INTRODUCTION. 



For a correct grouping and synonymy of the Genus Achatinella 

 two things are exceedingly necessary : first, an exact knowledge 

 of the geographical distribution of the shells, a knowledge only to 

 be gained by personal colledling ; and second, the ability to com- 

 pare shells colledled with the original types which have been 

 described. Figures and plates without exact habitat are compara- 

 tively of little value, and are often very misleading. Few writers 

 on this subject have been able to qualify in both of these requisites, 

 and as a result great confusion has arisen, and many so-called new 

 species have been added which have no right to be called such. 



The present writer does not lay claim to the ability to give a 

 correct grouping. It is simply the aim to present, as far as possi- 

 ble, the original descriptions, with such additional notes regarding 

 locality, varieties, etc., as have come from the collecfting and study 

 of many thousands of these beautiful shells. The writer's own 

 colledlion is one of the most extensive in the Islands, containing 

 over 30,000 examples. 



One of the greatest difficulties in giving the correct names to 

 shells found is the fact that very few of the earlier descriptions 

 give valley or place where the shell is found. It was only in later 

 years when the shells were studied in their native homes that this 

 point was found to be of first importance. vShells from one end of 

 the mountain range, which forms the backbone of the island of 

 Oahu, are quite different from .shells found in the middle of the 

 range or at the other end. A figure or description might not 

 always clearly show the difference, as often exceptional forms of 

 two widely separated groups may look similar. To one, however, 

 who is given a hundred shells from each di.stinct locality the differ- 

 ence would be very manifest. For example, a colle<5lor with fifty 

 A. cestus Newc. from Palolo (or, more exadlly, the near-by valley 

 of Wailupe) would never confuse them with fifty A . iunrida Newc. 



(V) 



