48 
translucent that it is easy to see through their outer walls the form 
of the enclosed roots of the molar teeth ; and if held up against a 
lamp, the light will readily pass through both their outer and palatal 
portions. A similar lightness of structure obtains everywhere. The 
supra-orbital process of the frontal bone is small and directed more 
backwards than in Pteropus; so small in #. schoensis that it can 
scarcely be called a process *. The zygoma throws up no process 
to meet that of the frontal, so that in those species where the process 
of the latter bone is wanting the orbit is continuous with the tempo- 
ral fossa, as in the generality of the Insectivorous genera, and as in 
other orders of Mammalia. Viewing the skull from beneath, it ex- 
hibits some other peculiarities. The auditory bullz are, as in Pachy- 
soma, more developed than in Pteropus, and the hinder margin of 
the palate is but very slightly curved, but has the appearance of a 
transverse ridge more or less raised from the level of the palate. 
The lower jaw, besides being exceedingly thin everywhere, has its 
alveolar or anterior part extremely narrow in a vertical direction. 
Its posterior part is remarkable for the almost entire absence of 
ridges or other roughness for the attachment of muscles, and for the 
form of the angular portion. The lower margin of each ramus is 
very nearly straight from the lower part of the symphysis menti to 
the angle, which forms a simple curve up to the condyle. In the dif- 
ferent species this curve is of different degrees of sharpness, most 
pronounced in Z. franqueti, n. s., and least so in Z. schoensis. In 
none of them does the angular region project so far back as the con- 
dyle. The coronoid process is elevated about as much above the 
condyle as the latter is above the lower margin of the ramus. Its 
anterior boundary runs obliquely forward with an easy descent to 
the posterior molar, constituting, in fact, more than half of the en- 
tire length of the upper margin of the jaw. 
I will now offer a few suggestions relative to the probable nature 
of the food of the Epomophorus. In the Desmodus, where there is 
absolutely no mastication required, the true molars are wholly want- 
ing; and the pre-molars, although not reduced to the minimum 
number, are diminished to a very rudimentary condition. It happens 
that in this genus the zoologist has the opportunity, rarely met with 
in this order, of comparing singularity of structure with habits known 
to be of a most extraordinary nature, so extraordinary as to be unique 
among Mammalia, and, as far as I know, among the whole of the 
Vertebrata; and he can at once discern the exact adaptation of the one 
to the other. But without information concerning the habits, would 
he by a mere inspection of the teeth have inferred them? I think 
not. He would indeed infer, from the absence of molars, that the 
creature did not eat food requiring mastication; and the form and 
character of the incisors and canines would clearly point to some 
food requiring to be cut or torn ; but it would scarcely occur to him 
that they were intended to puncture the skin of such animals as 
horses, and enable the creature by a suctorial operation to feed on 
* Being produced in a backward direction, it may be said to be adherent to 
the body of the bone, rather than to be wholly absent. 
