ISOLATION AS A CAUSE OF EVOLUTION 563 



know there may be directive forces inherent in the protoplasm, but 

 it is impossible to say that evolution might not have taken a dif- 

 ferent path, however determinate a given course may seem. As 

 for the paleontological evidence, the incompleteness of the fossil 

 record and our limited knowledge of environmental conditions in 

 the past make it unsafe to generalize. 



Isolation as a Cause of Evolution 



Isolation seems well estabUshed as a factor of some importance 

 in evolutionary causation. The classic example is the work of 

 GuHck upon snails of the family Achatinellidce, which occur in the 

 Hawaiian Islands. These snails Uve upon trees. Since they 

 do not easil}' move upon the ground and never travel any distance 

 over a land surface devoid of shade or moisture, their distribution 

 is restricted. In the island of Oahu there are, along the sides of 

 the principal mountains, small vallej^s in which these snails find 

 suitable habitations, but they can not cross the ridges between 

 neighboring valleys nor the crest of a mountain. Neither can they 

 move out upon the plain below. Thus the population of snails 

 that might happen to be established in any valley would tend to 

 remain isolated as long as similar climatic and topographical con- 

 ditions prevailed. In this instance almost every little valley has its 

 particular variety or varieties, differing in size, color, and shape 

 of the shell. These differences can hardly be accounted for by 

 anything in the environment, since conditions in the different 

 valleys must be almost identical. 



The most reasonable explanation of the facts is that when a 

 population is thus isolated the indi\'iduals become differentiated 

 because some types happen to be present originally and to survive 

 in one locahty, and others in another. There can hardly be any 

 selective action of features hke those mentioned under such uniform 

 conditions. Some of the more recent collectors of these shells 

 even report one variety which seems to have been restricted to a 

 single tree that happened to be sufficiently isolated to prevent 

 transfer. It appears, therefore, that evolution of non-useful 

 characters may occur merely as a result of isolation. Having 

 become estabhshed, such characters might later become subject to 

 selection if the en\'ironment changed and rendered them important 

 in the struggle for life. A similar case has been observed by Kincaid 



