PROFESSOR OWEN ON THE FOSSIL MAMMALS OF AUSTRALIA. 



559 



have supported the view first suggested by the flattened form of the femur next to be 

 described, if more instructive characters had not been shown deducible from the pelvis 

 in question. 



The ilia, though not quite entire along the " labrum " (Plate XLVII. figs. 1 & 2, c), 

 are suflficiently so to support the inference that they were short, broad or expanded, with 

 a flattened surface rather than a fossa, directed hsemad or downward, and in a minor 

 degree forward. Such a lamelliform ilium is not presented by any existing genus of 

 Marsupial, but is found, besides the Proboscidians, in Megatherioids, Sloths, Apes, and 

 Man. 



From the Elephant's the ilia of the present species difler in the much less production 

 of the angle terminating in the antero-superior spine («, a), which, in Proboscidians, 

 extends outward and bends down in an almost hooked form to near the parallel of the 

 acetabular outlet. 



In the Megatherium and Mylodon the ilia are proportionally more expanded and 

 outwardly extended than in the Elephant. The ilia of the Sloths [Bradypus, Choloepus) 

 come nearer to the proportions of those in Plates XLVII. ; but the antero-superior angle 

 is rounded off", and the position and aspect of the iliac planes are different. There is, 

 however, a more marked, definite, and weightier distinction between the present pelvis 

 and that of other Mammals with expanded lamelliform ilia. Leaving the human and 

 simial pelves out of the comparison, that of the Elephant includes four sacral vertebrae, 

 and the Sloths, both arboreal and terrestrial, have the sacrum unusually prolonged to 

 eff'ect the second junction with the innominate bones at the ischial tuberosities, thus 

 converting the " great sciatic notch " into a foramen. 



In the present pelvis the sacral vertebrae are but two in number. Now this, as a rule, 

 is the number to which the sacral vertebrae are restricted in Marsupialia ; and it strikes 

 me as the more significant of the aflflnity, so indicated in the present pelvis, because it 

 is associated with a modification of the ilium which, in the placental series, goes with at 

 least double that number, and commonly with many more sacral vertebrae, five or six, 

 e.g.^ in the Sloths and Megatherium, and as many as eleven vertebrae anchylosed in a 

 mass in the Mylodon. A still more decisive mark of Marsupial affinity in the pelvis in 

 question is the evidence of an ilio-pubic process (Plate XLVII. fig. l,e,e); and this also 

 points to the particular family of Marsupialia to which the large quadruped under con- 

 sideration is more nearly related. Only in the Kangaroos is this process so developed 

 as to be subject to such violence as has broken it away on both sides of the present 

 pelvis. In all other Marsupialia it is indicated, if at all, by a mere tuberosity. The 

 concurrence, therefore, of a bisegmental sacrum with the ilio-pubic process decides me 

 to restrict further comparison with the pelvis of the Kangaroos (Macropus). 



I take the difference of form of the iliac bones, which is very great, between Macropus 

 and IHprotodon — for if we arrive at the Marsupial genus with a diprotodont dental for- 

 mula by the pelvic route we may be absolved of rashness in drawing the obvious con- 

 clusion — to depend on the corresponding differences in the mode of locomotion deducible 



MDCCCLXX. 4 G 



