192 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 9 



Hubbs ( 1925) for the Japanese Congrid, Leptocephalus nystromi Jordan 

 and Snyder. The holotype of the genotypic species is in the collections of 

 the Natural History Museum of Stanford University, (Stanford 6469), 

 where it and three small paratypes of the species (Stanford 6969) were 

 examined by the author. The validity of the genus was questioned by 

 Parr (1932), when he provisionally placed it in the list of genera 

 inquirenda of the genus Ariosoma. Although Reid (1934) mentions that 

 he examined several Japanese genera, which according to Jordan and 

 Hubbs were distinct, he did not attempt to determine their systematic 

 position. In Reid's key the genus Rhynchocymba may be easily dis- 

 tinguished from Ariosomaj as the former does not possess the upturned 

 labial flange with a free superior border characteristic of Ariosoma. 

 Neither can Rhynchocymba be allocated to any of the other genera 

 included in Reid's key. 



Re-examination of the holotype of nystromi reveals several differences 

 between the specimen and the original generic description. Originally 

 described as without pores between the anterior nostrils, careful manipu- 

 lation of the folded skin with a needle reveals the presence of two enlarged 

 pores similar to those described for Rhynchoconger. The posterior 

 nostrils were described as horizontal slits with entire edges; actually 

 they are elliptical pores with raised rims that are very slightly fimbriate 

 along the inner edge. In the three paratypes of nystromi studied, the 

 posterior nostrils varied from a horizontal slit to an ovate pore with the 

 raised rim having entire edges. Nor is the pocket, found on the antero- 

 ventral surface of the snout of the holotype, present in any of the three 

 paratypes. The use of features so variable in character should be aban- 

 doned in favor of more stable features. It may also be doubted if such 

 characters are of sufficient importance to be of generic rank. 



Agreeing fully with Parr's statement that the arrangement of the 

 tooth bands is a matter of personal judgment and interpretation, the 

 author's analysis of the tooth bands of Rhynchocymba differs greatly 

 from that of Jordan and Hubbs ( 1925, p. 191 ). Repeated re-examination 

 of the available specimens of Rhynchocymba nystromi over a period of 

 time shows that the anterior ends of the maxillary tooth bands do not 

 separate the premaxillary-vomerine teeth, but instead are widely separated 

 by the continuous premaxillary-vomerine teeth. Pressure applied 

 anteriorly to the inner side of the maxillaiy bone forces the anterior end 

 of the maxillary band away from its close contact with the premaxillary- 

 vomerine teeth and reveals that the maxillary teeth end at that point. 



