380 HYATT: GENEALOGY OF ACHATINELLID^E. 



conditions be perpetuated as well marked genera or family 

 groups. It is obvious, also, that the isolated case of the 

 columella with three folds indicates another variation having 

 genetic possibilities. These have all, so far, been arrested in 

 their evolution, and we can speak of them with reference to 

 the phylogeny precisely as we describe the arrested develop- 

 ment of parts and characteristics in the life history or 

 ontogeny of the individual. 



The color patterns have a similar history in this group. 

 The individual mutations range from a normal uniform 

 brown to a very light brownish-white, or in the opposite 

 extreme towards black, and the patterns under certain con- 

 ditions may show localized zigzag bars or straight transverse 

 bars, or revolving bands. All of these have a fuller expres- 

 sion as permanent characteristics in some species, but shells 

 displaying the mutation of revolving bands are almost as 

 rare as those showing the double tooth described above; and 

 there are only about three species of Amastra that can be 

 characterized as banded. 



On the other hand, there is no comparison between the 

 tooth-folds and color patterns in other genera of this family. 

 In the arboreal forms, genus Achatinella proper, the colors 

 often vary in the same species from colorless shells, true 

 albinos, through intermediate grades of browns and greens, 

 often brilliantly banded, to dark uniform patterns often 

 brown or almost black. 



There is in most species a constantly reiterated tendency 

 to swing between these two extremes, one being the absence 

 of all color and the other a very dark uniform pattern, while 

 the more normal shells may display brilliant color patterns 

 often taking the form of alternating bands. Here and there 

 one or the other of these mutations become genetic, and they 

 are apt to become more or less permanent and predominant 

 in some species ; but in no case is any special pattern so 

 general as to exclude mutations more or less parallel with 

 those of other species. This is obviously variability taking 

 place without the interference of natural selection or ap- 

 parently any selective laws depending upon the advantage 



