the modern any, a. Translating, then, with due regard to syntax, and omitting for the moment the word sauerly, the passage would run: 'I think that no tongue could bear to say any . . word of that sight-it was so clean and bright and pure.' We should expect some word contrasting with 'clean' and meaning 'base, vile,' and no word could be more suitable than that found in Pur. 117 as soerly, spelt here in Pearl sauerly, the u not standing for v, but being part of the diphthong au. Phonologically this representation of ON. au is regular, since it appears in ME. as au, ou, or o (Björkman, Scand. Loan-Words 1. 68 ff.; generally au or ou in these poems, Knigge, p. 84). Finally, it may be added that the ME. derivatives of ON. saurr, etc., may well have been confused with ME. sore (OE. sār), ME. sour (OE. sūr), and it is hard to say whether such a spelling as sorge is a mere scribal error or due to folk-etymological confusion with sorge (OE. sorg). The ME. sory, adj. (ON. saurigr), which the Promptorium Parvulorum defines 'soory, or defowlyd yn sowr or fylþe . . cenulentus,' (NED. gives only this example) could hardly help being confused with ME. sorry (OE. sārig), which often means 'vile, wretched' (see NED., s. v. sorry, 5.).

cenosus

130. hym. 'them.' I have not changed such forms, nor by to be, 212, etc.

134. Hit watz, etc. For similar clauses, cf. 257, 981.

136. ne no festival frok. We should expect a preposition, with or in; but perhaps we must supply something like 'did he have on.' 144. on so ratted a robe. A pleonastic construction, which is apparently a combination of two idioms: (1) one (a) +so+ adj. e. g. 'þu eært a swa hende gome,' Layamon, ed. Madden, 1. 162, 1. 13; 'ane zuo greate emperur,' Ayenbite of Inwyt, ed. Morris, p. 100; and (2) the more usual order in Late ME., so+a+adj. as in 'so gaynly a God,' Pur. 728. Mätzner (Engl. Gram.2 3. 197) quotes one example of this combination (one +so+ a + adj.): ‘he yt wan of on so hey a kynge,' Robert of Gloucester's Chronicle 1166 (ed. Wright); but it is interesting that this construction appears in only two manuscripts, the other six having the more regular 'on so heie kynge.'

145. ungoderly. This word, which occurs again in 1. 1092, is not recorded in Bradley-Stratmann, nor is goderly in any of the dictionaries. The intrusion of the r may perhaps be explained as a development from godely by analogy with such words as hagherly (18), and perhaps also with the form goder, which occurs in the ME. expression goderhele, from OE. (tō) gōdre hæle (see NED., s. v. goderheal); similarly Emerson, Publ. Mod. Lang. Ass. 34. 510.

162. þat fele arn to called. 'Multi enim sunt vocati.' But the poet does not mention the 'pauci electi,' since the whole point of the parable for him is the impossibility of the 'man without a wedding garment,' the man defiled with sin, approaching God's pure presence.

164. fulzed in font. Cf. Erken. 299: 'fulloght in fonte'; Awnt. Arth. (Douce MS.) 225: 'folowed in fontestone.' 167. for aproch pou. If the text is correct, the sentence is very badly constructed: we should expect something like 'when you approach.' Perhaps the meaning is: 'See to it that your clothes are clean and fair for the holiday, lest you receive harm, for when you draw near that Prince of noble lineage (you will find that) he hates hell no more than those who are foul.' For 1. 168, cf. 577. 169. The interpretations of the 'wedding garment' in the Middle Ages generally go back either to Augustine, who explained it as 'Caritas' (Migne, Patr. Lat. 38. 562), or to Jerome, whose explanation may be quoted entire: 'Vestis autem nuptialis præcepta sunt Domini, et opera quæ complentur ex lege et Evangelio, novique hominis efficiunt vestimentum. Si quis igitur in tempore judicii inventus fuerit sub nomine Christiano non habere vestam nuptialem, hoc est, vestam supercælestis hominis; sed vestem pollutam, id est, veteris hominis exuvias, hic statim corripitur.'-Migne, Patr. Lat. 26. 160. The interpretation given here in Purity bears some resemblance to Jerome's, but it is without the odor of theological doctrine implied in the 'garment of the new man,' and the comparison of a man's clothes with his deeds is more direct. Cf. note on 162, and Introd., p. xl.

192. see hym with sy3t. This is the only one of the common pleonasms of the kind, such as 'tell with tongue,' 'hear with ears,' that the poet permits himself, and this he uses frequently. ‘See with sight (or eyes)' occurs in Pur. 576, 1710; Pearl 200, 296, 302, 985; Gaw. 197, 226, 1705. For other examples, see Fuhrmann, Alliterierende Sprachformeln, p. 15.

sour tornez. Probably 'evil devices'; see note on 117.

195. Cf. Erken. 161: 'Towarde pe prouidens of þe prince pat paradis weldes'; Winner and Waster 296: 'It es plesynge to the prynce þat paradyse wroghte'; and Death and Life 13: 'If thou haue pleased the prince that paradice weldeth.'

201. Bateson (Mod. Lang. Rev. 13. 378) proposed reading 'Ne never so sodenly so3t un (=on), soundely to weng.' This is awkward and unnecessary, as Gollancz shows (ibid. 14. 153), the word meaning here 'fatally.' G. compares unsoundyly, Gaw. 1438, and unsounde, Pat. 58.

204 ff. After this line M. placed a comma, as though the fall of Lucifer were intended as an illustration of the Lord's wrath. But the poet's purpose was just the opposite, a fact which is extremely important in the structure of the poem, which has often been unjustly regarded as a series of Biblical incidents loosely strung together. After enumerating the various sins for which a man may forfeit the bliss of Heaven (177-192), the poet declares he has often heard it said that God never took such fierce and

sudden vengeance on men as he did for fleshly sin (193-202); for only in avenging this sin did he abandon his customary demeanor and become really wrathful (203-4). The stories of Lucifer (205-34) and of the fall of Adam (235-48) are now introduced as illustrations of acts of vengeance in which God did not become angry. This is clearly emphasized at the end of the narrative of Lucifer's Fall, where the poet says 'And get wrathed not be Wy3' (cf. also 215). Similarly, after telling of the result of Adam's sin, he states expressly that this act of vengeance was carried out in moderation, and soon atoned for by means of a spotless maiden (247-8). But, in contradistinction to these two instances of God's vengeance, in the third instance, the destruction of the world by the Flood because of carnal sin, God showed merciless wrath (249-50). These last lines, then, continue the thought of 11. 192-204, and all that intervenes (204-48) is parenthetical or introductory, intended to make plain, by way of contrast, that God's most terrible vengeance is reserved for those who sin in the flesh.

211. tramountayne. 'North.' So Milton, Par. Lost 5. 755-7:

At length came into the limits of the North
They came, and Satan to his royal seat
High on a hill, far-blazing.

The tradition that Lucifer had his seat in the north of heaven, and, in fact, the whole legend of the fall of Lucifer, was based on Isaiah 14. 12-3: 'Quomodo cecidisti de cælo, lucifer qui mane oriebaris? Corruisti in terram, qui vulnerabas gentes? Qui dicebas in corde tuo: In cælum conscendam, super astra Dei exaltabo solium meum, sedebo in monte testamenti in lateribus aquilonis, ascendam super altitudinem nubiam, similis ero Altissimo.' This, and other Biblical passages, such as Luke 10. 18, led to the early formation of the legend of the fall of the angels, the beginnings of which may be found in Jerome, Augustine, and Gregory. As Carleton Brown notes (Publ. Mod. Lang. Ass. 19. 124), these lines in Purity repeat directly some of the phrases of Isaiah. For references to Lucifer in Old and Middle English, see Skeat's excellent note on

Piers Plowman C. 2. 105; and for the development of the legend of Lucifer, see H. Ungemach, Die Quellen der Fünf Ersten Chester Plays (Erlangen, 1890), pp. 18 ff.

215. Gollancz (Mod. Lang. Rev. 14. 153) adopts M.'s explanation that metz mese (G. mes), 'pity,' comparing the verb mese, 764, and amesying, Pat. 400. I have retained this reading in the text, though Bateson's suggestion (ibid. 13. 378) that metz is a scribal error for meth, had occurred to me also, and seems not unlikely in view of the other occurrences of meth (mepe) in 247, 436, 565.

222. [s]weved. Though I cannot distinguish the initial S, I adopt Gollancz's reading (Mod. Lang. Rev. 14. 153), since there is no instance of defective alliteration with s except at 958, where it is more easily explicable because of the proper names. G. compares Pat. 253, where the whale 'swayues to be se-boþem'; cf. NED., s. v. swayve.

as be snaw þikke. Cf. Fall and Passion 25-8 (ed. Mätzner, Sprachproben, p. 125), where it is said of the false angels:

Seue daies and seue nizt,

As ze seep þat falliþ snowe,
Vte of heuen hi alizt

And in to helle wer iþrow.

223. helle-hole. I have not been able to find any other instances of this word, which is not recorded by the dictionaries, except in Erken., where it is used twice (291, 307), and in Death and Life (386), which shows other traces of imitation of Purity (see Introd., p. xxvi).

as þe hyve swarmez. So Milton, in an elaborate simile, compares the swarming of the devils to 'bees in spring-time,' Par. Lost 1. 768-75. The grouping of similes in clusters, as in these lines (222, 223, and also 226), is a characteristic trick of the poet, and is used by Miss Thomas (Sir Gaw., p. 12) as a proof of the common authorship of The Pearl, Gawain, Purity and Patience. The other instances of this mannerism in Pur. occur at 554, 556; 790, 791; 1018, 1022; 1132, 1134; 1692, 1694, 1696, 1697.

224. forty dayez. The number is due to the alliteration. In the Fall and Passion (see quotation, 222 n.) the fiends fell seven days, and in Piers Plowman (B. 1. 119) and Milton (Par. Lost 6. 871) nine.

225. er ne. This very unusual use of a redundant negative after er occurs again in 1205 'er þay atwappe ne mozt,' which the NED. (s. v. atwape) emends to 'er þay atwappene mozt.' That the negative particle cannot be disposed of in this manner is

plain from this line, where it could not be attached to the preceding infinitive (stynt) as part of the ending. Einenkel discusses pleonastic ne after verbs of fearing, etc. (Anglia 35. 222 ff.), as in Chaucer's translation of Boethius 'he moot alwey ben adrad that he ne lese that thing' (2. prose 4. 1. 113, ed. Skeat); but students of syntax, and the NED., which does not even record the pleonastic ne discussed by Einenkel, do not seem to have noted the construction after er. Though it is possible that this use of ne developed independently in English, it is very likely that the poet borrowed it from Old French, where, as in modern French, avant que ne was a common construction (see Tobler, Vermischte Beiträge 4. 45 f.). It should be noted that in each case the ne appears before myzt (mozt). Cf. also the ne of Pat. 231: 'He watz no tytter outtulde bat tempest ne sessed.' Cf. Bateson, Mod. Lang. Rev. 13. 379, who cites an instance from Caxton.

226. smylt mele. 'Strained meal.' I believe that smylt is the past part. of an OE. *smyltan (OM. *smeltan) used here in the sense of 'filter through.' The original meaning of the Germc. stem *(s) melt is 'disintegrate, dissolve' (Falk u. Torp., Norweg.Dan. Etym. Wörterb., s. v. smelte; cf. Kluge, s. v. schmelzen; Skeat, Etym. Dict., s. v. melt, smelt). This sense of ‘breaking into pieces' may be traced in various ways in melt: WrightWülcker, Vocabularies 235. 33, has 'Fatiscit, . . . dissolvitur, mylt'; cf. Pur. 1566, where NED. defines 'filter in' (see other examples in NED., s. v., melt, v., esp. 2 and 5). NED.'s earliest example of smelt, except for the part. adj. smelt‘enamelled,' Destr. Troy 1667, is dated 1543, but this is in the specialized modern meaning, which is certainly a late borrowing from Scandinavia or the continent. Smylt, of which the NED. gives only this instance, without etymology, defining '?fine,' is the native word (only the derivative smilting in the specialized meaning 'amber' occurs in OE.). It is possible that smult, for which the NED. quotes, without definition, only 'With a smorther and a smoke smult through his nase,' Destr. Troy 911, is the same word, here meaning 'break through.' For the general meaning of the word the derivatives of the Germc. *(s)mel may be compared, e. g. Sw. mula, crumb; Mod. Engl. dial. in smill, in pieces (see Wright's English Dialect Dict., s. v. smuil).

228. Schumacher's emendation of worlde to erpe is unnecessary, as we may here have transverse alliteration; cf. Introd., p. lvii.

239. This line has offered great difficulty to the commentators. M. translated be wrech sagtled 'appeased the vengeance,' as though wrech wrache; but the next line shows that wrech must refer

=