THE ISLAND OF BALI. 
themselves. , f , 
A king is to have the accomplishments of the 8 gods ot the 
points of the compass, viz. In dr a , Yama, Surya , Chandra , 
Anila,Kuvera, Baruna, Agni (according to Ramayana,lontar- 
leaf 181 ) . pwii i 
TJtiaraknnda. This, as we have seen, is the last L7thJ 
division of the Indian Haw ay ana* T1 he author is likewise 
Balmiki (Walmikl) Up to this time 1 have not got it 
into my hands ; it is however the history of the brothers of 
Rama, and contains also stories altogether unconnected with 
the family of Rama. A more recent Kawi work is the Ar - 
juria-VYidaja, which borrows its subject from the Uttara- 
kanda; of which hereafter. Kanda, (compare the Kanda of 
Raffles I. pasre 373 et seq) division in India, is us^d in Bali 
like Parva for all sacred writings; those Kawi works how¬ 
ever, whose name is Kanda and Parva, are chiefly destined 
for the princes and nobles of the 2nd and 3rd caste in Bali, 
whilst the works written in S/okas are holy to the priests 
and Brahmans. The Ramayana and the Parvas (of the Ma- 
h&bh&rata) have not been long known to the whole people; 
they were a secret of the priests and chiefs, and contain rules 
for the latter in their government and for every action during 
their temporal life. In every undertaking and in every event, 
persons of rank are bound to conduct themselves in accord¬ 
ance with the precepts contained in those works. Contempt 
or indifference in following those sacred writings would bring 
disaster on princes and people, and the entire happiness of the 
country is indissolubly dependent on the imitation of those 
holy works. 
Parvas (of the Mahdbhdrata), The Second great Indian 
epos is the Mahdbhdrata, composed by the Muni Wyasa (Bal. 
Byasa.) The name of Mahabharata is not know on Bali, but 
its 18 divisions or Parvas are known. The names of those 
eighteen are correct. Six exist entire and two are incomplete* 
From the name of Mahabharata being unknown, it would 
appear that this work at the time it was brought from In¬ 
dia to Java, did not bear this name nor perhaps any gene* 
ral name at all, but that its divisions were already regarded 
as sacred writings. In that case the name Mahabhdrcta is 
only applicable to a small part of the whole work, since the 
war of the Bharater, that is of the Pandawas and Kurus oc¬ 
cupies not more than 20,000 Slokas, whereas the whole work 
contains above 100,000. The rest consists of interpolated 
narratives of various descriptions, which as occasion admits, 
are inserted loosely or annexed. How much the Balinese 
