sal and anal fins are both short-based, with 9 to 13 rays 

 each. 



Among the gymnodonts only Triodon has even a trace 

 of a spiny dorsal fin, as a minute structure composed 

 usually of two spines, occasionally only one but with pos- 

 sible rudiments of a third spine, borne on two basal 

 pterygiophores, present in most specimens of one of the 

 populations (Indonesia to Japan) of the single species, 

 with the second basal pterygiophore succeeded by two 

 supraneural elements. 



It is possible that the supraneural elements in gym- 

 nodonts are rudiments of the basal pterygiophores of the 

 now absent spiny dorsal fin. The structure and position 

 of the single supraneural element in tetraodontids give 

 no indication of its possible derivation from a spiny dor- 

 sal fin pterygiophore, but the structure and number of 

 the supraneurals in Triodon are reminiscent of the spiny 

 dorsal fin pterygiophores of triacanthids. An at least 

 analogous situation is present in ostracioids, for the 

 aracanids have a very long supraneural element which 

 extends throughout almost the entire distance between 

 the supraoccipital and soft dorsal fin, but in ostraciids 

 this supraneural is short and confined to a position just 

 in front of the soft dorsal fin. 



One should refer to Bridge (1896) and Rosen (1916a) 

 for descriptions of the cartilaginous distal pterygio- 

 phores in the dorsal and anal fins of several plectog- 

 naths. Hora's (1924, 1925) descriptions of Kanduka as a 

 genus of tetraodontid without a dorsal fin and with the 

 anal fin at least rudimentary, if not absent, are to be 

 looked upon skeptically. His two small (14 and 54 mm 

 SL) specimens were highly inflated, and the fins were 

 undoubtedly hidden beneath the skin. The fact that 

 Hora's radiographs did not show these fins probably 

 only means that the pictures were overexposed and that 

 the specimens were still weakly ossified. 



Much has been said in the literature about the general 

 structure of the bones of plectognaths. Some of the ear- 

 ly workers obviously thought the plectognaths to be 



similar to the "cartilaginous" fishes, probably because of 

 the relatively late ossification of the bones that is seen in 

 some plectognaths and because of misinterpretation of 

 the spongy nature of the ossifications in Mala and of the 

 large amount of cartilage that is present even in the adult 

 Mola. The consensus, however, seems to be that while 

 many plectognaths simply have a relatively late ossifica- 

 tion of the endochondral bones of the skull, the com- 

 position of the ossification is normal. This is more or less 

 what Cuvier (1817:144) said in his definition of the Order 

 Plectognathi. The principal references to the histology of 

 the bones of plectognaths are the following: Quekett 

 (1850-1855, bones of a number of plectognaths), Leydig 

 (1857, bones of Mola), Kolliker (1859, bones of 

 numerous plectognaths; 1860, bones of Mola), Harting 

 (1865, bones of Mola), Goette (1879, epipleurals of 

 Monacanthus), Trois (1883-1884, bones of Ranzania 

 and Mola), Goeldi (1884, general bone structure in 

 Balistes), Goeppert (1895, development of epipleurals 

 of Monacanthus), Stephan (1900, bones of Mo/a, Tetra- 

 odon, and Balistes), Supino (1904, bones of Mola), 

 Nowikoff (1910, bones of Mola), Kaschkaroff (1914a 

 and 1925, bones of numerous plectognaths; 1914b and 

 1916, bones oi Mola), Studnicka (1916, bones of Mola), 

 Rauther (1927b, epipleurals and vertebrae oi Monacan- 

 thus), Haines (1937, Meckel's cartilage in Tetrao- 

 don). 



The literature on the muscles of plectognaths has been 

 thoroughly reviewed by Winterbottom (1974) and dup- 

 lication of that laudable effort is avoided here. 



The smatterings of information available on other sys- 

 tems (such as the blood-vascular, digestive, renal, neu- 

 ral, chromosomal, cellular, physiological, developmental, 

 etc.) are too incomplete for plectognaths to be of im- 

 mediate value to the present studies, and this literature 

 will not be routinely reviewed here. It will be analyzed in 

 a forthcoming synopsis of the biological data on plectog- 

 naths, and specific papers will only be cited here if they 

 have direct bearing on other matters discussed. 



Relationship of the Plectognathi to the Perciformes 



As indicated in the historical review of the classifica- 

 tion of the Plectognathi, it is commonly thought that the 

 plectognaths are a derivative of the perciform fishes, and 

 perhaps most closely related to the acanthurid surgeon- 

 fishes. The features of the osteology of the Recent acan- 

 thurids most pertinent to an interpretation of their 

 classification and interrelationships have been reviewed 

 by Tyler (1970c), aiming especially at ascertaining what 

 is generalized versus specialized in that family. The com- 

 plementary analysis of the fossil forms of acanthurids is 

 not yet completed, and the necessary comptirisons 

 between them and the fossil plectognaths studied here, 

 especially those of both groups from the Eocene, must 

 await a subsequent publication under preparation by 

 J. Blot (Mus. Nat. Hist. Nat., Paris) and the author. 



However, as indicated by Tyler (1968), it is obvious 

 that the skeletal structure of acanthurids is not far 



removed from that of triacanthoids, the basal plectog- 

 naths, differing mainly in the presence of a number of 

 bones absent in triacanthoids and in a greater number of 

 certain meristic elements. In general appearance the 

 acanthurid skeleton is closer to that of triacanthids 

 among the plectognaths rather than to that of the more 

 generalized triacanthodids. This might suggest that if 

 the triacanthoids and thus the other plectognaths share a 

 common Upper Cretaceous ancestral stock (as sug- 

 gested by Patterson 1964:400) with the acanthurids, that 

 the Recent acanthurids are not as generalized a group as 

 the triacanthodids are relative to the triacanthids. 



Acanthurids differ most significantly from plectog- 

 naths by having: two or three anal fin spines (none in 

 plectognaths); a relatively better developed pelvic fin 

 varying from 1,5 to 1,3, with the rays cross-striated and 

 branched (one spine followed by, at the most, two much 



40 



