HYATT: GENEALOGY OF ACHATINELLID/E. 393 



a distinct shield 'Or callus. The peristome is apt to be ex- 

 panded but is not always thickened. The young have the 

 patterns of the young of the P. tappaniana series except in 

 degenerative or albino forms. The species are P. hawaiiensis 

 and horneri Bald., and physa Newc. 



The eastern variety of hawaiiensis has such lively colors 

 and coarse barred pattern that it suggests that this series 

 may have arisen from migrants like the existing forms P. 

 zebrina and zebra of East Maui. 



The relationship betiueen Partulinella dubia and the Mar- 

 morata group: The only clue that I at present possess with 

 regard to the origin of this group consists in the primitive 

 colors and characters of Partulinella dubia Newc. which is 

 found on Oahu. For a long time I regarded the species as a 

 migrant from some island to the eastward of its home that 

 had found its way there and 'become somewhat retrogressive 

 in colors and columella. This view can still be maintained, 

 but it seems more likely that Partulinella dubia is a remnant 

 of some primitive form. I was led to this conclusion by find- 

 ing in the Boston Society's collection a fossil Amastra, un- 

 fortunately without a label, which possessed decided resem- 

 blances to Partulinella dubia in its toothless columella and 

 general form. Its spire was, of course, Amastran and dis- 

 tinct, and the columella had a comparatively large perfora- 

 tion. It was similar to Am. antiqua Bald, in aspect, except 

 that the tooth was wanting. Fortunately Newcomb observed 

 in very old shells of dubia a tooth, was present and this en- 

 ables us to make the connection with the Marmorata series 

 more confidently. The transitional character of the habitat 

 'on bushes' is also notable. 



The colors of dubia are distinct from any that occur in 

 Amastra, but the reticulated pattern and zigzag lines that 

 occur over the entire spire of Partulinella dubia is a primitive 

 character and in some varieties there are large bars that ap- 

 proximate to those of marmorata and perdix. 



I have therefore provisionally supposed that dubia repre- 

 sents more nearly an ancestral stock of Partulinella than any 

 other species, and that the migrations of this series started 



