COOKE : ON THE GENUS CUMA. 1 69 



correct. Cuming's localities, as Dr. Gray has sufficiently shown, 

 are not to be relied upon as establishing, but only as confirming, 

 a question, of distribution. Its range is probably confined to 

 the second locality which Tryon gives, viz : Peru, and to North 

 Chili. 



Cuiiia coronata is placed by Tryon amongst the Cttmas, 

 without a word of explanation, except by saying that it resembles 

 kiosquiformis in the whorls being connected across the sutures by 

 laminae. But since this peculiarity is never stated, or supposed, 

 to be an index of generic distinction (see below), coronata had 

 better, in the absence of other evidence for removing it, stay 

 where it has hitherto been, with the Furpicroe. 



Cuma nmricata is admitted on the evidence of a single 

 specimen, dredged by Hinds at Panama in nineteen fathoms, 

 mud. It is only natural to remark that mud at nineteen fathoms 

 is not the usual habitat of shells of this group, and when we 

 further learn that this single specimen is a very young shell, 

 that it is only the wide difference of locality which prevents it 

 being identified as iiigosa Born (i.e. sacellum Chem), and that it 

 has also been described as a Trophon^ we need not have much 

 compunction in not letting it disturb us further. 



Thus then Tryon's list, purged of intruders, becomes ex- 

 actly identical with the reformed list of H. and A. Adams. 



Now let us see what are the recognised conchological dis- 

 tinctions, which separate these five shells, viz : carinifera Lam. 

 gradata Jonas, kiosquiformis Duel., rugosa Born, tectu7n Wood, 

 under the genus Cuma, from the genus Purpura, in other words, 

 why are the shells, or any of them, classed as Cuma, and not 

 Purpura ? 



The descriptions of Cuma and Purpura,^,?, given in H. and A. 

 Adams, and copied by Tryon, are — placing them for the sake of 

 convenience in parallel columns : — 



