Loanwords that show no expected syncope, umlaut, rhotacism or other characteristically Scandinavian changes are obviously pre-Old Norse. For example, the source form of <ansas> 'beam' must have been something very archaic like *ansaz (rather than <áss> or the like), and <kulta> 'gold' < *gulþa- shows an unassimilated cluster. Note also the characteristic difference between forms of the same word when borrowed twice: <patja> straight from *badja- and <peti> re-borrowed after i-umlaut and the loss of the unstressed syllable (Finn. -i was automatically appended to the final consonant of younger loans).
If by Old Nordic you mean Early Runic, then the distinction between that and proto-Germanic becomes very tricky. Early Runic can be regarded as the epigraphic manifestation of a group of dialects ancestral to North and West Germanic (rather than just "Proto-Nordic"). Those dialects were still very close to reconstructable Proto-Germanic; so close, in fact, that it may be impossible to decide whether a given item was borrowed, say, ca. AD 400 or ca. 100 BC. I suppose the lexical distribution of i/e and u/o could be used as a reliable criterion. For example, the vocalism of <tupa> 'living room' (ON stofa) seems to reflect a pre-metatony form (cf. ERun. horna < *xurnan), and <rengas> apparently predates the Common Germanic change *eNC
> *iNC in *xrengaz!
----- Original Message -----
From: keth@...
Sent: Wednesday, March 21, 2001 1:13 AM
Subject: Re: Finnish loans [norse_course]

How do you distinguish between Old Nordic and Old Germanic as sources ?