Baltic evidence for *sW-?

Miguel,
I'm afraid to seem importunate, but what do you think about the
Lithuanian river-name <S^eduva`> from msg#10809? IMO, it makes your
hypothesis on *sW a bit less speculative (cf. also Germanic _swim_-
words you mentioned). I'm going to screen the register of Lithuanian
hydronyms and some lexicographical sources on Lithuanian dialects
(where the root may still exist in appelatives) if this would be of
any relevance here.

Sergei