72 ILLINOIS BIOLOGICAL MONOGRAPHS [72 



name, orbitosphenoid, given to the bone by Owen (1848), is in general use at 

 the present time. 



In many of the lower teleosts, where a large interorbital septum is developed, 

 the orbitosphenoid is lacking. Among these are Scomber, the Loricati (Allis, 

 Gutberlet); Perca (Cuvier, Kallmann); Osteoglossum, Gonorhynchus, Chanos 

 (Ridewood, 1904); Gasterosteus (Swinnerton, 1902); and others not referred 

 to here. 



In Amia (Sagemehl, 1884) the orbitosphenoid is paired and lies on the 

 cartilage above the anterior margin of the optic foramen, separated from the 

 bones above and in front by cartilage. In Arapaima (Ridewood, 1904), it is 

 paired, but unlike Amia, these ossifications extend from the frontals to the 

 parasphenoid. In those species with a medium sized interorbital septum as 

 Salmo (Parker, 1872); Alepocephalus (Gegenbaur, 1878), and some of the 

 Cyprinidae (Sagemehl, 1891), the orbitosphenoid is unpaired and Y-shaped, 

 the arms of the Y extending up in the walls of the cranium and the basal part 

 down into the orbital septum. It always lies above the anterior margin of 

 the optic fenestra and usually the basal piece ossifies as far ventrally as the 

 trabecula communis. In the Characinidae the bone has various gradations in 

 structure from the U-shape corresponding to that of Amiurus, to the Y-shaped 

 Sahnonoid type. Of all the forms studied I think that the orbitosphenoid as 

 found in Homolopterus of the Characinidae is the closest in orbitosphenoid 

 relations to Amiurus. There is the same strongly ossified midventral piece, 

 and the same general relations to the surrounding bones and to the optic nerve. 

 The orbital foramen is also present between the orbitosphenoid and the ecteth- 

 moid of each side and considerable cartilage remains within the bone, a strip 

 of it persisting between the frontals and upper ends of the inner surface of the 

 bone. 



^ Vrolik (1873) maintained that, because the orbitosphenoid is an incon- 

 stant bone and because it could be developed from either the membrane of the 

 interorbital septum or from cartilage, that it could not be homologized through- 

 out the different groups. The fact of its relation to the nerve foramina, its 

 gradation in development in the lower forms all aid in proving that it was a 

 paired bone in the ancestral teleosts, and that it was originally developed from 

 cartilage. Cartilaginous cells are present in the interorbital septum of those 

 forms in which the septum is decidedly primitive, a circumstance which may 

 be taken to indicate that there has been a shortening of the ontogeny of the 

 septum in the higher forms where the septum ossifies directly. 



The parasphenoid (Figs. 6, 7), is a long, flat, unpaired bone extending on 

 the ventral surface of the cranium from the supraethmoid to the anterior end 

 of the basioccipital. Its anterior interdigitation with the supraethmoid is 

 hidden by the vomer. The median anterior end of the parasphenoid is kept 

 from contact with the ethmoid cartilage or intemasal septum by the supra- 



