297 



spine. Caudal forked. Scales smooth or with faint crenulations 

 at their hindborder, moderate, 32 39 in a row between head 

 and caudal. Lateral line absent or indicated by some scales 

 having a shallow pit. Six branchiostegals. 



Fig. 79. Rhombatractus kochi M. Web. X 3 /4- 



Distribution: Freshwater of New Guinea, Waigeu and 

 Aru Islands. 



Note: The genus Rhombosoma Tate Regan (Trans. Zool. 

 Soc. London XX. prt 6, 1914, p. 283) is synonymous with 

 Rhombatractus as described by M. WEBER (Nova Guinea V. 

 Livr. 2, 1908, p. 233). REGAN substituted the new name on 

 the following reasons : GlLL created the name Rkombatractus 

 to replace CASTELNAU'S name Aristeus, which was preoccupied. 

 The species described by CASTELNAU, and which is the type 

 of Aristeus Castelnau, is Aristeus fitzroyensis. According to 

 REGAN A. fitzroyensis is a Melanotaenia; Rhombatractus is there- 

 fore synonymous with Mclanotaenia, which name is the older. 

 DOUGLAS OGILBY (Proc. Linn. Soc. N. S. Wales 1896, p. 125) 

 is of the same opinion. GILL however (Amer. Naturalist 1894, 

 p. 709), in creating the name Rhombatractus, does not mention 

 a type, but says that it is a substitute for the group called 

 Aristeus by CASTELNAU. The diagnosis of Rhombatractus'. "Mela- 

 notaeniines with a much compressed rhombofusiform body, 

 emarginate dorsal-rostral contour, and no distinct lateral band" 

 is appliable as well to older specimens of Melanotaenia as of 

 RJiombatractus. There is therefore in our diagnosis, giving above, 

 nothing against the use of the genus name Rhombatractus in 

 the sense of M. WEBER for those species, which later are 

 called Rhombosoma by REGAN. 



