250 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 112 



The distribution of the samples does not, in some cases, coincide 

 exactly with that of certain key characters stipulated in the key and 

 diagnoses. For instance, specimens with a w/l ratio of 26 percent or 

 more occur across the western edge of South Carolina and into North 

 Carolina as far as Pisgah Ridge. This segment of the species is 

 regarded as intergradient rather than as georgiana because both color 

 pattern and gonopod details reflect the characters of ducilla. It does 

 seem probable, however, that migration to the northeast has carried 

 the wide body form more successfully than the invasion of the moun- 

 tains directly north of Georgia, although the typical lowland color 

 patterns have been lost in both cases (a matter discussed at greater 

 length on p. 251). 



The reader may be curious concerning the arbitrary selection of 

 value upon which the isophenes of figure 1 are based. For the most 

 part, the higher values (26-28 percent) occur throughout most of the 

 generic range with little or no geographic variation. The narrow 

 body form, however, coincides closely with the loss of tibio tarsal 

 marginal spurs and the appearance of dark pleurites, and the ratio 

 values were selected deliberately to include the population so charac- 

 terized. The subspecies diagnosed by the stipulated aggregate fea- 

 tures is thus seen to be restricted to the Great Smokies and the 

 adjacent Unicoi ranges. The mutual development of characters 

 within this area reflects the natural integrity of the population. 



It is anticipated that some critics of the subspecific category might 

 assume the decreasing body form to be a function of elevation. This 

 relationship is partly true, but is not the entire answer. Specimens 

 with a ratio of 25 percent or more are available from the vicinity of 

 Highlands, 3,800-4,000 feet; from Walnut Creek Gap, 4,800 feet; and 

 from Co wee Bald, 5,000 feet— all in the Cowee Range. Specimens of 

 ducilla with a ratio of 23 percent and less, are from Soco Gap, 3,800 

 feet; from Mount Sterling, 4,000 feet; and from Chimneys Campground 

 on the north side of Newfound Gap at less than 3,000 feet elevation. 

 The specimens from the last-named locality are the narrowest of the 

 genus, their ratio being 19 percent, and I think these figures prove 

 that the distribution is more horizontal than vertical. 



There can be no doubt that the extensive high country of the 

 Smokies has contributed to the ultimate differentiation of the localized 

 segment of the species. I contend, however, that these mountains 

 have received merely the culmination of the gradual northward pro- 

 gression of several variable tendencies which can be traced for more 

 than a hundred miles from the upper Piedmont of Georgia. Specimens 

 from the Tallulah River drainage in northeastern Georgia and those 

 from the basin of the Ocoee River in southeastern Tennessee agree well 



