AMI'HIDROMUS. 129 



To distinguish species in this group is a difficult matter, largely 

 dependent upon subjective considerations. What one would hold to 

 be merely an individual variation, sometimes proves, with more ample 

 knowledge, to be a species, well defined geographically ; and what 

 one might think from a single example to be a sharply-characterized 

 species, may appear, when it comes to be collected on the spot, to be 

 merely the extreme of an uninterrupted series of individual varia- 

 tions, living together, as in the forms of A. interruplus. Form, mark- 

 ings and tint of the ground-color may be constant in all specimens 

 from one locality, or they may vary. Thus, form and markings vary 

 in the specimens of A. interruptus from Maros, but the ground-color 

 remains the same ; but in A. purus from Bandjar, Java, the ground- 

 color varies. Only series of specimens from known localities can 

 permit conclusive opinions. 



An analogous case is found in the American genus Liyuus, where 

 coloration is equally variable. To some extent certain patterns char- 

 acterize local colonies ; and frequently the same pattern occurs in 

 several, probably many, separated localities. Thus, near Miami, 

 Florida, I found only green-striped specimens in one wood, while in 

 another, a hall' mile distant, many of the shells had broad brown 

 markings. To some extent the patterns are racial in some cases, 

 but these races are of but slight value, being analogous to 

 the slight characteristics common to human families of common 

 parentage. 



Of course in some species of Ampkidronnts, special color patterns 

 seem to be characteristic; but even in these, a wide variation may be 

 expected, when districts other than the lype locality come to be 

 searched. Under these circumstance* it is almost impossible to con- 

 struct satisfactory keys for the ready identification of species. 



Amphidromus was divided by von Martens into two groups: (1) 

 rather large and broad species, as a general rule indifferently dextral 

 or sinistral, and usually having varices, or periods of growth-arrest 

 marked by the deposition of an oblique pigmented stripe, and (2) 

 smaller and more slender forms, always sinistral, and without varices. 

 This is doubtless a natural division, though not indicated by any con- 

 siderable difference in the soft anatomy so far as known. Further 

 subdivision is largely a matter of opinion, and must be expected to 

 vary as the relationships of particular species become more definitely 

 appreciated. 1 have grouped the species into 20 groups, in part 

 9 



