No. i.] MORPHOLOGY OF THE PETROSAL BONE. 9 



enclosing bone and cartilage, but is separated from that surface 

 by a tough, fibrous membrane, which is the internal periosteum 

 of the vertebral column. The dura mater lies inside this mem- 

 brane, separated from it by an epidural space, and its external 

 surface is simply a hardened superficial portion of the general 

 tissue of the structure, and not at all the partially differentiated 

 and partly osteoblastic limiting membrane of the cranial dura 

 mater. The spinal and cranial durae are thus, from Sage- 

 mehl's own descriptions, not exactly similar structures, not- 

 withstanding his definite statement to the contrary (No. 27, 

 p. 470). It is perhaps not unimportant in this connection to 

 note that fibrous bands, the interclinoid ligaments, are normally 

 found beneath the dura mater in the pituitary region of the 

 human skull (No. 24, vol. ii, pt. i, p. 47) ; and that the cranial 

 dura mater, in many vertebrates, is subject to ossification, 

 sometimes extensive (No. 24, vol. iii, pt. i, p. 183, and No. 

 27, p. 85), which does not occur, so far as I find recorded, in 

 the spinal dura. 



If now the skull of Amia be considered, we find a tough, 

 glistening, fibrous membrane which forms the floor and sides 

 of the interorbital and pituitary parts of the cranial cavity, 

 closing at the same time the optic fenestrae, and forming the- 

 reof of the ventral portion of the eye-muscle canal, and the 

 median walls of its upper, lateral chambers. The hind edge 

 of this membrane is attached to the anterior bounding ridges 

 of the labyrinth recesses, and, between those recesses, to the 

 front edges of the median, horizontal processes of the petro- 

 sals. A separate and independent perichondrial or periosteal 

 membrane, the histological character of which I have not 

 attempted to investigate, lines the floor of the eye-muscle 

 canal, and the lateral walls of its upper, lateral chambers. 

 That this latter membrane is not simply a reflexed portion of 

 the outer pericranial membrane seems to be sufficiently indi- 

 cated by the fibrous tufts in which the free cartilaginous edges 

 of the orbital opening of the eye-muscle canal are always seen 

 to end in sections of larvae of Amia. Moreover, what is ex- 

 ceedingly important in this connection, a subpituitary or peri- 

 pituitary canal is found in Lepidosteus entirely closed, toward 



