SABELLIDES AND SERPULIDES 185 



Beaufort, found that they differed from Webster's description in this 

 same character (avicular uncini only in the abdominal tori) . As it is 

 hardly possible that two species would be found in the same two locali- 

 ties, which differ only in the same character, it is safe to assume that 

 the author's notes were at fault. It is therefore necessary to change this 

 character in the descriptions of both the genus and the species. This 

 change reveals the strong similarity between this genus and Hypsi- 

 comus Grube (1870) and Marenzeller (1884), non Ehlers (1887), 

 the two differing but little in form and arrangement of the setae, but 

 the collars are distinctly unlike. In Protulides it is of uniform depth, 

 like that of Chone and Euchone, and complete save the dorsal opening, 

 while in Hypsicomus it has a somewhat undulating edge and ends in 

 a ventral lobe on each side of the ventral fissure or cleft. Mclntosh in 

 his Challenger Report (1885) figures a seta and uncinus from a speci- 

 men {Laonome h&ckelii} from St. Vincent, Cape Verde Islands, of 

 which only the tail was found. The uncinus is given in a three-quarter 

 view, so that it is foreshortened. The same result was noticed in 

 mountings of the Bermuda species {Protulides elegans}, but pres- 

 sure turned the uncini, showing them in profile to have a posteriorly 

 elongated base. Ehlers (1887) and Saint-Joseph (1894) referred 

 Mclntosh's species to Hypsicomus; it is, however, identical with 

 Protulides elegans Webster. Notwithstanding the extended study 

 given by Saint-Joseph and the excellent results obtained, it has been 

 found impossible to place some of the new forms within the prescribed 

 limits of his analytical table. This is also true of several previously 

 described species. The genus Eudistylia, having equal spirally 

 coiled branchial lobes and two kinds of dorsal thoracic setae, should 

 combine with Distylia {Bispira) in his division I-A-3, but there no 

 eyes are mentioned, and the dorsal seta? in the type (D. volutacornis) 

 are superior ' limbate,' inferior * cimeter ' shaped, the latter com- 

 mencing on the fifth segment, while in the present form the inferior 

 ones are spatulate back of the collar fascicle, similar to those found in 

 Pseudopotamilla reniformis, as figured by Malmgren (1867). This 

 species has, however, simple branchial lobes, and is placed in his 

 second division under Potamilla. 



In my studies it has appeared impractical to place too much impor- 

 tance on the kinds of setae alone, as the same forms are repeated in so 

 many different genera. It has seemed desirable to give more consid- 

 eration to the form of the branchial lobes and the branchiae themselves. 

 In all the typical Sabellas studied the rachises of the branchiae are dis- 

 tinctly four-sided, connected along their posterior portions by a deli- 



