THREE EXTINCT SPECIES OF ELEPHANT. 265 
has been sawn across, I believe by Dr. Falconer; and this part is probably, to judge from 
analogy, not very far above the point of least circumference of the shaft. No indication 
of the nutrient foramen is to be perceived in the fragment; and its situation, therefore, 
was in all probability lower down the shaft than is usually the case in the Indian 
femur. The various dimensions afforded by the specimen are :—tr. d. at upper end 
2”-4., ap. d. about 07-95; tr. d. at lower end 1-25, ap. d. 0°85; circumference 34; 
whilst the outline of the transverse section at that point, which, as before said, cannot 
be very much, if at all, above the point of least circumference of the shaft, is shown 
in the accompanying figures, contrasted with that of the femur of Z. melitensis, taken, 
as nearly as can be judged, at the same part of the shaft. The anterior surface of the 
Re. 
PB 
E, falconeri. E, melitensis. 
bone presents a slight elevation in the middle, with a very shallow depression internally, 
and a much deeper and larger one (a pretrochanteric fossa, as it may be termed) 
externally, the outer boundary of which is formed by a well-pronounced, rough, elon- 
gated tuberosity. On the posterior aspect, at the upper end and outer angle, is seen 
the strongly projecting base of the ¢rochanter major, within which is the lower part - 
of a deep digital or posttrochanteric fossa. The inner and outer surfaces are of very 
nearly the same width from before backwards; and they both have the rectangular form, 
peculiar more especially to the African femur. 
This fragment, compared with the corresponding part of the femur of E. melitensis, 
exhibits such marked differences, in almost all respects, as to afford, perhaps as strongly 
as any other of the remains, as striking a proof as can be desired of the, at any rate, 
specific difference between the two dwarf Elephants. In the first place the transverse 
section of the shaft, shown in the two figures given above, is widely different at corre- 
sponding points. On the anterior aspect the surface is totally different in the two 
cases. In E.-melitensis it slopes obliquely backwards and outwards, from the anterior 
and internal angle, with an even, slightly convex curve ; whilst in Z. falconeri (owing to 
the anterior and posterior surfaces in the upper part, and till very near the lower end 
of the fragment, being parallel to each other, and the outer and inner faces conse- 
quently of equal width) the anterior surface is not oblique. But a still greater pecu- 
liarity in this respect, consists in the presence of the remarkable pre- or, more properly, 
