396 HYATT: GENEALOGY OP ACHATINELLIDJE. 



The connection between Newcombia and Perdicella is shown 

 through the resemblances in color and form of the young of 

 Newcombia to those of Perdicella. They are as a rule more 

 slender than the most primitive species of that genus, viz. 

 P. helena, but are quite similar to the more modified and more 

 slender forms of Newcombia itself, like N. perkinsii Sykes 

 of Molokai. 



The resemblances are better seen in the young of Newc. 

 cinnamonea and some of the smoother species than in the 

 more highly modified shells like sulcata, etc. This evidence 

 is strongly in favor of the direct derivation of Newcombia 

 from some form like the Perdicella helena of Molokai. 



The Relationship between Amastra, Eauaia, [Cyclamastra'], 

 Carelia and Armiella: Amastra, Kauaia, [Cyclamastra] 

 Carelia 'and Armiella in Oahu and Kauai have smooth shells 

 in the earliest stages and there is a common form in the 

 nepionic stage. This has somewhat stout and rapidly in- 

 creasing volutions, the surface is covered with fine transverse 

 ridges or growth-bands. The bases are separated from the 

 plano-convex dorsal sides by a more or less prominent sub- 

 angulation, but this is never developed into a keel nor are 

 there any coarse, transverse costae on the dorsum. There are 

 also no longitudinal markings of any sort. Keels, when pres- 

 ent, and longitudinal ridges are introduced in late stages in 

 [Cyclamastra'], Eaiiaia, Armiella, and Carelia. 



In Armiella the columella is only known in large shells 

 and in them it is solid. 



In Carelia the columella is known in the young of C. 

 dolei wherein it remains solid or incomplete until a compara- 

 tively late stage and is solid also in the ephebic stage, but in 

 one shell there was a small perforation -when the shell was 

 about one centimeter in length and was then just closing, 

 for it was entirely closed in the same shell when four mil- 

 lemeters longer. In some others it was present but very 

 minute at an earlier stage, and in some it seemed to be absent 

 altogether at every stage. It was present in two shells of 

 C. adusta when the shells were about six mm. in length (verti- 

 cal diameter) and closed at one centimeter. Its beginning 

 was not traced in these two shells. 



