716 THE TETEAODONTOIDEA GILL. 



have them " in small luimber." Various families of acantliopterygiaus 

 haye'beea distinguished on account of the numbers of vertebrae; the 

 Oarangids, for example, have been distinguished because they have 

 "ten abdominal and fourteen caudal vertebne." Fishes closely allied 

 in other respects to the Cvarangids have been excluded because they 

 have twenty five* or more vertebree.t 



Let it be borne in mind that the differences in the number of verte- 

 brae are not coordinated in the special cases of the acanthopterygiaus 

 had in mind, with any other modifications. Yet an eminent ichthyol- 

 ogist unites the Tetraodontids and Canthigasterids having " the verte- 

 brae in small number" (less than "ten abdominal and fourteen caudal 

 vertebrae") with the Chonerhinids having the vertebrse in increased 

 number ("more than ten abdominal and more than fourteen caudal 

 vertebme") in one and the same genus, although those differences are 

 coordinate with numerous other structural modifications. 



It is because President Jordan has probably been influenced by the 

 treatment of the group under review by an eminent authority, and not 

 allowed his own excellent and candid judgment to have full sway, that I 

 feel constrained to comment on the inconsistency and want of scientific 

 method involved in the examples in question. Far from fearing that 

 I have gone too far in subdivision of the Tetraodontoideans, I feel that 

 I have scarcely gone far enough. I should perhaps raise the Colomesines 

 to family rank, and if I do not so do it is because I am desirous to ap- 

 pear not to go to an extreme. The recognition of the family value of 

 the Colomesine group may be left to another or to some other time. 

 Perhaps Mr. Boulenger, the learned herpetologist of the British Mu- 

 seum, may so elevate it. 



That distinguished and really scientific herpetologist has employed 

 a character, analogous to the principal one which distinguishes the 

 Colomesines from the bulk of the Tetraodontids, to diagnose sev- 

 eral families of lacertilians.J The Colomesines have the " postfront- 

 als" so elongated and extended forwards as to unite with the "pre- 

 frontals " and thereby exclude the narrowed frontals from the orbits, 

 while the Tetraodontines have wide frontals entering into the roofs of 

 the orbits. Three families of lizards are distinguished chiefly on account 



* Many true Carangids have twenty-five or other than twenty- four vertebrae. 



t The strict accuracy and abseuce of exaggeration of the statements here made may 

 be interred from facts. Dr. Giinther, in his Catalogue of Fishes, referred Naucrates 

 to the Scomhriihe because it had Vert. 10-10 (p. 374), and referred Scriola (p. 462) and 

 Naucleriis (|). 469) to the Carangidi© because they were supposed to have Vert. 10-14. 

 Soon afterwards I demonstrated that two of the nominal species of Seriola and 

 the six nominal species of iVa«cZen/.s were based on different stages of the youth of Nau- 

 crates and that probably all belonged to a single -species. The correctness of this 

 statement has long been universally accepted — even by Dr. Giinther — and the genus 

 Naucrates is now retained by all in the Carangidte next to Seriola, where I placed it. 

 (See Proc. Acad, of Nat. Sc. Phila., 18()-2.) 



t See Boulenger, Cat. Lizards B. M., v. 1, pp. 1, 2. 



