32 MELVILL : ON THREE NEW SPECIES OF SHELLS. 



viz: E. attenuata (Sowb.), a native of Fiji, in its fewer whorls, 

 the latter possessing fifteen at least ; from E. teinostoina (Ad.), 

 also from Fiji, in its straighter build and slender body-whorl ; 

 and from the common form E. vitrea (Ad.), native of Fiji and 

 the south seas, in its much greater slenderness, and whorls more 

 elongate — E. vitrea being a smaller and stouter shell, with and 

 very numerous narrow whorls. 



Not very many species are yet described from Mauritius, 

 from which island, however, the large and conspicuous E. 

 Maitinii (Adams) comes, the second largest species in the 

 genus. The trivial name is derived from rjTrrjrrjpiov a needle, 

 in allusion to the shape. It is one of the most graceful in a 

 singularly beautiful genus, though to discriminate their beauty 

 of form to the full, a magnifying glass is almost imperatively 

 necessary, owing to the small size of the majority of the known 

 species. 



The type, at present the only specimen known, is in my 

 collection. 



Zizyphinus haliarchus (sp. nov.) PL ii, fig. 3. 



Testa erecto conica, pyramidali, anfraciu ultimo ad basin acut- 

 ijv angiilato, anfractihus pianis, dedivibus, sublxvibus, granis 



parvulis liratis, liris temiibus circuinanibientibus, ditabiis 

 ad suturas proxiiiiis crassioribus, granis majoribiis decoratis, 

 flavida, flammis purpureo-fuscis apud peripherain or/iatd, 

 obscure punctulata, basi liris sublcevibus punctato-ge/ninatd, 

 plaud, sub-convexd, columella callosd, fauce loevi striata. 

 Alt. 40, lat. j6 mill. 

 Hab. : Ad oras Australice ? 



This very large, conspicuous, and beautiful addition to the 

 genus Zizyphinus is at present unique, the specimen being 

 included in the collection now in the Museum, Owen's College, 

 Manchester, and formerly having formed part of that of 

 Reginald Cholmondeley, Esq., of Condover Hall, Salop. 



No very close connection is observable with the other 

 Australian species ; there is, however, owing to accidental trans- 

 position of labels, some little doubt as to the correctness of the 



J.C, vi., Jan., 1889. 



