166 AMASTRA, OAHU. 



the columellar and basal margins, quite unlike the rounded 

 curve of A. decorticata, etc., but similar to the reticulata 

 group. Parietal callus usually rather thick, but thin at the 

 edge and transparent. 



Fig. 3. Length 15.5, diarn. 7.8, aperture 7 mm. Palolo. 



Fig. 1. Length 15, diam. 8.5, aperture 6.9 mm. Palolo. 

 Length 14.5, diam. 8.7, aperture 6.9 mm. Palolo. 



Fig. 2. Length 14, diam. 9, aperture 7.2 mrn. Palolo. 

 Length 13.2, diam. 8.1, aperture 7 mm. Palolo. 



From the figures and measurements it will be seen that the 

 shape varies within very wide limits, from the more oblong 

 form of typical textilis to a globose-conic shape exactly like 

 A. breviata. Many of these short shells are really indistin- 

 guishable from the latter, though they may generally be sep- 

 arated by the greater solidity and smoother, glossy surface of 

 A. textilis. 



The specimens from localities mentioned above do not differ 

 materially, all the lots being variable. Several shells from a 

 large series from Nuuanu (pi. 30, figs. 4, 5) measure: 



Length 18, diam. 9.5, aperture 8.3 mm. 



Length 17.5, diam. 9.5, aperture 8 mm. 



Length 16, diam. 9, aperture 7 mm. ; whorls 6!/2- 



Length 15.2, diam. 8, aperture 7 mm. ; whorls 6y 2 . 



Length 13.8, diam. 8.5, aperture 6.8 mm. ; whorls 6. 



Length 13, diam. 8.5, aperture 6.8 mm. ; whorls 



Length 12.5, diam. 8.8, aperture 6.5 mm. ; whorls 



In his "Reprint of the original descriptions of the genus 

 Achatinella, " p. 143, under ventulus, Mr. Thwing gives part of 

 Reeve's description of his ventulus (= textilis), at the same 

 time quoting Newcomb's notes relating to Leptachatina ven- 

 tulus (see p. 54), thus confusing two distinct species belong- 

 ing to different genera. On p. 146 the original description 

 of textilis is given. 



Gould described this species twice ; in 1845 under the name 

 A. microstoma, no exact habitat, and in 1848 as A. ellipsoidea, 

 from Maui. The type of A. microstom-a has been lost, but it 

 has generally, and doubtless correctly, been considered a syn- 

 onym of tc.xt'dis. The original description follows. 



