448 REV. A. H. COOKE ON [May 1 7, 



in the 'Reisen'^ is admitted, even by its author, to be hardly 

 satisfactory. But there is practically a general agreement between 

 Semper and Pfeififer ^, since both agree in the limitations represented 

 by about 10 of the subgenera, the chief points of difference arising 

 with regard to the respective limits of Galocochlea and Helicostyla, 

 of Helicobulimus and Orthostylus, of Orthostylus and Canistrum, 

 and the restriction of Canistrum proper. The distinction, if any, 

 between some of these groups is, unless and until some definite 

 anatomical difference is established, at best arbitrary. After careful 

 consideration I have decided to abolish Helicobulimus altogether, 

 merging it in Ortliostylus. The species are in any case few (Pfeiffer 

 enumerates only 5, and one of these, grandis, Pfr., is better classified 

 as Galocochlea), and different authorities are much at issue with 

 regard to referring specific forms to one group or the other. This 

 seems sufficient reason for refusing to draw a line between them. 

 With regard to Orthostylus and Canistrum (Pfr.), there is a very 

 long series of forms ultimately connecting such typical Orthostyli as, 

 e. g., daphnis, Brod., and rufogaster, Fer,, with elongated shells like 

 camelopardalis, Brod., and nympha, Pfr. But the extremes are so 

 wide apart that it may be worth while to try to separate them, and 

 I do so by regarding pictor, Brod., as a sort of border-line form, 

 removing it from Orthostylus, and considering it and all the more 

 elongated forms as belonging to a separate group {Hypselostyla, 

 Mts.). This group is practically identical with Semper's 'Elongates,' 

 for there is strong ground for restricting, with Semper, the group 

 Canistrum to a pecuhar section of shells ^ the type of which is 

 ovoidea. Lam. {=lusonica, Morch, = ewrys'owa, Pfr.). 



The localities given in each case have been most carefully con- 

 sidered, and no species has been taken into account whose locality 

 is not regarded as authoritative. Thus the locality ' Philippines,' so 

 often given by the older writers, is useless for the present purpose, 

 and species not further localized (a considerable number) have been 

 neglected altogether. Recent investigation has been more exact in 

 its record of localities, and in the present paper 180 species in all are 

 brought to account. Further, it has been found necessary to neglect 

 Cuming's authority as establishing any locality whatsoever. Those 

 familiar with his method of preserving localities will understand this, 

 and it need only be added that Semper and von Mollendorff are 

 continually at issue with him. He may be taken as confirming, but 



^ (1) Corasia, (2) Callicochlias, (3) GlohoscB, (4) Hypomdanm, (5) Cinerem, 

 (6) Axina, (7) Helicostyla, (8) Orustia, (9) SphcsriccB, (10) Cochloclryas, (11) Or- 

 thostylus, (12) Elongate, (13) PTiengus, (14) Eudoxus, (lb) Canistrum, (16) Fro- 

 chilus, (17) Chrysallis, (18) PhcRnicobius. 



- Nomencl. Hel. Viv. pp. 202-212. Von Molleudorff's paper on the sub- 

 generic classification of Cochlostylcs (Jahrb. deutsch. mal. Gesell. xii. p. 72) 

 places the divergence at its maximum. 



Semper, ' Eeisen,' II. iii. p. 219, places in this section the following species : — 

 stahilis, Sby., euryzona, Sby., ovoidea. Lam., belcheri, Pfr., balanoides, Jon., 

 breviculus, Pfr., cinerosa, Pfr., dilatata, Pfr. I should agree with him with 

 regard to the first fiye ; cinerosa is probably a var. of satyriis, Brod., which is 

 a Hypselostyla ; the remaining two appear doubtful. 



