Comments:A Western isogloss: the PA antiquity is rather dubious because of late attestation and possible loans. Mong. may be < Turk., cf. TMN 2, 505, Щербак 1997, 159. The Turk. form itself is somewhat doubtful because of a strange variation -m/-p (see TMN 2, 504-505, with a suggestion of an Iranian origin). A possible source can be Chin. 土卵 tǔluǎn, MC thólwấn, OC thāʔrhōnʔ 'yam', lit. 'earth egg', whence also MKor. thóràn, Kor. thoran, Jpn. tororo 'yam'.