inquirere numne Pulmones haberent." It is possible that 

 Garden mistook either the swim bladder, kidneys, or dis- 

 tensible diverticulum of the oesophagus for lungs. Brous- 

 sonet (1780:680) believed that it was the distensible 

 diverticulum that had been misinterpreted, but Cuvier 

 (1817:146) thought that it was the kidney, and the par- 

 ticular source of Garden's error is impossible to identify. 

 Severe as was the elimination of the Branchiostegi, the 

 only significant change in the assemblage of species of 

 plectognaths described was the synonymizing of several 

 of his previously numerous species of Diodon, so that 

 only two species of that genus were recognized. 



Among the 29 Linnaean species were representatives of 

 the balistids, monacanthids, both subfamilies of os- 

 traciids, the tetraodontin tetraodontids, diodontids, and 

 molids. As discussed in the preceding section on 

 Willughby, the triacanthids were known before 1758, 

 even though Linnaeus did not recognized them, and they 

 were not binomially described until Bloch (1786), while 

 triacanthodids were not described until Schlegel (1850). 

 As discussed in the succeeding section on Cuvier, the 

 aracanids and triodontids were not formally described 

 until well after Linnaeus, respectively by Shaw (1798) 

 and Cuvier (1829), but see Tyler (1967) for a history of 

 the impressionistic pre-Linnaean illustrations of TVt'o- 

 don. Several names for canthigasterin tetraodontids 

 were inadvertently listed by Linnaeus as synonyms for 

 several species of tetraodontin tetraodontids (e.g., see 

 Shipp 1974 for Lagocephalus) , but the first binomial 

 description of a canthigasterin was in a now mostly 

 forgotten publication by Paterson (1786), as Tetrodon 

 electricus, and the currently used specific name of one of 

 the more common Indo-Pacific species of Canthigaster 

 may have to be changed. 



Gmelin's 13th edition (1788) of the "Systema Naturae" 

 reestablished the Branchiostegi in the Pisces and entirely 

 eliminated the Amphibia Nantes, while at the same time 

 the "Systema Naturae" introduced the Chrondropte- 

 rygii. Fishes were thus all included in six categroies of 

 Pisces: the Apodes, Jugulares, Thoracici, Abdominales, 

 Branchiostegi, and Chondropterygii. 



This general classification was a distinct improve- 

 ment over any of those in previous editions of the 

 "Systema Naturae" and was largely the synthesis of 

 Gmelin himself. Within the Branchiostegi the genera ap- 

 peared in the following order: Mormyrus, Ostracion 

 with 10 species of ostraciids, Tetrodon with 12 species of 

 tetraodontids and 2 species of molids, Diodon with 6 

 species of diodontids and 1 species of molid, Syng- 

 nathus, Pegasus, Centriscus, Batistes with 18 species of 

 balistoids, Macrorhamphosus, Cyclopterus, and 

 Lophius. Thus, the treatment of the individual genera of 

 plectognaths was inferior to that of previous editions, for 

 molids were placed in both Tetrodon and Diodon, while 

 Balistes was spatially separated from the other plectog- 

 naths by the intervening Syngnathus, Pegasus, and 

 Centriscus. Even if this separation of Balistes was 

 deliberate, its significance, if any, is unknown. 



The classification adopted by Goiian (1770:94) in his 

 "Historia Piscium" divided the Pisces into three main 



groups, the Acanthopterygii, Malacopterygii, and 

 Branchiostegi, each with the same four sub- 

 divisions: the Apodes, Jugulares, Thoracici, and Ab- 

 dominales. 



Such subdivisions of the three major categories have a 

 deductive simplicity that is truly alluring, but, unfor- 

 tunately, fishes are not constructed to fit into such 

 logical pigeonholes. The Acanthopterygii and Malacop- 

 terygii contained the expected genera, and the Branchi- 

 ostegi sorted out to: 



Apodes: Syngnathus, Balistes (with a footnote 



saying that certain Balistes have ven- 

 tral fins and thus must be placed in 

 the Abdominales), Ostracion, Tetrao- 

 don, Diodon. 



Jugulares: Lophius. 



Thoracici: Cyclopterus. 



Abdominales: Centriscus, Pegasus (and, according to 

 the footnote, a few Balistes, which are 

 not specifically named, but may imply 

 Macrorhamphosus). 



Only genera were described, but the species of Balistes 

 or of other plectognaths that would have to have been 

 placed in the Branchiostegi Abdominales are the triacan- 

 thoids, which had not yet been described binomially. 

 The above treatment of the Branchiostegi is basically 

 that of Gmelin's 13th edition of the "Systema Naturae," 

 except that Mormyrus was placed by Goiian in the 

 Malacopterygii Abdominales. 



In Bonnaterre's (1788) "Tableau Encyclopedique," the 

 plectognaths reverted to the cartilaginous fishes, for five 

 classes were recognized: the Cartilagineux, Apodes, 

 Jugulares, Thoracici, and Abdominales. These five 

 groups are those of the 12th edition of Linnaeus' (1766) 

 "Systema Naturae," the only difference being the sub- 

 stitution of the term Cartilagneux for Amphibia Nantes. 

 The Cartilagineux contained exactly the same genera as 

 the Amphibia Nantes, differing only in that Petromyzon 

 was called Lampetra. The value of Bonnaterre's work to 

 the study of plectognaths rests on the fact that of the 41 

 species described, the great majority were figured with 

 some accuracy, and outline cross sections of the body 

 were given for many. 



If fishes were ever to fit into a deductively logical 

 scheme of classification, they were given a chance to do 

 so in Lacepede's (1798, 1800, 1802a, b, 1803) cat- 

 egories, as outlined in his "Histoire Naturelle," even 

 more so than in Goiian's work. The "Classe des Pois- 

 sons" was divided into the Sous-classe Poissons Car- 

 tilagineux and the Sous-classe Poissons Osseux. Each of 

 these subclasses was split into the same four divisions, 

 and each of these divisions was split into the same four 

 orders. Thus, there was a dichotomy followed by two suc- 

 cessive sets of quadrichotomies, to give a total of 32 

 orders in eight sets of Apodes, Jugulaires, Thoracins, and 

 Abdominaux (following Gouan). The balistoids were in 

 the third order (Thoracins) of the second division of the 



