^°im"'] PROCEEDINGS OF THE NATIONAL MUSEUM. 163 



rays readily perceptible tlirough the skiu ; dorsal variable, commencing 

 behind or in front of the anus and mostly low ; anal deeper and com- 

 mencing close behing the anus ; caudal prominent. 



Pectorals well developed, near the breast, with the rays distinct. 



Branchial arches nearly complete,with slender glossohyal and urohyal, 

 and with the lirst, second, and third basibrauchials ossified, first and 

 second hypobranchials ossified, third cartilaginous, ceratobrauchials and 

 epibrauchials of four ])airs ossified ; pharyngobrauchials of second pair 

 rod-like, of third pair developed as dentigerous epipharyngeals; hypo- 

 pharyngeals oblong, closely apposed to and superincumbent on the rudi* 

 mentary fifth arch. Interbrauchial fissures extended. 



The osteological characters supporting those enumerated are many 

 and important. It need only be added here, however, that the ectop- 

 terygoid is developed as a slender rod bordering the membranous palatal 

 roof, and that the hyomandibular is very large and directly obliquely 

 backwards, and, inasmuch as the facet for articulation of the operculum 

 is near the distal extremity, the opercular apparatus is thereby thrown 

 far away from the cranium. The corneous appearance of the dentary 

 is striking, but of more importance is the extension forward of the artic- 

 ular bone. Perhaps the most obvious character is the abbreviation of 

 the branchiostegal rays entailing a striking contrast to the other eels 

 of the families iSimenchelyidw, Congridce, AtiguilUdw, Oplmuridw, and 

 Murccnidw, in which they are much elongated and sweep around the 

 opercular apparatus, often intervening between it and the cranium. 



There are two external features which are interesting on account of 

 their bearings on the relationship of the family. They are (1) the char- 

 acter of the scales and (2) the position and approximation of the bran- 

 chial apertures. 



The form and mode of distribution of the scales (elongated and ellip- 

 tical, distant, and arranged in patches at right angles to neighboring 

 ones) are so marked that it might well be thought at first that they 

 must indicate genetic relationship among the fishes so distinguished. 

 If the eels alone were to be considered, it might be argued, with consid- 

 erable plausibility, that the forms so characterized were descended 

 from common ancestors so distinguished. It would even be difficult to 

 present cogent arguments against such a postulate. But the struct- 

 ural differences between the several types so distinguished {Anguil- 

 lidce, SynajyhobrancMda', and Simenchelyidcc) are very great, and that 

 the same mode of squamation may originate independently is evidenced 

 by the fact that among the ophidiids a like form and arrangement of 

 the scales is to be found, although among others of the same family the 

 usual form and imbrication of the scales occurs. 



The approximation of the branchial ai)ertures beneath the throat and 

 their cincture by a common border recall the symbrauchiate fishes, and 

 pu account of this character the present family has been compared 



