1896.] NATURAL SCIENCES OF PHILADELPHIA. 421 



extinct species, it is higbly probable tliat this island was never 

 wooded and has always been much dryer than St. Helena. 



It would carry us too far afield to undertake a discussion of the 

 characteristics of the terrestrial mollusk fauna of those Pacific 

 islands which by their elevated and volcanic character and geo- 

 graphic situation might be comparable with those we have already 

 reviewed. A comparison of other highland subtropical faunas where 

 the situation is complicated by seasonal or general aridity, will 

 throw much light on the principles involved. I have elsewhere 

 examined the Lower Californian Bulimxdi (Proc. U. S. Nat. Mus., 

 1893) a group which, like that of the highlands of Peru and Chili, 

 offers an excellent field for study. But the absence of detailed knowl- 

 edge of the situs affected by the several species is a great drawback 

 to safe generalization. A species which spends its existence bur- 

 rowing in the succulent fronds of cactuses can hardly be said to be 

 subject to arid conditions, even if the cactus stands in a desert, and 

 similar doubts and difficulties are encountered at every turn, when 

 one would investigate a general question of this kind. On isolated 

 islands like the Galapagos and St. Helena, the conditions are com- 

 paratively simple, but on the continents it is different, and there the 

 complexity of conditions is too great to allow us with safety to take 

 much for granted. 



Fischer has pointed out that existing faunas are most nearly 

 related to the antecedent tertiary faunas of the same region (Man. 

 Conch., p. 118), the writer has shown that this is true for the Amer- 

 ican and Antillean regions, and others have recognized the same 

 truth in other parts of the world. In pursuance of the same idea, 

 the writer believes that, in the majority of cases, a circumscribed 

 local fauna of land shells will be found in the main to be most 

 nearly related to geographically adjacent groups from which it has 

 probably been derived ; that the conditions of the environment are 

 capable of inducing directly and without the aid of natural or any 

 other kind of selection, certain changes in the form and surface 

 characters which, on the present basis of classification, are generally 

 taken as of systematic value ; that these characteristics may be so 

 loosely worn as to disappear in the individual or in the whole group 

 if the pressure of the environment inducing them be altered or re- 

 moved ; that in time, and especially if the characters be of useful 

 nature, they may become fixed by hereditary, transmission or natural 

 selection, or both combined ; that similar factors in the environ- 



