THE COOK'S OE HARVEY ISLANDS. 397 



Now, it would be most interesting to ascertain how this migration has been 

 effected. It cannot, I think, be attributed to human agency, for the simple reason 

 that in the Society Group, where the islands are not very far apart, the natives 

 are almost daily conveying fruits and vegetables from one to another, and have 

 probably done so for hundreds of years ; yet not one species of Partula has been 

 introduced from one island to another through their agency. Moreover, the gene- 

 ral diffusion of hyalina over Tahiti, which exceeds that of any of its endemic 

 species, is opposed to the human agency theory. 



As stated in my paper on the Rurutu shells, migration might have taken place 

 in some remote period when land communication existed between the three 

 islands. But the most serious objection to this hypothesis is, if the Tahitian 

 endemic species Avere contemporary at the time of migration, some would probably 

 have extended their range so far to the southward and westward, that, when com- 

 munication was interrupted by the subsidence of the land, some of their descend- 

 ants would have been found at this day either on Magaia or the Austral Islands. 



The existence of several genera and numerous species of marine shells in the 

 above island, which do not occur on the Tahitian coast, are opposed to the above 

 theory. The avifauna of the islands is also unfavorable to the hypothesis. 



The hypothesis of independent creation of the same species in different places, 

 I do not feel competent to discuss; but will leave it to others who are better quali- 

 fied to elucidate the difficult problem. 



Examples of hyalina from the three distant localities are precisely alike; prov- 

 ing that food, temperature, and difference in station have not developed a single 

 varietal character to distinguish the species of the different groups. The local 

 variation, which is exactly the same throughout its range, is so slight as not to be 

 worthy of notice. 



GENUS TORNATELLINA, Beck. 

 T. Philippii, Pfeiffer. 



Tornatellina Philippii, Pfeiffer, Zeitsch. Malak., 1849, p. 93; Mon. Hel., iii. p. 524; iv. p. 651 ; 



vi. p. 263 ; viii. p. 311. Pease, Proc. Zool. Soc, 1871, p. 413. Garr., Proc. Phil. Acad. Nat. 



Sci. p. 22. 

 Pupa Philippii, Kuster, PL 18, fig. 20, 21, 

 Leptinaria Philippii, H. & A., Adv. Gen. Moll., p. 141. 

 Achalina Philippii {Leptinaria), Pfeiffer, Vers., p. 110. 

 Cionella Philippii, Martens. 



A few examples of this species were found under decaying leaves in forests 

 near the sea-shore at Aitutaka, It also occurs, but not plentiful, in the Austral, 

 Society, and Marquesas Islands, 



