SCIENCE 



[N. S. Vol. XXV. No. 649 



may indicate a former southern, rather than 

 northern connection. The southern route 

 would give a continuous lowland habitat, and 

 although it might mean competition with re- 

 lated species now, it does not follow that 

 formerly there was such competition. Geo- 

 logically speaking, both the glaciated region 

 and the Coastal Plain are recently populated 

 areas. 



The geographic relations of G. propinquus 

 and yar. sanborni, and ohscurus are discussed 

 as a unit on account of their close affinities. 

 Their history is quite involved on account of 

 the complex histories of the streams they 

 occupy. These closely related forms are rep- 

 resentative of different parts of the Ohio sys- 

 tem : ohscurus of the upper Ohio, propinquus 

 sanhorni of the middle, and propinquus of the 

 lower Ohio drainage. These respective sec- 

 tions of the Ohio are considered by physiog- 

 raphers as formerly parts of independent 

 drainage systems which later became fused to 

 form the Ohio, and consequently there was an 

 opportunity for differentiation while the 

 streams were separated. Even now while oc- 

 cupying the same system they show only a 

 limited tendency to fuse. Their glacial pre- 

 serve was apparently in the more south-central 

 parts of the Ohio system, not far from the 

 ice margin. 



The species hartoni has the most extensive 

 range in Pennsylvania, which is in decided 

 harmony with its preference for small, rapid, 

 and cool streams — such as abound in the Ap- 

 palachians. This is apparently a preglacial 

 species which has extended its range in post- 

 glacial times to the northeast along the Ap- 

 palachians, across various drainage lines — 

 rather than along them. This is probably due 

 to its tendency to frequent head waters, where 

 divides themselves frequently migrate, and on 

 account of the tendency and ability of this 

 species to wander overland and thus ' migrate ' 

 the divides and around obstacles in streams, 

 such as cascades or low falls. Such facts as 

 these clearly emphasize the need of a de- 

 tailed knowledge of the ecological relations 

 of such animals before due weight can be 

 given to the biological evidence suggesting 

 physiographic changes. 



This species becomes dwarfed in eastern 

 Pennsylvania, a fact of much interest. 

 There are so many conditions which may 

 cause dwarfing that it would be of consider- 

 able interest to know what conditions have 

 been operative here. The variety hartoni ro- 

 hustus is larger than the typical form and in 

 Pennsylvania is confined to the extreme north- 

 western part, where the two forms occur 

 together or separately. The general relations 

 might be expressed thus: the largest form 

 (var. rohustus) occurs in the extreme north- 

 western part of the state, associated with or 

 separately from a smaller form (hartoni) 

 which alone occupies the intermediate eastern 

 area of the state, and a still smaller or 

 dwarfed form occupies the eastern part of the 

 state. Thus there is a more or less progres- 

 sive dwarfing to the eastward, if these forms 

 prove to be intimately related. 



0. caroUnensis is restricted to the south- 

 western part of Pennsylvania, is a species 

 of southern Appalachian range, and appears 

 to prefer the higher altitude and clay soil 

 of the Old Tertiary baselevel. 



0. monongalensis also occurs only in the 

 southwestern part of the state. This is a sub- 

 terranean or burrowing species, even avoiding 

 small streams, and thus it is not lemarkable 

 that the Ohio and Allegheny rivers have 

 proved a barrier to its northward extension. 

 The Youghiogheny and Monongahela rivers 

 seem to have been traversed indirectly by the 

 migration of a divide during the glacial 

 period. 



The range of diogenes is exceptional in that 

 it occupies a narrow strip of the Coastal Plain 

 along the Delaware river, is absent from all 

 the central portion of the state, and occura 

 again in the southwestern part. Such discon- 

 tinuity certainly suggests divergence as to 

 origin. Ortmann is inclined to consider this 

 species of Allegheny Plateau origin. A com- 

 parison made between the present range of 

 diogenes and limosus shows that both exhibit 

 a discontinuity of range between the eastern 

 and western parts of the state, and both 

 occupy the Coastal Plain. These facts sug- 

 gest to Ortmann a retreat from the north be- 

 fore the glacial ice. Here again it may be 



