Somehow that IE root, which seemed to play a role in the name of the Bastarnae,
http://tech.groups.yahoo.com/group/cybalist/message/10334
http://tech.groups.yahoo.com/group/cybalist/message/12006
http://tech.groups.yahoo.com/group/cybalist/message/49779
etc, migrated to the Caucasus too:
Georgij A. Klimov
Etymological Dictionary of the Kartvelian Languages
'Common Kartvelian(?) *band- 'to interweave, plait':
Georg. band- 'to interweave, plait';
Megr. bond-; Svan ba:nd- 'to darn, patch up'(?).
According to S. Orbeliani, the Georgian derivative band-ul- denotes a kind of bast shoe. At present the base in its initial meaning occurs in Xevs. dialect, whereas in Gur. it means an incoherent talk (cf. G^onT.i 1984: 53). The Megrelian cognate is represented by the nouns bond- 'suspended bridge (wattled with living plants)' and bondul- 'seine'. The Svan stem may, however, be a simplification of a Georgian borrowing blandva- 'to patch up'.
|| Georgian, Megrelian: Illic^-Svityc^ (1971: 194).
Georgian-Zan *bandG- 'to twist, tie together':
Georg. bandG- 'to interlace';
Megr. bondG- 'to net, spin, web'.
Apparently the verb stem does not occur in Old Georgian texts. Cf. the Georgian dialectal (Imer.) noun correlate bandy- 'cobweb' and the Megrelian action noun bondGua- alongside gobondGil- 'cobweb' (literally past participle 'interlaced'). The Laz equivalent is probably lost. Note a special similarity to Indo-Aryan bandh- 'to tie' (< PIE *bhendh-).
|| Z^Genti (1940: 225). Cf. Fähnrich (1982: 34).'
It seems the -G- is not a suffix in Kartvelian, since Klimov chose not to merge the two entries. Considering their phonetic and semantic similarity the two must be related by derivation, but then in a non-Kartvelian language . So it might be a loan.
Torsten