NO 45—1894.] ARCHEOLOGY OF THE WANNl. 153 



The tanks were in many cases provided also with artificial 

 spills or flood escapes, some of which remain. The floors and 

 sides of these spill waters were often covered with large 

 wedged stone pitching, as at Pavatkulam, where the spill is 

 125 ft. long and 60ft. wide ; Pandarakulam, where it is 250 ft. 

 long and 21 ft. wide ; and Vavunikulam. The spills were 

 usually built as kalingulas, i.e., they were provided with a 

 series of pillars, upon the framework of which a dam was 

 erected for holding up an extra depth of about 2 ft. of water. 

 An example is to be seen at Pavatkulam. 



The pillars, which are very irregular in size and shape, are in pairs, 

 a short one in front of a tall one 5 or 6 ft. high, and a few inches 

 distant from it, their line crossing the spill-water slightly in front, i.e., 

 on the tank side of the centre line of the embankment. Between these 

 pillars a temporary dam of sticks and earth would be raised when it 

 was desired to retain extra water at the cessation of the north-east 

 monsoon rains. For this purpose the extra height of the tall pillars 

 would be of no use, and it may possibly have been used as the pier 

 of a temporary foot-bridge, by which travellers could cross the flood 

 that occasionally poured over the spill water." 



There is another smaller kalingula 300 yards lower down 

 the stream from the spill. Examples of kalingulas are 

 to be seen at Maha Rambaikulam,t Mamadu, Erupotane, 

 Matavuvaittakulam, and Pandarakulam. At Mahakachchat- 

 kodi the kalingula has a single pillar only. There was a 

 sort of kalingula at the northern end of the bund of the 

 Vavuniya tank before its restoration, but it is not visible.^ 



With regard to the date of the construction of these tanks, 

 there is perhaps more certainty as to the abandoned tank of 

 Vavunikulam than any other. Mr. Parker says : — 



Yavunikulam may be included among the earliest reservoirs in the 

 Island. It is mentioned in the Mahdwansa. 



* Sessional Papers, 1886, p. 110. 



f Id., p. 183. The pillars at Maha Rambaikulam appear, however, to have 

 been in threes and not in twos, which Mr. Parker has not noticed (diary 

 of September 28. 1889). 



t Id., p. 178. 



