704 THE GENUS GNATHANACANTHUS GILL. 



Cirritids. The eyes are near the proximal ends of the preopercular 

 boues, the proscapular bones are ver^' much inclined backwards, and 

 must consequently connect with the posterotemporals at decided 

 angles, the pectorals are comparatively narrow and consequently the 

 actinostsaud coracoid elements must be modified, and the veutrals are 

 subbrachiai. Notwithstanding Bleeker's remarks, I must regard it as 

 doubtful whether the uormal cataphract structure is manifest, that is, 

 whether the third suborbital is developed as a stay. In view of the 

 combination of characters exemplified the genus appears to be refera- 

 ble to neither the Scorpnenids nor the Cirritids, and it probably repre- 

 sents a peculiar family to be called Gnathanacanthidw. It may be 

 most nearly related to the Congiopodids or Agriopodids and the Pa- 

 tiiecids. It is very desirable that the questions thus submitted should 

 be investigated, and to rectify the nomenclature and to direct attention 

 to a peculiarly interesting type, unduly neglected, this article is pre- 

 sented. The author would be greatly obliged to any one who would 

 favor him (or rather the U. S. National Museum) with specimens or 

 with any bones of the fish. 



VI. 



Inasmuch as Gnathanacantlms (or Holo.renus), Congiopodus (or Agri- 

 opus), and Patacus have been widely separated and associated with very 

 different forms, I venture to express my belief that that they will be 

 found to be related, and may even constitute a single superfomily. 

 They agree superficially in the pauciradiate simple rayed pectorals, ad- 

 vanced spinous dorsalis, and position of eyes. I trust that skeletons 

 or specimens to be skeletonized of the several types may be sent to the 

 U. S. National Museum for examination. 



