Classification of the Ganoids. 335 



other allied genera, such as Plianeroi^leuron^ a Devonian genus 

 differing from the Cyclodipteri by its undivided dorsal fin 

 occupying the posterior half of the hack, a naked notochord is 

 combined with ossified ribs, apophyses, and rays, as in the 

 ancient Lepidosteidre and Lejndojileuridai. 



The great extent of time which separates the Pala30zoic 

 Dipteridce from the living Polypteridie is filled up in part by 

 the remarkable group of the Coelacanthi, presenting a very 

 peculiar combination of unique zoological and anatomical 

 characters (for example, the structure of the tail, the peculiar 

 interspinals of the anal fin and of the two dorsals, the ossified 

 swimming-bladder, &c.), with less anomalous featm-es bor- 

 rowed from the other Gano'idei Grossopteri (such as the gular 

 plates, the fringe-like paired fins, the scaly base of the vertical 

 fins, the duplicity of the dorsal, &c.). This group originated 

 in the Carboniferous period, and maintained itself with rare 

 persistence of type throughout all geological periods down to 

 the Cretaceous, when it became extinct. But as I can refer 

 the reader to the admirable works of Prof. Huxley, to whom 

 belongs the inestimable merit of having so perfectly seized 

 and so admirably developed the relations of the different types 

 belonging to the great polymorphic series of the Gano'idei 

 Grossopteri^ I shall abstain from speaking of them at greater 

 length, so as to abridge this summary as much as possible. 



Here concludes the representation of the true Ganoids^ as to 

 the nature of which there is no doubt, thanks to our method 

 of synthesis. But what is to be done with all the other types 

 which have been referred to the Ganoids by a greater or less 

 number of authors ? I will not speak here of the Siluro'ideiy 

 which are true Physostome Teleosteans, nor of the Lopho- 

 hranchii and Plectognathi, belonging to the suborder of Aphy- 

 sostome Teleosteans, nor of the Dercetiformes or Hoplopleuridw^ 

 a very remarkable tribe characteristic of the Cretaceous pe- 

 riod, if we omit the Triassic genera Belonorhynchus and 

 IchthyorliynchuSy the place of which in the system is uncertain 

 (perhaps they ought to be arranged among the Aphysostomi) , 

 but which have no relationship to the Gano'idei. But I must 

 express a more decided opinion upon the other types generally 

 regarded as Ganoids — namely, the Lepidosirens^ the Sturgeons^ 

 the Amiidce, the Jurassic Teleostei, the Acantliodei^ and the 

 so-called cuirassed Ganoids — types to which I have not yet 

 been able to assign a place in the picture of the Ganoids, 

 seeing that the synthetic method has not yet proved those in- 

 timate bonds, those relations of structure, those intermediate 

 forms — in one word, that filiation which alone would allow us 

 to place them there. Nevertheless we must not deny the possi-. 



24* 



