172 UPPER CRETACEOUS TELEOSTS 
den smx q iop 
Fic. 79. Alepisaurus ferox. Skull in lateral view. 
124). Regan (1911 : 120) has noticed the similarity between Apateodus (Text- 
fig. 11) and Alepisaurus (Text-fig. 79) in both of which no great rostral elongation 
has occurred. The composition of the snout in these two genera is also similar 
with an extended mesethmoid, palatine and vomer with the palatine supporting 
several enlarged, laterally compressed teeth. One difference is that in Apateodus 
both palatine and ectopterygoid bear teeth, whereas in Alepisaurus all of the 
teeth are borne on the palatine. The dermal upper jaws are superficially similar 
but Apateodus is more primitive in that the premaxilla does not have an ascending 
process, and the maxilla is a simple strut. In Alepisaurus the premaxilla and 
maxilla have been derived by simplification from a much more complex pattern. 
Both genera have lost the antorbitals and nasals due to the elongation of the snout. 
The caudal skeleton has already been mentioned as being in a condition com- 
parable to the elopoids and salmonoids in respect of the number of separate ural 
vertebrae. The caudal skeleton of Ichthyotringa furcata (Text-fig. 4) shows a neural 
arch without a neural spine on the first preural vertebra which is like that seen in Elops 
(Text-fig. 80) and Sedenhorstia (Goody, 1969, fig. 2). In the Salmoniformes the 
neural arch of this vertebra is incorporated with the first uroneural (and probably the 
neural arch of ural vertebra one) to produce the stegural (Text-fig. 81). In Ichthyo- 
tvinga the neural arch of preural vertebra one is distinct and the anterior uroneural is 
expanded anteriorly. This expansion probably represents the neural arch of ural 
vertebra one and thus an earlier stage in the production of a stegural, with the neural 
arch of preural vertebra one not having yet been incorporated. In Elops there is also 
