296 TRANSACTIONS OF THE AMERICAN PHILOSOPHICAL SOCIETY 



even though its fin rays are mostly soft, since the skull characters rather definitely remove 

 it from truly soft-rayed fishes and ally it with the berycoids. In Velifer, according to 

 Frost (1927fl, pp. 440, 444), the principal otolith is fairly generalized and resembles that 

 of the berycoid Polymixia. The sagitta of Trachypterus may be a reduced and simplified 

 derivative of the P'elifer type. In Lampris extreme specialization of the sagitta has wiped 

 out all resemblances to the other allotriognaths; the second otolith (asteriscus) is as high 

 as the sagitta — a very unusual condition (pp. 439, 444). 



The remarkable modifications of the mouth parts and of the cranium in the different 

 families have been fully described by Tate Regan. In Trachypterus and Velifer (Fig. 173) 



Velifer h/pselopterus 



Fig. 173. Velifer hypseloplerus. 



the protractile mouth has a peculiar type of maxillse which can be protruded along with 

 the premaxillae; the lower forks of the maxillse meet below the ascending processes of the 

 premaxillse and slide backward and forward on each side of a median keel on the vomer or 

 on the preethmoid cartilage (Regan, 1907a, p. 639). The protrusility is lost in the highly 

 specialized lophotid Eumetichthys, in which the posterior ends of the premaxillary processes 

 are attached to the anterior face of the vomer (p. 639). 



The cranium of Lampris luna (Fig. 174) strongly resembles the primitive berycoid 

 type as described by Starks (1904^) except that it bears a huge sagittal crest on the supra- 

 occipital and frontals which extends forward above the prefrontals (Tate Regan, 1907a, 

 p. 635). In Felifer (p. 636) the cranial elements are arranged according to the same plan. 



