170 
MEMOIRS OF THE QUEENSLAND MUSEUM. 
spines ; whether the posterior border was rounded or excavated is uncertain. 
There are at least 14 abdominal vertebrae, probably more, the last three 
compressed. The last vertebra is continued as a urostyle, to which is apposed 
a cover of fused neural arches, and there is a number of hypurals ( see Fig. 13). 
The scales are ctenoid, and there are about 17 in a transverse series. Neural 
and haemal spines are relatively strong. 
Remarks. — In the absence of the head, the fish is indeterminable. It is 
not certain whether it is a pigmy perch or the fry of a large fish, but I incline 
to the view that the latter is the case and it may be that, as the general 
similarity suggests, it is a young specimen of Percolates sp. 
This specimen was labelled “ Ipswich or Upper Trias Jura.” Mr. Ball 
points out, however, that when it was collected the Tertiary beds in the 
Bundamba district had not been mapped, and thus there is no necessary 
implication that the fish came from the Triassic Ipswich Series as now known. 
The well is situated within the Tertiaries, and from these the fish was obtained. 
Order indet. 
Locality. — Cooper’s Plains. 
Material. — A few fragments, the bone being actually preserved, in soft 
mudstone. 
Description. — Fragments of the axial skeleton and fin spines of a 
Teleostean fish, indeterminable. 
Remarks. — There is an extremely close resemblance between the fossil 
species P . antiquus and the living P. colonorum , the greatest differences being 
in the number of vertebrae — 25 (11 + 14) in the living species and 30 (12 + 18) in 
the fossil — and the variability and less regular development of the spines on the 
pre-operculum of the fossil species. This close similarity is somewhat unexpected 
in view of the presence in the same beds of a distinct genus of Osteoglossidae from 
that now living in the same region, and of the Gonorhynchid Notogoneus, of 
which no living relatives inhabit fresh waters. 
It seems, however, that the more primitively organised Percoid genera 
such as Perciththys and Dates , have changed but little throughout Tertiary 
time. Percichthys, still living in South America, occurs fossil in the Tertiary 
lignites of laubate, Brazil (Smith Woodward, 1898), and Cyclopoma from 
the Upper Eocene of Monte Bolca, Italy, is scarcely distinguishable from 
Dates , still living in Northern Africa (Smith Woodward, 1901). Percichthys 
and Percolates are indeed extremely alike. Boulenger (1895) separates them 
in his key to the family Serranidse [s. 1.] solely on the presence of cycloid 
scales in Percolates as against ctenoid in Percichthys , The specimens of 
