TERTIARY FRESH WATER FISHES FROM SOUTHERN QUEENSLAND . 171 
Percolates in the National Museum, Melbourne, however, all possess delicately 
ciliate ctenoid scales, and Ramsay and Ogilby in their definition of the genus, 
state that ciliated scales are present. A means of distinction must therefore 
be sought in other characters. The number of vertebrae, for instance, in 
the living species of Percichthys, varies from 33-35, as against 25 in Percolates , 
although it will be noted that the fossil Percolates ontiquus approaches 
Percichthys in this respect, possessing 30 vertebrae. 
GEOLOGICAL AGE OF THE FOSSIL BEARING BEDS. 
Redbank Plains Series . — Of the forms occurring here, Phareodus is the 
most restricted in time, being found in the Lower Eocene (Paleocene) Green 
River Shales of Wyoming, and the (?) Eocene of the Padang Highlands, Sumatra. 
The age of the Sumatran beds is certainly Palaeogene, if not Eocene (de 
Beaufort, 1931, and personal communication), but it may be that the Osteo- 
glossids from there, at present referred to Phareodus, are really a new, though 
closely allied genus. Notogoneus ranges from the Paleocene (Green River 
Shales), through the Upper Eocene (Gyps de Montmartre) to the Lower 
Oligocene of Aix en Provence and the Upper Oligocene of the Mainz Basin. 
Percolates, by analogy with Cyclopoma [Lates] and Percichthys , may be expected 
to range throughout Tertiary time. Epiceratodus is known to occur in the 
Upper Cretaceous of White Cliffs, New South Wales, but E. denticulatus is much 
closer to the Pleistocene and Recent E. forsteri than is the Cretaceous species. 
The evidence, therefore, is strong that the Redbank Plains Series is of 
Palaeogene age — Eocene or Oligocene — certainly not Cretaceous, as has been 
suggested. Beyond that it is not possible to particularise with certainty, 
although if, as seems possible, Phareodus queenslandicus should prove to be 
closer to Scleropages than to the North American species, especially P. acutus, 
an Oligocene, rather than an Eocene age would be indicated, and this would 
be supported by the resemblance of Percolates ontiquus to the living species, 
and of E. denticulatus to E. forsteri. The Redbank Plains Series may therefore 
be tentatively (in view of the paucity of comparative material) referred to the 
Oligocene, and with this Series the beds at Cooper’s Plains (with Phareodus and 
other Teleosts) are clearly to be linked {see Jones, 1926). 
Bald Hills Basin . — The evidence is insufficient to indicate more than that 
the oil shales with fragmentary fish remains are of Kainozoic age. 
I am indebted to Mr. G. Mack, of the National Museum, Melbourne, for the 
opportunity of studying and dissecting specimens of Percolates , Gonorhynchus , Scleropages , and 
pigmy Perches in his care, 
