MUSTELUS LiEVIS. 120 



internuSj and somewhat anterior to the surface of origiu of 

 that muscle, as Tiesing (62) shows it in his figures. Three 

 branches arising close together, if not as parts of a single 

 branch, are immediately sent to the ventral portion of the 

 rectus internus, and then a fourth branch to the dorsal 

 portion of the same muscle. A fifth and wholly independent 

 branch is then sent upward and backward to the rectus 

 superior. The main nerve, as it gives off these five branches, 

 is here running backward between the skull and the rectus 

 internus, the muscle being so pressed against the nerve that 

 its mesial sui-face is grooved to receive it. The muscle runs 

 forward immediately dorsal to the eye-stalk. The nerve 

 issues from the skull slightly anterior to the eye-stalk, and 

 dorsal to it, and runs backward above it. 



The eye-stalk at this age is of cartilage, but this cartilage 

 is not directly continuous with the cartilage of the skull, an 

 intervening space of procartilaginous tissue separating them. 

 This procartilaginous part of the stalk curves downward, and 

 is continuous with the ventral edge of a defect or perforation 

 in the cartilaginous part of the side wall of the skull, the 

 defect of the cartilage being, however, entirely closed by the 

 perichondrial membranes. The external perichondrial mem- 

 brane extends outward ou to the base of the stalk. The 

 defect lies almost directly anterior to, and not far from, the 

 external opening of the canalis transversus. What the 

 significance of the defect is I could not determine, but it 

 might receive an explanation under the assumption that the 

 eye-stalk is a remnant of a visceral arch, as Gegenbaur has 

 suggested. 



As the oculomotorius gives off the five branches above 

 described it passes backward beyond the eye-stalk, and 

 beyond that structure tlie remaining portion of the nerve, 

 which is now its ventral division only, turns downward 

 between what look like two independent heads of the rectus 

 internus. The ventral and much larger head of this muscle 

 arises from the side wall of the skull posterior to this point, 

 the surface of origin of the muscle lying immediately ventral 



