PEOFESSOR OWEN ON THE FOSSIL MAMMALS OF AUSTRALIA. 



75 



upper ends 1 inch 3 lines. The foramen magnum is 1 inch in transverse, and 7^ lines in 

 vertical diameter : it looks backward and a little downw^ard. About 4 lines in advance of 

 the foramen the basioccipital (Plate III. i) begins to show a median longitudinal ridge, 

 formed, or left apparently, by a pair* of smooth excavations of this part of the under 

 surface of the basi-occipito-sphenoid, the anterior half of which is bounded externally 

 by the thick obtuse ridges developed from the sides of the basisphenoid {ib. s). The 

 depressions and dividing ridge are present, but more feebly marked in Felis ; the basi- 

 sphenoida ridges, which prolong backward the pterygoid walls {ih. 24) of the post-narial 

 canal, are not developed in Felis : the breadth of the basis cranii between the tympanies 

 is much greater in Thylacoleo than in Felis. The suture between the basioccipital and 

 basisphenoid is obliterated. The pterygoids [ib. 24) are of great length ; their bases 

 almost meet beneath the presphenoid, but gradually diverge as they extend backward, 

 uniting suturally with the sides of the basisphenoid as far back as the low thick ridges 

 [ib. 5), which are the sole representatives of the " pterygoid processes " of the sphenoid. 

 The bodies of the cranial vertebrae describe the same curve, convex lengthwise downward, 

 as in the first specimen. The basisphenoid anterior to the median ridge is gently 

 concave transversely, and this concavity becomes deepened by the junction of the ptery- 

 goids, as the centrums advance, the pterygoid plates arching from the junction down- 

 ward and outward, to form the sides of the posterior continuation of the bony nostrils 

 or respiratory passage. The conformity in this part of the cranial structure of Thyla- 

 coleo with that in Tliylacinus (Phil. Trans. 1859, PI. XIV. fig. 3) is very close. 



The paroccipital (Plates II. & III. 4) is broken, but seems not to have extended far 

 below the level of the occipital condyle; it articulates externally with the mastoid (s), 

 which is not coextensive therewith vertically. Together they form a low, thick, obtuse, 

 rough, subtrihedral pyramid, with the base downward and the outer side excavated. 



The median vertical superoccipital ridge (Plate IV. fig. 2, 3) and the lateral depressions 

 are better developed in the present specimen, with the more abraded carnassial teeth, than 

 in the one first described; indicating the longer subjection of the bone to the action of 

 the nuchal muscles. 



The cranial nervures and vascular foramina are as in the first specimen, with which 

 also the tympanic (Plate III. agrees in its small size and relative position. 



Nearly the whole of the articular surface for the mandibular ramus (Plate III. 27) is 

 preserved on the right side of the present specimen ; it is 1 inch 6 lines in transverse, 

 and nearly 1 inch in antero-posterior extent ; concave transversely, moderately convex 

 from behind forward at its horizontal part; this facet changing or deepening to the 

 concave only, where it is continued upon the post-glenoid plate : this is deeper than in 

 the Lion, and much thicker ; it is on the same transverse line as the tympanic, and is 

 divided by the narrow fissure below the meatus auditorius extemus from the mastoid. 



There was sufficient of the articular surface in the first-described fossil to show that 

 it had a greater antero-posterior extent than in Felis, with some minor characters of 

 difference in which it more resembled that part in the Ursine Dasyure. So much of 



M 2 



