168 U.S. NATIONAL MUSEUM BULLETIN 259 



one-half of the range covers an area underlain by limestones, these 

 formations are mostly noncavernous, and practically nowhere does 

 geological interpretation allow for the presence of extensive, inte- 

 grated, subsurface drainage systems. I would tend to interpret the 

 range of S. clantoni as representative of an old pattern of distribution, 

 which presently reflects the former range of a more vagile ancestral 

 form. To what extent any two populations along this range can 

 presently exchange genes is largely open to conjecture. 



Conclusions 



1. The combined ranges of species in the genus Stygonectes form a 

 geographic pattern which correlates closely with the old shoreline of 

 the Mississippian embayment. On the basis of this distributional 

 pattern, it has been concluded that the most feasible time for invasion 

 of continental fresh waters by ancestral stygonectid stock would have 

 been during the early to middle Cenozoic (Eocene and Miocene). 

 The possibility that some invasion took place as early as in the Upper 

 Cretaceous and as late as in the late Tertiary and early Quatenary 

 cannot be ruled out, however. If it is assumed that ancestral forms 

 were phreatobitic from the onset of fresh-water colonization, then the 

 most acceptable route of subsequent inland migration was through 

 interstitial habitats developed in flood plains along major rivers. 



2. On the bases of morphological, geographical, and, to some 

 extent, ecological data, three major lines of evolution within the genus 

 Stygonectes have been recognized : (a) emarginatus and spinatus groups 

 of the central Appalachians, (b) tenuis and pizzinii groups of eastern 

 and south-central United States, and (c) jlagellatus and possibly 

 hadenoecus groups of central Texas. 



3. Stygonectids occupy a number of distinct but closely related 

 biotopes which fall into three major categories: (a) caves and solution 

 channels developed in limestone, (b) interstices developed in meta- 

 morphic and noncarbonate sedimentary rocks and in overlying 

 mantle deposits, and (c) interstices developed in loosely consolidated 

 sedimentary deposits. Although, to some extent, occupancy of a 

 particular type of ground-water habitat may depend on areal dis- 

 tribution, few species are associated with more than two of the three 

 major types given above. Moreover, species that occur in limestone 

 terrains are seldom found outside of cave waters. 



4. Patterns of insular speciation are characteristic of species in the 

 emarginatus and spinatus groups of the central Appalachians and in 

 the jlagellatus and hadenoecus groups of the Edwards Plateau. In the 

 Appalachians, ranges are restricted to limestone-floored vaUeys 



