THE CHONDROCRANIUM IN THE ICHTHTOP8IDA. 117 



foramen traversed by a nerve (apparently the palatine). 

 This foramen does not appear in the later stages modelled 

 but I am unable to give the details of the disappearance. 



After passing forward separately to the orbital region, 

 the dorsal and ventral trabecular rods of each side are 

 connected by two narrow bands of cartilage, a post- 

 orbital {pob) and a pre-orbital {prb) between which the 

 elongated optic foramen {of) is enclosed. Beginning in 

 the region of the post-orbital band the dorsal and ventral 

 trabecular rods of each side, which have thus far been 

 approximately parallel, diverge in a horizontal direction. 

 The dorsal rod curves first outward and then inward, 

 giving off ventrally a plate-like lamina cribrosa near its 

 anterior end. The two ventral rods bend inward to the 

 median line where they unite to form a small ethmoid 

 plate. Just behind their point of fusion each ventral rod 

 gives off a ventro-lateral process which underlies the 

 posterior end of the olfactory vesicle. From the anterior 

 margin of the pre-orbital band of cartilage an antorbital 

 process (Fig. 23, anj)) extends outward and forward to- 

 wards the ventral portion of the lamina cribrosa. 



The quadrate is composed of a l)ody, and ascending and 

 pterygoid processes. The ])ouy is small and stands out 

 from the side of the ventral trabecular rod below the an- 

 terior end of the ear capsule, with which it is not directly 

 connected. The ascending process passes upward and 

 forward and unites with the dorsal trabecular rod a little 

 in front of the ear capsule. The pterygoid process {pt) 

 is composed of two parts, a short proximal portion which 

 projects forward from the body, and an isolated portion 

 which later becomes the distal end of the process. This 

 method of development of the pterygoid is the same as that 

 mentioned by Gaupp ('91) for Siredon. A short dis- 



