= 9'^] STEVENSON— THE FORMATION OF COAL BEDS. 549 



deposited by the flood, divides during the dry season into pentagonal 

 columns. Remedi, in the same magazine for 1897, says that the 

 Bermejo drops its heavy load, fills its channel and is compelled fre- 

 quently to take a new direction. Trunks of trees, lodged on the 

 bottom, soon lead to formation of dams and to diversion of the 

 stream. 



The Chaco Oriental is north from the Pilcomayo River, which 

 enters the Paraguay near Asuncion, somewhat more than 100 miles 

 above the mouth of the Bermejo. This part of the area was studied 

 by Smith. ^^- The Paraguay River, north from Asuncion, is very 

 near the highland at the east, but west from the river the Chaco is 

 always low and the plain extends far inland to the high table land of 

 Bolivia, which is said to fall ofif abruptly toward the east. The 

 Chaco is covered with water during heavy rains. Above the mouth 

 of the Bermejo, the region is forested, but farther north are great 

 areas with only scattered Carauda palms and no other vegetation. 

 In lat. 21° 26' 40" S., the river issues from a gorge through rocky 

 hills, which, toward the east, are well connected with the Brazilian 

 highlands; but, west from the river, there seems to be a series of 

 isolated hills rising from the Chaco, whose relation to the Bolivian 

 highlands is unknown, as the region has not been explored. Above 

 this gap, known as Fecho dos Morros, vegetation changes, the palms 

 disappear and one sees only open grass land with bushes and a 

 forest fringe on the river banks. These upper lands are covered 

 almost wholly when the river, which rises 50 feet, is in flood. All 

 this flat country from Fecho dos Morros to Villa Maria, a distance 

 of 400 miles, is subject to floods, which are greater toward the 

 north. At the mouth of Sao Lourengo, the area flooded during 

 high water is not less than 150 miles wide. The whole is a labyrinth 

 of lakes, ponds, swamps and channels in a grassy plain, there being 

 forests only near the river bank. It is more remarkable in this 

 respect than the Amazon flood plains, for even at low water one 

 fourth of the area is covered. When the river is highest, the whole 



"- H. H. Smith in J. B. Hatcher, " Origin of the Oligocene and Miocene 

 Deposits of the Great Plains," Proc. Amer. Phil. Soc, Vol. XLL, 1902, pp. 

 113-131. The communication by Smith is pp. 128-130. 



