Comments:The Jpn. diphthong is not quite clear. It is interesting also to mention Turkic data: cf. Karakh. tolun in tolun aj 'full moon' (MK, see EDT 501), with tolun formally being a participle from *dol- 'to be filled', but in fact used only in this phrase in the sense of "full moon". Modern Turkish has dolunaj 'full moon' with d- (just as in dol- 'be filled'), but the verb tolun- 'to become full (of moon)' - which would correspond quite regularly to PA *t`i̯ṑlo. One should reckon with a possibility of rather archaic contamination in Turkic (since most languages do not distinguish between the reflexes of *t- and *d-).
Comments:SKE 34. A Kor.-Jap. isogloss, with Kor. -ph- < *-bV-k- and usual verbal low tone. In TM cf. perhaps Ewk. tūwkē 'stick in a cross-bow' (cf. the Kor. form), Ulch. tuj-ǯe- 'to make ready a cross-bow', possibly pointing to PTM *tub(i)- 'stick in a cross-bow'.
Comments:АПиПЯЯ 287. A Turk.-Tung. isogloss. TM also has variants *tilpa- / *tilba- (see ТМС 2, 180) meaning both 'dirt, smear' and 'snowstorm' and having arisen due to contamination with *t`elbu q. v.
Comments:The root seems reliable, although the vocalism is not quite certain: see above the notes on variation *e/i in Turkic; Kor. -ɨ- is also unexpected.
Comments:EAS 112, KW 415, Владимирцов 361, Poppe 112, Дыбо 6, Мудрак Дисс. 39, Лексика 286, Doerfer MT 69-70. Despite АПиПЯЯ 282, Mongolian and Tungus forms should be still considered as genetically related to PT *dɨ̄ŕ (*dīŕ), and the reconstruction must be changed accordingly. This is one of the few cases of secondary voicing in PT (before *ŕ, *r): the original voiceless reflex is preserved within the Common Turkic derivative *tir-sgek 'elbow' (Лексика 247-248) ( = TM *türē-kse; the latter form, despite Rozycki 213, cannot be borrowed < Mong.).
