LAND AND FKESH-WATER SHELLS. 337 



Had 1 been aware of the extent of the influence of locality in tliis 

 region, I might have made my collection of greater value. It would 

 therefore seem necessary for future collectors in this group to make 

 in eveiy small island and in different districts of the larger islands 

 special indejtendent collections, disregarding the fact that they may 

 have apparently met with the same shell very frequently before, 

 because many of the varieties and some of the species can only be 

 distinguished by the practised eye of the specialist, and a new 

 locality for a previously well-known species may be often unwit- 

 tingly found. 



As an instance of the unexpected results, which may fall to the 

 lot of others in this group, I may here add, that out of eleven land 

 and fresh-water shells that 1 collected in the small island of Santa 

 iVnna, whicli is onl}^ 2^ miles in length, four were new species, and 

 besides there were some new varieties. The stations of these four 

 species may be suggestive. Two of them — Helix ( Videnci) sanctcs 

 aniKB and Helix {iianind) solidiiisciila — were generall}^ found on 

 the trunks of the cocoa-nut palms at the coast ; whilst the other 

 two occurred in situations far more likely to yield new species, 

 Melania sanct(^ annce being obtained from a small stream in the 

 interior of the island, and Melania guppyi being- found dead in the 

 stomach and intestines of a fish that frequents the fresh- water lake 

 of Wailava. This last shell would appear to live in the deeper 

 parts of the lake, as I only found one living specimen, all the others 

 being obtained from the stomach and intestines of these fish. Mr. 

 Smith describes it as " a very remarkable and distinct species." Its 

 length is about \\ inches ; and its sharp-pointed spire was to be 

 sometimes seen protruding through the vent of the fish, which 

 evidently digests the animal and ejects the shell. These fish were 

 usually 9 or 10 inches long ; but the full-grown shells were found 

 also in fish half this size, when the relation between the lenafth of 

 the shell and the size of the fish was trul}" alarming. Since the 

 little fish actually swallow sharp pointed shells measuring a fifth of 

 their own length and pass them out through the vent after they 

 have digested the animal, we must credit them with a remai^cable 

 capacity for adapting their diet to circumstances. 



To exemplify the variation which some species of sliells display in 

 this group, I will take the instance of Helix (Geotrochus) cleryi, 

 Recluz. Tills species is probably distributed through the wiiole 

 group ; but considerable variation prevails in different islands. 



