paper on plectognaths might be noted, since he worked 

 out the 36 specific names that had been applied to the six 

 species of diodontids that occur off the American coasts, 

 while that of Goode (1880) did the same for ostraciids. 



No mention has yet been made of works on fossil plec- 

 tognaths, other than Agassiz's species descriptions, for 

 the simple reason that until part IV of Woodward's 

 "Catalogue of Fossil Fishes" (1901) appeared, there was 

 no systematic account of them. The order Plectognathi 

 was not recognized by Woodward; rather, the plectog- 

 naths occurred in the families Balistidae and Gym- 

 nodontidae, which, along with the Chaetodontidae and 

 Acronuridae (= acanthurids), composed the Division 

 Chaetodontiformes of the Suborder Acanthopterygii. Ad- 

 mittedly the higher categories adopted by Woodward for 

 the purpose of his pioneering catalogue were somewhat 

 artificial, but it is significant that the one person who 

 had reviewed the fossil plectognaths felt that they were 

 so closely related to the acanthurids and chaetodontids 

 that all of these groups should be placed in the same 

 higher category. 



The plectognaths were the first of the long series of 

 orders that Regan (1903a) revised; his "On the Classi- 

 fication of the Fishes of the Suborder Plectognathi" was 

 in sharp contrast to Gill's plectognath work. Whereas 

 Gill concentrated on examining the literature, and par- 

 ticularly Hollard's figures, Regan made direct use of the 

 skeletal material of the British Museum and nearly com- 

 pletely shunned nomenclatural matters. Gill having al- 

 ready straightened out most of the latter. Not only were 

 the research styles much different but the classifications 

 arrived at were equally dissimilar. 



Regan recognized the Plectognathi as one of the sub- 

 orders of the Acanthopterygii in close relationship with 

 the acanthurids. He (1903a:285) felt that the "feature of 

 most importance in diagnosing the suborder Plectog- 

 nathi is the absence of ribs, although in some well-os- 

 sifed epipleurals are present which have been mistaken 

 for ribs," but he pointed out that, like everyone else since 

 Dareste and HoUard, he had not had the opportunity to 

 examine Triodon and only supposed that it also lacked 

 ribs. On the other hand, Regan believed that "the 

 coalescence of the teeth in the jaws is a feature of little 

 importance, and has, as probably as not, originated in- 

 dependently . . ." in the gymnodonts and scleroderms. 

 One could not do justice to Regan's classification without 

 quoting in full his brief but excellent osteologically 

 grounded diagnoses for the various subdivisions, but 

 space forbids and in outline his system was as follows: 



Suborder Plectognathi 

 Division Sclerodermi 



Family Triacanthidae: Triacanthus, Triacanthodes, 



Halimochirurgus 

 Family Triodontidae: Triodon 

 Family Balistidae: Balistes, Monacanthus, Para- 



luteres, Pseudaluteres, Pseudomonacanthus, 



Alutera, Psilocephalus 

 Family Ostraciontidae: Aracana, Ostracion, Lac- 



tophrys 



Division Gymnodontes 

 Family Tetrodontidae: Tetrodon, Ephippion, 



Tropidichthys, Chonerhinus, Xenopterus 

 Family Diodontidae: Diodon, Lyosphaera 

 Family Molidae: Mola, Ramania. 



The most obvious feature of Regan's classification is 

 that it was much more conservative than that of such 

 predecessors as Bleeker and Gill, but the differences were 

 more basic than a simple and inevitable "lumper-split- 

 ter" disagreement. The unique Triodon had always been 

 placed in the Gymnodontes because of the fusion of its 

 teeth, following the lead of Cuvier (1829) in the second 

 edition of the "Le Regne Animal." Regan (1903a:285), 

 however, thought that "the structure of the pectoral arch 

 and vertebral column, as well as the presence of a pelvis 

 and well-ossifed epipleurals" related it to the triacan- 

 thoids and balistids, and thus Triodon for the first time 

 appeared in the Sclerodermi. Contrary to Miiller and 

 Bleeker, who recognized three subdivisions of the Plec- 

 tognathi, Regan followed Hollard and Giinther in recog- 

 nizing only the original two Cuvierian subdivisions, for 

 the "Ostraciontidae do not seem to me to differ suf- 

 ficiently from the Sclerodermi to rank as another 

 division— Ostracodermi" (Regan 1903a:285). Whereas 

 Gill recognized three families of tetraodontids, Regan 

 preferred a single family. Regan pointed out that Gill 

 based his diagnoses of the tetraodontid families on the 

 figures given by Hollard of the top of the skull of various 

 puffers, but that "the figures of Hollard have met with 

 too ready an acceptance, that author having mistaken 

 ridges on and fissures in the frontal bones for sutures" (p. 

 293). Regan's objection to Hollard's figures and Gill's use 

 of them was entirely correct. In short, Regan gave 

 excellent diagnoses for the limited number of plectognath 

 subgroups that he recognized, and corrected numerous 

 osteological errors that had prevailed up to that time. Un- 

 fortunately, as was typical of most of Regan's revisions, 

 only a few figures were given to accompany the great 

 amount of osteological information that was discussed in 

 the text. It is impossible, for example, to tell what Regan 

 meant by the presence of an "ossified praeorbital" in the 

 Balistidae. But these are relatively minor criticisms of 

 the work'that became the standard treatment of the ma- 

 jor plectognath subgroups. 



Regan's only other effort devoted solely to the plectog- 

 naths was a brief, but useful at its time, revision of the 

 genus Triacanthus (1903b). In the only complete fish 

 classification that Regan ever presented, his "Fishes" 

 (1929) in the 14th edition of the Encyclopaedia Britan- 

 nica, the plectognaths were discussed and diagnosed as 

 in his 1903 revision of the order. But now the Plectog- 

 nathi were an order of the Neopterygii rather than a 

 suborder of the Acanthopterygii. Thus, in the years that 

 intervened between his revision of the plectognaths and 

 his general fish classification, the plectognaths would 

 seem to have grown in stature in Regan's view. 



After Regan's (1903a) plectognath classification, the 

 diagnoses that were there given were used to a greater or 



