MYXINE. 343 



Son diametre est alors le dix-luiitieme ou a peu pres de sa longueur totale." 

 From Comnierson he quotes directly " Conger olivaceo-virens, iniiuaculatus, 

 lac et gluten pluriniuin fundens." The name Murtenobleima was not 

 approved by Rafinesque, 1815, who promptly changed it to Anopsiis, with- 

 out adding anything to a knowledge of tlie object to which the term was 

 to be applied, and which he in the Analyse, page 9o, placed in his sub- 

 family Apicridia of his family Opludia. On the next page, 94, he credits 

 himself with 3If/.vinc, and with it and Gastrohninchus of Bloch forms his 

 subfiimily Mf/xbiia of his family Cydostoniia. 



The specimens on which five of the species characterized below are 

 based are those used by Putnam, 1874, for his Notes on the M3xinoids, 

 where they Avere arranged as three varieties of a single species. The differ- 

 ent types resemble one another so closel}^, and individual variation is so fre- 

 quent that determination is a matter of some difficulty. In fact it is only in 

 comparisons of many individuals that the distinctions become really appar- 

 ent. Even when the averages are quite distinct the species may appear to 

 overlap, through variations occurring in particular cases. In Myaine gluti- 

 iiosa the dentition formula most conmion may be indicated by g^, eight 

 teeth on each side both above and below, while variations to one more or 

 one less occur in the lower series and less frequently in the upper. M. H- 

 mosa is better shown by the formula li, lo, nine in each upper series and ten 

 in each lower, while the most frequent variations add one more to each 

 upper or subtract one from each lower series. M. acidifrons has tji, varied 

 by one more in each upper series or by one less in each of the lower. Either 

 ioTo 01" lui will represent M. austrcdis, and the variations include all pos- 

 sible additions and subtractions of a single one from each series between the 

 two formulas. All of these have the anterior two teeth of each series con- 

 fluent at their bases. M. iridcntlger has ten or eleven teeth in each series 

 but differs from the preceding in having the anterior three teeth confluent in 

 their bases. 31. circlfrons has the formula iitq of which the anterior three 

 of each upper and the anterior two of each lower series are confluent. 

 Again, if the mucous sacs, or the pores, are compared it is found that 31. 

 gliitinosa has not as many in either the pectoral (from head to gill opening) or 

 abdominal series (from gill openings to vent) as 31. limosn ; while 31. ausfralis 

 is rather close to 31. Uniosa but has fewer pectoral and more abdominal pores 

 than 31. acutifrons. The species may approximately be distinguished by the 

 followino; : 



