Addendzm. 



[No. S2, 



holy place," &c. I would cheerfully submit, were it accurate ; but 

 there is the dative case, and north is simply the nominative. Talam 

 has many meanings in Tamil ; but I admit the sense of "place" (the 

 general Sanscrit meaning) with the superinduced sense of fane, or 

 " holy place ;" because such evidently is the meaning. I cannot 

 admit the " clerical error" of substituting v, for zh. If a letter w^ere 

 changed it would be the latter into L But " the easy clerical er- 

 ror" would hardly have got into two stanzas, and a prose document. 

 Hence, as far as that evidence is concerned, B'havani-lcudal cannot 

 be the place meant by Pazhani : neither was Trichengodi, in Salem 

 district, intended by Chengodu^ or Ten-Casi. Two of Mr. Dowson's 

 boundaries are, I conceive, quite wrong. 



The boundaries, in so far as the three authorities are concerned, 

 I conceive to be as follows : 



On the extreme south, for certain, the sea. 



To the north, or on the north, the modern Pyney. To the eststCTien- 

 godu or south Casi, or the great town. To the west Ko%hi-Kudu 

 or Koli'Kudu^ or the great mountain. 



When there is no desire to force a meaning these bounds are very 

 plain : Chengodu is the same as Ten-Casi; and Koli-Kudu (Calicut) 

 and the great mountain near it, are one. 



Look now at a map, and you have South Travancore narrowed in 

 by the ghats to Ten-Casi, moving thence north by the line of ghats 

 to Pyney, west of Dindigul ; and a line carried thence through Paul- 

 ghat wdll cut the sea-shore at some distance north of (Calicut,) the 

 western boundary. 



Such I conceive to be the ancient limits of Chera-desa proper. 

 It does not follow that such are the bounds of the Congu-ndd. 

 These I take to be the same line prolonged to Coorg, for a western 

 boundary ; a line running from Pyney to Caroor and the Cauvery 

 for an eastern boundary; and the Cauvery with its line prolonged 

 up to Coorg, for a north eastern boundary : the enclosure being tri- 

 angular, or nearly so ; but possibly with a boundary on the north, 

 and fr»m the north, or north-eastern, boundary the Carnataca-desa 

 apparently commences. 



That the Chera-desa and Congu-ndd were not one, or always one, 

 appears from the existence of two different capitals. Koli or Calicut 

 was the metropolis of the former ; and, as we find in our Manuscript 

 Scanda-puram, or TaUcadu, or Dalavan-purajn, of the latter.*" 

 Here I may advert to the note at foot of p. 12 in which, as r 



