238 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 112 



extremes, there is an area of intermediate types, and body form can be 

 given a value no higher than that of a subspecific character. 



Nearly the same summary can be made concerning the distribution 

 of the tibiotarsal spur of the gonopod, which becomes obliterated in 

 a more or less sporadic fashion throughout the range of the species 

 georgiana, but which is invariably wanting in the region occupied by 

 the slender-bodied subspecies. From a normal absence of about 

 25 percent in most of the generic range, the tibiotarsal spur is absent 

 from 36 percent of a sample from Towns County, Georgia, from 44 

 percent of the population around Highlands, North Carolina, and 

 from 66 percent of specimens collected in the Cowee Mountains. It 

 is missing in 100 percent of the millipeds from the Great Smokies. 

 Here is a case of variational coincidence which certainly reinforces the 

 desirability of recognizing the slender-bodied population as sub- 

 specifically distinct. 



The variation of color pattern hardly needs review. The tendency 

 for localized color races is pronounced, and if subspecies were recog- 

 nized on this basis alone, one could name about eight. Since most 

 such populations would not correlate with groupings made on more 

 fundamental structural bases, it is felt that in Cherokia, at least, no 

 practical ends would be served by the wholesale application of names 

 to color forms. This view is particularly true when we recall that one 

 distinctive phase has arisen independently at two different and widely 

 separated localities, and that elsewhere two or more color phases 

 occur together both typically and with various intermediate forms. 



At only one place does a color pattern phase coincide with other 

 local structural divergence — again in the Smokies, where, for instance, 

 the typical pale color of the pleurites is replaced by dark burnished 

 brown or dilute black. The combination of at least three distinct 

 localized character variants makes it impossible to discount the import- 

 ance of evolution in this region, even though all the characters merge 

 into the more typical phases of C. g. georgiana by gradual gradations, 

 which however do not progress at the same rates of modification. 



In summary, working with geographic variation in paranotal form, 

 body proportions, and coloration, we can divide Cherokia into three 

 mutually exclusive populations, a partition which is supported by such 

 other characters such as those of the gonopods and cyphopods (fig. 5). 

 However, either direct or presumptive evidence indicates that all 

 these populations merge where their ranges meet, and should be re- 

 garded as geographic races of a single species, Cherokia georgiana 

 (Bollman.) One is recognizable by a very distinct modification of the 

 paranota which holds uniform over a rather extensive geographic 

 range. The other two, which share another paranotal form, differ 

 from each other in three or four characters. The lowland population 



