoubtedly associated with a functional articular-
quadrate jaw articulation as well as a mammalian dentary-
squamosal jaw joint.
The incomplete third molar from the rear preserves only the
last cusp (cusp d), which is inclined posteriorly; tooth size rela-
tive to the two complete teeth cannot be determined. The lengths
of the penultimate and last molars are 1.05 mm and 0.9 mm,
respectively. The last two molars each have four cusps aligned
parallel to the long axis of the jaw. The cusps in the two teeth are
similar in relative height, but those of the last molar are more
anteroposteriorly compressed. The second cusp, cusp a, is the
principal cusp and is much larger than the cusps anterior (cusp b)
and posterior (cusp c) to it. Behind cusp c is a small cusp d, which
is continuous with the internal cingulum. The complete cingulum
of the penultimate molar is narrow and without cuspules, being
highest internal to the main cusp and concave upward lingual to
cusps b and c. Posteriorly, it rises behind c to form the small cusp
d. The cingulum of the last molar is similar, though it is incom-
plete internal to the main cusp.
In both molars, cusp a is moderately recurved, with the ante-
rior margin being strongly recurved and the rear margin only
slightly so. Cusp b is nearly symmetrical, with a dorsally directed
FIGURE 3. Bocaconodon tamaulipensis gen. et sp. nov., IGM 6617,
right partial dentary with three teeth, in medial view. Abbreviations:
a,b,c,d, lower molar cusps a, b, c, d; cor, facet for coronoid bone; d.l,
groove for dental lamina; i.g, internal mandibular groove; m.t, internal
mandibular trough.
MONTELLANO ET AL.—LATE EARLY JURASSIC MAMMALIAFORMS 1133
apex. It is much smaller than cusp c, being about half the height
of the latter. In the penultimate molar, the anterior margin of the
tooth forms a rounded bulge below the base of cusp b; in the last
molar this region forms only a slight anterior swelling. Cusp c is
also nearly symmetrical, with slightly convex margins, but its
apex is directed slightly backward. The posterior cingulum cusp
d is small and pointed, inclining more posteriorly than cusp c.
Discussion—The preserved molars of Bocaconodon resemble
those of morganucodont-grade mammaliaforms (Morganu-
codon, Megazostrodon, Erythrotherium) in having the b cusp
much smaller than c and lying at the front margin of the crown.
They differ from those of “morganucodonts,” and resemble
more derived groups of “triconodonts,” in having a narrower
internal cingulum lacking distinct cusps, including the enlarged
cusp at the level of the notch between cusps a and c, called the
Kühnecone by Parrington (1971) and cusp g by Crompton and
Jenkins (1968).
Dinnetherium nezorum from the Kayenta Formation of Ari-
zona (Jenkins et al., 1983) resembles IGM 6617 in lacking a
distinct cusp g on the lingual cingulum, but differs in having
essentially symmetrical crowns with cusps b and c of equal height
and a series of well-developed cusps on the lingual cingulum.
Bocaconodon resembles all of the above taxa in retaining an
internal mandibular trough for postdentary bones. In Middle
Jurassic and later “triconodonts,” an internal mandibular trough
is absent, suggesting, though not demonstrating (see Luo et al.,
2007), that the postdentary bones have become definitive ear
ossicles no longer contacting the lower jaw in the adult.
The molars of Bocaconodon show significant differences from
those of Amphilestes and Phascolotherium, the oldest and best-
known “amphilestids,” from the Middle Jurassic Stonesfield
Slate of England. “Amphilestid” molars are nearly symmetrical
from front to back, with each of the three main cusps being
symmetrical or only slightly recurved and cusps b and c being
nearly equal in size; also, the internal cingulum bears accessory
cusps both anterior and posterior to the main cusps, giving a
quinquecuspid character to the crown.
Simpson (1925b) attributed two lower jaws from the Late Ju-
rassic Morrison Formation of Wyoming to the “Amphilestidae”:
Comodon (“Phascolodon”) gidleyi and Aploconodon comoensis.
Comodon differs from Bocaconodon in that its cusp b is taller
than cusp c and a lingual cingulum is lacking, except for a promi-
nent anterointernal cusp (cusp e) not present in the Mexican
specimen. The cusps of Comodon are also shorter and more
rounded, with convex margins and more symmetrical outlines.
The penultimate molar of Aploconodon is a symmetrical tooth
with a tall, anteroposteriorly long, main cusp with convex mar-
gins, small b and c cusps, and lacking distinct anterior and pos-
terior cingulum cusps.
Bocaconodon shows no special resemblances to members of
the family Triconodontidae, in which the three main cusps are
either equal in height or, as in Priacodon, cusps b and c are
slightly lower than cusp a (Simpson, 1925a). Nor does it resemble
members of the Gobiconodontidae, in which the lower molars
are essentially symmetrical, with cusps b and c subequal in size
and anterior and posterior cingulum cusps are present.
This new specimen cannot be placed in a known family of
mammaliaforms. Although it resembles Morganucodon, Eryth-
rotherium, and Megazostrodon in the small size of cusp b, it more
closely resembles Middle Jurassic and younger “triconodonts” in
the reduction of the internal cingulum.
Family incertae sedis
Gen. and sp. indeterminate
(Fig. 4)
Referred Material—IGM 6622, portion of right dentary with
last two molars. (Described by Clark et al., (1994) as part of
“Taxon C” in association with IGM 6619, described below as
Huasteconodon wiblei, with which it almost surely is not closely
related.)
Locality and Horizon—Casa de Fidencio, Huizachal Canyon,
Municipio de Ciudad Victoria, Tamaulipas, Mexico. Lower part
of the La Boca Formation.
Age—Late Early Jurassic (Fastovsky et al., 2005).
Description—The fragmentary dentary preserves a small por-
tion of the horizontal ramus holding the last two molars, part of
the anterior border of the coronoid process, part of the condylar
process, and part of the lower margin including a small angular
or pseudangular process (the homology of the “pseudangular”
process with the angular process of therians is uncertain and we
prefer to use the latter term). The coronoid process ascends at an
angle of about 45 degrees lateral to the last molar, becoming
slightly more vertical above the level of the tooth row. It has a
distinct anterior ridge, which bounds a concave masseteric fossa
on the lateral surface of the ascending ramus. Above the level of
the angular process is a low rounded ridge that extends forward
FIGURE 4. Mammaliaformes, family, genus and species indetermi-
nate, IGM 6622, right dentary with last two molars, in A, lateral and B,
medial views, and C, molars in occlusal view. Abbreviations: a,b,c, lower
molar cusps a, b, c; a.p, angular (“pseudangular”) process; i.g, internal
mandibular groove; m.t, internal mandibular trough.
JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 28, NO. 4, 20081134
from the condylar process and curves upward toward the last
molar; it forms the lower and anterior border of a deeper portion
of the masseteric fossa.
On the medial side of the dentary is a large fossa bounded by
the anterior border of the coronoid process and a ridge that
descends behind the last molar to merge into a prominent hori-
zontal flange extending back on to the condylar process. Below
the overhanging flange is a broad trough that extends forward
into a narrower groove underlying the tooth row on the hori-
zontal ramus. The internal trough and overhanging flange are
similar to those of Morganucodon, indicating the retention of the
postdentary bones in the lower jaw. The low horizontal ridge on
the lateral surface of the dentary is a reflection of the trough on
the medial surface.
The angular process lies below the trough, extending only
slightly ventral to the more anterior lower border of the jaw. The
floor of the internal trough curves down on to the posterior
border of the angular process, leaving a depression like that in
Morganucodon that houses the proximal part of the reflected
lamina of the angular (Kermack et al., 1973, fig. 7).
The two molars have divided roots, which slightly diverge ven-
trally. Neither possesses an external cingulum. The crown of the
penultimate tooth is damaged but preserves a main cusp (cusp a)
slightly anterior to its midlength and a small accessory cusp (b) in
front of it. Distinct posterior accessory cusps cannot be distin-
guished. Damage obscures features of the lingual surface of the
tooth. The last molar is complete and well preserved. It is 0.8 mm
long. It has been shifted slightly out of its alveolus so its crown
lies above the level of the preceding molar. Its main cusp (a) is
very tall and distinctly recurved, with its tip lying anterior to its
midlength. The anterior accessory cusp (b) lies low on the crown
and extends forward to overhang the front of the anterior root.
The crest behind the cusp tip is horizontal. Behind the main cusp
is a single posterior accessory cusp (c), which is about twice as
long anteroposteriorly as the anterior cusp. Its marginal crest is
slightly convex and extends only a short distance above the level
of cusp b. A slight internal crest extends down and back from
cusp b and another down and forward from cusp c, both crests
merging into the rounded lingual face of the main cusp.
Discussion—This specimen preserves too few details for
meaningful comparison with other taxa. It resembles Middle Ju-
rassic and later “triconodonts,” and differs from “morganu-
codonts” in having reduced internal cingula that lack cusp g or
smaller cuspules and in having cusps b and c of about equal
height. However, it is primitive in having a recurved cusp a lo-
cated anterior to the center of the crown and in retaining an
internal trough for postdentary bones and an angular process, fea-
tures present in “morganucodonts” and absent in “triconodonts.”
Family incertae sedis
Gen. and sp. indeterminate
(Fig. 5)
Referred Material—IGM 6855, partial right dentary preserv-
ing the posterior part of the horizontal ramus, which contains the
last two molars and half of the next more anterior molar, and a
large portion of the ascending ramus.
Locality and Horizon—Cementerio, Huizachal Canyon, Mu-
nicipio de Ciudad Victoria, Tamaulipas, Mexico. Lower part of
the La Boca Formation.
Age—Late Early Jurassic (Fastovsky et al., 2005).
Description—This lower jaw is more complete than IGM
6622, to which it shows the greatest resemblance. Its ventral
profile is slightly convex below the horizontal ramus, slightly
concave below the anterior margin of the masseteric fossa, and
then slightly convex below the ascending ramus. The coronoid
process curves smoothly up and back from the level of the tooth
row, much as in IGM 6622. The crest forming the anterior border
of the masseteric fossa extends down the anterior border of the
coronoid process to about a tooth’s length behind the last molar,
then curves down and back to merge into the ventrolateral sur-
face. The masseteric fossa is partially divided by a low ridge that
extends forward, and slightly downward, from the region of the
condyle (not preserved) to merge into the surface of the fossa.
The specimen does not preserve evidence of an angular process.
On the medial surface of the horizontal ramus is a distinct
internal (meckelian) groove, which parallels the lower border of
the jaw and extends back beyond the tooth row. A shallow
groove for the dental lamina, as described in Morganucodon by
Parrington (1971) and Kermack et al. (1973), lies internal to the
alveoli and, posterior to the last molar, curves gently downward
toward the rear of the internal groove. The pterygoid fossa is
covered by matrix, so it is not possible to determine whether an
internal mandibular trough for postdentary bones was present.
However, the presence of a ridge on the lateral surface suggests
the presence of a trough.
The molars are proportionally larger with respect to the depth
of the horizontal ramus than in IGM 6622. All are slightly dam-
aged labially and all but the first are well-preserved lingually.
The posterior root of the incomplete anterior-most molar is ex-
posed on the lateral break; it extends at least three-fourths of the
distance to the lower border of the ramus. The crown preserves
a damaged posterior accessory cusp (c) and a small posterior
cingulum cusp (d), which overlaps the anterolateral surface of
the next tooth almost to the level of the tip of cusp b. A cingulum
cannot be distinguished on this damaged tooth. The penultimate
molar is 0.7 mm long and has a low, rounded anterior accessory
FIGURE 5. Mammaliaformes, family, genus and species indetermi-
nate, IGM 6855, right dentary with last two molars and incomplete third
tooth from rear, in A, lateral and B, medial views, and C, molars in
occlusal view. Abbreviations: a,b,c,d,e, lower molariform cusps a, b, c, d,
e; d?, possible cusp d; i.g, internal mandibular groove.
MONTELLANO ET AL.—LATE EARLY JURASSIC MAMMALIAFORMS 1135
cusp (b), anterolingual to which is a prominent cingulum cusp,
probably homologous to cusp e of morganucodontids. The main
cusp is damaged, so its height is unknown, but it was narrow and
elongated. Behind its damaged base is a large posterior cusp that
seems to be derived from the cingulum and is, therefore, cusp d.
Under this interpretation, there is no c cusp. Possibly this cusp
lay closer to the damaged main cusp and is now unrecognizable
because of damage and/or wear. The posterior cusp overlaps the
labial side of the anterior cusp b of the last molar. A complete
lingual cingulum extends from the tip of the anterointernal cusp
e to the tip of the posterior cusp d. It rises slightly on the internal
surface of the main cusp, and encloses a continuous shelf, which
is narrowest internal to the main cusp and widens anterior and
posterior to this cusp.
The last molar is 0.58 mm long, about two-thirds the length of
the preceding tooth, and has a single root. The main cusp is
rather small, less than one-third the length of the crown. It ap-
pears to be slightly recurved, although its tip is truncated due to
damage. The anterior accessory cusp (b) is lower than cusp a, but
is as long. It is higher than the posterior accessory cusp (c), which
is very low, extending only slightly above the level of the internal
cingulum. The posterior cingulum cusp (d) is not clearly present,
though its lack of differentiation from cusp c may be due to
damage or wear. A distinct internal cingulum is present extend-
ing from the lingual side of cusp b, where a small cusp e lies
posterior to the anterior margin of b, back to the posteromedial
margin of the crown. The cingulum has a slightly undulating
dorsal border and encloses a continuous elongate shelf.
Discussion—The dentition of this specimen shows some re-
semblance to that of IGM 6622 in having a relatively prominent
cusp b and inconspicuous cusp c, but it differs in having a smaller
main cusp and a more prominent lingual cingulum on the last
molar. Also, the last molar has only a single root rather than two.
The crowns overlap one another, as in Sinoconodon (Crompton
and Sun, 1985) rather than interlocking as in morganucodontids,
triconodontids, gobiconodontids, and “amphilestids,” It is not
known whether this specimen possessed an internal trough for
postdentary bones, though the presence of a lateral ridge, which
is seen in all taxa possessing an internal trough, suggests that it
was possibly present.
In the cladistic analysis, IGM 6855 is the sister taxon of Bo-
caconodon though they show few dental similarities. They form
a clade between Dinnetherium and the docodontids Haldanodon
and Docodon.
Family incertae sedis
Gen. and sp. indeterminate
(Fig. 6)
Referred Material—IGM 6856, left lower jaw, measuring 8.4
mm in preserved length, with evidence of seven teeth: five pre-
served teeth (second, third, and fifth to seventh), the root of an
anterior-most tooth, and paired alveoli of a fourth tooth.
Locality and Horizon—Cementerio, Huizachal Canyon, Mu-
nicipio de Ciudad Victoria, Tamaulipas, Mexico. Lower part of
the La Boca Formation.
Age—Late Early Jurassic (Fastovsky et al., 2005).
Description—This is the most complete, and smallest, lower
jaw described here, consisting of most of the horizontal ramus
and the lower part of the masseteric attachment area. The length
of the tooth row is 5.6 mm and the greatest depth of the jaw is 1.3
mm. Unfortunately, the teeth are heavily damaged, except for
the tiny last molar. For this reason we have chosen not to name
this specimen, even though the dentary possesses a unique com-
bination of features that might permit attribution of new speci-
mens to the same taxon.
The jaw is very shallow, especially anteriorly, where it tapers
toward the incisor/canine region. Its ventral border forms a
nearly smooth curve, deepest below the penultimate tooth. Be-
low the posterior half of the second preserved tooth is a very
large mental foramen that opens anteriorly into a deep groove.
The groove passes forward and slightly upward to fade out just
anterior to the first preserved tooth. The masseteric fossa is
bounded below by a rounded ridge and in front by a ridge that
extends upward on to the anterior border of the coronoid pro-
cess. The fossa extends forward to a point below the rear of the
last tooth. The coronoid process is missing, but its broken base
lies a short distance behind and slightly lateral to the last tooth.
In section, it is transversely widest in its anterior half and tapers
to a thin lamina posteriorly.
On the medial side of the dentary, the symphysis is a long, low
surface that extends back to a point below the space between the
first two preserved teeth. Its contact surface with the opposite
jaw consists of a shallow oblique trough bounded below by an
oblique ridge. The horizontal ramus lacks all traces of an internal
(meckelian) groove; the only features on the medial surface are
several striations on the lower half of the jaw ventral to the last
two teeth and the base of the coronoid process. Likewise, an
internal mandibular trough for postdentary bones is lacking, in-
dicating that postdentary bones no longer attach to the lower
jaw. Near the center of the preserved length of the coronoid
process is an oval depression behind which is a narrow horizontal
groove that continues to the broken posterior edge of the jaw.
The mandibular foramen, which would be expected to lie pos-
teroventral to the last molar (in the vicinity of the oval depres-
sion), is not preserved on the specimen.
Immediately behind the anterior break across the symphysis is
a small narrow oval root that may represent an anterior premo-
lar, perhaps p1, or a very small canine. Behind the root the jaw
is slightly constricted, as seen in dorsal view, and possesses a
short diastema. The first two preserved teeth are interpreted
here as premolars, perhaps the second and third, or, if the ante-
rior root represents a canine, the first and second. The anterior-
most tooth has a small rounded crown with a slight anterior
bulge, which probably represents an accessory cusp, and a pair of
widely splayed roots. The crown of the second tooth is slightly
more than twice the length of the first and is transversely very
narrow; the remnant of the rounded main cusp lies above the
anterior root (which extends obliquely down and forward) and
FIGURE 6. Mammaliaformes, family, genus and species indetermi-
nate, IGM 6856, left dentary with five preserved cheek teeth, in A,
lateral, B, occlusal and C, medial views. Abbreviations: a,b, lower molar
cusps a, b; m.f, mental foramen; r, root of presumed first premolar.
JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 28, NO. 4, 20081136
its posterior surface slopes down and back into a low rounded
posterior cusp.
Behind the second tooth is a pair of alveoli that we interpret as
probably representing the first molar; this would yield a molar
count of four, which we believe is more likely for an early “tri-
conodont” than three. The next preserved tooth, the probable
m2, is heavily damaged and worn, preserving no useful features
of the crown. The penultimate tooth, the probable m3, has a low
anterior accessory cusp and a larger main cusp, which lies pri-
marily above the anterior root. The posterior half of the crown is
damaged. Both of these molars are very compressed from side to
side. The last molar is the smallest preserved tooth; it is single-
rooted with a symmetrical crown, the main cusp being centrally-
placed and the small anterior and posterior accessory cusps being
of equal size. None of the molars shows evidence of a lingual
cingulum.
Discussion—The most notable feature of this specimen is the
absence of an internal groove below the tooth row and the ap-
parent absence of an internal trough for postdentary bones. The
dentition is poorly preserved and the numbers of premolars and
molars are uncertain, so useful comparisons with other early
mammaliaforms are not possible. The possible dental formula of
p3m4 is also seen in the triconodontid Priacodon, but otherwise
the dentition shows no special resemblances to that of trico-
nodontids. The location of the principal cusp on the anterior half
of the crown is a primitive condition seen in “morganucodonts”
but not in “amphilestids” and gobiconodontids.
Mammalia Linnaeus, 1758 (sensu Rowe, 1988)
Triconodonta Osborn 1888
Definition—A clade consisting of the common ancestor of the
Triconodontidae (sensu stricto) plus any taxa more closely re-
lated to the Triconodontidae than to “morganucodonts”, spala-
cotheriids, tinodontids, and multituberculates (Kielan-
Jaworowska et al., 2004:216). Because “Amphilestidae” is cur-
rently regarded either as “a structural grade of primitive
eutriconodontans” (Kielan-Jaworowska et al., 2004:218) or as a
paraphyletic group basal to trechnotheres (Rougier et al., 2001),
and Gobiconodontidae appears to be part of this same paraphy-
letic series (Rougier et al., 2007), Triconodonta at present con-
tains only the family Triconodontidae. Whether continued use of
the term Triconodonta is useful or not merits further consider-
ation.
Diagnosis—As for Triconodontidae.
Family Triconodontidae, Marsh 1887
Diagnosis—“Triconodonts” with molars in which the princi-
pal cusps are usually of nearly equal height; lower molars char-
acterized by reduction or loss of cusp e and by a unique tongue-
in-groove interlocking mechanism that lacks involvement of cin-
gulum cuspules; cusp d fits into a basally placed embayment on
the mesial face of cusp b on the succeeding molar (Kielan-
Jaworowska et al., 2004:244). In our cladistic analysis, unequivo-
cal synapomorphies of Triconodontidae are: 1(2) fewer than
three lower incisors; 8(0) last lower premolariform taller than
first molariform; 36(1) mesiodistal size of lower molariforms in-
creases posteriorly; 45(1) masseteric ridge forms prominent
shelf; 57(1) coronoid bone absent; 58(1) flat occlusal surface on
upper molariforms; 64(3) interlocking mechanism between cusp
d and embayment or groove of cusp b.
Victoriaconodon gen. nov.
(Fig. 7)
Type and Only Species—Victoriaconodon inaequalis.
Etymology—After Ciudad Victoria, the closest town to the
locality; and from the Greek conos (cone) and odontos (tooth).
Diagnosis—As for the species, by monotypy.
Victoriaconodon inaequalis sp. nov.
Holotype—IGM 3493, anterior portion of a right mandible
with p2-4, m1, and partial alveolus of canine and alveolus of p1.
(Figured as “triconodont mammal” in Fastovsky et al., 1987, and
described as “Taxon B” in Clark et al., 1994.)
Locality and Horizon—Rene’s Roost, Huizachal Canyon,
Municipio de Ciudad Victoria, Tamaulipas, Mexico. Lower part
of the La Boca Formation.
Age—Late Early Jurassic (Fastovsky et al., 2005).
Etymology—From the Latin inaequalis with reference to the
unequal height of the molar cusps, a primitive feature relative to
the condition in most Late Jurassic and Cretaceous triconodon-
tids.
Diagnosis—Resembles triconodontids (especially Priacodon)
in having a great disparity in height between the last lower pre-
molar and first molar (character 8(0)); premolars showing an
anterior to posterior graded size increase (as in Triconodon and
FIGURE 7. Victoriaconodon inaequalis gen. et sp. nov., IGM 3493,
right dentary with p2-m1, in A, lateral, B, medial and C, occlusal views.
Abbreviations: a,b,c,d, lower premolar cusps a, b, c, d, and lower molar
cusps a, b, c; m1, first lower molar; m.f, mental foramen; p1, alveolus of
first lower premolar.
MONTELLANO ET AL.—LATE EARLY JURASSIC MAMMALIAFORMS 1137
Trioracodon, as well as in Morganucodon, where the disparity in
height between p4 and m1 is much less); and asymmetrical pre-
molars with a well-developed anterior accessory (b) cusp and, in
the more posterior premolars, a prominent c cusp and a small d
cusp. Differs from all triconodontids in having a cusp e and from
all but Priacodon in that molar cusp a is distinctly taller than the
subequal b and c cusps. It differs from Priacodon in having four
rather than three premolars, and from Triconodon and Triora-
codon in that p1 is single- rather than double-rooted. It differs
from Morganucodon in having symmetrical molars and a well-
developed b cusp on the premolars, and from “amphilestids” and
gobiconodontids in having a taller last premolar relative to m1. It
differs from Morganucodon, “amphilestids”, and some Jurassic
triconodontids (Trioracodon and perhaps Triconodon (Simpson,
1928), but not Priacodon) in lacking an internal (meckelian)
groove below the first molar (and usually the posterior premo-
lars).
Description—The specimen is broken across the middle of the
canine alveolus and immediately behind m1. Its preserved length
is 5.7 mm and its depth below p1 is 1.65 mm and below p4 is 2.1
mm. The jaw bears four double-rooted teeth, p2-4 and m1, and
the oval alveolus for a single-rooted p1. A recessed mental fo-
ramen lies below the alveolus for p1. A prominent groove runs
forward from the foramen. The medial surface shows no evi-
dence of an internal (meckelian) groove or a groove for the
dental lamina.
The premolars increase in size posteriorly with p4 being much
larger than p3. All possess anterior and posterior basal cusps,
which are continuous lingually with a slight cingulum. The cusps
are aligned anteroposteriorly. The p2 is 0.7 mm long, with a low,
asymmetrical central cusp flanked anteriorly and posteriorly by
small, rounded accessory cusps. A narrow, relatively straight,
cingulum extends across the lingual face of the crown between
the accessory cusps, the latter of which resembles the d cusp of
the following premolars. The p3 is 0.95 mm in length. Its central
cusp (a) is broken, but it appears to have been relatively tall and
slightly asymmetrical. A small but distinct b cusp inclines slightly
forward to overhang the anterior root. The narrow internal cin-
gulum extends from the b cusp to a small rounded posterior cusp
that overhangs the posterior root, the d cusp. Above the level of
the cingulum, at the base of the posterior ridge of cusp a is a
small, vertical cusp c about equal in diameter to cusp b, but about
a third taller. The p4 is 1.3 mm in length, at least half again as
long and tall as p3. The central cusp is damaged apically, but was
clearly taller than long. It appears to have been slightly asym-
metrical, with its anterior ridge more convex than its posterior.
The small b cusp is continuous with the anterointernal cingulum,
which drops steeply down from the cusp and curves upward to
merge with a vertical ridge that bisects the lingual face of the
central cusp. On the rear margin of the main cusp is a small
rounded cusp c that lies above the small cusp d, which is conflu-
ent with the posterointernal cingulum. This cingulum extends
forward below cusp c and upward a short distance on the inner
surface of the main cusp, where it fades out before reaching the
vertical ridge.
The sole preserved molar is ∼1.1 mm long, slightly shorter than
p4 and much lower in height. The main cusp of m1 is broken
apically, but was probably about twice the height of the pre-
served portion of its base. It is flanked by much smaller cusps b
and c, each of which is about half its basal length. The anterior
cusp (b) is slightly smaller than the posterior cusp (c). There is a
narrow cingulum anterior to cusp b and posterior to cusp c. It
extends on to the lingual side of the crown below the accessory
cusps, where it is slightly cuspidate, and extends upward on to
the inner face of cusp a, where it fades out. A small cusp e lies
anterointernal to cusp b. A distinct d cusp is not present, but the
rear of m1 may be slightly damaged.
Discussion—In the strict consensus tree, Victoriaconodon in-
aequalis is nested within Triconodontidae, being more derived
than Priacodon, and forming the sister group to the remaining
triconodontids on the basis of three unequivocal synapomor-
phies: 5(0) number of lower premolariforms increased from
three to four; 16(1) crown height less than half the jaw height;
and 41(1) depth of the jaw between canine and last molariform
increases posteriorly. None of these characters inspires confi-
dence. The plesiomorphic number of premolariforms among Tri-
conodontidae is uncertain. Furthermore, the dentary in the ho-
lotype and only specimen of V. inaequalis is lacking posterior to
m1 and may have borne taller molariforms and been shallower
than suggested by the part preserved.
Victoriaconodon resembles “morganucodonts” and trico-
nodontids, and differs from “amphilestids” and gobiconodontids,
in having a size gradient in the premolars, with a gradual size
increase posteriorly from p1, and with p4 being taller than both
p3 and m1, although the discrepancy in height of p4 to m1 is
greater in Victoriaconodon and triconodontids than in “morga-
nucodonts”. Victoriaconodon differs from triconodontids in that
molar cusps b and c are much lower than cusp a, whereas in most
triconodontids the height differential between the main and ac-
cessory molar cusps is much less. Only in Priacodon are cusps b
and c “a little shorter relative to a” (Simpson, 1925a). However,
Victoriaconodon is more like Triconodon, Trioracodon, and Cre-
taceous triconodontids in having four lower premolars, and un-
like Priacodon, which has only three (Simpson, 1925a; Rasmus-
sen and Callison, 1981), presumably having lost p1 (though the
analysis suggests that the condition in Priacodon is primitive for
the family). The single root of p1 in Victoriaconodon suggests
that this tooth may be in process of reduction and possible loss.
Comparison with morganucodont-grade taxa (e.g., Morganu-
codon, Megazostrodon, Dinnetherium) and Amphilestes suggests
that the primitive lower premolar number was at least four.
However, the docodontid outgroup to the “triconodont” plus
therian clade has three premolariforms (four to three in
Docodon), and the Mexican specimens forming the outgroup to
docodonts lack preserved premolariforms, so the loss of one
premolariform is indicated in most trees to have occurred at the
node above Dinnetherium.
Triconodontidae are not very diverse. Three genera of trico-
nodontids—Priacodon, Triconodon, and Trioracodon—have
been described from the Late Jurassic, at least six forms from the
Early Cretaceous in North America, Asia, and Africa, and one
genus from the Late Cretaceous of North America (Cifelli et al.,
1998). This new possible triconodontid would be the oldest mem-
ber of the group by far and shows the primitive condition in
which the height of molar cusp a is relatively much greater than
that of b and c.
Family Gobiconodontidae Chow and Rich, 1984
(⳱ Gobiconodontinae Chow and Rich, 1984;
Gobiconodontidae Jenkins and Schaff, 1988)
Diagnosis—“Triconodonts” with last upper molar lying lin-
gually and somewhat obliquely to more anterior molars (Jenkins
and Schaff, 1988; Kielan-Jaworowska and Dashzeveg, 1998), re-
sulting in a lingually curving tooth row (Rougier et al., 2001) that
extends medial to the base of the zygomatic process of the max-
illa (in contrast to Triconodontidae, in which the tooth row ex-
tends on to the zygomatic process). Three main upper cusps
appear to form a slight triangle in the more posterior molars,
with cusp A lying lingual to cusps B and C (though the degree of
angulation may be exaggerated by wear). Upper molar roots
transversely widened. Palate lingual to the molars with promi-
nent fossae for reception of tall cusp a of lower molars.
JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 28, NO. 4, 20081138
Huasteconodon gen. nov.
(Fig. 8)
Type and Only Species—Huasteconodon wiblei.
Etymology—After the Huastecos, a Mesoamerican culture
that existed between 750 BCE and 800 CE in La Huasteca, a
region to the southeast of Ciudad Victoria, plus Greek conos
(cone), and odontos (tooth).
Diagnosis—As for the species, by monotypy.
Huasteconodon wiblei sp. nov.
Holotype—IGM 6619, fragment of tiny left maxilla with two
molars and alveoli for a third.
Locality and Horizon—Huizachal Canyon, Municipio de Ciu-
dad Victoria, Tamaulipas, México. Lower part of the La Boca
Formation.
Age—Late Early Jurassic (Fastovsky et al., 2005).
Etymology—For Dr. John R. Wible, in recognition of his im-
portant studies of Mesozoic mammals and mammalian phylog-
eny.
Diagnosis—Possible gobiconodontid “triconodont” of very
small size (last preserved (penultimate?) upper molar is .55 mm
in length) in which the upper molars are labiolingually narrower
than in Cretaceous gobiconodontids and have a more prominent
ectoflexus. Cusp C appears to be small and cusp D greatly en-
larged. The roots are labiolingually wider than the crown, being
visible internal and external to the crown in occlusal view.
Description—This tiny fragment of a left maxilla consists of a
portion of the lateral surface including a small part of the zygo-
matic process, the alveolar margin containing two molars and a
pair of alveoli for a third (ultimate) molar, and a portion of the
secondary palate internal to the first two molars. The distal-most
alveolus, which appears to be much larger than the mesial alveo-
lus, lacks its posteromedial rim, but it is bounded posterolaterally
by a sheet of bone with a natural rear margin that extends lat-
erally on to the medial surface of the zygomatic process. Thus,
the posterior border of the maxilla lay immediately behind the
last alveolus, indicating that the preserved alveoli held the last
molar. The ventral surface of the palatal plate contains three oval
depressions for the reception of the tips of the main cusps of the
last three occluding lower molars. The first depression lies inter-
nal to the anterior half of the first preserved molar, the second
internal to the gap between the first and second molars and the
anterior half of the second molar, and the third centered on the
gap between the second and third molars.
The teeth are heavily worn, but show some details of the
crowns. The molar roots are exposed only slightly below the
labial margin of the maxilla, but on the lingual side the alveolar
margin lies much higher so that much more of the roots of both
teeth is exposed. The roots of the second molar appear to be
transversely wider than the crown, as they are visible both labial
and lingual to the crown in occlusal view.
The precise pattern of molar cusps is not totally clear. The
central cusp (cusp A) is preserved in both molars as an antero-
posteriorly elongated, low prominence with a rounded apex
(damaged in the second preserved molar). The anterior acces-
sory cusp (B) is preserved in the second molar, where it lies close
to the anterior margin of the crown as a worn rounded projec-
tion. The identity of the posterior accessory cusp is not totally
clear. In the first molar, a definite ridge-like rounded cusp lies on
the posterior slope of cusp A and anterior to what appears to be
a posterior cingulum cusp; this would be cusp C, or the meta-
cone, and the posterior cusp would be cusp D, the metastyle. In
the second molar, however, the presumed C cusp is much smaller
and the presumed cingulum cusp (metastyle) is very large, pro-
jecting posterolaterally as a rounded lobe. In crown view, it is
seen that the main cusp (A) lies lingual to the accessory cusps so
that the line of four cusps forms a lingually-directed obtuse angle
of about 120 degrees, with cusp A at the apex.
An internal cingulum is lacking, possibly due to the heavy
wear the molars have suffered. A well-developed external cin-
gulum with an irregular margin is present on the posterior half of
both molars, extending between the main cusp and the metasty-
lar cusp. The cingulum appears to be absent on the anterior half
of the second preserved molar. When viewed from below, the
labial profile of the crown is deeply indented at the level of the
main cusp and expands outward fore and aft. The lingual wear
surface extends posterolingually from cusp B to the apex of cusp
A and curves posterolabially from A to the tip of the metastylar
cusp.
Discussion—The molars of Huasteconodon most closely re-
semble those of Gobiconodon (Jenkins and Schaff, 1988; Kielan-
Jaworowska and Dashzeveg, 1998) in the angulation of the cusps,
the deep indentation of the outer surface labial to cusp A, and
the transversely broad roots. In occlusion, the main lower cusp
appears to have inserted into the V-shaped embrasure between
the two upper molars. This is as occurred in “amphilestids” (in
which the upper dentition is unknown) and gobiconodontids, but
not in triconodontids, where principal lower cusp a occluded
between upper cusps B and A. Huasteconodon also resembles
Gobiconodon, but apparently not Repenomamus, in having a
medial curvature of the rear part of the upper molariform row
and resembles both genera in having prominent depressions in
the palate for reception of the tips of the main lower cusps.
In our cladistic analysis (Fig. 8), Huasteconodon shares char-
acter 59(1), labial margin of posterior molariforms bilobate, with
Repenomamus and Gobiconodon and character 63(1), lingual
curvature of rear of upper molariform tooth row, with Gobico-
nodon. Though Huasteconodon shows similarities to gobico-
nodontids of the Early Cretaceous of Asia and western North
America, and we place it in Gobiconodontidae on the basis of
our cladistic analysis, we believe that the only known specimen
of Huasteconodon is too fragmentary and too distant in time to
inspire complete confidence in this familial allocation.
DISCUSSION
The fossil record of mammals and their near mammaliaform
relatives prior to the Late Jurassic is poorly known, especially for
the interval from the middle part of the Early Jurassic through
FIGURE 8. Huasteconodon wiblei gen. et sp. nov., IGM 6619, frag-
ment of left maxilla with two molars and alveoli for a third, in A, lateral,
B, occlusal and C, medial views. Abbreviations: A,B,C,D, upper molar
cusps A, B, C, D; alv, alveolus; z.p, zygomatic process of maxilla.
MONTELLANO ET AL.—LATE EARLY JURASSIC MAMMALIAFORMS 1139
the Middle Jurassic. The Huizachal fauna samples the early part
of this interval and clearly represents a more derived group of
mammaliaforms than was present in the “morganucodont” fau-
nas of Late Triassic and earlier Jurassic time. It is also clearly
more primitive than the fauna of the Middle Jurassic Stonesfield
Slate. The “morganucodont” faunas are characterized by a vari-
ety of taxa in which the postdentary bones are invariably re-
tained as part of the lower jaw, an angular (“pseudangular”)
process is usually present, and cuspidate lingual cingula are usu-
ally prominent features of the lower molariforms. The Stones-
field fauna is characterized by mammals, particularly “amphiles-
tids,” in which the postdentary bones appear to be entirely or
largely freed from contact with the adult dentary, an angular
process is absent, and molar cingula are reduced and usually
without individual cusps. The Huizachal fauna is evolutionarily
intermediate in that some taxa (e.g., IGM 6856) appear to have
lost the postdentary bones from the lower jaw, whereas others
(e.g., IGM 6617, 6622, and possibly 6855) retain them (as indi-
cated by the persisting internal mandibular trough), the angular
process is present with certainty only in IGM 6622, and molari-
form cingula are reduced and usually lack individual cusps.
The taxonomic composition of “morganucodont”-dominated
faunas of the Late Triassic and earliest Jurassic is fairly diverse,
with a variety of taxa with cusp-in-line Morganucodon-like mo-
lars (including Megazostrodon, Brachyzostrodon, Erythrothe-
rium, Dinnetherium), as well as sinoconodontids with cusp-in line
molars of a simpler morphology, kuehneotheriids with reversed-
triangle molars, and haramiyids with multiple rows of enlarged
cusps. Likewise, in the Middle Jurassic are “amphilestids” with
cusp-in-line molars, Amphitherium, peramurids, and dryolestids
with a reversed-triangles molar pattern, and docodontids,
haramiyid- or multituberculate-like taxa (Kermack et al., 1998),
and possibly true multituberculates (Butler and Hooker, 2005)
with multiple cusp rows. The Huizachal fauna of late Early Ju-
rassic age contains only taxa with cusp-in-line molars, taxa that
would conventionally be called “triconodonts.”
Although all of the Huizachal mammaliaforms are dentally
“triconodonts,” only two can be placed in known families,
though with less than complete confidence. Although some share
dental synapomorphies with Middle Jurassic and younger “tri-
conodonts” (“amphilestids,” triconodontids, gobiconodontids),
such as reduction of cingula and loss of cingular cusps (including
cusp g), some retain the internal mandibular trough, a feature
otherwise unknown in Middle Jurassic and younger “trico-
nodonts.” This suggests that the loss of the trough, and freeing of
the postdentary bones/ear ossicles from the lower jaw, may have
occurred independently in a number of separate “triconodont”
lineages (see also Rich et al., 2005a,b; Luo et al., 2007). Were it
possible to confidently ally some of the specimens possessing an
internal mandibular trough with later “triconodont” taxa, this
suggestion would be greatly strengthened.
Few late Early and Middle Jurassic localities bearing mamma-
lian remains are known in the world and what material is known
consists mainly of teeth and relatively few jaw fragments. The
mammals from Tamaulipas appear to represent new taxa of “tri-
conodont” mammaliaforms that are not closely similar to Early
and Middle Jurassic mammaliaforms described from elsewhere,
although some show tantalizing resemblances to Late Jurassic
triconodontids and Early Cretaceous gobiconodontids. They in-
dicate the presence of a radiation of early mammaliaforms, some
appearing to be true mammals, that show variations on the
primitive cusps-in-line (“triconodont”) molar pattern, but differ
widely in lower jaw morphology.
ACKNOWLEDGMENTS
We sincerely thank the Instituto de Geología, Universidad
Nacional Autónoma de México, Consejo Estatal para la Cultura
y las Artes del Estado de Tamaulipas for their support. The work
could not have been done without the friendly cooperation of the
Huizachal Ejido, and we specially thank Fidencio Moreno and
his family for advice, assistance, and hospitality. Some of the
fossils were prepared by W. Amaral, D. Chaney, and C. Vander-
slice. Assistance in the field was rendered by René Hernández,
Víctor-Hugo Reynoso, and Paul Sereno. The illustrations were
prepared by Claire Vanderslice. The advice of Drs. Richard Ci-
felli, Zhe-Xi Luo, Yaoming Hu, and an anonymous reviewer
greatly improved the manuscript.
The work was supported by National Geographic Society
grants 3017-85 and 3695-87 to J. A. Hopson and J. M. Clark, and
by National Science Foundation grants EAR 8917386 and
EAR9218971 to D. E. Fastovsky and EAR9218871 to J. M.
Clark.
LITERATURE CITED
Aguillón-Martínez, M. C., D. Zelenistsky, and D. Eberth. 2004. Eggshell
fragments from the Uppermost Cerro del Pueblo Formation (Upper
Cretaceous) Mexico. Journal of Vertebrate Paleontology 24(3,
Supplement):33A.
Archibald, J. D., A. O. Averianov, and E. G. Ekdale. Late Cretaceous
relatives of rabbits, rodents, and other extant eutherian mammals.
Nature 414:62–65.
Butler, P. M., and J. J. Hooker. 2005. New teeth of allotherian mammals
from the English Bathonian, including the earliest multitubercu-
lates. Acta Palaeontologica Polonica 50:185–207.
Chow, M., and T. S. Rich. 1984. A new triconodontan (Mammalia) from
the Jurassic of China. Journal of Vertebrate Paleontology 4:226–231.
Cifelli, R. L., J. R. Wible, and F. A. Jenkins, Jr. 1998. Triconodont
mammals from the Cloverly Formation (Lower Cretaceous), Mon-
tana and Wyoming. Journal of Vertebrate Paleontology 18:237–241.
Clark, J. M., and R. Hernández-Rivera. 1994. A new burrowing diapsid
from the Jurassic La Boca Formation of Tamaulipas, Mexico. Jour-
nal of Vertebrate Paleontology 14:180–195.
Clark, J. M., and J. A. Hopson. 1985. Distinctive mammal-like reptile
from Mexico and its bearing on the phylogeny of the Tritylodonti-
dae. Nature 315:398–400.
Clark, J. M., J. A. Hopson, R. Hernández, D. E. Fastovsky, and M.
Montellano. 1998. Foot posture in a primitive pterosaur. Nature 391:
886–889.
Clark, J. M., M. Montellano, J. A. Hopson, and D. E. Fastovsky. 1994.
An Early or Middle Jurassic tetrapod assemblage from the La Boca
Formation, northeastern Mexico; pp 295–302 in N. C. Fraser and
H. D. Sues (eds.), In the Shadow of the Dinosaurs: Early Mesozoic
Tetrapods. Cambridge University Press, New York.
Clark, J. M., M. Montellano, J. A. Hopson, and R. Hernández. 1991.
Mammals and other tetrapods from the early Jurassic La Boca For-
mation, Northeastern Mexico. Journal of Vertebrate Paleontology
11(3, Supplement): 23A.
Clemens, W. A. 1980. Gallolestes pachymandibularis (Theria, incertae
sedis; Mammalia) from Late Cretaceous deposits in Baja California
del Norte, Mexico. PaleBios, Museum of Paleontology, University
of California, Berkeley 33:1–10
Crompton, A. W., and F. A. Jenkins. 1968. Molar occlusion in Late
Triassic mammals. Biological Reviews 43:427–458.
Crompton, A. W., and A. L. Sun. 1985. Cranial structure and relation-
ships of the Liassic mammal Sinoconodon. Zoological Journal of the
Linnean Society 85:99–119.
Fastovsky, D. E., J. M. Clark, and J. A. Hopson. 1987. Preliminary report
of a vertebrate fauna from an unusual paleoenvironmental setting,
Huizachal Group, Early or Mid-Jurassic, Tamaulipas, Mexico; pp.
82–87 in P. M. Currie and E. H. Koster (eds.), Fourth Symposium on
Mesozoic Terrestrial Ecosystems, Short Papers. Occasional Papers
of the Tyrrell Museum of Paleontology 3.
Fastovsky, D. E., J. M. Clark, N. H. Strater, M. Montellano, R. Hernán-
dez, and J. A. Hopson. 1995. Depositional Environments of a
Middle Jurassic terrestrial vertebrate assemblage, Huizachal Can-
yon, Mexico. Journal of Vertebrate Paleontology 15:561–575.
Fastovsky, D. E., O. D. Hermes, N. H. Strater, S. A. Bowring, J. M.
Clark, M. Montellano, and R. Hernandez R. 2005. Pre-Late Jurassic,
fossil-bearing red beds of Huizachal Canyon, Tamaulipas, Mexico.
JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 28, NO. 4, 20081140
In Anderson, T.H., Nourse, J.A., McKee, J.W., and Steiner, M.B.
(eds.), The Mojave-Sonora megashear hypothesis: Development, as-
sessment, and alternatives. Geological Society of America Special
Paper 393:401–426.
Imlay, R. W., E. Cepeda, M. Álvarez, and T. Díaz. 1948. Stratigraphic
relations of certain Jurassic formations of eastern Mexico: American
Association of Petroleum Geologists Bulletin 32:1750–1761.
Jenkins, F. A., A. W. Crompton, and W. R. Downs. 1983. Mesozoic
Mammals from Arizona: new evidence on mammalian evolution.
Science 222:1233–1235.
Jenkins, F. A., and C. R. Schaff. 1988. The Early Cretaceous mammal
Gobiconodon (Mammalia Triconodonta) from the Cloverly Forma-
tion of Montana. Journal Vertebrate Paleontology 8:1–24.
Kermack, K. A., W. M. Kermack, P. M. Lees, and J. R. E., Mills. 1998.
New multituberculate-like teeth from the Middle Jurassic of En-
gland. Acta Palaeontologica Polonica 43:581–606.
Kermack, K. A., F. M. Mussett, and H. W. Rigney. 1973. The lower jaw
of Morganucodon. Zoological Journal of Linnean Society 53:87–175.
Kielan-Jaworowska, Z., R. L. Cifelli, and Z.-X. Luo. 2004. Mammals
from the Age of Dinosaurs: Origins, Evolution, and Structure. Co-
lumbia University Press, New York.
Kielan-Jaworowska, Z., and D. Dashzeveg. 1998. Early Cretaceous am-
philestid (‘triconodont’) mammals from Mongolia. Acta Palaeonto-
logica Polonica 43:413–438.
Lillegraven, J. A. 1972. Preliminary report on Late Cretaceous mammals
from the El Gallo Formation, Baja California del Norte, Mexico.
Contributions in Science. Los Angeles County Museum of Natural
History 232:1–11.
Lillegraven, J. A. 1976. A new genus of therian mammal from the Late
Cretaceous “El Gallo Formation”, Baja California, Mexico. Journal
of Paleontology 50:437–443.
Linnaeus, C. 1758. Systema naturae per regna tria naturae, secundum
classes, ordines, genera, species, cum characteribus, differentiis, syn-
onymis, locis. Editio decima, reformata. Laurentii Salvii, Stockholm,
Vol. 1, 824 pp.
Luo, Z.-X., P.-J. Chen, G. Li, and M. Chen. 2007. A new eutriconodont
mammal and evolutionary development of early mammals. Nature
446:288–293.
Luo, Z.-X., A. W. Crompton, and A.-L. Sun. 2001. A new mammalia-
form from the Early Jurassic of China and evolution of mammalian
characteristics. Science 292:1535–1540.
Luo, Z.-X., Z. Kielan-Jaworowska, and R. L. Cifelli. 2002. In quest for a
phylogeny of Mesozoic mammals. Acta Palaeontologica Polonica
47:1–78.
Marsh, O. C. 1887. American Jurassic mammals. American Journal of
Science, ser. 3, 3:326–348.
Mixon, R. B., G. E. Murray, and G. Díaz. 1959. Age and correlation of
Huizachal Group (Mesozoic), state of Tamaulipas, Mexico. Ameri-
can Association of Petroleum Geologists Bulletin 43:757–771.
Montellano, M., J. A. Hopson, J. M. Clark, D. F. Fastovsky, and R.
Hernández-Rivera. 1995. Mammals from the Middle Jurassic of Hu-
izachal Canyon, Tamaulipas, Mexico. Journal of Vertebrate Paleon-
tology 15(3, Supplement):45A.
Osborn, H. F. 1888. On the structure and classification of the Mesozoic
Mammalia. Journal of the Academy of Natural Sciences, Philadel-
phia 9:186–265.
Parrington, F. R. 1971. On the Upper Triassic mammals. Philosophical
Transactions of the Royal Society of London (B) 261:231–272.
Rasmussen, T. E., and G. Callison. 1981. A new species of triconodont
mammal from the upper Jurassic of Colorado. Journal of Paleon-
tology 55:628–634.
Reynoso, V. H. 1996. A Middle Jurassic Sphenodon-like sphenodontian
(Diapsida: Lepidosauria) from Huizachal Canyon, Tamaulipas,
Mexico. Journal of Vertebrate Paleontology 16:210–221.
Reynoso, V. H. 2005. Possible evidence of a venom apparatus in a
Middle Jurassic sphenodontian from the Huizachal red beds of
Tamaulipas, Mexico. Journal of Vertebrate Paleontology 25:
646–654.
Reynoso, V. H., and J. M. Clark. 1998. A dwarf sphenodontian from the
Huizachal Formation of Tamaulipas, Mexico. Journal of Vertebrate
Paleontology 18:333–338.
Reynoso-Rosales, V. H. 1992. Descripción de los esfenodontes del Ju-
rásico temprano o medio del Cañon del Huizachal, Tamaulipas,
México. M.S. thesis, Facultad de Ciencias, Universidad Nacional
Autónoma de México, Mexico City, México, 86 pp.
Rich, T. H., J. A. Hopson, A. M. Musser, T. F. Flannery, and P. Vickers-
Rich. 2005a. Independent origins of middle ear bones in mono-
tremes and therians. Science 307:910–914.
Rich, T. H., J. A. Hopson, A. M. Musser, T. F. Flannery, and P. Vickers-
Rich. 2005b. Response to comments on “Independent origins of
middle ear bones in monotremes and therians.” Science 307:1492c.
Rougier, G. W., Isaji, S., and Manabe, M. 2007. An Early Cretaceous
mammal from the Kuwajima Formation (Tetori Group), Japan, and
a reassessment of triconodont phylogeny. Annals of Carnegie Mu-
seum 76(2):73–115.
Rougier, G. W., M. J. Novacek, M. C. McKenna, and J. R. Wible. 2001.
Gobiconodonts from the Early Cretaceous of Oshih (Ashile), Mon-
golia. American Museum Novitates 3348:1–30.
Rowe, T. 1988. Definition, diagnosis and origin of Mammalia. Journal of
Vertebrate Paleontology 8:241–264.
Simpson, G. G. 1925a. Mesozoic Mammalia, I. American triconodonts,
Part 1: American Journal of Science, ser. 5, 10:145–165.
Simpson, G. G. 1925b. Mesozoic Mammalia, I. American triconodonts,
Part 2: American Journal of Science, ser. 5, 10:334–358.
Simpson, G. G. 1928. A Catalogue of the Mesozoic Mammalia in the
Geological Department of the British Museum. Trustees of the Brit-
ish Museum, London, 215 pp.
Strater, N. H. 1993. The origin of pre-Late Jurassic strata of Huizachal
Canyon (Tamaulipas, Mexico) and their relationship to the tectonic
evolution of northeastern Mexico. M.S. Thesis, University of Rhode
Island, Kingston, Rhode Island, 169 pp.
Wible, J. R., G. W. Rougier, M. J. Novacek, and R. J. Asher. 2007.
Cretaceous eutherians and Laurasian origin for placental mammals
near rhe K-T boundary. Nature 477:1003–1006.
Submitted October 10, 2007; accepted March 28, 2008.
APPENDIX 1. Character List.
(Characters 1–62 are from Rougier et al. (2007); characters 63–64 are
new. Rougier et al. (2007) used the terms “premolariforms” and “mo-
lariforms” for the teeth we refer to as premolars and molars.)
Lower Dentition:
1) Number of incisors: four or more (0), three (1), fewer than 3 (2);
ORDERED.
2) Large mesial incisor: absent (0), present (1);
3) Canine shape: caniniform (0), premolariform (1), incisiform (2);
4) Canine roots: one (0), two (1);
5) Lower premolars: four or more (0), 3 (1), two (2); ORDERED.
6) Anterior premolars tall, peglike, and single-rooted: absent (0), pres-
ent (1);
7) Premolars: asymmetrical (0), symmetrical (1);
8) Last lower premolar: taller than first molariform (0), shorter or sub-
equal to first molariform (1);
9) Cusp c in last premolar: present in crown (0), connected to cingulum
or absent (1);
10) Number of lower molars: fewer than five (0), five (1), more than 5
(2); ORDERED.
11) Occlusion of molars: A cusp occludes between a and c cusps (0), A
occludes either posterior to c and/or on anterior face of b (1);
12) Crown cusps on first lower molar: arranged in line (0), forming an
open triangle (>90 degrees) (1), forming a more closed triangle (<90
degrees) (2); ORDERED.
13) Crown of lower molar: higher than long (0), longer than high (1);
14) Locking mechanism between lower molars: poorly developed, at-
tained by interlocking cusps or notched cingula (0), locking involves
shallowly excavated anterior root (1), locking involves deeply exca-
vated anterior root (2);
15) Lower molar lingual cingulum: straight cingulum with cuspules (0),
smooth and undulant (1), absent (2);
16) Crown height relative to jaw height under tallest molar: greater than
1/2 jaw height (0), less than 1/2 jaw height (1);
17) Recumbent molar crown cusps: absent (0), present (1);
18) Position of cusp a in molars: mesially placed on crown (0), centrally
placed on crown (1);
19) Bulging of labial slope of cusp a: absent (0), moderate (1), forming
a rounded eminence at the base of the crown (2); forming the apex
of an acute triangle (trigonid) (3); ORDERED.
20) Height of cusp a: tallest (0), subequal in height to b and c (1);
MONTELLANO ET AL.—LATE EARLY JURASSIC MAMMALIAFORMS 1141
21) Relative height of cusp b to c: b lower than c (0), about equal to c (1),
b taller than c (2),
22) Cusp b: connected to cingulum (0), in crown, independent from
cingulum (1);
23) Cusp b position: aligned with a along the mandibular axis (0), lingual
to a in posterior molars forming angle of 20 degrees or less with
mandibular axis (1), b lingual to a forming angle greater than 20
degrees (2); ORDERED.
24) Cusp c position: aligned with a in mandibular axis (0), lingual to a in
posterior molar forming angle of 25 degrees or less with mandibular
axis (1), lingual to a forming angle greater than 25 degrees (2);
ORDERED.
25) Cusp c: present (0), absent (1);
26) Conspicuous shelf formed by cusp d on rear of molar: absent (0),
present (1);
27) Cusp d: is small cingular cusp (0), is tall and incorporated into the
crown (1);
28) Tall, narrow, finger-like cusp d: absent (0), present (1);
29) Accessory cusp f: absent (0), present (1);
30) Cusp e; present(0), absent (1);
31) Kühnecone (cusp g): present (0), connected to cusp a by crest (1),
absent (2);
32) Size of last lower molar: about 50% or less of preceding tooth (0),
subequal to or larger than preceding tooth (1);
33) Last lower molar: crown and accessory cusps present (0), cingulum
or cingular cusps greatly reduced or absent (1);
34) Shape of molar cusps in cross section: symmetrical or slightly buc-
colingually compressed (0), asymmetrical, with lingual face flat or
concave; ⳱ (1);
35) Roots of last molar: two (0), one (1), incipiently bifurcated (2);
36) Mesiodistal change in size of molars: all subequal (0), increases pos-
teriorly (1) decreases posteriorly (2);
37) Section of molar roots: oval or circular (0), mesiodistally compressed
(1);
38) Alignment of last molar with respect to coronoid process: medial to
(0), in line with (1);
39) Cutting surfaces on crown: shaped by extensive wear (0), present
upon eruption, with only limited remodeling by wear (1);
Lower jaw:
40) Large mental foramina: fewer than three (0), three or more (1);
41) Depth of lower jaw between canine and last molar: sub-uniform (0),
becomes deeper posteriorly (1);
42) Ventral edge of lower jaw: convex ventrally (0), nearly straight (1);
43) Coronoid process: mesiodistally narrow (0), mesiodistally broad (1);
44) Coronoid process: rear slope straight and directed posteroventrally
(0), approaches vertical or is concave (1);
45) Masseteric ridge: small to absent (0), forms prominent shelf (1);
46) Medial flange: present (0), absent (1);
47) Pseudangular process: present (0), absent (1);
48) Dentary: Not expanded posteriorly to support condyle ventrally (0),
expanded posteriorly to support condyle ventrally (1);
49) Remnant groove for dental lamina: present (0), absent (1);
50) Meckelian groove: confluent with ventral edge of jaw at about level
of first molariform (0), does not reach ventral edge of jaw (1), absent
(2);
51) Pterygoid fossa: absent (0), present (1);
52) Broad groove for attachment of postdentary elements: present (0),
absent (1);
53) Dentary-squamosal articulation: supplemented by articular-
quadrate articulation (0), only load-bearing articulation between
skull and lower jaw (1);
54) Pterygoid crest: absent (0), strong, straight crest connecting ventral
edge of jaw with condyle (1), curved around ventral edge of jaw (2),
ends at angular process (3);
55) Shelf-like lateral ridge on dentary: present (0), absent (1);
56) Angular process; absent (0), present (1);
57) Coronoid bone: present (0), absent (1);
Upper Dentition:
58) Flat occlusal surface on molars: absent (0), present (1);
59) Labial margin of posterior molars: straight, or slightly concave/
convex (0), bilobate (1);
60) Number of principal cusps on last molar: three (0), two (1);
61) Outline of last molar in occlusal view: oval (0), triangular (1);
62) Position of cusp A relative to cusps B and C: A aligned with B and
C along the axis of the tooth row (0); A lies lingual to B and C to
form an angle of 90 degrees or more (1), A lies lingual to B and C
to form an angle of less than 90 degrees (2); ORDERED.
New Characters:
63) Lingual curvature of rear of upper molar tooth row: absent (0),
present (1);
64) Interlocking mechanism between adjacent lower molars: absent (0),
cusp d fits between cusps e and b (1), cusp d fits between cusps e and
f (2), cusp d fits embayment or groove of b (3), posterior cusp of
preceding tooth overlaps anterior cusp of succeeding tooth (4);
JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 28, NO. 4, 20081142
APPENDIX 2. Matrix
10 20 30 40 50 60
Morganucodon 0000000000 0000000000 0000000000 0000000000 0000000000 0000000000 0001
Erythrotherium A000000100 0000000000 0000000000 0??0?00000 0000000000 0000000000 00?1
Megazostrodon 0000000101 1000000000 0100010000 0100000000 0000000110 0000?00010 0001
Kuehneotherium 0000000??2 1A00000110 B111000010 21?00?0000 0000000010 000000001? ?1?2
Woutersia ?????????? 1?000?0110 B101000010 0??0??0?0? ?????????? ???????0A? ?1??
Dinnetherium 0000000101 0000100100 1100000010 0100000001 0011000110 0000110010 0002
Haldanodon A001100101 ?110010110 ?10?110000 1100000000 0011000011 1000000000 0000
Docodon A001A00102 ?110010110 ?10?110000 1100000000 0011000011 1000000000 0000
Jeholodens 01?020011A 1000100110 11000000?0 2??0?0000? 0011?1111? 1112100000 ?1?3
Priacodon 2000100000 0010100100 1100000001 2100010100 0011111111 1111101111 1003
Triconodon B000000000 0010110101 1100000001 2100010100 1111111111 1111101100 1003
Trioracodon 2000000000 0010110101 1100000001 2100010100 1111111111 1111101101 1003
Astroconodon ??00000001 0012111101 1100001101 2101010000 10??11??12 111?101100 10?3
Alticonodon ?????????? ??121?0101 1100001101 2111010?0? ????????1? 1?1????1?? ?0?3
Corviconodon ??00000?01 0012111101 1100001101 201101000? 101?111?12 111B??1??? ???3
Aploconodon ?????????? 1?00200110 1100000000 201020010? ?????11?1A 11??1?0??? ????
Phascolotherium 0001101111 1000100110 1100000010 2100?00101 0011011110 1112100??? ???2
Amphilestes 0001001011 1000100110 1100000010 2100000101 0011011110 1112100??? ???2
Amphidon ????10??01 1000100110 1100000010 201012010? 001?011110 11121?0??? ????
Hakusanodon ??01200101 1000100120 2100000010 2010120101 00??011?10 1112?????? ????
Comodon ?????????? 1?00100120 2100000010 2010120?0? ????????1? 11???0???? ???2
Gobiconodon 11202101?1 10011A0100 1100000010 2100020101 1011011110 1112100010 0112
Tinodon A011101110 1100100130 1111000010 2100000100 0011011110 111210000? ?102
Zhangeotheriids 1110100112 1200100130 1122000000 2000101001 000101111? 1112100000 12??
Spalacotheriids A001100102 1200100130 1122000000 2100001011 0011111110 1112100010 0204
Amphitherium 0001000112 1210200130 1122000010 2110000011 0011011110 11131100?? 1??4
Repenomamus 2120210111 1101110100 1100000010 2100000101 1111011110 1112100010 010?
Arundelconodon ????00000A 0012111101 1100000001 2??1?10?00 1???????11 ?1???????? ???3
Bocaconodon ?????????? ?00?1010?0 01000000?1 210?0200?? 00???0??01 0000??0??? ????
Victoriaconodon ??0000000? ?01?110100 1100000??0 2?????0??0 10??????1? ?????????? ????
IGM6855 ?????????? ?0101001?0 2100000?00 210?1200?? 00??0???01 ????0????? ???4
Huasteconodon ?????????? 1????????? ?????????? ????????0? ?????????? ???????01? 011?
A⳱0+1, B⳱!+2
MONTELLANO ET AL.—LATE EARLY JURASSIC MAMMALIAFORMS 1143
