92 PARTULINA, SECTION BALDWINIA. 



throw a flood of light upon the vexed question of specific re- 

 lationship. The writer believes that a study of a carefully- 

 selected series of the one hundred and more species of the 

 Oahu Achatinellas by a student familiar with the local geo- 

 graphic conditions would reduce the number by at least one- 

 half. 



Partulina horneri Baldwin. 



" This species, described in 1895 from specimens collected 

 by Mr. J. Lewis Horner, appears to possess a present rather re- 

 stricted range in the sparsely timbered region above Kukui- 

 haele, Hamakua, and on the ridges above the Waipio and Wai- 

 manu valleys, an area of perhaps three or four square miles. 

 Except for a few specimens obtained above Honakaa (on the 

 doubtful authority of a Kanaka's statement) and Kukaiau 

 (introduced or * planted : there by Eugene Horner) ten to 

 twenty miles distant, this shell has never been found else- 

 where. This restricted habitat, in a region where the compe- 

 tition is almost nil, represents all the territory the species has 

 been able to acquire since the unknown date of its occupancy 

 of the island. That the period of its occupancy has been con- 

 siderable appears from the striking unlikeness of the species 

 to all its congeners. 



" Although occasionally living upon ohia (Metrosideros 

 polymorpha) , kolea (Myrsine lassertiana) and, rarely, upon 

 kawaao (Bryonia sandwichense) , the species is chiefly con- 

 fined to a small berry-bearing tree called the ahakea (Bo~bea 

 elatior) which is rather common at an altitude of 1,800-2,500 

 feet. This species, and in fact practically all the Achati- 

 nellas the writer is familiar with, is often found on half-dead 

 trees, and not; infrequently a thriving colony may inhabit a 

 tree which boasts of but a handful of foliage. This appears 

 to indicate that once a colony is established on a tree it per- 

 sists, as long as it can obtain food, and I have found small 

 colonies of shells on stubs of trees that apparently had been 

 dead twenty-five or thirty years. 



" This and the following species belong to the sub-genus 

 Baldwinia of Partulina, and they are much more closely re- 



