FALSE INFERENCES FROM FOSSIL SHELLS. 399 



allies in Celebes, tlie Moluccas, Java, and elsewhere, exhibit all 

 the typical Indian characters both of the shell and of the animal, 

 both the species that have migrated into the Philippines are so 

 completely metamorphosed as to the form and colour of both 

 animal and shell that it would have been impossible to guess, 

 without the closest anatomical investigation, that they did not 

 belong to the genei'a proper to the Philippines. With regard 

 to Xesta Gumingii the explanation is available that here we 

 have a case of true protective mimicry ; but how can we account 

 for the even more striking resemblance between Xesta minda- 

 naensis and a Rhysota ? This resemblance cannot certainly be 

 . regarded as protective mimicry, since the species of Rhysota 

 have absolutely no peculiar property by which they are better 

 protected against their enemies than other mollusca, and conse- 

 quently an imitation of their appearance can be no sort of 

 advantage to the other imitative species. 



The diiEciilty which is evident in the case just described is 

 still further enhanced by the fact that other similar cases are 

 known, and actually in the Philippines. By far the greater 

 number of the brilliantly and variously coloured land-snails of 

 these islands live on trees and belong to the highly charac- 

 teristic genus Cochlostyla. The forms of their shells, too, are so 

 extraordinarily various that they have hitherto been included 

 in three or four genera (or sub-genera), and I am convinced that 

 any palaeontologist to whom the various species — above 200 — 

 should be submitted in a fossil state would distribute them into 

 at least six, or more, genera. Anatomically, however, they are 

 so nearly alike that it may be confidently asserted that there is 

 no other genus of land mollusca at once so rich in species and 

 so exclusively distinct as this of Cochlostyla; also, with the 

 exception of six, or at most eight, forms, which occur on the 

 small islands in the vicinity, they are confined to the Philip- 

 pines. 



With them, there occur there only two small groups of true 

 Helicidae ; one of which, Chlorma, lives on trees, while the other, 

 Dorcasia, is found among grasses and low-growing plants, or 

 even half buried in the soil. Neither genus includes many 

 species, and the greater number of them occur in the northern 



