NO. I ACHATINELLA APEXFULVA DIXON— WELCH 189 



a Study of the species of Par tula on Tahiti and Moorea, have pointed 

 out that tree snails vary from valley to valley in various character- 

 istics such as size, color pattern, and form of the shell. Welch (1938) 

 pointed out that A. mustelina from the Waianae Mountains of Oahu, 

 Hawaii, not only varied from valley to valley in different character- 

 istics, but also at different elevations on the same ridge or in the same 

 valley. 



Before discussing the variation of Achatinella apexfulva I will 

 explain what is meant by a highland and lowland form. On a basis 

 of a group of characteristics such as color pattern of the embryonic 

 and postembryonic whorls, shape and size of the shell, I have divided 

 the material into highland and lowland forms. On figure 7 (p. 194) 

 two broken lines drawn across the Koolau Range outline the three 

 zones of shell variation. Zone I contains extreme lowland forms and 

 is probably the region where many of the extinct Gulick forms were 

 collected. The zone occupies all the region below the lowest broken 

 line, which extends from just above areas 117, 97A??, 37?, to just 

 below areas i and 2.^ Zone II contains the usual lowland forms. 

 Gulick collected or obtained material from the lower part of this zone. 

 Zone II occupies all the region between the two broken lines. Zone 

 III is inhabited by highland forms and extends from above zone II 

 to the backbone ridge of the Koolau Range. Along the border of 

 zones II and III forms occur in certain areas which can be classed 

 as either highland or lowland shells on a number of characteristics. 

 The placing of these borderline areas into highland or lowland zones 

 has been a matter of judgment and might be done differently by 

 another worker. Areas 5 ? ?, 6? ?, 7 ? ?, 14 ? ?, 65 ? ?, 66 ? ?, 78 ?, found on 

 figures 3, 4, and 5, have been omitted from figures 7 and 8 because I 

 am uncertain concerning the exact location of these areas. 



In the following account of shell variation in each area, the usual 

 form ^ and color pattern of the shell, generally exhibited on plates i, 

 2, and 3, is taken as a basis for discussion. In some cases it was not 

 possible to obtain a shell having both the usual form and color pattern. 

 In that case only one of the group of characteristics is shown on one 

 of the first three plates, such as the usual form of the shell or the 

 usual color pattern, and the other characteristic is shown on another 

 plate. In one case (pi. i, figs. 27, 28) rare color patterns of a race 



*An area is made up of a single locality or a group of localities containing 

 similar forms or the same subspecies. 



"See Welch (1938) for an account of how the usual form of the shell is 

 selected. 



