4 Carl L. Hubbs 



7. Finally, a fact perhaps most conclusively indicative of a more advanced 

 organization, the branchiostegal rays of the Cyprinodontes are of the acan- 

 thopterygian type, vi^hereas those of the Haplomi are of a distinctly more, 

 generalized type.^ 



II 



The writer agrees with Jordan' that the old family Poeciliidae or Cyprino- 

 dontidae should be dismembered. While as a whole Jordan's division of the 

 group seems sound, in certain details it apparently requires modification. 



The subfamily Fundulinae is equivalent to the Haplochilinae (properly 

 the Aplocheilinae) of Garman, a name of earlier date. The Orestiidae might 

 better be regarded also as a subfamily (Orestiinae) of the Cyprinodontidae. 

 The Characodontidae and Goodeidae should not be separated, for to do 

 so would destroy the extreme naturalness of the combined group. The name 

 Characodontinae is synonymous with Goodeinae, which is the older; the 

 family should therefore be named Goodeidae. This family may best be 

 divided into two groups (quite unlike those recognized by Jordan), namely, 

 the Zoogoneticinae (new subfamily name), containing Zoogoneticus and 

 Girardinichthys, and characterized by the simple teeth and short intestine, 

 and the Goodeinae proper, consisting of Characodon, Chapalichthys, Goodea 

 and Skiffea, and characterized by the bicuspid teeth and more or less elongate 

 intestine. 



The Poeciliidae (misprinted Poecilidae) may be accepted as outlined by 

 Jordan, except that four generic names listed under this family should be 

 transferred to the Cyprinodontidae. These names follow : Gambusinus, 

 Cynodonichthys, Ilyodon (relationship doubtful) and Cynopoecilus. 



The Fitzroyidae and Anablepidae should probably be reduced to subfam- 

 ily rank, for the latter group was very probably derived from the former. 

 The name Anablepidae holds for the combined group, being the older. The 

 name Fitzroyinae must be altered to Jenynsiinae, according to the interna- 

 tional rules as at present interpreted (Fitzroyia and Jenynsia were proposed 

 in the same work by Giinther; Fitzroyia has page priority, but the name 

 Jenynsia was accepted by the earlier writers, who regarded the two genera 

 as synonymous). 



Ill 



The families of Cyprinodontes may therefore stand as follows: 



Cyprinodontidae Poeciliidae 



Adrianichthyidae Phallostethidae 



Goodeidae Amblyopsidae 

 Anablepidae 



2 See Hubbs, Jour. Morph., 33. 1919. P- 67. 



3 A classification of fishes including families and genera so far as known. Stan- 

 ford Univ. Publ. {Biol. Sci.), 3, 1923, PP- 77-243, i-x. 



