58 NATURAL HIStORT 



you lefe the woqd altogether, and find ycurfelf In a beautiful 

 deep canal, winding through a fpacious wet favannah, which 

 is fometimes many leagues in circumference. The firfl time 

 we went up one of thefe creeks, (called Camouni), I was fur- 

 prifed at this appearance, and thought it muft be a mere local 

 circumflance peculiar to it. We found afterwards the fame in 

 one or two more inftances, and were fatisfied upon enquiry, 

 that it is common to them all. It was natural to look for an 

 explanation of this phenomenon, and 1 foon found it in one 

 of thofe laws, which probably extend to all rivers fubjedl to 

 frequent inundations. It has been obfcrved, in particular, of 

 the Ganges *, that the banks of that river are higher than 

 the adjacent lands at a diftance from the ftream, owing, no 

 doubt, to the annual depofitions of nuid, l^c. during the fwell 

 of the river. Apply the fame rule to the Demerary, and 

 the difEculty will be folved. The wet favannah behind, and 

 the fwampy woods around them, are the body of the low 

 country at its natural level, fcarcely a foot or two above 

 the fea. Whatever additional height the land has in the vi- 

 cinity of the river, from the time you have afcended about 

 twenty miles or fo, is all acquired. It has arifen from the fedi- 

 ment of the river during the rainy feafon, when the country is 

 overflowed fo as that all the lower part of it is under water. 

 This depofition mufl be always more copious, in proportion as 

 it is neai'er the ftream, where additional quantities are always 

 brought, and where it is kept in motion both by the current 

 and the tide. Every thing which we afterwai'ds faw confirmed 

 this theory, and notliing more diretflly than the canals which 

 rmi out at right angles from the river. Some of tliefe extend 

 four miles inward, and they prove to a demonftration, that 

 the land becomes lower and lower the farther you recede from 



the. 



• Account of the Ganges, &c. Phil, Tranf. 1781, by M. Rennell. 



