390 Mr. G. Dollman on 



systems is a low and very ill-defined one, and it is doubtful 

 whether during the times of highest flood the Okovango 

 maishes are not connected with tlio?e of the Ohobe and 

 Zambesi. 



*'At the present day tlie importance and capacity of Lake 

 Ngami is infinitesimal wiien compared with the huge extent 

 of the Okovango marshes and periodically flooded area to the 

 N. and N.E. of the lake ; and it is important to realize that 

 the origin and only source of all tiie intricate maze of streams 

 and marshes of Ngamiland is the great Okovango River, 

 which rises in the Mosamba Mountains in Portuguese W. 

 Ahica and drains an enormous area with a very heavy rain- 

 fall from October to March. The result of this is a huge 

 periodical flood which flows down the Okovango into the 

 marshes of Ngamiland. These gradually rise and overspread 

 hundreds of square miles of the surrounding country, which 

 is extraordinarily flat, the inundation reaching its highest 

 point not during the rainy season, but towards the end of tiie 

 dry season, about August or September. None of this 

 enormous volume of water finds its way out to the sea, but 

 after filling the marshes N. of the lake, and formerly the lake 

 itself, flows on down the Botletle or Zonga, and is at length 

 lost by evaporation and percolation. No doubt formerly on 

 many occasions some of this flood has reached the great 

 Makarikaii salt-pan which is the Ngamiland basin, but 

 apparently no flood has been large enou^i to reach the 

 Makarikari for many years, although an old dry river-bed 

 can be traced. Now there is no doubt that it is only compa- 

 rativel}^ quite recently that the water-supply of Lake Ngami 

 has failed and the lake partially dried up ; for although the 

 processes which brought about this result must have been in 

 progress long before Livingstone's visit in 1849, his descrip- 

 tion of the lake and illustration clearly show it to have been 

 then an imposing sheet of water to a great extent open. 

 To-day Lake Ngami is just a great reed-bed, which dries up 

 almost entirely by the beginning of the periodical flood. 



" The explanation of this failure of the water-supply of the 

 lake is to be found iu the fact that previously one of the 

 many large channels of the Okovango Eiver called the 

 T^onghe ran into the lake at the N.W. corner, but by a 

 natural process of reed-growth and silting up this channel 

 has gradually become choked, till now no water at all finds 

 its way into the lake from the N.W., and its only source of 

 supply is at the S.E. corner, where it is connected with the 

 Boiietle by a kind of backwater or arm, through which it 

 receives a certain amount of water when the floods have risen 



