114 MR. E. PHELPS ALUS, JUN., OX 



pi'aecevebralis, and hence not partxS of tlie Pra^frontallUcke of 

 Gegenbaur's (1872) descriptions of the Selacliii. The eavuni 

 praecerebrale, which in the Selachii lies directly anterior to the 

 fenestra prfecerebralis, cannot be represented in any part of 

 the ethmoidal canal, for, aside from the origin of that canal as 

 above explained, the floor of the canal is not formed by the 

 trabeculse, while the floor of the caviim prtecerebrale is. The 

 hind end of the ethmoidal canal is shown, in Schaninsland's 

 figure 165, lying but slightly anterior to the anterior surface of 

 the mid-brain. If this figure is correct, the membranous mesial 

 walls of the orbits must accordingly, even in the oldest embrj^os 

 considered by Schauinsland, be separated from each other by 

 a considerable interval. 



The chondrocranium of embryos of Callorhynchus thus ap- 

 parently owes the several points in which it difl^ers from that of 

 the Selachii mainly, if not wholly, to the fact that that portion 

 of the central nervous system that lies anterior to the plica 

 encephali ventralis not only projects ventrally or antero-ventrally 

 at the time when the trabeculse are laid down, but that, for some 

 reason, it has continued to lie in that position instead of later 

 gradually curving forward or forward and upward. The tra- 

 beculfe still apparently seek to curve upward into the line 

 prolonged of the parachordals, but, because of the interference of 

 the overlying brain, this iipward curve is found immediately 

 anterior to the lobi olfactorii instead of, as in the Selachii, in the 

 pituitary I'egion. 



The mandibular branchial bars were, doubtless laid doAvn 

 primarily at right angles to the trabeeulsR, and hence, as in the 

 Selachii, in a nearly horizontal position, but as the trabeculse 

 later grow downward and forward instead of, as in the Selachii, 

 curving gradually forward or forward and upward, the mandibular 

 bars have been distoi'ted. The dorsal ends of the processus 

 oticus and basilaris, the latter representing the primitive dorsal 

 end of the arch, apparently remain appi^oximately in their 

 pinmitive positions in relation, respectively, to the lateral wall of 

 the otic capsule and the trabeculae, but they have been stretched 

 out into long cartilages by the marked ventro-anterior growth of 

 the trabecidfe, and the epal (quadrate) and ceratal (mandibula) 

 elements of the arch lie in the region of the antorbital process. 

 This will be again referred to when describing the conditions in 

 ChimcBva. Because of this distortion and change in position of 

 the mandibular branchial bars, the branchial bars of the more 

 posterior arches have also been carried forward and somewhat 

 downward without having been previously pushed backward to 

 the extent that they were in the Selachii, and the dorsal ends 

 of the anterior branchial bars are shown lying in the cranial 

 region in Schaninsland's oldest embryos, The sigma form of arch 

 has, however, been impressed upon their doi'sal ends, but not 

 upon their ventral ends. 



