NO. I ACHATINELLA APEXFULVA DIXON WELCH 205 



nantly dextral, while ridge complexes 1-14 contain both dextral and 

 sinistral shells with the exception of ridge complexes 2 and 4, which 

 have only sinistral forms. 



F. The Effect of the Habitat on Subspeciation 



Gulick (1905) does not believe that natural selection due to external 

 conditions explains the diversity of species in Achatinella. He points 

 out that climate, soil, and vegetation are essentially the same from 

 valley to valley. He discounts the effect of differences in the amount 

 of rainfall in a locality on speciation, because forms from the south- 

 east slope of the Koolau Range show more divergency among them- 

 selves than those on opposite sides of the Koolau Range, even though 

 the rainfall is greater on the northwestern or windward slope than 

 on the leeward slope. Gulick collected most of his material at a low 

 elevation below 1,500 feet. He had no material that showed any 

 possible variation between a lowland dry locality and a highland more 

 humid locality in the same valley or on the same ridge. The diversity 

 of forms on opposite sides of the Koolau Range is certainly as great 

 as, if not greater than, that found between forms on the same side 

 of the mountains. 



Crampton (1932, p. 208), in his study of Partula from the Island 

 of Moorea, states : 



With reference to the problem of the possible effects of the environment upon 

 structural or other qualities the only conclusion warranted by the facts is that 

 congenital factors are solely responsible for the diversities exhibited by the 

 several varieties, by the numerous colonies, and by the individual components 

 of the colonies. 



Dobzhansky (1937, p. 136) believes that while it is difficult to prove 

 that a given trait is not, or has not been, influenced by adaptation to 

 the environment, nevertheless the facts given by Crampton and 

 Gulick are explainable on the assumption that racial differences are 

 merely due to random mutations and to random changes of gene 

 frequencies in isolated populations. In discussing Wright's papers on 

 evolution Dobzhansky (1937, p. 134) explains how species may be 

 differentiated into subspecies. Such an explanation applied to A. apex- 

 fulva would satisfactorily account for valley-to-valley variation. 

 However, it does not explain why there is a definite vertical change 

 in color pattern along a definite line such as the dotted line in figure 7 

 which divides highland dominantly white color patterns from lowland 

 dark color patterns. If variation is purely a matter of the random 

 combination of genes and mutations within a locality, both highland 



