W. S. MacLeay on the Natural Arrangement of Fishes. 1 99 



NORMAL GROUP. 



MALACOPTERYGII, Artedi. Spines in dorsals soft.— Qwere. Are all 

 these Cycloidians of Agassiz ? 



4. PLEURONECTINA. Ventral fins, when existing, inserted 



under the pectorals, and directly 

 suspended to the bones of the 

 shoulder. 



5. CLUPEINA. Abdominales, Cuv. Ventrals suspended behind the 



pectorals, and not attached to the 

 bones of the shoulders. 

 Ohs. — The Balistina, by the confluence of the bones of their jaw, 

 and by the tardy induration of their skeleton, evidently lead off to the 

 Sturgeons, with which they agree in having their free branchiae 

 opening by a perforation in the skin behind the temple. The Fis- 

 tular'ma evidently lead off to the Lophobranchii by Fistularia. Un- 

 fortunately I have not been able to find a near character to separate 

 Fistularina from Perchia ; but they are natural groups, because each 

 forms a circle. The following groups appear to be nearly those into 

 which the above tribes may be naturally divided : — 



1. Balistin.\. 2. Percixa. 3. Fistularina. 



1. Balistidae? 1. Chsetodontidse. 1. Scombridas. 



2. Ostraciontidse ? 2. Percidas. 2. Fistularidae. 



3. Cephalaspis ? Ag. 3. Scorpsenidae. 3. Gobioidae. 



4. Orthogoriscidse ? 4. Cirrhitidse. 4. Lophiidae. 



5. Diodontidae? 5. Sparidae. 5. Labridae. 



4. Pleuronectina, 5. Clupeina. 



1. Anguillidae. 1. Siluridse. 



2. Echeneidas. 2. Cyprinidae. 



3. Cyclopteridae. 3. EsocidjE. 



4. Pleuronectidse. 4. Clupeidae. 



5. Gadidse. 5. Salmonidae. 



Ohs. — I do not believe the above places of the families of Balistina 

 to be correct : besides I only know four. I shall say little there- 

 fore respecting them, except that I suspect some undiscovered family 

 of BALISTINA leads off to the genus Monocentris, among the Scor- 

 pcenidce. I shall begin therefore with the true PERCINA and the 

 family Scorpcenidce. The following are the probable genera of Scor- 

 panidee, which family agrees with the group called Buccce Loricatce 

 by Cuvier ; but it is rather a stirps than a family, and the following 

 genera ought to be deemed of the rank of families : — 



ABERRANT GROUP. 



Head either tuberculous or spinous. 



1. Monocentris, Linn. Free spines in lieu of first dorsal. 



2. Trigla, Linn. Two distinct dorsal fins. 



3. ScoRP^NA, Linn. Two dorsals more or less confluent. 



NORMAL GROUP. 



Head neither tuberculous nor spinous. 



4. Orcosoma, Cuv. Ventrals complete. Free cones in 



lieu of first dorsal. 



5. Gasterosteus, Linn. Ventrals reduced to a spine or spines. 



Free spines in lieu of first dorsal. 



For subgenera, 1 must always refer to Cuvier and Valenciennes. 



