268 



the adductor arcus palatini of Ameiurus are wanting, unless they are 

 represented in one or both of the third and fourth divisions of the 

 levator maxillae superioris; the recorded innervations of the muscles 

 in the two fishes being against this supposition. And in Amia, also, 

 as in Polypterus, there is no separate muscle corresponding to the 

 posterior division of the adductor arcus palatini of Ameiurus, unless 

 this latter muscle be represented, in Amia, in the entire adductor 

 hyomandibularis of that fish. The adductor arcus palatini of Teleosts 

 is generally assumed to be represented, in Amia, in an undefined 

 anterior portion of the large adductor hyomandibularis of that fish; 

 this latter muscle becoming separated, in course of development, into 

 two parts, one of which becomes the adductor arcus palatini while the 

 other remains as the adductor hyomandibularis. It seems however 

 possible, or even probable, that the entire adductor of Amia may re- 

 present the adductor arcus palatini of Teleosts, and that muscle alone; 

 a separate adductor hyomandibularis being developed, in these latter 

 fishes, from the adductor operculi of Amia. This subject evidently 

 needs further study. 



One further point should be briefly referred to. In Polypterus 

 there is a large so-called sphenoid (ahsphenoid?) bone. Posteriorly 

 this bone has invaded the region of the prootic in Teleosts, and the 

 bones of opposite sides are there connected by a prootic bridge. 

 Anteriorly the sphenoid invades the orbitosphenoid region, and the 

 bones of opposite sides may there meet in the mid- ventral line; but 

 they there always remain separated by suture. There is said to be 

 neither basisphenoid nor orbitosphenoid, as independent ossifications. 

 In Ameiurus, McMurrtch describes an alisphenoid, a basisphenoid 

 anchylosed with the parasphenoid, and an orbitosphenoid that is a 

 single median bone forming part of the base of the skull. But this 

 latter bone has, in many respects, the position and relations of a 

 basisphenoid. In my young larva, as in the 20 mm larva examined 

 by Mc MuRRiCH, the bone has three distinct portions ; two lateral 

 portions, one on each side, that are of perichondrial origin, and a thin 

 median connecting plate that is wholly of membrane origin. The lateral 

 portions lie in the inclined plane of the side wall of the skull, and are 

 each developed in perichondrial relations to, and wholly enclose a bar 

 of cartilage that forms the ventral portion of the side wall of the skull 

 and that extends longitudinally backward beneath the opticus foramen, 

 the two bars apparently representing the trabeculae. Anterior to the 

 opticus foramen a flat process of membrane bone arises from the ex- 

 ternal surface of the dorsal end of the bone, and projects upward 



