26 HARPER 



does not indicate that they are higher than the country on both 

 sides, but merely that they have a decided slope on one side 

 (toward the stream). They are most delightful places to explore, 

 being free or nearly free from mud, dust, briers, snakes, mosqui- 

 toes and other discomforts, and on them the botanist continually 

 encounters pleasant surprises in the way of rare plants. Their 

 continuity lengthwise of the stream is interrupted by occasional 

 tributary streams, but otherwise one may walk for miles on them 

 almost without any trouble. 



The origin of these fluvial sand-hills, and the reasons why 

 they are so largely confined to the Altamaha Grit region and to 

 one side of the streams, are as little known as the analogous 

 problems in connection with the fall-line sand-hills. It happens 

 that most of them lie off the main highways of travel, and con- 

 sequently have been little studied by other persons than the 

 writer. One may travel by the usual routes from Macon to 

 Savannah, Brunswick, Valdosta, or Thomasville, right across the 

 Altamaha Grit region, without seeing a sand-hill. On the two 

 most direct routes from Savannah to Jacksonville, sand-hills are 

 seen only at the Altamaha River, and going from Savannah to 

 Way cross and Bainb ridge, a distance of 237 miles, the only 

 sand-hills crossed are those of the Altamaha and Satilla rivers. 

 But from the newer railroads of South Georgia (four or five 

 hundred miles of which have been built since 1900), sand-hills 

 are visible at many points. 



There seem to be very few unmistakable allusions to stream 

 sand-hills in literature dealing with other states, so but little 

 idea can be had of their total geographical distribution. The only 

 "sand-hills" mentioned as such in Dr. Mohr's Plant Life of 

 Alabama (p. 195) cannot be definitely correlated with those 

 under discussion here, but in Dr. Smith's Report on the Geology 

 of the Coastal Plain of Alabama (pp. 56, 57, 84, etc.,) there are 

 brief descriptions of such features, located in parts of the state 

 which Dr. Mohr probably never explored. There are a few meager 

 evidences of the same sort of thing in South Carolina. Elliott, 1 



iBot. S. C. & Ga., 2: 676. 1S24. See also Curt. Bot. Mag, 54: pi. 

 2758. 1827. 



