the posttemporal and forms of the pharyngeal ap- 

 paratus." In the same year Cope (1871b) again ex- 

 pressed the idea that the Plectognathi and Lopho- 

 branchii were natural groups, but that Miiller had given 

 them too high a rank in comparison with the other 

 teleosts. Cope's opinion is almost unanimously accepted 

 today. 



Winther (1877), in part three of his "Fiskenes Ansigt: 

 en Comparativ-Anatomisk Unders«igelse," briefly de- 

 scribed the general anatomy, especially of the head 

 region, of such representative plectognaths as 

 Triacanthus, Batistes, Monacanthus. Ostracion, 

 Tetraodon, Diodon, and Mola. He was the last person 

 with a knowledge of the plectognath fishes as a whole 

 who did not believe that they were a natural group. Like 

 others before him, Winther criticized Cuvier's character 

 of the intimate fixation of the upper jaw bones to one 

 another as being neither unique to, nor characteristic of, 

 all of the plectognaths. He believed the plectognaths to 

 represent three families: Triacanthini, Sclerodermi (in- 

 cluding balistoids and ostracioids), and Gymnodontes. 

 He was unsure of the relationship of the triacanthoids to 

 his Sclerodermi, but he concurred with Dareste that the 

 balistoids were related to the acanthurids. For some 

 unaccountable reason, Winther thought that the Gym- 

 nodontes were related to the Discoboli (Cyclopterus). 



A good summary of the then prevailing feeling about 

 plectognaths by knowledgeable ichthyologists is found 

 in Gill's (1882) "Arrangement of the Families of 

 Fishes . . .," published shortly after Bleeker's "Atlas," 

 Giinther's "Catalogue," and Cope's "Systematic Relation- 

 ships." While Gill's arrangement was drawn up to serve as 

 the fish filling system to be used at the Smithsonian In- 

 stitution, it was more than a mere list of families, for Gill 

 gave numerous comments on the phylogeny of the higher 

 categories. After reviewing the characteristics of the plec- 

 tognaths. Gill expressed the contemporary view that 

 (1872:XLI) "the many common characters justify their 

 association together, and the characters that are peculiar 

 to them sanction their isolation as a group," while "Their 

 differences sink into comparative insignificance." Gill 

 agreed that the scleroderms "have been deemed more 

 related to ordinary Acanthopterygian types than to the 

 other admitted Plectognaths. And it is quite true that 

 they (and especially the Triacanthids) are much more 

 similar to the ordinary fishes than are the typical Plectog- 

 naths. This, however, is quite explicable by the sup- 

 position that they are the most generalized, and repre- 

 sent the immediate line of descent . . . ." It was thus 

 agreed that the work of Dareste and others related the 

 scleroderms to the acanthurids and that the gym- 

 nodonts were obviously highly specialized offshoots of 

 the scleroderms. This was then, as it is now, the more or 

 less accepted view. In the actual arrangement used by 

 Gill, the Plectognathi were divided into three major sub- 

 groups, as had been done by Muller and Bleeker, in- 

 stead of into two major subgroups, as had Cuvier, 

 HoUard, and Giinther. The families that Gill recognized 

 in each of the three subgroups were those as recognized 

 by Bleeker. 



Several years later Gill (1885) published his "Synop- 

 sis of the Plectognath Fishes." This was a skillful diag- 

 nosis, based almost entirely on the literature, of the 

 characteristics of the various plectognath subgroups, and 

 of the nomenclatural labyrinth that surrounds the order. 

 Gill relied heavily for the osteological information in his 

 diagnoses on the various publications of HoUard, whose 

 papers he thought had been unwisely neglected. Gill's 

 classification of the plectognaths was much more 

 elaborate than that of his "Arrangement of the Families 

 of Fishes," and although it has much in common with 

 Bleeker's system, it is sufficiently different to merit 

 citation: 



Order Plectognathi 

 Suborder Sclerodermi 

 Family Triacanthidae 

 Subfamily Triacanthodinae 

 Subfamily Triacanthinae 

 Family Balistidae 

 Subfamily Balistinae 

 Subfamily Monacanthinae 

 Subfamily Psilocephalinae 

 Suborder Ostracodermi 



Family Ostraciontidae 

 Suborder Gymnodontes 

 Superfamily Triodontoidea 



Family Triodontidae 

 Superfamily Tetrodontoidea 

 Family Tetrodontidae 

 Subfamily Tetrodontinae 

 Subfamily Colomesinae 

 Family Psilonotidae 

 Family Chonerhinidae 

 Superfamily Diodontoidea 



Family Diodontidae 

 Superfamily Moloidea 

 Family Molidae 

 Family Molacanthidae. 



The above scheme differs only slightly from Bleeker's 

 handling of the scleroderms, the difference being that the 

 elongate monacanthid Psilocephalus was elevated by 

 Gill to familial rank. In regard to the gymnodonts. Gill 

 recognized a great many more higher categories than had 

 Bleeker, but few of Gill's extra divisions have continued 

 to be recognized. To Gill's credit, however, was the fact 

 that all of the subgroups were diagnosed, and his sys- 

 tematic handling of the maze of names that have been 

 applied to the various plectognath groups was a boon to 

 all subsequent researchers. 



A number of years later Gill (1892a) published his 

 "Notes on the Tetraodontoidea," in which he expanded 

 on his nomenclatural treatment of the tetraodontids, 

 without any basic change in the classification he had 

 previously used. In a series of notes, Gill (1888a, b, 

 1889, 1892b, 1897) dealt effectively with a number of 

 other nomenclatural problems which need not be dis- 

 cussed here. In the same vein, Eigenmann's (1885) only 



