14 ALLIS. [VOL. I. 



perichondrial or periosteal layer of the same membrane, which 

 in Amia forms a separate and independent lining of the inner 

 surface of the lateral wall of the chamber. The dorsal portion 

 of the tall orbital opening of the eye-muscle of Amia is thus, in 

 these teleosts, closed toward the orbit; and the upper lateral 

 chamber of the eye-muscle canal on each side of the head is re- 

 duced to certain intradural spaces, or to certain canals or chambers 

 which traverse, or lie in, the side wall of the skull, and transmit 

 the several nerves. The same is apparently true not only of all 

 other teleosts, but also of all elasmobranchs, as the membranous 

 interorbital walls of Chimaera and Callorhynchus plainly show 

 (No. 13, Figs, i, 2). 



The two separate layers of the dural membrane of Amia 

 having fused in other fishes, or never having separated to the 

 extent found in Amia, whichever it may be, chondrification or 

 ossification invades them from different sides and to different 

 extents in different fishes. 



In most teleosts the petrosal in particular undergoes a great 

 development, often extending forward in the dural membrane, 

 and also in the adjoining cartilage of the side wall of the skull, 

 beyond the lateral edge of the pituitary fossa to the hind wall 

 of the orbit, a part of which it forms. This marked character- 

 istic of the teleostean skull, found already indicated in Amia, is 

 accompanied by a relatively small development of the alisphe- 

 noid, that bone in teleosts seeming never to invade the dural 

 membranes or adjacent cartilage to any great extent. The tele- 

 ostean basisphenoid, where it is developed, also extends its 

 horizontal wings on each side in the dural membrane, invading 

 that part of the membrane that lies in front of the pituitary 

 fossa, and even reaching and articulating with the anterior 

 extension of the petrosal, or with the adjacent ventral edge of 

 the alisphenoid. Esox represents an intermediate stage in this 

 development (No. 32, PL XVIII, Fig. 9). 



In Polypterus, in marked distinction to Amia and teleosts, it 

 is the alisphenoid and not a petrosal bone that tends to occupy 

 that antelabyrinthian region of the skull that lies between the 

 trigeminal and facial foramina. In man, the complete invasion 

 of this region by the former bone is plainly shown by a com- 



