Comments:EAS 51, KW 102, Владимирцов 174, Poppe 23, 75, JOAL 85, ОСНЯ 1, 221-222, Murayama 1962, 108, АПиПЯЯ 51, 72, 284, Дыбо 11, Лексика 21. Despite Щербак 1997, 124, Mong. cannot be explained as a Turkic loanword, and despite Doerfer MT 72 the TM and Mong. forms should be regarded as genuinely related. Note that several forms reflect an original derivative *di̯ūl(u)-gV (Turk., Mong. and Jpn. *dù < *di̯ul-gu).
