22 



JOURNAL, R. A. S, (CEYLON). [Vol. VIII. 



are dinamaha (Tiss. 67, 18); dunamaha (W. P., D. 14); 

 danamhayi (Gp. B 21.) 



droghya sidhi: — is a tatsama. 



3. — Inscription at Wandaru'pa wihara (No. 153.) 



paemini : — comp. paemunu'm inscription at the Ruwan- 

 waeli Dagaba (145) B 25 ; pamini (Gp. B 15.) This is 

 derived from pa-V >J dp, Pali pdpundti, with change of p 

 to m ; like Jaina Prakrit mandma for mandpa. The ae shows 

 that the forms with i, pamini and paemini, are the original 

 ones; although pamunu agrees better with the Pali form. 

 We find two other forms of the same word besides, where the n 

 of the derivation is changed to I, viz., paemili (Mayil. 20, A 4) 

 and pamili (Dambulla, 1). These forms would be difficult 

 to explain by themselves ; but the context shows that they 

 must be identical with those above named. For the change 

 of n to I, comp. piriheld, pirihelima a verbo pirihenawd 

 (Nam. 62 ; Parakramabahu's inscription 137, 20, 25.) 



dew an a :~~ " The 2nd," modern deweni (Childers I., 4). 



pat an : = prasthdna, "since." 



si sard: — gerund of sisaranarcd = sancarati. Comp. 

 Gp. C. 2, Ruan. D. 23. 



Samanola :— older form for Samanela = Samanakuta 

 " Adam's Peak" (Mahav. 68, 6). 



dic/antarayehi i — -from dig and antara, with change 

 of g to g* 



satra :— comp. Dambulla, 21 ; Ruan. D. 14; generally 

 translated " resthouses." At Gp. B 22, we have ddnasatra, 

 of which a synonym is kudaddnasdla (Inginimitiya, A 22); 

 another form is saetraya (Thuparama VII L), and the Tamil 

 equivalent cattiram. About the origin of the word I am 

 doubtful ; but I believe it has nothing to do with Skt. chattra, 

 "canopy," of which the genuine Simhalese form sat occurs 

 at Ambasthala, A 3 ; Ellawaewa Pansala, A. 12. 



nanwd :— only used with satra, and once with dewdlayak 

 (Gp. B 20), As it only occurs in inscriptions of Nicc,anka 

 Malla we are justified in considering it as a corrupted tat- 

 sama— similar to ruswd for rucitvd (Amb. A. 6) — which 

 replaced the old laengu in the inscriptions of the 10th and 

 11th centuries. At Amb. B 23 we find a form nangd, which 



