GREGORY: FISH SKULLS 



307 



Rachvcentron -This fish (Figs. 186, 187) is classed by Boulenger among the Scombri- 

 formfs an'd i^ amily i placed nL to the Carangid. and ahead of the Scombnd. m h.s 

 sTem But an examina'tion of its skull and vertebral column suggests to me that ^..Ay- 

 cenZls merely a somewhat mackerel-like offshoot of the true percoids, J^^ch has hkew se 

 been derived from a short-bodied type that has later become elongated. The side view 

 of the skull in general appearance somewhat suggests the scombroid type especially in 

 1 genera contour, broad curved preopercular, rather -small opercular^ ^\7JJ:: 

 Fig 187), however, affords a very wide contrast to the scombroid type m the small develop- 



isthrrius 



Poronotus triacanthu5 



Fig. 184. Poronotus. 



ment of the supraoccipital, very much flattened, crestless ^^^^^\^''\'P'''''\ ^^J^l' 

 Absence of parietals, etc. In short, while it seems highly probable that the scombroids are 

 much mo e^neTrly rdated to the carangids than to any other modern type, it does not seem 

 7^^RachyZrol serves to connect the scombroids either with the percoids or with the 



"'' Tale Regan (1909^, p. 68) notes that '< Rhachicentrum- shares a number of primitive 

 skull characters with both the normal Perciformes and the Carangid. and that its pectoral 

 arch vertebral and rib characters conform to the perciform type. •.,,,!,„ 



In brief the hypothesis that seems to accord best with the known facts is that the 

 scombroid represent^ highly progressive but now isolated derivative of ^ojneCre^c^^ 

 carangoid that has rapidly become long-bodied as its predaceous powers and swiftness 



increased. 

 16 



