AMPHIDROMUS, GROUP IV. 141 



white, the margins joined by a very thin parietal callus, the outer 

 and basal margins expanded and reflexed, columellar margin dilated, 

 nearly straight, nearly forming an angle with the basal margin. 

 Length 43, width 19, alt. of aperture 22 mill. (T. (?.) 



Celebes: Kandari, on the southeast peninsula. (Beccari.) 



Bulimus (Amphidromus] beccarii T. C., Annali Mus. Civ. Stor. 

 Nat. di Geneva, xx, 1884, p. 170, pi. 1, f. 10, 11. A. beccarii FUL- 

 TON, t. c., p. 74. 



A peculiar species, resembling the A. winteri group of Java in its 

 plicate surface, but distinguished by the narrow, elongate' mouth, 

 oblique sutures and slender contour. Possibly belongs to the second 

 division of the genus, which includes thin, exclusively sinistral species. 



4. Group of A. perversus. 



Considerations of utility no less than probable phylogenetic rela- 

 tionship induce me to segregate the species of the Philippines, Bor- 

 neo, Celebes and Java from those living north of Sunda Strait, 

 though the latter are not separated from these by any characters of 

 more than specific weight. The senior species of the group, A. per- 

 versus, lias been understood to cover many and diverse species or 

 varieties by Pfeiffer and others, and not without reason ; but it is 

 now restricted to a solid, yellow or black- varixed form of southern 

 Celebes and Java, around which a number of closely related forms 

 group themselves. There seem to exist more or less complete tran- 

 sitions between perversus, interruptus, sultanus, etc.; and even the 

 more northern forms, melanomma, atricallosus, leucoxanthus, which I 

 group separately, are united to perversus by various authors, as Mr. 

 E. A. Smith and Mr. H. Fulton. Others, as Prof. E. von Martens 

 and Dr. 0. von Mollendorff, take a more analytical view. The fact 

 seems to be that while most specimens are readily assignable to one 

 or other of these names, there exist transitions between perversus and 

 the interruptus group inhabiting the same area ; and there may be 

 transitions where the ranges of the several closely allied forms meet 

 or overlap. In general, I consider it best in a genus like this, where 

 specific values are notoriously discounted, to recognize as species 

 those forms which have a consistent or harmonic distribution, and 

 readily recognizable characters in the vast majority of specimens. 

 Exactly similar cases are encountered in the American genera Oxy- 

 styla and Cerion. 



