612 Bulletin 4J, United States National Museum. 



bauds on intermaxilla, nianclible, and vomer ; in narrower bands on max- 

 illa and palatines. End of tongue barely free. Integumentary flaj) pro- 

 jecting beyond margin of suboperculum. Dorsal beginning about over 

 end of the base of ventral, its distance from tip of snout a little more 

 than twice length of head ; length of its base nearly equal to greatest 

 depth of head ; basal half of the fin profusely covered with scales ; it 

 consists of 2 simple and 9 divided rays. Longest dorsal ray (1,^ inches) 

 equals f greatest height of body, last ray ^ as long as third. Ventral dis- 

 tant from the snout a space about equal to twice head; it consists of 2 

 simple and 7 divided rays, the longest equal to base of dorsal; pectoral 

 placed above middle of body, not far from head, its length ecjual to \ 

 distance of its base from origin of ventral, and reaching to about seven- 

 teenth row of scales. Vent under sixty-third scale of lateral line ; about 

 12 rows of scales between origin of dorsal and lateral line and about 2^ 

 rows between lateral line and origin of ventral ; lateral line composed of 

 enlarged and modified scales, becoming obliterated abont middle of tail; 

 15 or 16 of these enlarged scales between head and origin of ventral; 

 about 27 between vent and head ; distance from origin of ventral to anus 

 (3 inches) nearly equal to head. Anal rays about 166, their origin not far 

 behind vent; caudal, which is long and slender, with about 4 rays; pec- 

 toral 13. Membrane covering anal rays scaled throughout almost its entire 

 length. Gill rakers rather short and few, 3 + 12. Color of scales a light 

 silvery gray, through Avhich the body color appears as a light orange- 

 brown ; branchiostegal membrane and opercular bones bluish ; inside of 

 gill covers very dark blue. The type, a specimen two feet long, taken 

 in the Gulf of Mexico, 21° 36^ N., 84° 05^ W., by the BUU at Station 173, 

 at a depth of 955 fathoms. Many others were obtained in the Gulf Stream, 

 in depths ranging from 679 to 1,430 fathoms, by the Albatross. (Goode 

 & Bean.) (jpalUdus, pale.) 



AhJrovandia jialUrla, Goode & Bean, Oceanic Ichthyology, 13-1, IS'jr,, Gulf of Mexico. (Co-type, 

 No. 38140. Coll. Albatross.) 



Order V. HETEROMI. 



(The Spiny Eels.) 



The 2fotacaHthi(l(v* are separated from the other physostomous fishes by 

 Gill, to form an order which he thus defines: 



* Some recent American authors have placed the Notacanlhidir with the iihysoclystous family 

 of Maslacemhelidse in the order Opislhovii of Cope, orisiually framed for the latter. There seems, 

 however, to be no evidence, as Dr. Vaillant bus Uitily iinteil, of atlinity between the Mustacendie- 

 lidx and the NotcwanOiidit. Vaillant calls attention tu the presence of an air duct in XoUtcanOins, 

 and places the family among the Ganoids, between the Sturgeons and the Eels. According to 

 Vaillant (Exped. 8ci. Traviulleur et Talisman, 324) the dorsal spines in Nolacnidhiis differ widely 

 from those of Teleostean tishes jn having but a single nourishing canal insteadof two. Ilefinds 

 these spines homologous with the plates of sturgeons or rather with the jiinnules of the Ganoid 

 genus P(%p/erH.<!, instead of with the spines of the pliysoclystous fishes. Vaillant also findsmany 

 of the skeletal characters of Nolncoilhus very primitive, suggesting Ganoid affinities. At the 

 same time, the simple air bladder, the absence of sjiiral valve in the intestinesand nf mnltiplied 

 valves in the heart seem to ally Noldcan/hns rather to the Inospoiidi/li or the Eels. Giinther 

 places it among the Pliiisoslumi between the Ualosauridx and the eels, but does not regard it ae 

 having close affiuities with either. 



