A still more speculative thought:
*saru-mant-/-mat-, with common dialectal developments such as *sauru-...
(Iranian "umlaut") ~ *sar-... (syncope) could account for "Sauromatai" and
"Sarmatai" at the same time. Old Indo-Aryan s'aru-mant- 'equipped with arrows'
('well-armed') is actually attested, and I think this
attractive connection must have been noticed before (though I
can't remember seeing it published). The only question is whether there is a way
of getting *-ma-ta as the collective of -mant- at a sufficiently early date
(this seems to be the main obstacle). I'll investigate this possibility further;
it might be a rewarding exercise, since there are several similarly formed
ethnonyms.
Piotr
----- Original Message -----
Sent: Friday, November 30, 2001 6:54
AM
Subject: Re: Rosomoni (was[tied]
Dirmar)
Sarus from *saru- has the advantage of
being in the right (u-stem) declension.