Though the poet frequently rearranged his original, he omitted comparatively little in the Biblical passages which he followed. Some omissions are made to avoid the repetition that is not uncommon in Scriptural style: for instance, the entrance of the animals into the ark (Gen. 7. 7-9, 14-6), where the phraseology is the same as in God's, command to Noah (Gen. 6. 19-20). So the sending out of the third dove is not mentioned, perhaps because the poet thought it would be an anticlimax after the second had brought back the green branch of olives. To avoid an apparent inconsistency, the poet omits the passage in which Daniel says 'Let thy gifts be to thyself, and give thy rewards to another' (Dan. 5. 17), since Daniel later accepts the reward offered by Belshazzar.

Far more is added to the Biblical account than is taken away. The apocryphal incidents attached to the Biblical stories were traditional, and therefore throw no light on the poet's invention or art. Of the poet's own additions hardly any serves to modify or contribute to the characterization of the personages. The only exceptions are the speech of Nebuchadnezzar (1663 ff.), where the poet has prefixed to the Biblical words the line (1623),

I am god of þe grounde, to gye as me lykes,

which may be an echo of the Herod of the mystery plays; and the behavior of Belshazzar, whose physical manifestations of terror (1542-3) and blustering rage (1583-5) are apparently emphasized in order to represent him as a typical 'boaster on bench.' Again, the poet introduces some passages for the sake of making his story clearer or more coherent. To Abraham's intercession for the faithful in the cities of the plain, he adds a special intercession for Lot (771-6), and in order to make the connection plainer, he makes the angels who appear to Lot refer to Abraham's prayer for him (924).

By far the largest part of the lines not dependent on the

Vulgate are elaboration of detail and ornamentation. It is significant that the poet does not modify or add to the narrative in order to enforce his moral, or call attention to the particular application of his story.1 To be sure, he tells us the significance and point of each narrative before he begins it; but once embarked on it, he depends for his moral effect solely on the vividness of his presentation of the doom of the wicked. The brief outline of Belshazzar's story that he found in the fifth chapter of Daniel, he fills out with all the trappings of medieval chivalry, the sound of trumpets and gay revelry, the lavish ornament of precious jewels, and all the splendor that God shatters in a single night. Out of the few verses which relate the destruction of Sodom and Gomorrah, he forms a picture that in its terrible grandeur is unsurpassed by any poet

of his time.

It is, in fact, the gorgeous color and swift movement of such passages of Purity that make the poem worthy to endure. For, in spite of the fact that the outline is clear and carefully worked out, the poem is not well proportioned. The story of the destruction of the cities need hardly have had so long a preamble as that which includes the announcement of an heir to Abraham. And the introduction to Belshazzar's Feast, which narrates the siege of Jerusalem and the seizure of the sacred vessels, both elaborated with much irrelevant detail, is also unnecessarily long. The poet apparently found such excellent material for story-telling in the events recorded that he could not refrain from inserting it. He becomes so engrossed in his narrative that he forgets that his stories are not being written for their own sake, but as illustrations of a particular theme. Abraham's prayer for Lot illustrates another fault of the poet. His

1 The only instance I have noted is 1. 302, which the poet adds in order to bring out his point that God was really wrathful; cf. note on 1. 204.

remarkable resources of language, his scorn of repeating the same thought in the same words, is often a virtue, distinguishing him from many lesser poets who monotonously repeat the same alliterative phrase. But it becomes a fault when he attempts to find a different phrase for each of Abraham's questions and God's answers; and the whole passage, which is effective in the Bible because of its very simplicity and brevity, becomes in Purity a tedious tour de force, so obviously is the poet striving to avoid repetition.

It is also possible to discover faults in the poet's style; for the frequent abruptness and obscurity of his lines can hardly be ascribed to the defective state of the text. Anacoluthon and involved constructions are to be found in no small number in all his poems. But at his best this boldness in syntax and style results in extraordinary vigor and originality of expression. And with this virility of style is combined a sense of beauty that flashes out in brief simile, or unfolds itself in description of the wild aspects of nature, and the gorgeous pageantry of medieval life in peace and in war. Yet the poet is capable also of tender feeling, such as is revealed in the strain of almost lyric sweetness in which he hymns the praise of the Virgin.1 It is such qualities as these that enable us to recognize in Purity, in spite of its unevenness and lack of proportion, the author of the greater poems, The Pearl and Sir Gawain and the Green Knight.

VII. METRE AND ALLITERATION

The poetry of the Middle English alliterative school is probably, as we have seen, not a learned revival of Old English alliterative poetry, but simply the continuation of a tradition of which the intermediate stages have been lost. In any case, it is certain that the long alliterative line of

Old English poetry forms the basis for that used by the Middle English writers. But though the essential principles are the same, Middle English alliterative verse differs strikingly from Old English in several respects. The most important difference for the general movement of the verse is the tendency to make each line a rhythmic entity, and to avoid altogether the enjambement which is so marked a characteristic of Old English poetry. The only means of defining the limits of the line in Old English was the alliteration, since it was usual to begin a sentence in the middle of one verse and to end it in the middle of another.1 But as soon as the lines came to be all end-stopped, as in Middle English alliterative verse, the line was defined by the pause at the end, and the alliteration, since it was now no longer structurally indispensable, became much less strict.2 The results of this decrease in the structural importance of alliteration will be seen below in the more detailed discussion of the use of alliteration in Purity. The modification of the rhythm of the line was accompanied by many changes3 in the types of half-line common in Old English, some of which were abandoned in Middle English, and others developed and modified.

Two entirely different theories are held concerning the form of the Middle English (and also Old English) alliterative line. Without venturing into a discussion of the evidence for either one, the orthodox view may be given here, that each half-line contains two stresses.* The first four lines of Purity would accordingly be read thus:

2 Cf. J. Thomas, Die Alliterierende Langzeile des Gawaindichters, p. 9.

3

For details, see Deutschbein's dissertation, and Luick in Paul's Grundriss, 2d ed., 2. 2. 162; for the form the types take in Purity, Thomas' dissertation should be consulted.

4

This theory, which has the support of Sievers, Skeat, Schipper, and Luick, was followed by J. Thomas in his valuable dissertation

Clánnesse who so kýndly - cówþe coménde,

And rékken up alle þe résounz - þat ho by rízt áskez,
Fayre fórmez my3t he fýnde - in fórþering his spéche,
And in þe cóntrare kárk and cómbraunce húge.

As in most poems of the alliterative school, there was a strong tendency to avoid a masculine ending of the line, and the final syllable was generally a weak e, which was probably always pronounced.1

As in Old English, the alliteration is generally on the two stresses of the first half-line, and the first stress of the second half-line, but the presence of only one alliterating word in the first half-line, which is permissible in Old English, is rare in Middle English. In general, however,

on the alliterative lines of the Gawain-poet. The dissertation of Fischer, a pupil of Trautmann, attempts to prove the existence of four stresses in the first half-line and three in the second (four in Old English). The great number of rash emendations that Fischer finds necessary in order to provide enough words for seven stresses to the line, inclines one to extreme distrust of his work and of the soundness of his theory-see, for example, the liberal besprinkling of ful's and mony's, p. 19 and passim. Kuhnke, whose dissertation is limited to Gawain, also follows the theory of four stresses for the half-line.

1 Thomas (pp. 21 ff.) tries to show that all the lines of the original had weak ending, and those which do not he makes conform to the majority by adding e's, organic and inorganic, and by several emendations. Though undoubtedly certain spellings, such as gette at the end of 867, lead to the suspicion that the original may have had zette, not get, at 815, 1021, 1049 (Thomas notes only 815), and though it is naturally impossible to prove in most cases that there could not have been analogical and inorganic e's, certain cases, such as þerwyth of 1501 and hygt Cam of 299, where Thomas is obliged to resort to unlikely emendations, show that the weak ending cannot be accepted as a rule without exception. For this reason, and because of the considerable number (8%) of masculine endings in the manuscript, I have not emended any of the final words of the line except for other than metrical reasons.

2 The following lines of Purity contain only one alliterating word in the first half: 105, 175, 315, 427, 770, 779, 958, 993, 1073, 1518,