206 A. Smith Woodward — On a Liassic Fish. 



just described, but also seem to necessitate the reference of two other 

 specimens from Lyme Regis to Osteorachis, and probably to 0. macro- 

 cephaliis itself. These are imperfect fishes bearing the names of 

 Sarpactira velox ^ and Seterolepidotus grandis^^ each about a metre in 

 length, the first in the British Museum and the second in the Davis 

 Collection, Halifax. They not only exhibit the teeth, but also remains 

 of the typical squamation ; and they are important as making known 

 the greater part of the fins. 



Harpactira velox has never hitherto been satisfactorily determined ; 

 and, remembering that the unique example of this fish displays 

 neither vertebral elements nor ossified vertebral arches, there may 

 still be some hesitation in accepting its identification with Osteorachis. 

 The great amount of disturbance, however, to which the specimen has 

 been subjected is indicated by the fact that the left suboperculum is 

 displaced to the hinder part of the caudal region. Moreover, we may 

 cite another example of a fish from the Lias of Lyme Regis in which 

 the fins remain in position, while nearly the whole of the well-calcified 

 series of vertebral arches is wanting ( Chondrosteus acipenseroides, Brit. 

 Mus. No. P. 3367). The present writer is thus convinced that the 

 evidence of teeth and scales must outweigh all other considerations ; 

 and a careful comparison of the original specimens compels him to 

 relegate the name Harpactira velox to the synonymy of Osteorachis 

 macrocephalus. This unique type fossil exhibits the imperfect head 

 obliquely crushed, and the greater part of the right mandibular ramus 

 is exposed from the inner aspect. The dentary is much fractured in 

 front, though showing the coronoid elevation behind ; while the 

 splenial is displaced, but sufficiently complete to display its clustered 

 small slender teeth and the excavation of its hinder margin. Some of 

 the characteristic teeth are also scattered below the jaw. The right 

 hypohyal occurs ; and the large gular plate is completely preserved, 

 this being two-thirds as broad as long, obtusely rounded in front, 

 truncated behind. The exposed portion of the suboperculum is slightly 

 more than twice as broad as deep, ornamented with a very fine and 

 close tuberculation ; and there is a small ascending process at its 

 antero-superior angle. Scales occur sparsely in the region of the 

 pelvic, dorsal, anal, and caudal fins, all exhibiting a fine tubercular 

 ornamentation. On the few principal flank- scales which are preserved, 

 the tubercles have a tendency to elongation and pass in the hinder half 

 into conspicuous striae. The peg and socket for articulation are very 

 large, as usual. The ventral scales are shown to have been narrowed, 

 as also are those immediately covering the base of the caudal fin-rays. 



Scales of the form just described are much better preserved in 

 the so-called Seterolepidotus grandis of Davis ; and the fins of both the 

 specimens in question are apparently identical in general characters and 

 disposition with those of the genus Eugnathus. It is not yet known, 

 however, whether the fulcra in Osteorachis were biserial or uniserial. 



The result of these further comparisons therefore is, that Osteorachis 

 macrocephalus belongs to the same family as Eugnathus, and is 

 generically distinguished from this fish by the enlargement of the 

 inner clustered teeth, and the comparative tenuity of its scales. 



1 Egerton, Geol. Mag. Dec. II, Vol. Ill (1876), pp. 441, 576. 

 . 2 J. w. Davis, Journ. Liim. Soc. Zool. vol. xviii (1885), p. 293, pi. vii. 



