Thylacoleo carnifex, which is commonly referred to as the "marsupial lion", was a rather robustly built Pleistocene animal approximately the size of a large leopard or medium-sized African lion.  It had a broad skull with a blunt rostrum and forward-facing binocular eyes.  The dentition of Thylacoleo is among its most peculiar features.  Its dental formula is expressed as I ³/1, C ¹/0, P ³/3, M ¹/2.  This dentition is essentially similar to that of the phalangerid marsupials, except that there is a reduction in the size and number of molars. Thylacoleo's enormous caniniform maxillary first incisors, semi-procumbent, laniariform mandibular first incisors, and the gigantic upper and lower slicing third premolars (P ³/3) are sure to attract the attention of even those who lack a specific interest in mammalian dental anatomy.

    The maxillary canine (C¹) along with the first two premolars (P¹/1, P²/2) are significantly reduced and generally considered to be functionless. Thylacoleo's dentition is so unique that it has created a number of different views on the phylogenetic position of the animal and generated great debates regarding its feeding habits.  Features of the skull, particularly the dentition, were virtually the only evidence for early discussions of Thylacoleo, as until quite recently, the postcranial skeleton was only known from a few isolated limb bones which had been recovered in the proximity of Thylacoleo skulls.

    To follow is a summary of the discovery and interpretation of the various fossils which have been found of Thylacoleo carnifex.  A functional interpretation of the feeding habits and locomotion of Thylacoleo will be presented based upon examination of the known, as well as some new, material.

The First Discoveries and Descriptions:
 

. A plate from an 1838 work by T. L. Mitchell depicting some of Owen's drawings of fossils from the Wellington Caves of New South Wales.  Among them are teeth which would later be classified as those of Thylacoleo.

The first evidence for the existence of Thylacoleo came from material collected in the early 1830s from the Wellington Valley region, New South Wales, by Major (later Sir) Thomas Mitchell.  Incisor and premolar teeth, now known to be of Thylacoleo, were depicted in Mitchell's 1838 work "Three Expeditions into the Interior of Eastern Australia", but were not detailed in the Zoological Appendix to volume 2 contributed by Professor Richard Owen, the well known English palaeontologist.  Owen's opinion (1877, vol. 1: 107) was that the premolar, the "large, two-fanged sectorial tooth", was so extraordinary in form that it was quite impossible to make a determination of the animal's characters without an examination of more substantial material.     Owen later obtained fragments of left and right maxillae, each including P3 and M1, and a section of a left dentary with P3 intact but bearing only the roots of M1 and M2.  These specimens had been discovered in the bed of a tributary of the Condamine River, Queensland, but no date of when they had been found was recorded.  Lydekker's 1887"Catalogue of the Fossil Mammalia in the British Museum (Natural History)" lists M3653 and M3654, the right and left maxillary fragments of Thylacoleo from Queensland, as having been figured by Owen in 1887 and bestowed by him to the Museum.

    Owen described the material from the Condamine River and commented on the position of the "working surfaces" of both the maxillary and mandibular sectorials in his summary of research on the fossil mammals of Australia (1877).  He made note of the vertical striae and concluded that these teeth functioned as shearing blades used in the cutting of meat.  However, he could find no indication of how the animal could be reconciled taxonomically among its marsupial or placental relatives.

    Studies conducted by M. E. Finch (1982) of the specimens M3653 and M3654 have shown the presence of a well-defined ridge near the posterior end of the wear facet of each maxillary and mandibular "carnassial" and a second, more definite ridge at the interrootial level on the occlusal surface of M3654.  The latter ridge is not developed in other specimens available for study and the former is usually only a low boundary between the wear facets which result from contact of the lower carnassial and the adjacent lower first molar.  Owen made mention that the ridges were similar to

those seen in the sectorial teeth of Felis spelaea and Hyena spelaea, however, on present evidence is can be considered quite probable that the Condamine animal had been afflicted with a deformity of, or damage to, its mandibular dentition during life, leading to a peculiar wear pattern on the maxillary premolar.     Fragmentary skull material of Thylacoleo was collected by W. Adeney in 1845 from his property at Lake Colongulac, some eighty miles southwest of Melbourne, Victoria, and was forwarded a year later by Dr. H. Hobson of Melbourne to Owen at the British Museum.  Included was a cranium with an incomplete posterolateral surface and missing its zygomatic arches.  Also in this series was . . Views of the right side of a Thylacoleo mandible.  Source: Owen (1887). .

a maxillary fragment which retained the carnassial premolar and the single molar, which was shown to be part of the fragmentary cranium.  Owen concluded that any valid interpretation of the similarities and functionality of this amazing material could only be based upon detailed comparisons of Thylacoleo with a variety of other mammals.  Such examinations were not completed until 1858, when Owen presented his research to the Royal Society.  Included in his account was information about a cast of a mandibular section which he had received from Samuel Stutchbury, Geological Surveyor of New South Wales.  This cast consisted of much of the ramus and included the carnassial premolar (P3), M1, and the alveolus of M2, and it easily indicated the nature of the characters of the symphysis.  This initial account of Thylacoleo was based upon fossil material from two widely separated locales, the skull from Victoria and the mandible from King Creek, Darling Downs, Queensland.  The definitive description of the species was not published until 1859, as only an abstract and an article "Odontology" appeared in the 1858 edition of the Encyclopaedia Britannica.

    An even earlier reference to Thylacoleo is to be found in Gervais' "Zoologie et Paleontologie Francaises" published in sections between the years of 1848 and 1852, but no dates are stated for the individual sections.  Apparently, when Gervais visited Owen in London in 1848, he was shown the lake Colungulac fossils during a discussion on the affinities of Hyaenodon, a placental creodont which at the time was thought by many authorities to be a marsupial.  Gervais states in his book that Owen had named the Australian material Thylacoleo but provided no description of it.  This reference to Thylacoleo was apparently unknown to Stutchbury who, in a technical report (1853) to the Colonial Secretary, had given the name Schizodon to the original specimen of the mandibular cast which he had provided to Owen after making a comparison of the carnassial with that figured by Mitchell (1838).  Owen (1859) makes note that even if Schizodon Stutchbury had priority over Thylacoleo, it was a homonym of Schizodon Agassiz and Schizodon Waterhouse, names which had been assigned to a fish in 1829 and to a mammal in 1842, respectively.

In his paper of 1859, Owen gave a condensed yet comprehensive description of the Lake Colongulac and Darling Downs material.  However, he stated Hodgson's Creek rather than King Creek as the source of the mandible.  In the paper, he used the species name carnifex but did not specify a holotype which has led Mahoney and Ride (1975) to record as syntypes both the cranium and the mandible upon which Owen's description is based.  Based upon extensive comparisons between Thylacoleo and and the marsupial carnivores Dasyurus (Sarcophilus) ursinus (harrisii) and Thylacinus harrisii (cynocephalus), and also with Felis spelaea, the placental cave lion, Owen had no doubt  that Thylacoleo was a marsupial.   He pointed out its large posterior palatal vacuities, the position of the basicranial foramina, the participation (though a small one) of the alisphenoid in the auditory bulla whilst the tympanic process was essentially free, the extension of the lachrymal bone on to the face, the extraorbital position of the lachrymal foramen and the rather small size of the brain in relation to skull size.  He proposed that Thylacoleo was most closely related to the dasyurids and to Dasyurus ursinus in particular "although the interval be still very great between them" (Owen, 1859: p. 321).

. The famous British vertebrate palaeontologist Sir Richard Owen, circa 1846, with the leg bones of an extinct New Zealand Moa.