ON THE VERTEBRATE SKELETON. 211 



dismemberments of the human temporal bone ; and we cannot climb to the 

 higher generalizations of anatomical science, except by the firm steps of true 

 and assured special homologies. There are more important subjects than 

 homologies, no doubt ; but nothing is more important than truth, in whatever 

 path we may be in pursuit of her. 



Orbitosphenoid. — As evidence will be given in the section on ' General 

 Homology' that both squamosal and tympanic belong to a quite distinct 

 category of bones from the parts of the 'temporal' which have just been 

 discussed. I shall proceed next to the neurapophyses that precede the 

 alisphenoid. 



As the determination of this bone (6 in all the figures) involves that of 

 the orbitosphenoid (10), which has rarely been mistaken* for any other bone 

 than 6, there remains little to be added in proof of its homology after 

 what has been advanced respecting the alisphenoid. The most constant 

 character of the orbitosphenoid is its relation to the optic nerve, which either 

 perforates or notches it, whenever the ossification of the primitive cartilage 

 or membrane holding the place of the bone is sufficiently advanced, which 

 is not always the case in fishes, especially those with broad and depressed 

 heads, and still more rarely in lacertine saurians. The recognition of the 

 orbitosphenoid is also often obscured by another cause, viz. the tendency in 

 the class Reptilia, and especially in ophidians and chelonians, to an extension 

 of ossification downwards into the primitive membranous or cartilaginous 

 neurapophysial walls of the brain-case, directly from the parietal and frontal 

 bones. 



In the fishes with ordinary-shaped, or with high and compressed heads, 

 the orbitosphenoids are usually well-developed : they are, however, repre- 

 sented by descending plates of the frontal in the garpike ; and they are, like the 

 alisphenoids, mere processes of the basisphenoid in the polypterus, which thus 

 offers so unexpected a repetition of the human character of the correspond- 

 ing parts f. In the cod (fig. 5, io) they are semielliptic, raised above the pre- 

 sphenoid (9), suspended, as it were, between the alisphenoid (6) and the 

 frontal (11), and bounding the sides of the interorbital outlet of the cranium : 

 the optic nerves pierce the unossified cartilage closing that aperture, imme- 

 diately beneath the bone itself. In the malacopterous fishes with higher 

 and more compressed heads, the orbitosphenoids are more developed ; they are 

 directly pierced or deeply grooved by the optic nerves, and are pierced also 

 by the ' nervi pathetici' in the carp. The crura of the olfactory ganglions 

 (rhinencephala) pass out of the interorbital aperture of the cranium by the 

 upper interspace of the orbitosphenoid, into the continuation of the cranial 

 cavity which grooves the under surface of the frontal, in their course between 

 the orbits to the prefrontals. The orbitosphenoids protect, more or less, the 

 sides of the prosencephalon ; and this function, their transmission of the optic 

 nerves, their anterior position to the alisphenoids, and their articulation 

 above with the frontals, establish their special homology from the fish up to 

 man. 



In certain fishes a distinct centre of ossification is set up in the median 

 line of the fibrous membrane or cartilage, closing the interorbital aperture 

 of the cranium, below the orbitosphenoids, and extending forwards as the in- 

 terorbital septum. The bone (represented in outline in fig. 5, at 9') extends 

 downwards to rest upon the presphenoid (ib. 9), and bifurcates, as it ascends, 



* Geoffroy in his memoir on the skull of birds (Ann. du Mus. x.), indicates the orbitosphe- 

 noid at P, fig. 2, pi. 27, as the 'rocher': and Cuvier describes it as part of his 'os en cein- 

 tui - e' in anourous batrachia. 



t Agassiz, Recherches sur les Poissons Fossiles, ii. p. 38. 



