XI IE S K U bL () F ( ' J 1 1 M _K RA . 1 J D 



slielf wliicli transmits the arteria cavotis externa, jiud tlie lamus 

 palatiuus facialis, tliese two foramina, being jiersistiug remnants 

 of the palatoquadrate tissure of Dean's (1906, p. 108) descriptions 

 of embryos. Tliat tissure of embryos lies, as sliown in Dean's 

 figures, between tlie palatoquadrate, the trabecula, and'tlie 

 anterior edge of a shelf of cartilage that projects laterally from 

 the neurocranium beneath tlie hind end of the orbit and is 

 fused, in its antero-lateral portion, with the palatoqua,(h-ate. 

 The vena jugularis accordingly lies definitely ventral to this 

 laterally projecting shelf -like process of the neurocranium, and 

 hence also morphologically ventral and internal to the dorso- 

 posterior portion of tlie palatoquadrate. If then this shelf-like 

 process is an outgrowth of the neurocranium, the dorso-posterior 

 end of the palatoquadrate, as shown by Dean, must be the 

 processus oticus of that cartilage, but it seems much more 

 probable tha.t the shelf itself is the processus oticus, or extra- 

 branchial element of the arch. Dean (I. c. p. 129) considers this 

 slielf to be ,the posterior portion of the pharyngeal element of 

 the mandibular arch, but in that case this element of the arch 

 -has fused with the neurocranium dorsal to the vena jugularis, 

 Avhich would be in mai-ked exception to the conditions found in 

 all other fishes that I know of (Allis, 1915). It accordingly 

 seems to me, as above sta,ted, that this shelf must represent 

 the extrabranchial of the arch, or processus oticus qnadi-ati, 

 apparently found only partially fused with the palatoquadra,te in 

 Dean's six-month embryo. The vena jugulaiis, ramus liyoideo- 

 uiandibularis facialis, and arteria carotis externa would then all 

 lie ventral and internal to this element of the arch, as they 

 normally should, and the foramina for these several structures 

 ■would represent the trigemino-facialis cha.mber of the fish, as 

 I have already suggested in an earlier work (Allis, 1914). 



The orbit has postorbital and antorbital processes. The 

 postorbital process is a pronounced and slightly curved ridge 

 which is fused ventrally with the suborbital shelf. The dorso- 

 lateral antorbital process is a short but tall ridge which lies at 

 the dorso-anterior edge of the orbit, and its base is perforated 

 by a foramen which transmits the ramus ophthalmicus supei"- 

 ficialis trigemini from the orbit to the ethmoidal canal. 

 Immediately ventral to this process there is a foramen which 

 .tra,nsmits a vein and arteiy, and antero-vential to this latter 

 foramen there is another foramen which transmits the nervus 

 ophthalmicus profundus; both these foramina leading into the 

 median ethmoidal canal. A small foramen dorsal to these 

 foramina also leads into the ethmoidal canal and gives exit to 

 a small branch of the ophthalmicus superficialis. A small 

 foramen o» the dorso-mesial surface of the nasal capsule gives 

 passage to a, branch of the orbital branch of the external carotid, 

 and anterior to it there is another small foiumen for a delicate 

 branch of the ophthalmicus nerve. The ventro-lateral antorbital 

 process is simply a ridge which traverses the suborbital shelf 

 and is continued onto the palato(]uadrate, Avhere it turn.s 



