Apparent sandhi irregularities in the Bhagavad Gita

[Harry Spier]

I'm using Harard-Kyoto (HK) convention in the following.

Verse 2.5

Van Buitenen and Purohit Swami in their editions of the Bhagavad Gita
apply the sandhi rule changing final "n" to palatal nasal "J" before
palatal sibilant "z".  i.e. in verse 2.5 they have:
 "...mahAnubhAvAJ zreyo...".

Radhakrishnan, S.R. Sastri, Warrier, Tapasyananda, Winthrop Sargeant do
not change the final "n". i.e. they have
 "...mahAnubhAvAn zreyo...".

I do not see anything in either Whitney's or Macdonell's grammars that
say this sandhi rule is optional.  Are the sandhi rules optional between
lines in a pada?  In the very next verse sandhi rules are applied to the
ends of lines in a pada.  Or is there some other reason for this
variation?


Verse 10.41

Purohit Swami, Warrier and Radhakrishnan read:

"tejoMzasambhavan".

i.e. no avagraha for the elided "a".

The other editors cited above follow the rule of inserting the avagraha.
i.e. they read:

"tejo'MzasaMbhavan".

Is their a grammatical or usage justification for their not having an
avagraha to indicate the elided "a"?



Many thanks,



Harry Spier



______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com