John Jackson (1881-1952), Marginalia Scaenica
(Oxford: Oxford University Press, 1955), p. 30 (defending his conjecture εὐνᾶν
at Euripides, Andromache
As for εὐνᾶν, it may be admitted that σ and ν have normally little resemblance; but it must also be admitted that the contour of a character traced by a fallible man, under a flickering light, with a reed pen and evanescent ink upon paper not imperishable, may after the lapse of fifteen centuries be deciphered erroneously, if at all, by a fellow creature working under like handicaps with like materials. The possibility is regrettable, and disconcerting to the friends and enemies alike of conjectural criticism in ancient texts, but it is necessary to remember it.