Annelida of the Bermudas. 65 7 



ments, but much less developed than in C. torqiiata, when the latter 

 has been equally contracted by alcohol. The collar is doubtless 

 much narrower in life than in the latter, but it is of the same nature. 



As Axiothea was in prior use in Coleoptera (Pasc, 1864) it must 

 be abandoned for this genus, and Glymenella now seems to be its 

 equivalent, both types being essentially alike in all generic characters. 



This genus has the following characters: Number of segments 

 variable. A limbate cephalic plate; a funnel-shaped anal plate bor- 

 dered with papillae; a thoracic collar on the 4th setigerous segment, 

 and sometimes on the 3d and 5th ; rows of ventral, rostrate, uncinate 

 sette, having a series of apical hooks and a beard, on all the anterior 

 setigerous segments; both pennate and smooth bilimbate capillary 

 setae in the upper fascicles (pennate ones overlooked or perhaps 

 accidentally absent in some described species); usually 2 or 3 preanal 

 segments without setae. 



Besides the type, at least two other East American species are 

 known : 



C. elongata (Webst.) 1879, as PraxiUa, from New Jersey and 

 Connecticut. It has thirty-seven to thirty-nine segments (thirty- 

 six setigerous in the larger ones). Mr. Moore, 1893, has also 

 described, as a new species, Glymenella elonqata from New Jersey, 

 which is probably identical, though the coincidence in name was 

 accidental. 



C. mucosa (Andrews) as Axiothea, Proc. U. S. Nat. Mus., 1891, 

 has twelve uncini on the 1st setigerous segment, and thirty farther 

 back. The anal papillae are of various lengths. 



These three species all have small pennate setae mixed with the 

 bilimbate ones, but in C. torquata the pennate setae ai-e very small, 

 slender and fragile, so that they are easily broken off and over- 

 looked. 



Axiothella, sub-gen., nom. nov. Type, A. catenata (Mgn.). 

 Axiothea Malmgren, 1865; St. Joseph and others (non Pasc, 1864). 

 The name Axiothea, as above shown, is untenable, but I propose 

 to establish a subgenus, Axiothella for the typical species of Axio- 

 thea, making the smaller or rudimentary condition of the collar* the 



*St. Joseph, op. cit., p. 181, objects to the use of the existence of a thoracic 

 collar as a generic character, because it has been found to exist in species 

 of other genera {Rhodine, etc.). But the same objection would apply to 

 the limbate head, and to the infundibuliform an^l plate, which exist in several 

 genera. In fact it is probable that in those cases where it exists it will be found 

 to be associated with other truly generic characters. (See p. 654.) 



