HYATT'S CLASSIFICATION, ETC. 



are only two species on the inner watershed of the east range, 

 A. turritella and A. nigrolabris, connected in a line accord- 

 ing to their geographic distribution, turritella occupying the 

 more southern valleys and overlapping nigrolabris, that is 

 found only in the more northern valleys. These are both 

 radical of two subseries that were evolved in the western 

 range. The turritella subseries is the most interesting and 

 most highly modified, not only in turritella itself but in its 

 apparent descendants. The shells are narrower and more 

 elongated than in the other subseries, and in their evolution 

 on the western range the banded pattern becomes probably 

 of specific value in A. cylindrica. In other series, as stated 

 above, this pattern appears occasionally on this range, but 

 not at all perhaps on the eastern range, and always in a few 

 shells appears to be distinct only as varieties of accepted 

 species. There is a decided trend towards albinism in tur- 

 ritella, that, however, tends as in the nigrolabris subseries, to 

 affect the exposed sides and leave the base dark, if any color 

 is retained. Shoulder-bands are also formed as in spirizona 

 and some shells of nigrolabris. When, however, the western 

 range is reached the variations are different, and in rudis the 

 base of the last volution fades out, often leaving sides or 

 dorsuru comparatively dark. 



The forms of the nigrolabris subseries are larger and darker 

 than the above. Their affines on the western range are also 

 large shells, but, except for the decisive shoulder-band in 

 spirizona, do not exhibit similar tendencies to become banded. 

 There is, however, a decided tendency to the evolution of 

 albinoid variations but these are apt to retain a dark base 

 and have the dorsal sides lighter, just the reverse of the other 

 subseries. 



The wide gap between turritella of the southern valleys and 

 the Kauaian A. brevis is not so satisfactory evidence as exists 

 in other series of the probable derivation of this genetic 

 series from migrants that come from that island to Oahu by 

 way of the southeastern valleys of the Konahuanui [i. e., 

 Koolau] range. The collateral evidence is, however, all in 

 this direction. It appears to be certain that the most highly 



