FISHERY BULLETIN: VOL. 79, NO. 1 



Total sum of slime pores on left side. 



Total sum of gill apertures (not counting the 

 pharyngo-cutaneous aperture if the left pos- 

 teriormost gill aperture opens separately). 



Many specimens were cut open to determine the 

 number of gill pouches and their position relative 

 to the tongue muscle and branching of aorta. 



Sex was determined by examining the gonad 

 through a cut in the lateral right side body wall 

 anterior to the cloaca. 



The usefulness of different counts and mea- 

 surements for systematic studies in Paramyxine 

 and Eptatretus have been discussed (Dean 1904; 

 Bigelow and Schroeder 1952; Strahan 1975). We 

 agree with Dean's (1904) statement that ". . . in the 

 case of myxinoids it is peculiarly necessary to base 

 specific determinations upon the average charac- 

 ters of as great a number of individuals as practica- 

 ble." Dean (1904) and Strahan (1975) stressed the 

 importance of the relative position of the gill aper- 

 tures and body proportions, which we also find 

 useful. Dean (1904) tended to disregard the 

 number of slime pores as a systematic character, 

 but we stress the value of that count and point out 

 that both Bigelow and Schroeder (1952) and Stra- 

 han (1975) arrived at ranges for this character 

 which are far too broad because they include two 

 composite species (P. springeri Bigelow and 

 Schroeder andP atami Dean). This inclusion of an 

 undescribed species {E. minor, see below) in P. 

 springeri also led to the erroneous conclusion 

 (Bigelow and Schroeder 1952) that the number of 

 slime pores increased with length of the animal. 



As indicated earlier (Dean 1904; Strahan 1975), 

 the gill aperture counts vary slightly within the 

 five- to seven-gilled species of Eptatretus , hut when 

 a larger sample is available, the character is, of 

 course, quite valuable and is easily examined. 



Tooth counts were considered of little value by 

 Dean (1904) and Strahan (1975). We find, however, 

 as had Regan (1912) in his admirable yet terse 

 synopsis of the multibranchiate myxinids, that 

 this is a very useful character. Especially we find 

 the pattern of fused cusps (Figure 2) in the inner 

 and outer row of teeth to be constant within 

 species in all of the hundreds of specimens of Epta- 

 tretus and Paramyxine we have studied (unpubl. 

 data). 



The movements of the rasping lingual tooth 

 plates are rather elaborate in hagfishes and com- 

 plicate the terminology used to express positional 

 relationships. In agreement with some previous 



authors (Dean 1904; Strahan 1975), we have cho- 

 sen to call the row of teeth with larger teeth the 

 outer row and the other row the inner. The fused 

 tooth or multicusp we regard as the anterior in 

 each row, and to designate the pattern of fused 

 cusps or teeth in the multicusps of the outer/ inner 

 row we write 3/2 or 3/3, which are the two patterns 

 found in the Atlantic species of Eptatretus. 



The shape of the gill apertures, as well as exten- 

 sion and shape of the relatively low ventral fin fold, 

 have been used as systematic characters, but we 

 find it difficult to assess these characters in pre- 

 served specimens. 



The pattern of fused teeth (3/2 or 3/3), 

 supplemented with counts of gill apertures and 

 slime pores, appears to suffice to distinguish the 

 western Atlantic species of Eptatretus. 



GENERIC ALLOCATION 



The generic allocation of the polybranchiate 

 species of hagfishes has been considerably dis- 

 cussed (for summary, see Holly 1933 and Strahan 

 1975). To us it is obvious that the name with prior- 

 ity, Eptatretus Cloquet, 1819, should be used. As 

 stressed earlier (Bigelow and Schroeder 1948; 

 Adam and Strahan 1963; Hubbs 1963; Strahan 

 1975), there is no obvious advantage in dividing 

 the genus into subgenera. 



It has been argued that Paramyxine as a genus 

 should be treated as a junior synonym of Eptatre- 

 tus (Strahan 1975). It is true that tendencies to 

 shortening of the gill aperture area can be found in 

 E. springeri (Bigelow and Schroeder 1952) and in 

 E. burgeri (Strahan 1975), but we believe this rep- 

 resents a convergent trend of development in the 

 Atlantic and Pacific Oceans. We retain the generic 

 name Paramyxine for the western Pacific species. 

 Dean (1904) defined the genus Paramyxine on the 

 basis of a single specimen of P. atami. Now that 

 much more material is available, it is somewhat a 

 matter of choice whether one wants to retain 

 Paramyxine for the Asian species. We choose to do 

 so for the following reasons: 1) Dean's concept of 

 crowdedness of gill apertures has been 

 strengthened by the description of Taiwanese 

 species of Paramyxine (Teng 1958; Shen and Tao 

 1975), which are more extreme in this character 

 than is the type-species; 2) as a further charac- 

 teristic of the many Asian Paramyxine species, we 

 point to the absence of slime pores in the branchial 

 area (with the exception that in P. atami Dean and 

 P cheni Shen and Tao, there may be a single pair of 



72 



