But the artificial character of these endings is startlingly revealed when we find _-en_ added in an impossible position, shewing that its true grammatical use was quite dead. Yet we find such examples. A serious error (hardly the scribe's) occurs in l. 347: 'Wheder that she me _helden_ lefe or loth.' _Hold_ being a strong verb, the pt. t. is _held_; we could however justify the use of _held-e_, by supposing it to be the subjunctive mood, which suits the sense; but _held-en_ (with _-en_) is the _plural_ form, while _she_ is singular; and really this use of _-e_ in the subjunctive must have been long dead. In l. 684, we have a case that is even worse, viz. _I kepen in no wyse_; here the use of _-en_ saves a hiatus, but the concord is false, like the Latin _ego seruamus_. In l. 928, the same thing recurs, though the scribe has altered _greven_ into _growen_[40]; for this present tense is supposed to agree with _I_! A very clear case occurs in l. 725: _For if by me this mater springen out_; where the use of _-en_, again meant to save a hiatus, is excruciatingly wrong; for _mater_ is singular! This cannot be the fault of the scribe. Other examples of false grammar are: _thou serven_, 290; _thou sene_, 499. But the climax is attained in l. 526, where we meet with _thay kepten ben_, where the _-en_ is required for the metre. _Kepten_, as a _past participle_, is quite unique; let us drop a veil over this sad lapse, and say no more about it[41]. We may, however, fairly notice the constant use of the Northern forms _their_ and _thaim_ or _theim_, where Chaucer has _hir_ and _hem_. The use of _their_ and _them_ (not _thaim_) was well established by the year 1500 in literary English, as, e.g., in Hawes and Skelton. Caxton uses all four forms, _hem_ and _them_, _her_ and _their_. § 73. I add a few notes, suggested by an examination of the rimes employed.