128 U.S. NATIONAL ]VIUSEUM BULLETDT 2 59 



One of the greatest problems encountered in fully developing this 

 hypothesis is that of providing a satisfactory explanation for the means 

 of dispersal from coastal areas into adjacent mountainous regions. 

 The upstream migration of ancestral stock into areas of rugged and 

 elevated terrain such as those of the present-day, folded Appalachians 

 and the central Ozarks is difficult to conceive. Assuming that enough 

 time elapsed to allow for coastal to upland migration, however, several 

 considerations can be introduced that offer a feasible solution to this 

 problem. First, there is no evidence that ancestral forms were either 

 more or less agile and better adapted for dispersal than contemporary 

 forms. Even with present species, little is known about their powers 

 of dispersal, and although most subterranean species superficially 

 appear to be more or less restricted in their movements, they may 

 possess greater migratory potential than presently anticipated. 



A second consideration is one of geomorphology and historical 

 geology. At least three times diu-ing the Cenozoic, the Appalachians 

 are believed to have been generally sculptured by erosion into a 

 broad peneplain (Schuchert and Dunbar, 1950; Kummel, 1961). 

 Although it has never been established precisely what consititutes a 

 peneplain, the most satisfactory definition for purposes of this dis- 

 cussion appear to be that of Thornbury (1954, pp. 188 and 189). 

 During any three of the extended periods of peneplanation, stream 

 gradients would have been much reduced and rugged landscapes would 

 have been eroded into gently undulating terrains perhaps similar to 

 those exemplified by the present topographic features of the Piedmont. 

 Theoretical barriers to dispersal, created by elevated terrains and 

 steep stream gradients, would have been drastically reduced and in- 

 land migration greatly facilitated. The Ozark Plateau has also 

 apparently undergone a history of peneplanation dm-ing the Cenozoic 

 (Bretz, 1956; Fenneman, 1938), and one may speculate that similar 

 geological changes there similarly facilitated inland migration. The 

 invasion of subterranean waters of central Texas would have been 

 much easier: the Mississippian embayment reached nearly to the 

 present boundary of the cavernous Edwards Plateau, and during 

 part of the Cretaceous this area was covered by a shallow sea. 



The third consideration depends on whether inland migration took 

 place thi^ough subterranean or epigean routes. If marine ancestral 

 stock invaded coastal areas through interstitial habitats as implied 

 above, then presumably this fauna was already of a hypogean facies, 

 and therefore, from the onset of fresh-water occupancy stygonectids 

 were inhabitants of phreatic waters. Assuming this to be the case, 

 inland migration would of necessity have been largely restricted to 

 subterranean waters. It is, however, logical to assume that sub- 

 terranean dispersal is inherently more difficult than epigean dispersal, 



