418 BULLETIN OF THE 



Class SCAPHOPODA. 

 Order SOLENOCONCHIA. 



Family DENTALIID^E. 



Genus DENTAL.IUM Linne. 



The species to be considered may be conveniently divided into smooth 

 species, delicately striated species, strongly sculptured species, laterally com- 

 pressed species, and dorsally compressed species. 



I am inclined to believe that the differences in the soft parts reported in 

 regard to various Scaphopods will be found to be more or less connected by 

 intermediate forms when more species have been examined. To some extent 

 the differences will prove to be due to the point of view of the observer. Dif- 

 ferences in the notch or slits at the posterior orifice are often due to erosion or 

 pathologic causes. 



An annelid, Pomatoceras, simulates some forms of Dentalium or Cadulus. 

 Other worms adopt the dead shells as a residence, and grow to fit them exactly. 

 A very pretty hermit crab, not asymmetrical like his brethren who dwell in 

 Gastropod shells, makes his home in the dead abyssal Dentalia, or in the shells 

 of Cuvieria. The large claw is modified to fit the aperture of the shell exactly, 

 like an operculum. Though the shells are white and opaque, and the species 

 is dredged at great depths, the crab is prettily colored with pink and yellow, 

 and has large well pigmented eyes. 



A. Shells circular in section. 



a. Shells smooth and polished. 



Dentalium agile M. Sars. 



Dentalium agile M. Sars, Bull. M. C. Z., IX. p. 37, 1881. 



Antalis agilis G. O. Sars, Moll. Reg. Arct. Norv., p. 102, t. xx. fig. 9, 1878. 



Dentalium entalis Linne' var. agile Watson, Chall. Gastr., p. 6, 1885. 



Habitat. Station 100, off Morro Light, Havana, in 400 fms. 



Most of the specimens originally referred to this species appear on more 

 thorough study, and after comparison with typical specimens of agile, to be im- 

 perfect or young specimens of D. pcrlongum,. A single undoubted specimen of 

 agile was, however, taken as above. The original agile is very likely connected 

 with some other nominal species, as suggested by Mr. Watson, but I have not 

 had time to make a careful study of the question. 



