44 



ROLAND M. HARPER 



Taeda,^'^ bearing much the same relation to the neighboring 

 prairie that P. rigida does to the Long Island prairies above 

 referred to. 



No satisfactory solution for the whole prairie problem has yet 

 been found, and the occurrence of prairie vegetation in what has 

 evidently been a flood-plain in the recent geological past would 

 seem still harder to explain than the better known upland prairies 

 of the upper Mississippi Valley. It appears from the colored 

 geographical map in Bulletin 494 of the U. S. Geological Sur- 



Fig. 2. Prairie landscape just south of the slough shown in fig. 1, looking 



west. 



vey, on the New Madrid earthquake, by M. L. Fuller, published 

 some time in 1912, that more or less similar prairies are rather 

 widely distributed in the coastal plain of Missouri and Arkansas, 

 and there are meagre records of the same sort of thing in the 

 ''delta" of Mississippi. 11 I do not know whether an^^ of these 

 other prairies have ever been \'isited by botanists, or even whether 



" This is the only place where I saw this pine between Little Rock and Mem- 

 phis. 



1' See Hilgard, Tenth Census U. S. 5: 242, 1884. 



