50 CLASSIFICATION OF FISHES. 



this question a little closer ; the determination of three 

 new types in the GymnetridcE, now for the first time 

 characterised^ completes that gi'oup as a perfect circle; 

 inasmuch as its turns out, on further investigation, to 

 contain representations not only of the primary divi- 

 sions of the Gymnetres, but also of the CoryplicBnidcB. 

 Let us therefore examine the analogies of 



The Gymnetrid^, the Gymnetres, and the CoRYPn-a:- 



NID.E. 



rjpnpra nf thp Families Sub-families 



r;,,„«^/.vf^ Analogical Characters. of the of the 



i:.ymneinace. GymxVetres. CoaYPHiNiDiE. 



Gymnetrus Bl. {^'jSue circfis/^" ] <^^'""^''''^^- Stromatirue. 



r Caudal fin more or 'J 

 Xiphichthes Nob. -j ^^*^ fif/longer than f Trachypterida.CoryphcBnirue^ 



t the pectoral. 3 



r V entrals entirely T 

 Nemotherus Raf. < wanting j ta.i\> OphidonidiS. Trachiurhus. 



L smooth. J 



rTail armed withT 

 Gy?nnogaster Brun. < spines; \entxa.\s>Stylephorid{B. Acanthurince 



C none. 3 



r Ventral fins verjO 

 Lopkotes Giorn. < small; forehead >P/e;-ac//rf«E, Astroder mince. 



t elevated. j 



We have seen how prevalent is the spiny armature of 

 the tail in all types representing the order Plectognathes. 

 Yet with aU this, we could hardly have expected to have 

 found it in a gTOup of such delicate fishes, and so unlike 

 all others, as are the Gymnetres. Nevertheless, the re- 

 cent light that has been cast on the Gymnngaster of 

 Brunnich, or the deal-fish of INIr. Yarrell, has established 

 this fact on the most conclusive evidence ; so that by 

 comparing this type with Stylephorus, through the Acan- 

 thiu'idce, we can no longer doubt of the true course of 

 the two circles of the Gymnetres and the Gymnetridcs. 



(50 a.) ^I. Cuvier has placed the genus Cepola along 

 with that of Lophotes. Now, notwithstanding the very 

 different structure and shape of the two forms, there 



