25 



convex, while their interstices become reduced to narrow, shallow grooves between broad, 

 approximate, rounded ribs. 



The number of ribs at the posterior end is very variable. If multipHcation of costce 

 occurs, it is plain that the older the shell and the more truncated, the greater will be the 

 number of ribs at the posterior extremitv; and if such multipHcation ahvays began at the same 

 stage of growth and was equally rapid, the number would always be greater with a greater 

 truncation. But such is not the case, hence the number of ribs at the hinder end varies widely. 

 Six is the fewest I have found. But there may be any number beyond this up to fourteen, 

 which is the most yet observed. These larger numbers are by no means restricted to examples 

 with much truncation, nor is there any definite proportion between the number of ribs and the 

 diameter of the shell ; some of large diameter at the truncation have but few, and vice versa. 

 A diagnosis framed upon the number of ribs would be baseless. Pilsbry suggests the typical 

 form is hexagonal; probably he is right, but usualh' there are more than six costse. 



The anal appendical tube is wanting in most cases, even during life. When present it 

 may be two or three millimetres long. It may exist when the shell is young and narrow, and 

 be wanting when okl and wide-, possibly it ma}- have been broken off. It is central and most 

 frequentlv in the axis of the shell. But it may be distinctly out of the axis, joined at an angle 

 so as to point markedly towards the concave side, or slightly towards the convex, and in one 

 it is funnel-shaped instead of cj'lindrical. These circumstances confirm the suggestion of its being 

 an outgrowth subsequent to truncation, and not merely a residual inner layer of the shell after 

 the outer portion has been absorbed. 



The radula (pi. XXVI, figs ïj^.a, d, c, d) is comparatively large, and contains fifteen 

 rows of teeth, with the formula i . i . i . i . i. The rachidian tooth is about twice as wide as 

 hiorh, is thickened alone its free edee, and thinned alono- its attached border. The sincjle lateral 

 is stout and rather short, and has one obliquely placed cusp without serrations. The marginals 

 are trapezoidal flat plates, thickened along their inner end, and the whole or larger part of 

 their upper margin. Mr Kesteven, in executing the drawings, detected a small accessory plate 

 of chitin (fig. 14^'). It is somewhat pyriform, stouter at its narrow, attached end, and thinner 

 and slightly striate at its free, expanded extremity. Its height is about one-half that of the 

 rachidian tooth, outside of w'hich it stands, with its base about half-way between this and the 

 lateral. As the laterals overlap the outer fourth or third of the rachidian, this plate lies behind 

 or between the laterals, and being comparatively thin it cannot be seen through the much 

 denser laterals; but in a dismembered radula it can be certainly recognised. 



*2 3. Dentalhun macandrewi n. n. PI. V, figs 25 — 34. 



1885. Dentaliiun lineolatuin Cooke, Ann. Mag. Nat. Hist. (5), XVI, p. 274. 

 1885. Dcntalimn aratoruiii Cooke, Ann. Mag. Nat. Hist. (5), XVI, p. 274 — 275. 

 1885. Dentalium claviis Cooke, Ann. Mag. Nat. Hist. (5), XVI, p. 275. 



1897. D. lineolatimi Cooke, aratonim Cooke, claviis Cooke. PiLSBRV & Sharp, Man. of Conch., 

 vol. XVII, p. 10, p. II and p. 55. 

 ? Dentalium reevii Desh. I\ISS. Mac-Axdrew in collection. 

 ? 1871. Dentalium j-eevei Desh. MSS. FiSCHER, Journ. de Conchyl. XIX [(3), XI], p. 212. 

 ? 1894. Dentalium laugieri Jousseaume, Buil. Soc. Philom. de Paris VI, p. 103. 



SIBOGA-EXPEDITIE LIV. 4 



