The second hemistich of the Rco akSare -verse goes
like this:
yas tan na veda kim Rcaa kariSyati
ya it tad vidus ta ime samaasate
I believe the pada-paaTha ("sandhi-free reading")
of the second line would be something like this:
ye; it; tat; viduH; te; ime; samaasate
Griffith's translation:
Who knows not this, what will he do with praise-song? But they who
know it well sit here assembled.
The "Diirghatamas-trick" here seems to be using such an order
of words that the plural form of the relative pronoun
'yaH', that is 'ye', change by the rules of sandhi to 'ya'.
The rule here is something like this: if final 'e' is followed
by any other vowel than a short 'a', it changes to 'a'. In the
case above the 'i' of 'it' causes the 'e' of 'ye' to be changed
to 'a', thus 'ye + it' -> 'ya it'.
The "kewl" thang is that 'ya it' could equally well be sandhi
for 'yaH + it'. Well, 'yaH' is "of course" the nominative *singular*
(masculine) for that relative pronoun. Thus, that D. is referring to a
bunch of chaps, is revealed only by the verb forms 'viduH' and
'samaasate' which are plural verb forms.
Checking out some grammatical details revealed that
that was not so cool as we originally thought, but who cares.
