Jordan and Hvermann. — Irishes of North America. 779 



Order BB. ACANTHOPTERI. 



(The Spinv-raved Fishes.) 



Anterior vertebra' unnioditied and without ossiculii anditns; no meso- 

 coracoid and no interclavicles (so tar as known). Border of month formed 

 l»y premaxillary ; maxillary normally distinct from it and always present, 

 but sometimes coossified with it. Gills laminated. Shoulder girdle 

 attached to the skull by a post-temporal, which is normally furcate and 

 usually not coossified with the skull. Hyjjercoracoid and hypocoracoid 

 distinct, ossified, the former usually perforate. Pharyngeals well devel- 

 oped, the lower rarely united, the third upper pharyngeal largest, the 

 fourth often wanting. Pectoral actinosts always present, opercular 

 apparatus complete; gill openings in advance of the pectorals; pec- 

 toral fins above the plane of the abdomen ; ventral fins more or less 

 anterior, normally attached by the pelvis to the shoulder girdle, typically 

 with 1 spine and .5 rays, sometimes wanting, sometimes without spine or 

 with many rays, or otherwise modified. Anterior rays of dorsal and anal 

 typically simple or spinous, but all the fin rays often articulate. Air 

 bladder typically without duct in the adult. Scales various, typi- 

 cally ctenoid; lateral line usually running high. To this group are 

 referable the great body of existing marine fishes. It may be not a per- 

 fectly natural order, inasmuch as the members of the group diil^er very 

 widely among themselves, and at the same time some of them ajiproach 

 very closely to neighboring orders as the H(q)Iomi, Syneniofjnaihi, Hemi- 

 branchii, etc., groups related to the parent stock of some of the Acan- 

 thopteri. It is not at all certain that the AcantJiopteri have all developed 

 from a common stock, nor in many cases can we indicate the genetic 

 relations. 



Among the Acanihopieri a certain number of well-defined suborders 

 exist — branches more or less separable from the parent stem. It is impos- 

 sible, however, to divide the whole group into suborders, as many of the 

 constituent groups cannot be defined. Thus the Fercusoces, Taniosomi, 

 Heterosomata, Sclerodcrmi, Gii)iniodo)itefi, etc., can be easily segregated and 

 defined, while other groups clustering around equally well-marked forms 

 can not receive any sort of concise and inclusive definition. Examples 

 of this are the Scomhroidci , rercoklei, Trdchinoidei, and the like. But these 

 large groups cannot be all referred to a single suborder, as they differ 

 as much intei- se as the well-defined suborders do. 



We have here adopted the anomalous system of recognizing two soi'ts 

 of subordinate divisions within the order of Acanthopteri, the " suborders " 

 or categories susceptible of definition, and " groups" which can only be 

 defined through alliance with some central or some parent form. Thus 

 the varied group of Scomhroidei center about the mackerels, seeming to 

 diverge in various directions from the type of Scomber. The difficulty 

 of definition may arise from any one of several sources: («) from the real 

 iutergradatiou of forms; (h) from our ignorance of the real characters: 

 or, (c) from the intrusion of unrelated forms which obscure the true 



