24 



JOURNAL, R.A.S. (CEYLON). 



[Vol. VIII. 



of the alphabet, with the attached vowels, has been met 

 with, cut by several persons who had quite different styles 

 of writing. Much of this writing evidently forms part of 

 sentences inscribed round the outsides of Chatties' or on the 

 rims of plates ; but owing to the fragmentary state of the 

 pottery no complete sentence has been obtained. Some 

 letters, which are large and angular, are plainly the work of 

 men who had not very much practice in such writing ; 

 others are small and of very good shape, and are evidently 

 such as might be written on ordinary leaves with a style. 

 No one seeing the different kinds of writing or engraving 

 could attribute all to one person ; yet we find that, without 

 exception, in instances met with at varying heights in the 

 lowest stratum of remains, the shape of the characters is 

 exactly that of the Asoka alphabet, as found in the oldest 

 rock inscriptions in the Island. This agreement includes the 

 letter sa, which in all cases has the angular form resembling 

 the modern Koman F, or rather the Greek digamma, and 

 the letter ma, which is always rounded. No letter of a 

 later shape has been met with, nor a single instance of the 

 rounded form of vowels, or lengthened k or r, which indi- 

 cates the beginning of the transition period of Sirhhalese 

 palaeography. If, then, the oldest inscriptions yet dis- 

 covered in the Island, which contain no letter older in 

 shape than these, go back to the time of Wattagamiui, 

 there is not room for great error in assuming the most 

 recent of these letters to have been made not later than 50 

 B.C. Bat the stratum in which this engraved pottery is 

 embedded is quite four feet thick ; and if the upper part of 

 this dates from 50 B.C., the bottom layer (the tiles and 

 pieces of earthenware are in more or less distinct layers in it, 

 separated by thin layers of soil, and sand, and fine gravel) 

 must be admitted to be of considerably older date. It 

 cannot, I think, be assumed that the whole ground-level at 

 the site (although it is in a hollow) has been raised four 

 feet in much less than 150 years ; and, if not, the earliest 

 remains appear to date from a period not much later than 

 the construction of the dagabas and tank. Only by the 

 assumption that the artificers, the carpenters, and stone- 

 cutters settled at this spot were engaged in the erection of 

 houses in the city on the opposite side of the tank, or in 



