The above classification was rarely accepted in its en- 

 tirety by subsequent researchers on the plectognaths, 

 and it was not a radical departure from previous ar- 

 rangements of the order. By present-day standards, 

 Hollard's osteological observations were often incorrect, 

 with two inaccuracies being of particular prominence. 

 The bone now called the parasphenoid was thought by 

 Hollard to be composed of two entities, a large anterior 

 piece or "sphenoide anterieur" and a small posterior 

 piece or "sphenoide posterieur." What is presently term- 

 ed the frontal was recognized as such by Hollard, except 

 that he believed a small portion at its posteromedial end 

 to be separated from it, and called this small piece the 

 "parietal." But Hollard had to rely, for the most part, on 

 dried skeletons, and if he occasionally saw a few too 

 many bones in the cranium, it is readily under- 

 standable. With the present techniques of clearing and 

 staining, the sutural regions between bones are much 

 easier to define, and surface sculpturing of individual 

 bones is less likely to be confusing. Hollard's con- 

 tribution, then, was that he systematically made known 

 for the first time the general morphological structure of 

 all of the primary types of plectognath fishes. 



Coming after Dareste's (1850) critique of Cuvier's or- 

 dinal recognition of the Plectognathi but before the 

 general plectognath classification of Hollard (1860) was 

 the "Ichthyologie Analytique" by Andre Dumeril (1856), 

 father of the August Dumeril who had edited Bibron's 

 manuscript. In probable deference to Dareste, the elder 

 Dumeril did not recognize an Order Plectognathi, but, on 

 the other hand, his families Sclerodermes and "Gym- 

 nognathes" followed one another in his subclass Chon- 

 drostichthes, as outlined in the large folding sheet 

 between pages 92 and 93: 



Famille Lophobranches (syngnathids) 



Famille Podopteres (Cydopterus, Lophius, anten- 



nariids, ogcocephalids, etc.) 

 Famille Sclerodermes (scleroderms) 

 Famille Gymnognathes (gymnodonts) 

 Famille Hypostomates (climaerids, Polyodon, Aci- 



penser, Pegasus). 



Whereas the two plectognath groups were listed one 

 after the other in the folding sheet, in the text they were 

 described discontinuously. Dumeril did not seem to have 

 any confidence in the validity of the Plectognathi. In the 

 section on "Les Chondrostes Gymnognathes," Dumeril 

 (1855:159-160) simply and conservatively listed the four 

 genera he recognized: Diodon, Triodon, Tetraodon, 

 and "Cephale" (= Mola). "Les Chondrostes Sclero- 

 dermes" (p. 173-182) were treated in only a slightly 

 more detailed manner. Two groups were recognized, the 

 Ostracides and the Balistides. There were four genera of 

 trunkfishes {Ostracion, Aracana, Cibotion, and 

 Doryophrys) and the standard four genera (Monacanthe, 

 Alutere, Triacanthe, and Baliste) in his Balistides. One 

 will notice that the term "Gymnognathes" (of Bleeker) 

 replaced Cuvier's "Gymnodontes." But the term Gym- 

 nodontes was not simply stricken from the record, for 



Dumeril used it as the name of one of the families of his 

 Ordre Hemisopodes. This "Famille Gymnodontes" in- 

 cluded such items as Gerres, sparids, Upeneus, mullids, 

 etc. 



In Dumeril's (1806) much earlier "Zoologie Analyti- 

 que," the fish classification of Lacepede was followed 

 with only the addition of a large number of familial and 

 ordinal names. 



Perhaps the fish classification that appears in ths vari- 

 ous editions of the Encyclopaedia Britannica can be 

 taken as an example of the standard or average condition 

 of our knowledge of that subject. Richardson's (1856) 

 article on "Ichthyology" (8th edition) recognized the 

 Order Plectognathi (p. 312-314) within the teleosts. For 

 the arrangement of the families of plectognaths Richard- 

 son followed Kaup (1855). Kaup, however, had not fin- 

 ished his major groupings of plectognaths in that paper, 

 and Richardson filled in the missing categories with 

 what he thought to be Kaup's logic, as follows: 



Family Balistidae 

 Sub-Family Balistini: Pyrodon, Melichthys, 



Xanthichthys, Canthidermis, Batistes, Balistapus. 

 Sub-Family Monacanthini: Monacanthus, Aluterius, 

 Triacanthus. 

 Family Ostraciontidae 



Cibotion, Doryophrys, Ostracion, Aracana (with 

 four subgenera: Aracana, Capropygia, Kentrocapros, 

 Anoplocapros) , Centaurus. 

 Family Diodontidae 



Sub-Family Diodontini: Diodon, Dicotylichthys, 



Cyclichthys, Cyanichthys, Chilomycterus. 

 Sub-Family Tetraodontini: Tetraodon (with four 

 subgenera, after Miiller: Physogaster, Chelono- 

 don, Cheilichthys, Arothron). 

 Sub-Family Orthagoriscini: Orthagoriscus. 



There are two obvious errors in the above scheme, one 

 of omission and the other of commission. The unique 

 Triodon is not even mentioned, and all of the triacan- 

 thids (Triacanthus) were placed in the Monacanthini. 

 The explanation is that when Kaup treated the 

 Balistidae he included all the balistid species he de- 

 scribed under the Sub-Family Balistini, but did not 

 mention £uiy other subfamily since he was not describ- 

 ing any triacanthoids or monacanthids. Richardson as- 

 sumed that if in the Balistidae there was a Sub-Family 

 Balistini, then there should be a counterbalancing sub- 

 family, so he set up the Sub-Family Monacanthini to in- 

 clude all the other nonostracioid scleroderms. This treat- 

 ment of the Plectognathi was later adopted by Fitzinger 

 (1873), but after that it fortunately disappeared from 

 use. 



After the turn of the 18th century there was an almost 

 unanimous, with the few exceptions to be discussed later, 

 belief that the plectognaths of Cuvier were indeed a 

 natural group, although there would continue to be much 

 discussion about the rank at which they were to be recog- 

 nized. Attention was then turned to better descriptions, 

 diagnoses, and eirrangements of the species within the 



