238 AMASTRA, LANAI. 



tended, as long as the shell ; base of foot and mantle of a 

 bluish-gray " (Newcomb). 



A. magna differs from the related A. violacea of Molokai 

 by its embryonic shell (pi. 26, fig. 5, length 6 mm., 3% 

 whorls). The initial half whorl is smooth; next two whorls 

 strongly oostate, and earinate above the suture ; on the third 

 whorl the ribs are finer than on the second. After the third 

 whorl they become weak, persisting longest near the suture. 

 The fully-formed embryo, as figured, is imperforate, acutely 

 carinate, the carina white. The columellar fold is moderate 

 and very oblique. In A. violacea the costation of the embryo 

 is less strong and no supersutural carina is visible. 



The last whorl varies from closely, subregulaiiy striate to 

 irregularly striate and malleate, the subperipheral portion 

 often having revolving, flattened facets. Sometimes the peri- 

 phery is angular, as in fig. 3. The blackish cuticle remains in 

 shreds or lines between the striae on the last whorl or two. 

 There is great variation in the color of the aperture. It may 

 be white (fig. 1) or cream-white (fig. 3), salmon-colored (fig. 

 2), or purple-brown or leaden^purple (fig. 4). The parietal 

 callus is very thick in adult shells (thereby differing from A. 

 violacea) ; it is pure white or dark-brown. The columellar 

 fold is thin. 



Fig. 4. Length 36.5, diam. 17, aperture 16.2 mm. ; 7 

 whorls. 



Fig. 1. Length 30, diam. 16.2, aperture 15.5 mm.; 6y 2 

 whorls. 



Fig. 2. Length 30.5, diam. 15, aperture 13 mm.; 6% 

 whorls. 



Fig. 3. Length 29.5, diam. 13.2, aperture 12.2 mm.; 7 

 whorls. 



Fig. 6. Length 35, diam. 19, aperture 16.8 mm. ; 7 whorls. 



Length 35.5, diam. 15.5, aperture 1.5 mm. ; 7% whorls. 



Length 25, diam. 12 mm. 



The size varies widely. In shape it may be subcylindric, 

 ovate or oblong-conic. It is likely that several subspecies will 

 eventually be recognized, but a knowledge of the distribution 

 of the forms is requisite for any step in this direction. Figs. 



