288 J. H. ASHWOETH. 



author, viz. Jolmston (1865, p. 221), who describes a speci- 

 men of E. crassa wliich bears gills and is only one inch and 

 a quai'ter (aboufc 31 mm.) long. I am indebted to Dr. Theel 

 for two specimens of E, crassa, 29 and 35 mm. long respec- 

 tively, each of which bears four pairs of gills. Another 

 specimen in my possession, obtained off the south coast of 

 Nova Scotia, is 31 mm. long, and also bears the same number 

 of gills. It would, I think, be better to consider the gills ns 

 one of the diagnostic characters of the genus Eumenia, and 

 to consider that Theel's specimens, named by him E. longi- 

 setosa, do not really belong to this genus, from which it 

 differs in other respects (see p. 292). 



S. Joseph (1894, p. 103) has evidently resolved to maintain 

 the branchiate character of the genera Scali bregma and 

 Eumenia, for he divides the family of the Scalibregmida3 

 into two sections : 



(1) Those with gills — Scali bregma, Eumenia. 



(2) Those without gills — Sclerocheilus, Lipobran- 

 chius. 



In the last-named genus he would place L. jeffreysii, 

 Cunn. and Ram. (= Eumenia jeffreysii, Mcintosh), and 

 two other abranchiate species of Eumenia, viz. E. reti- 

 culata, Mcintosh, and E. glabra, Ehlers. The genus Lipo- 

 branchius in his hands thus becomes a somewhat hetero- 

 geneous assembly, the members of which agree in general 

 shape and absence of gills, but differ in other characters 

 quite as important, e. g. the prostomium and parapodia. S. 

 Joseph's classification is therefore not entirely satisfactory, 

 and requires some modification. 



Before proceeding further it will be advisable to review 

 the chief characters of each of the genera already known, 

 beginning with the best known genera and species. 



The genera Scalibregma and Eumenia are very similar, 

 and it is not easy to find many characters by which they may 

 be distinguished. 



Taking the species S. inf latum and E. crassa as typical 

 of the respective genera, we may say that the former is more 



