165 



pliyses to which the ribs are attaclied. There are fourteen caudal 

 vertebrae. The caudal has se\eiiteeii ravs, two of whicii aie ()ro- 

 bablv siuiple. The dorsal has .r -|- 3 spines and 17 soft rays, tiie 

 anal has 11 and the ventrals have 6 rays. 



Part of the scales are extremely well pre^erved. They are more 

 or less rectangular, with a convex posterior border. From the centre 

 about ten diverging grooves run to the anterior border. A great 

 number of crowded parallel rows of extremely small flat spinelets 

 run to the posterior border. For the rest, llie surface of the scales 

 is covered with delicate small lines, concentrically arranged round 

 the middle of the scale and scalloped where they cross the grooves. 



With these scanty particulars a further determination had to be 

 tried. A first indication was the great difference between the number 

 of soft dorsal and anal rays. In consequence a number of forms, 

 in which the soft dorsal and the anal are of nearly equal length, 

 could be excluded. Farther on the structure of the scales put me 

 on the right track and brought me to the recent genus Lutjanns. 

 It is true that most species of this genus have fewer D-rays than 

 my specimen, but in some of them the number is about the same, 

 f.i. in Lutj'inus sebae, which species I, therefore, selected for closer 

 comparison. A skeleton of last named species shows so much likeness 

 with my fossil, even in details, that I scarcely doubt that this too 

 belongs to Lutjanus. 



What can these two fossils now teach us about the age of the 

 deposits, in which they were fossilized: 



As far as 1 know, no other tertiary Teleosts are known from the 

 Indo- Auslralian Archipelago, than those from a freshwater-deposit 

 in the Padangsche Bovenlanden formeily described by Gcnther 

 (Geol. Mag. (2) III, 1876). As far as I know, forms related to our 

 fossils are lacking too in the tertiary fish-fauna of the neighbourhood. 

 Neither amongst the tertiary fishes from Australia (Ch.apman and 

 Pilchard, Pr. Roy. Soc. Victoria (2) XX, 1907) nor amongst those 

 of Siam (Andersson, Upsala Bull. Geol. Inst. XIII, 1916) a species 

 of S'in/inellii or Lutjanus has been described. The Clupeid, recently 

 described by Jokda.v (Proc. Cal. Acad, of Sciences IX, 1919) from 

 Japan, is not related to our specimen. It is even uncertain, if it is 

 a Clupeid at nil. 



Among the many herrings, described from the tertiary of Europe 

 and America, I do not know of any species, related to Sardinella. 

 Smith Woodward (Cat. Fossil Fish British Mus. IV, 1901, p. 152) 

 gives the following description of the scales of Cliipea numidica, 

 from the Upper Miocene of Algeria: "Scales sometimes pitted in 



