260 
POPULAR SCIENCE REVIEW. 
will not be out of place to refer here to a remarkable fact in 
geographical distribution which I have omitted in my pre- 
vious communications on Ceratodus. The division of fresh-water 
fishes offers not a few instances in which two or more natural 
families, much differing in their structural characters, have 
exactly the same geographical distribution. We shall see sub- 
sequently that Ceratodus , Protopterus , and Lepidosiren , are 
members of the same natural Ganoid family ( Sirenidce ). Now 
this family coincides in respect of its geographical range with 
a Teleosteous family which I have called Osteoglossidoe , and 
which comprises the genera Osteoglossum , Arapaima, and 
Heterotis , as will be seen from the following table : — 
Ganoid. Teleosteous. 
Tropical America. 
Lepidosiren paradoxa. Osteoglossum bicirrhosura. 
Arapaima gigas. 
Tropical Australia. 
Ceratodus Forsteri. Osteoglossum Leichardti. 
Ceratodus miolepis. 
East Indian Archipelago. 
xx. Osteoglossum formosum. 
Tropical Africa. 
Protopterus annectens. Heterotis niloticus. 
Thus it is only in the East Indian archipelago that we have 
not yet found the G-anoid representative of the Teleosteous Os- 
teoglossum formosum. That it will be found there I have no 
doubt. 0. formosum has hitherto been obtained in Sumatra, 
Banka, and Borneo ; and of the inland fishes of the latter island 
scarcely anything is known at present. 
The body of the Barramunda (fig. 1) is eel-shaped, but con- 
siderably shorter and thicker than a common eel, and covered 
with very large scales. The head is nearly entirely naked, covered 
with a porous skin, flattened and broad, the eye lateral and 
small, the mouth in front of the broad snout comparatively 
narrow, and provided with thick and soft lips. The gill open- 
ings are a narrow slit on each side of the head, immediately 
in front of the fore-paddle. There are no external nostrils. 
The foremost portion of the trunk is depressed like the head, 
but it soon passes into the compressed remaining portion, 
the boundary between trunk and tail being externally in- 
dicated by the vent only, which is situated between the hind- 
paddles. The tail varies in length ; it is sometimes shortened, 
and it appears that mutilations of this part, particularly when 
