3. The product of OE. _a_ before _l_ + cons. seems to be _a_, or with lengthening before _-ed_, _ō_. e.g.; _wal_ : _cristal_ 273-4 C (also 609-10 T), _wolde_ : _golde_ 208 T, _tolde_ : _holde_, _sholde_ : _holde_ 435-6 T (also 77-80 C), _wolde_ : _beholde_ 751 T (also 449-50 C), 769-70 T (also 471-2 C). Exceptions occur in the Southern MSS. e.g.; _elde_ : _helde_ 102 Cott., _halle_ : _welle_ 230 C, _welle_ : _alle_ 224 C, 280 C, but are not paralleled in MS. T.[I-21] 4. The inflectional endings of the pres. indic. seem to be _-e_, _-est_, _-eþ_ for the singular. There are rimes to prove the 3rd sing. in _-eþ_; _seith_ : _withe_ 106 T, _he sit_ : _nabit_ 40 C, _geþ_ : _deþ_ 200 C (also T and Cott.), 422 C (also T). The plural ending is less evident. The Cambr. MS. has rcularly _-eþ_. e.g.; _habbeþ_ 20, _serueþ_ 1256, _beoþ_ 294, 295, _weneþ_ 314, _leteȝ_ 448, _chaungeþ_ 510, _goþ_, _seoþ_, _spekeþ_ 708 C, _crieþ_ 526; the T. MS. _-en_, e.g.; _seruen_ 590 T, _cryen_ 815 T, _ben_ 909 T, etc. That the _-eþ_ ending did not belong to the original we may probably infer from the fact that while the _-eþ_ of the 3rd sing. counts metrically, the _-eþ_ of the plural usually does not. Cf. 20 C, 256 C, 448 C, 526 C, 708 C, etc. But cf. _springeþ_ 296 C, _bisecheþ_ 765 C, _falleþ_ 786 C. These endings, then, point to an East Midland dialect. Cf. also the rimes; _wepinge_ : _bringe_ Cott. p. 105, _cusseþ_ : _blisse_ 549-50 C. 5. OE. _æ̆_ (_e_) and shortened OE. _ǣ_, umlaut of WG. _ai_, or WG. _ā_. The OE. short _æ_ appears regularly as _a_. _trespas_ : _was_ 1043 T, _orgas_ : _was_ 102 T, _Cesar_ : _bar_ 182 T, are probably to be explained as due to one of the Southern scribes of MSS. C and Cott. e.g. _vnderȝet_ : _set_ 166 C (but cf. _vnderȝat_ : _sat_ 98 C). The shortening of OE. _ǣ_ (umlaut of WG. _ai_) also appears regularly as _a_. e.g.; _glade_ : _ladde_ 480 T, _ilast_ : _cast_ 338 C, _glad_ : _ilad_ 114 C. But cf. _lasse_ : _wytnesse_ 952 T. In the rime, _rest_ : _mest_ 120 C, 384 C, it is impossible to determine whether the _ǣ_ is shortened to _ĕ_, as in parts of the South, or the _ĕ_ is lengthened to _ē_. The shortening of OE. _ǣ_ (WG. _ā_) does not occur in rime often enough to permit any safe conclusion. The rimes _radde_ : _madde_ 826 T and _radde_ : _hadde_ 1025 T, seem to show that the product of shortening was _a_. That the representative of WG. _ā_ was the Saxon _ǣ_ rather than the Anglian, and Kentish _ē_, seems probable from the rimes _rede_ : _seide_ 21-2 T, 51-2 T, 215-16 T, 263-4 T (66 Cott.); _reede_ : _deede_ 45-6 T, 53-4 T. From these criteria, which seem to be the best available, we may infer that Fl. and Bl. belongs further north than King Horn. Further, the inflections seem to point to the Eastern rather than to the Western Midland, so that we may feel fairly safe in attributing Fl. and Bl. to the East Midland.