ESPECIALLY MYCTOPHOIDS 173 
a small anterior expansion in advance of the first uroneural, which may well represent 
the neural arch of ural vertebra one. If this is the case then the forerunner of a 
stegural is present in the elopoids. In the Salmoniformes the epurals are associated 
with the dorsal edge of the stegural. In Elops and Ichthyotringa the epurals contact 
the dorsal edge of the anterior expansion of uroneural one. This possibly affords 
further evidence that the anterior expansion is the forerunner of a true stegural. 
n.s negate ii esi url 
n.a. pul 
r.n.s.pu2 
ul 
pu4 
hyl 
Fic. 80. Elops sauvus. Caudal fin skeleton in lateral view. 
Preural vertebra two in Ichthyotringa has a complete neural spine. In Elops this 
neural spine is reduced (Text-fig. 80), as it is in certain salmonoids (Norden, 1961; 
pl. 14, fig. G, Coregonus artedi1). In other salmonoids the neural spine is complete 
(Text-fig. 81). In those forms in which a reduced neural spine occurs then three 
separate epurals are normally seen (the primitive teleostean number according 
to Patterson, 1968a : 220). Ichthyotringa, with its complete neural spine on preural 
vertebra two, has only two epurals. It would seem possible, then, that the anterior- 
most epural, when three are present, comes from the neural spine of preural vertebra 
two, the remaining two epurals being the neural spines of preural vertebra one and 
ural vertebra one. Patterson (1968a : 221) has drawn attention to the fact that the 
foremost epural can “‘ fuse with a neural arch so that it is indistinguishable from a 
neural spine’’. It is equally possible that this is what has occurred in Ichthyo- 
tringa. 
