Gloss: ‘ice’ (item 64 in Bjørn 2017)
Attestations: Hit. eka; OIr. aig; ON jaki ‘piece of i.’; Lith. yže; (?)Wakhi yaz ‘glacier’
Notes: Mallory & Adams suggest the Iranian comparanda (1997:286), and Turner further connects Kati, a Kafir language, yūċ ‘cold’, possibly related to Vedic śyai– ‘to freeze’ (1966: 38, 601), which would require prefixation, and the Indo-Iranian branch remains unsecure. This does not change the fact that the strong evidence in Anatolian with sound cognates elsewhere establish the item for early PIE.
Uralic: FU *jäŋe > e.g. Fin. jää, Hun. jég
Discussion: Formal and semantic identity makes this connection attractive. Although the comparison usually is included in the Indo-Uralic hypothesis (cf. Čop 1970: 158 and Collinder 1965: 124), the limited distribution of a geographically important item in Uralic suggests a later entry. Semantically this item may be compared with *wed-r ‘water’ (item 128) with pan-Uralic distribution.