tissue primarily to the vomer, usually to a slight con- 

 cavity on the surface of the latter. The palatine is so 

 small and closely apposed to the ectopterygoid that most 

 authors have been noticeably vague about its articula- 

 tion or even about is presence, although Hollard (1860: 

 fig. 6 of pi. 3) presented an accurate figure of it. 

 The palatine in the gymnodonts is a much larger bone 

 than it is in the scleroderms. The ventral edge of the 

 palatine in gymnodonts always immovely articulates, 

 usually by interdigitation, with the ectopterygoid. 

 Posteroventrally the palatine articulates with the meso- 

 pterygoid, and in some cases with the metapterygoid as 

 well. The anterodorsal edge of the palatine supports the 

 upper jaw, primarily through a flexible fibrous tissue ar- 

 ticulation with the concave facet on the maxillary. There 

 is some variability, however, in the immovable ar- 

 ticulation of the palatine with the ethmoid region of the 

 skull. In Triodon the medial surface of the palatine ar- 

 ticulates with the ethmoid, vomer, parasphenoid, and 

 prefrontal. Rather unexpectedly, the shape and ar- 

 ticulation of the palatine in molids is strikingly similar to 

 that in Triodon. The characteristic feature of the 

 palatine in tetraodontids is that its posterodorsal region 

 possesses an anteriorly directed cleft which fits closely 

 around some type of flange on the lateral surface of the 

 vomer. With the almost total reduction of the ethmoid 

 region in diodontids, the palatine becomes extensively 

 interdigitated with the frontals and makes only a sec- 

 ondary and functionally unimportant contact medially 

 with the ethmoid-vomer. 



Gregory (1933:287) presented highly diagrammatic 

 figures of the palato-pterygoid region of the skull in 

 Triacanthus, Balistes, Alutera, and Psilocephalus to il- 

 lustrate the progressive elongation of the snout in that 

 series. Fraser-Brunner (1941a:423) showed, with more 

 detailed illustrations, the changes that take place in the 

 palato-pterygoid region when the same type of elonga- 

 tion of the snout takes place in the triacanthodid genera 

 Johnsonina, Tydemania, and Macrorhamphosodes, 

 while Tyler (1968) did so for all the genera of that family. 

 The figures and descriptions given by Thilo (1920) of the 

 jaws and palato-pterygoid region of Triacanthus, 

 Balistes, and Tetraodon are too crude to be of any value, 

 but those of Winther (1877) are much more useful. 

 Others have presented accurate descriptions and figures 

 of the palato-pterygoid region in particular species, e.g., 

 Steenstrup and Liitken (1898) for Mala, and Supino 

 (1905, 1907) for Mola and Balistes. 



In the opercular region the length and posterior 

 articulation of the interoperculum have been used as 

 diagnostic familial characters. A summary of the inter- 

 operculum in plectognaths with brief comments on the 

 literature follows. The interoperculum is best developed 

 in the triacanthoids, in which it is relatively broad and 

 rounded posteriorly and only gradually tapers to a point 

 anteriorly. Posteriorly the interoperculum articulates 

 closely with the suboperculum. In balistoids the inter- 

 operculum is a short rod which never extends posteriorly 

 past the level of the epihyal. The posterior end of the 

 interoperculum in balistoids has its main articulation 



with the epihyal, while more posteriorly it connects by a 

 long band of fibrous tissue with the operculum, or with 

 the region of articulation between the operculum and 

 the suboperculum. In ostracioids the interoperculum 

 is even slightly shorter than it is in most balistoids, and 

 similarly it has its main posterior articulation with 

 the epihyal. In contrast to the balistoids, however, 

 the long band of fibrous tissue running posteriorly from 

 the ostracioid interoperculum attaches exclusively to 

 the suboperculum, rather than to the operculum or to 

 both the operculum and suboperculum. Among the 

 gymnodonts the interoperculum is relatively long and 

 well developed in all groups, except in the molids. In 

 the latter group the interoperculum is reduced to a thin 

 but sometimes long splint of bone entirely embedded in 

 the long ligament which runs between the angular and 

 suboperculum. In tetraodontids and diodontids the rod- 

 like interoperculum always has a ventral flange in about 

 the middle of its length wWch articulates by fibrous 

 tissue with the epihyal. The bifurcate posterior end of 

 the interoperculum in Triodon is unique among the 

 plectognaths, with it being hypothesized that the ventral 

 process of the bifurcate portion corresponds to the 

 ventral flange of the interoperculum in tetraodontids and 

 diodontids. The interoperculum articulates posteriorly 

 with the anteriorly directed process of the suboperculum 

 in Triodon, diodontids, and molids, but in tetra- 

 odontids it articulates with the operculum. 



While many early workers described the unusual shape 

 of the interoperculum in this order, and Dareste (1850) 

 said that it was one of the few characters that the plec- 

 tognath fishes had in common, it was not until Regan 

 (1903a) that the structure of the interoperculum was 

 used diagnostically for various plectognath subgroups. 

 Regan incorrectly said that in the Sclerodermi the inter- 

 operculum was attached to the suboperculum, while in 

 the Gymnodontes it was attached to the suboperculum in 

 all groups except the tetraodontids. Fraser-Brimner 

 (1943) similarly used this supposed difference in the pos- 

 terior articulation of the interoperculum as a diagnostic 

 character. 



In his various publications on the respiratory ap- 

 paratus of plectognaths and related forms, Willem (1941, 

 1942, 1944, 1945, 1947, 1949) pointed out the correlation 

 between the presence of a distensible diverticulum of the 

 oesophagus and the size of the suboperculum (i.e., the 

 suboperculum is relatively well developed in the tetra- 

 odontids and diodontids) and discussed the role of the in- 

 teroperculum in coordinating the opening of the lower 

 jaw with the expansion of the opercular apparatus at the 

 beginning of the inspiration phase of the respiratory cy- 

 cle. 



The structure of the jaw8 of plectognaths has played 

 such an important role in the diagnoses of various 

 families since the founding of the order that it must be 

 summarized here. Cuvier's (1817) term Plectognathi 

 (from the Greek: plektos, plaited or twisted together; 

 gnathos, jaw) refers to the intimate and inflexible union 

 of the premaxillary and maxillary in all families except 

 the triacanthoids. 



