ICHTHYOLOGY OF THE LESSER ANTILLES, 



451 





Basal radii much more numerous than the interneural bonos radiating from their extremity and oontinuod into 

 branching distal radii ; interneurrls much reduced in number. Acttnistia, fam. CoelacanUndw. 



Tho basilar segments are entirely wanting in the higher Actinopleri, but are variously present in the lower. 

 They are rudimental in the dorsal fin of Amiurus, and are more distinct in tho dorsal fin of GymnaroJius and Carpiodes. 

 They are well developed, but short in the dorsal and anal fins of Goragonus, Salno, Slops, Mormyrus, OHeoglossum, 

 and oven in tho spinous rayed Batrachus. They are small in the anal fin of Ilydrocyon. In Amia they arc long in 

 the anal, but wanting in tho dorsal fin. Finally, in the sturgeons they arc well marked, but imperfectly ossified in 

 both dorsal and anal fins. In most Clupeidaj and Characinidaj they are not visible, and are wanting in Ileteroiis, 

 Butyrinus and ITyodon. In all cases whore they are present, however, they are greatly more numerous than tho 

 vertebra;, as are also tho interneural bones. 



The origin of this increase suggests itself, and leads back to tho proposition discussed with referonco to tho 

 increase seen in tho Coelacanthidce. Should a type of this group bo discovered with dorsal flnsor pinnules as numerous 

 as those of Polyplsrus and with the structure of those of Tristichopterus, wo would then bo with reference to our 

 increase in tho number of interneural bones, nearly whore Polypterus places us with reference to tho increase of the basal 

 segments. Wo may then suppose that the most divergent of tho supernumerary basilars became lateral on tho inter- 

 neural adjacent to each, by the shortening of the axial basilar. That this shortening has taken place, is obvious from 

 tho condition in which wo find it in the many genera mentioned, and that its effect would be to throw the lateral 

 basilars on to tho interneural is almost a necessity. Wo have thus the high number of interneurals seen in tho 

 Actinopleri accounted for, and though as yet largely hypothetical, the explanation seems extremely probable. 



5. On the AcTiNorTEiu. 



In determining tho primary typos of this sub-class, we recur to some characters already mentioned, in which 

 they approximate tho Grossopterygia, and adding others, follow the various divergences to their specialized terminations. 



Thus in Aecipc.nser and allies the ventral fins possess a complete scries of basal radial bones, and the pectorals each 

 a largo mosopterygium.. In Amia and Lepidosleus, the mesoptorygium is small and tho basal radii of tho ventrals are 

 reduced to their lowost number. In none of them are the basihyals fully developed. Most of the eels retain a char- 

 acter which wo have only observed heretofore in the Selachii. 



Wo pass by a number of tho lower fishes boforo we find tho mandibular arch furnished with a symplectic. Oho 

 of tho most important modifications, which is more or less coincident with a number of others, is that which formed 

 tho basis of Bonaparte and Muller's order of Physoslomi. Tho prcsonco of tho ductus pneumatious which character- 

 ized it is always associated with an abdominal position of ventral fins and cycloid scales, and mostly with tho presenco 

 of the prjocorocoid arch, tho entrance of the maxillary bone into tho border of the mouth and tho non separation of 

 the parietal bones by tho supraoccipital. Yet none of those characters are precisely associated at the point of ehango 

 in each, for there are Physostomous fishes witli separated pariotals and ctenoid scales (some Gyprinodontidm) and there 

 arc Physoclysti with abdominal ventrals. Nevertheless thrco prominent types stand out in tho Aclinopteri, the 

 sturgeons or Ghondrostei, tho Physoslomi, and tho Physoalysli, and may bo considered as tribes. 



An entire series Of basilar segments of the abdominal ventral fin; no branchiostogal rays. 



CIIONDROSTEI. 



Basilar segmonts of ventrals rudimental, position of fins abdominal, parietal bonos usually united ; branchiostegal 

 rays ; swim bladder connected with the stomach or oesophagus by a ductus pneumaticus. piiysostomi. 



No ductus pneumaticus ; parietal bones separated by tho supraoccipital ; ventral fins usually thoracic or jugular ; 

 no basilar segmonts. niYSOCXYSTi. 



CnONDKOSTEI. 



There are two orders in this division, as follows : 



A prajcoraooid arch ; no symplectic bone ; promaxillary forming mouth border ; no suboperoulum nor prascoper- 

 culum ; mosoptcrigium distinct ; basihyals and superior ceratohyal not ossified ; intcrclaviclcs present, 

 operculum or maxillary bones ; branchihyals cartilaginous. 



No inter- 



1 SELACIIOSTOMI. 



A 



A 



