The verb ei)/llw is in the first place the Attic form of ei)le/w, i)/llw (*vel-ne/-w) "wind, turn round". But probably as a result of a confusion between such forms (which are close in meaning), and/or because of iotacism, there are also forms conjugated from a present ei)/llw, i)/llw that correspond to ei)le/w (*vel-new) "shut in". This confusion (also present in the old manuscripts and some modern lexica) is apparent in the examples given by the Suda. See also iota 310 and iota 322.
[1] cf. epsilon 1815, epsilon 1817, and Hesychius epsilon906.
[2] Aristophanes, Clouds 761, with scholion. ei)/lle is the reading of the Suda and most mss of Aristophanes. Against them, modern editors like Coulon prefer i)/lle (with Cod. M).
[3] Homer, Iliad 13.572; cf. iota 298 (q.v.), Eustathius 3 p.165.
[4] Thucydides 2.76.1, abridged (siege of Plataiai, 429 BCE); cf. epsilon 1282 and tau 130. The Suda and almost every manuscript of Thucydides have e)nei/llontes Only Pi [= Codex Parisinus Graecus 1638 (the second hand)] reads e)ni/llontes, accepted by Alberti.
[5] These words (Aristophanes, Clouds 762-3) come immediately after the line already quoted, and somehow have been inserted in the text here. See again at mu 933, the "cockchafer" entry.