There seems no reason to suppose that stre, sle had ẹ̄, as has been concluded for Chaucer’s language because of such rhymes as sle : he, stre : she, stree : we, see ten Brink, Chaucers Sprache, §23.
It has already been shown that see (sea), which we have supposed to have unstable vowel quality, very seldom rhymes with words having ẹ̄, notwithstanding the frequent opportunity for such rhymes, and ȝee, ‘yea,’ never. It may be questioned whether the rule laid down by ten Brink for Chaucer is a sound one, and whether Chaucer’s practice does not really depend simply upon the larger supply of rhymes in ẹ̄, such as he, she, me, thee, be, se (verb), tre, three, &c. It is at least possible that the difference here between Gower and Chaucer arises from the fact that the latter was less strict in his rhymes, and certainly the later developments of sle, see, stre, ȝee supply no confirmation of the idea that they had ẹ̄ regularly in Chaucer’s language.
O. E. ēo. By the side of sek (siek) there is occasionally sik.
The form fil, fille for fell, felle, pret. sing. and pl. from falle, are not used by Gower. He rhymes fell (fēoll) : hell (hyll) and felle, pret. pl.: felle (fyllan).
The personal pronoun ȝow (ȝou) from ēow rhymes with thou, now, &c.
O. E. ī. Fahrenberg’s instances of ī : ē, i. 177 f. and iii. 413 f., are both founded on mistakes.
O. E. ū. The personal pronoun from O. E. ūs is always written ous, but rhymes in some instances with -us in Latin names, e.g. Tricolonius : ous, Tereüs : ous.
būtan is shortened to bot, not but. It occurs also as a dissyllable in the form bote.
O. E. ȳ. The only example of ȳ as ē is fer from fȳr, which occurs in rhyme with ȝer, iii. 694, (elsewhere fyr). Chaucer has fere, dat., rhyming with here, adv., Troilus, iii. 978, and also afere in rhyme with stere, ‘stir,’ Troilus, i. 229.
The cases of hedde, hed, pret. and past participle (from hȳdan), are examples of shortened ȳ passing naturally to e, and so also fest from fȳst, felthe from fȳlþe, threste from þrȳsta.