70 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 



disturbed anteriorly, scarcely at all so posterior to the line of flexure. All the 

 fins with the exception of the pectorals are beautifully preserved, but both 

 pectorals are very defective. Notwithstanding the thickness of its separate 

 plates, the cranial box yielded to pressure of the overlying matrix, and became 

 irregularly flattened prior to fossilization. Most of the external head-bones are 

 displaced, and the only ones escaping serious injury are the opercular apparatus 

 and jaws of the right-hand side. The cranium, therefore, is in a verj' unsatis- 

 factory condition for study, and it is fortunate our knowledge of its osteology 

 is supplemented by a second specimen, which is described below. 



Cranium. — Turning our attention first to the right-hand side of the head, 

 we find that the operculum, suboperculum, interoperculum, preorbitals, maxil- 

 lary and mandibular ramus all occupy their normal position with respect to 

 one another, being simply flattened out, not displaced. The opercular plates 

 have practically the same configuration and arrangement as iu recent species, 

 but are many times mcwe massive, thus harmonizing with the powerful armor- 

 ing of the trunk. The postero-inferior angle of the interoperculum is developed 

 into a stout, blunt process overlapping the suboperculum. The latter plate, 

 together with the operculum, has a slightly different ornamentation from the 

 remaining bones of the head, in that the ganoine tubercles are fused into 

 more or less continuous and radiating ridges. On the jaws and bones form- 

 ing the roof of the head the tubercular ridges are ramifying and irregularly 

 confluent. 



The maxillary is preserved in its entirety and measures 19 cm. in length 

 Anteriorly it shows a fontanelle as in recent forms, but traces of its segmenta- 

 tion are now nearly obliterated. In the Washington cranium described by 

 Mr. Lucas the segments are very distinct, and are seven in number. (See 

 infra, p. 73). As the oral aspect of the maxillary is not exposed, nothing can 

 be affirmed of its dentition. Considering its extreme narrowness, however, 

 and the fact that only a single dental groove is opposed to it in the lower jaw, 

 it is improbable that more than one series of large teeth was present In this 

 character a noteworthy diff'erence is to be observed between the species under 

 consideration and the recent L. trisicechus (= L. viridis Gthr.), with which it 

 stands other\\ase in close agreement ; and incidentally it proves the artificial 

 nature of Rafine.^que's subgenus Atractosteus. For if we emend the definition 

 of the latter so as to include its nearest allied fossil species, no characters are 

 left by which it can be distinguished from Lepidosteus s. str. Hence it seems 

 best to discard altogether the subgencric terms Atractosteus and Cylindrosteus.^ 



The mandibular ramus is 25 cm. long and composed of the usual parts, den- 

 tary, angular, and coronoid. Immediately behind the last-named element are 

 two circumorbital plates, but all the remaining circumorl)itals and suborbitals 



1 According to Jordan and Evermann (Bull. 47, U. S. Nat. Miis., pt. 1, p. 100), 

 "The name Lilholcpis, Kafinesq\ie, .applied by him to a gigantic gar, Lithohpn ada- 

 ninntiiiiis, the 'Devil-jack Diamond Fish,' is based on a drawing by Audubon, not 

 intended by Audubon to represent any possible fi.sii." 



