naturalness of the triacanthoid-balistid-monacanthid 

 line, but from what stock the tetraodontoids and ostra- 

 cioids arose is an open question. After examination of the 

 larvae of a triacanthodid, Fraser-Brunner suggested that 

 both the tetraodontoid and ostracioid lines may have 

 been derived neotenously from the triacanthodids. 



Whereas Miiller and Bleeker had started the trend 

 toward the recognition of three major subgroups of plec- 

 tognaths, rather than the original two of Cuvier, Berg's 

 (1940) "Classification of Fishes" went one step further 

 and recognized four. Berg came to the following classi- 

 fication on the basis of his analysis of the works of Regan, 

 Rosen, Kaschkaroff, Gregory, and Fraser-Brunner: 



Order Tetraodontiformes (Plectognathi) 

 Suborder Balistoidei (Sclerodermi) 

 tFamily Spinacanthidae 

 Family Triacanthidae 

 Subfamily Triacanthini 

 Subfamily Halimochirurgini 

 Family Triodontidae 

 Family Balistidae 

 Subfamily Balistini 

 Subfamily Monacanthini 

 Subfamily Psilocephalini 

 Suborder Ostracioidei (Ostracodermi) 



Family Ostraciidae 

 Suborder Tetrodontoidei (Gymnodontes) 

 Family Tetrodontidae 

 Family Diodontidae 

 Suborder Moloidei 

 Family Molidae. 



Other than the fact that the molids were given subor- 

 dinal rank, this was a very conservative classification, 

 based entirely on works already mentioned. 



Breder and Clark's (1947) "A Contribution to the Vis- 

 ceral Anatomy, Development and Relationships of the 

 Plectognathi" is a useful, and interesting, comparison of 

 the anatomy of the acanthurid-plectognath line. It is in- 

 dicative of the then current state of our knowledge of 

 plectognath anatomy (Breder and Clark 1947:312-313, 

 table 2) that these two authors, on the basis of the litera- 

 ture and their own research, were unable to state whether 

 the palatine was movable in either Triodon or the trunk- 

 fishes. One also sees in table 2 that even greater gaps 

 exist in our knowledge of the characteristics associated 

 with reproduction. It has been mentioned several times 

 in this historical review that the rarity of museum 

 specimens of Triodon had hampered the phylogenetic in- 

 vestigations of all workers after the time of Dareste and 

 Hollard. Breder and Clark were able to make a partial 

 dissection of one of the Stanford University specimens of 

 Triodon obtained by Herre in the Philippines in 1931. 

 Their report on the intestine was the first addition that 

 had been made in 1(K) years to our knowledge of Triodon. 

 This same specimen has been utilized for the present 

 monograph. 



The classification adopted by Breder and Clark was 

 modified from that of Fraser-Brunner: 



Order Plectognathi 

 Suborder Sclerodermi 

 tFamily Spinacanthidae 

 Family Triacanthidae 



Subfamily Triacanthodinae 

 Subfamily Triacanthinae 

 Family Balistidae 

 Family Monacanthidae 

 Subfamily Monacanthinae 

 Subfamily Aluterinae 

 Family Triodontidae 

 Suborder Ostracodermi 

 Family Ostraciidae 

 Subfamily Aracaninae 

 Subfamily Ostraciinae 

 Suborder Gymnodontes 

 Family Tetraodontidae 



Subfamily Canthigasterinae 

 Subfamily Lagocephalinae 

 Subfamily Colomesinae 

 Subfamily Tetraodontinae 

 Subfamily Chonerhininae 

 Family Diodontidae 

 Suborder Moloidei 

 Family Molidae. 



On the basis of their research, Breder and Clark (1947) 

 felt that the trunkfishes deserved subordinal rank rather 

 than inclusion in the Sclerodermi: "In view of the 

 marked differences in the ontogeny from that of Mona- 

 canthus and its resemblance to that of Spheroides, we 

 cannot feel justified in holding these fishes within the 

 Sclerodermi almost solely on the basis of cranial charac- 

 ters" (p. 311). Their justification for following Berg's 

 handling of the molids was that "Although these fishes 

 represent an offshore modification of the gymnodontid 

 type, they are so much further modified as, in our judge- 

 ment, to warrant such a separation. The fact they have 

 lost their tail and its associated musculature ... in- 

 dicates a fundamental change in the whole plan of or- 

 ganization" (p. 311). Breder and Clark's views were sum- 

 marized in a phylogenetic tree (p. 315, fig. 8) that is one 

 of the most sensible and understandable that has been 

 presented for the plectognaths, and the reader is referred 

 to it. 



The extensive work by Y. Le Danois (1954, 1955, 1956, 

 1959, 1961a, b) on the osteology, myology and other 

 soft anatomy, taxonomy, and phylogeny of the plectog- 

 nath fishes, including monographs on the tetraodon- 

 toids and ostracioids, has not had a favorable reception. 

 The osteology and systematics were criticized by Tyler 

 (1963b) and the myology by Winterbottom (1974). To 

 more fully comment on the controversial aspects of Le 

 Danois' work would require hundreds of pages, and it 

 must suffice to summarize Le Danois' conclusions: the 

 triacanthoids and balistoids are of acanthurid origin; the 

 other plectognaths (Orbiculati) are not even of percoid 



