ACHATINELLA PH^OZONA. 185 



about half a mile from the shore, southeast of Kailua Bay, in 

 humus of plowed fields (Spalding). 



Achatinella phaeozona GULICK, Ann. Lye. N. H. of N. Y., 

 vi, p. 214, pi. 7, f. 40, December, 1856. 



"May be grouped with A. buddii and A. fulgens Newc., but 

 differs in its more solid structure, its thicker lip and colu- 

 mellar fold, and in the more convex outline of its spire. It 

 also lacks the black tip which characterizes A. buddii. The 

 typical varieties found in Keawaawa are rare, that sterile re- 

 gion affording but few trees, which occupy the ravines near 

 the summit of the ridge." (Gulick.) 



Professor Hyatt looked upon A. phaozona as the common 

 ancestor of Achatinellastrum, Bulimella and Apex (Science, 

 viii, p. 395). Later he claimed for A. ph&ozona a relationship 

 to Kauaia. Our investigations lead to totally different re- 

 sults, and we believe the affinities claimed by Hyatt to be alto- 

 gether erroneous. 



Some specimens of phaozona are hardly separable from cer- 

 tain shells of the plumata pattern of A. fulgens ; but other 

 patterns of both differ widely. A. ph&ozona never has green 

 or yellow varieties. 



Although this species is at present restricted to a very 

 small area, it formerly inhabited an extensive district on the 

 northern or Koolaupoko side of the island, which was then 

 wooded down to the shore. About sixty years ago, when Mr. 

 Gulick discovered the species, it was still to be found in a 

 few scattered colonies in Waimanalo and Kailua. On his 

 labels Mr. Gulick indicated that it was almost extinct in these 

 valleys. 



The Keawaawa lot in Gulick 's collection consists wholly of 

 dead shells. A series is shown in plate 24, figs. 10-13, pi. 36, 

 f. 10, 10a. The shell is commonly white with chestnut bands 

 0230 or 0030, but the bands vary in width, sometimes nearly 

 covering the last whorl. There is almost always a white band 

 at the periphery, and the sutural margin and summit are 

 always white. Often the bands are split, giving rise to nu- 

 merous band-forms, mentioned by Gulick. 



Another pattern is closely streaked with fleshy-brown, cut 



