78 SPOLIA ZEYLANICA. 



The Kalu-ganga, having been gradually headed by a bar working 

 up from tlie south, has now broken out at the northern end ; and 

 the Bentota-ganga bar is a complete instance of a northward 

 trend. 



The mouth of the Randomba lake (connected with the Madu- 

 ganga), which meets the sea at Balapitiya on the western coast, has a 

 northward pointing peninsula of beach fronting it, and so also has 

 the Madampe lake, not far to the south of it, at Ambalangoda. 



The mouths of the Hikkaduwa-ganga and of Ratgama lake (near 

 Dodanduwa) have sand spits directed to the southward ; and the 

 Gm-ganga, the next outlet of fresh water along the coast to the 

 southward, has a very decided sandy bar pointing to the south-east. 



This introduces another proposition as to the formation of the 

 lagoon and river bars, namely, that when the material is brought 

 down the rivers, it is at the time when the monsoon is blowing 

 strongly ; and there is not only a " wind-slop " of waves driving 

 before the wind, but also a heavy swell running in the same direction. 

 It is not at all improbable that these forces combined may form a 

 current of no great width — a forced current, as one may say — run- 

 ning along the beach to the northward up the western coast, and to 

 the eastward and southward past Galle, towards the Basses. This 

 theory may possibly suffice to explain the whole phenomenon of the 

 growth of the coast line, but until properly systematized observations 

 of currents with a current meter are taken at various depths over 

 the plateau the problem must remam in doubtfulness. 



On the eastern side there is less difficulty of exiDlanation, for the 

 current is running most strongly to the southward at the time of the 

 north-east monsoon and its rains, and thus all the requisite condi- 

 tions are fulfilled. 



It cannot be predicted how long it will be before the coast lines 

 will have worked outwards all round to the 100-fathom line, for 

 there is no reason to suppose that the coast-forming work has now 

 reached a "standstill. On reaching the edge of the plateau it will 

 practically cease, from the great depth that would have to be filled 

 up with detritus. This is probably beginnmg to be felt in the case 

 of the Mahaweli-ganga at Koddiyar Bay, where the edge of the great 

 alluvial plain which has been formed by the river now approaches 

 the head of the deep bight that there occurs in the 100-fathom line, 

 within little more than half a mile. 



On fig. IV. I have included the more important of the artificial 

 tanks scattered over the low-country. These certainly indicate 

 depressions, which m former days may have held swamps or small 

 lakes, whose presence suggested to the tank builders the improve- 

 ment of their depths or dimensions. These tanks need not neces- 

 sarily have once been coastal lagoons, though it is quite within the 

 bounds of probability to suppose so, and especially those that occur 

 near the courses of rivers. 



