Lithuanian <ri'es^utas>, dial. <ri'es^as>, <ruos^uty~s> 'nut',
Lettish <rie~ksts> 'id.', Old Prussian <bucca-reisis> 'beechnut',
Slavic *ore^xU (/ *orIxU > S-Cr. <o`rax> ?) 'id.' are obviously
related, but a common proto-form can hardly be reconstructed (-
*re:'is- with Piotrian merger *ei and ai in Slavic?). Usually the
word is compared to Greek (n. pl.) <a'rua>, glossed as <ta`
He:rakleio:tika` ka'rua> by Hesychius and translated as 'walnuts' in
Liddell-Scott (but. cf. ka'ruon He:rakleo:tiko'n 'filbert') and
considered a non-IE substrate word or a Wanderwort. Gamkrelidze-
Ivanov insist in PIE origin and involve Greek ka'ruon 'nut, esp.
walnut' and Punjabi karua: (< *karuk) 'nut', from alleged PIE *qHar-
'(wal)nut' (they postulate PIE *qH, reflexed as 0:k in specific
branches). Any ideas?
Sergei