HYATT: GENEALOGY OF ACHATINELLID^E. 379 



in very few examples, all belonging to the genus Amastra. 

 One single example has 'been found in which the columella 

 had three definite tooth- folds, evidently an isolated, sporadic 

 variation. This last case appears never to have become 

 genetic in any series of forms so that one could call these a 

 distinct group or species. The case of the two or double 

 tooth-fold does not, however, appear in this way, but only 

 on the island of Lanai and in the one group of Amastra 

 biplicata Newc. It occurs, however, in so many shells and 

 in such definite and invariable succession that they are uni- 

 versally recognized as one species, Amastra biplicata. This 

 has long been known, but the remarkable interest and signi- 

 ficance of this fact has not been noted. It is one species in 

 the genus Amastra, or if another view be preferred, a new 

 genus in the family of Achatinellidce departing in a new 

 direction so widely that it must be cited as an exception to 

 diagnostic, analytical description of that family.* Another 

 exception occurs in two species on one island, but in this 

 case the deviation consists in the absence of the single 

 spiral tooth-folds so universal otherwise in this family group. 

 The variation in this case is also heralded by an anticipatory 

 mutation which occurs in individuals of other species oc- 

 curring on other islands. 



Another matter of interest in this connection is that the 

 shells showing mutations of the double tooth-folds, and the 

 absence of folds have no obvious or traceable connection 

 with the species characterizable by the possession of similar 

 and permanent variations. They appear to be of indepen- 

 dent origin in different genetic series occurring on different 

 islands. Thus while homogenetic in their own series in their 

 own locality, they are simply homoplastic repetitions of 

 common tendencies when compared with the mutations oc- 

 curring in other genetic series in other islands. 



This indicates the possible beginnings in each case of en- 

 tirely distinct genetic groups that might under favorable 



* Several species of Amastra and Laminella have a second columellar 

 fold. H. A. P. 



