182 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 9 



by Jordan and Hubbs are subject to personal interpretation and judgment 

 and must be used with discretion. The wide range of possible variation 

 in the interpretation of Jordan and Hubbs' tooth characters is discussed in 

 detail by Parr (1932, p. 25). 



At the same time, while adequately reviewing the Congrid literature, 

 Parr did little to abate the taxonomic confusion in which the Congrids are 

 so deeply mired. Without specimens for comparison, his work, based upon 

 a study of the inadeqviate literature, cannot be accepted without reserva- 

 tion. Subsequent investigation by Reid (1934) has shown more exactly 

 the relationships of several of these genera, and it is hoped that this paper 

 will assist somewhat in the clarification of the confusion surrounding this 

 family. It is well to remember, as Reid suggests, that Congrids can be 

 properly classified only when studied upon a world-wide basis, as 

 their generic relationships are widely distributed. Until such time as 

 adequate material is available, their generic concepts must remain provi- 

 sional and subject to change. 



Reid, using characters hitherto overlooked by most investigators, based 

 his generic differences upon the bony structure of the labial canal, the 

 presence or absence of a free, upturned, labial flange, and the shape, 

 position, and spacing of the dental bands. The bony structure of the 

 labial canal was first described and figured by Bleeker (1864, p. 29) for 

 JJroconger lepturiis, and was recently more fully described by Schmidt 

 (1929). Careful examination of his Congrid material convinced Reid 

 that Schmidt was inaccurate in his description of the structure of the bony 

 raj's which project downward from the inferior edge of the canal. He 

 found that these bony rays are articulate with the lower edge of the 

 canal, and not free elements as Schmidt had described. Nor are the pore- 

 like slits in the upper lip invaginations between the bony supports for the 

 purpose of facilitating the expansion of the membrane, as Schmidt had 

 supposed, but seem rather to be muciferous pores. 



These bony processes extending downward from the inferior edge of 

 the labial canal are present in all the genera studied by Reid, reaching 

 their greatest development in JJroconger lepturus, where they serve as 

 distendable supports for the membrane of the upper lip. The structure of 

 the labial canal, as shown by Reid, is a longitudinal canal lying just above 

 the edge of the upper lip, and extending posteriorly from the rostral 

 cavity at the tip of the snout to beneath the eye and upward around its 

 posterior edge. The inferior, superior, and interior walls of the canal are 

 of bone, the external surface is covered by skin only. From the inferior 

 edge of the canal, several small bony processes extend downward into the 



