Chapter II: Part 2

It still remains for us to compare the «Sowdan» with the Provençal version.

In most cases where «S» differs from «F», it also differs from «P», therefore «S» cannot have taken those variations of readings from the Provençal poem.

The account of the knights sent on a mission to Laban, in «S» 1663–1738, considerably differs from the corresponding passage in «P» 2211 ss.

In «P» the scene of the whole poem is placed in Spain, there is no mention of the combat before Rome,[69] as in the first part of the «Sowdan».

The game of blowing a coal, «S» 1996 ss., is not mentioned in the Provençal version.

From these variations, taken at random out of a greater number, [p-xxxiii] it becomes evident that the Provençal poem has not been the original of the «Sowdan».

If now we compare the «Sowdan» with Caxton’s version, which we know to be simply a translation of the French prose romance of «Fierabras»;[70] the few following instances of differences between «C» and «S» will show at once, that also that version from which the prose romance was copied or compiled[71] cannot have been the original of the «Sowdan».

There are several variations in the names contained in the two versions. Thus we find «Ballant» in «C» for «Laban» in «S»; «Fyer»a«bras» in «C» for «Fe»rum«bras» in «S»; «Garin», «C» 55/3 = «Generyse», «S» 1135; «Amyotte», «C» 176/26 = «Barrokk», «S» 1135, &c. The game of blowing a coal is told with more details in «S» 1998, and somewhat differently from «C» 118/24; the incident of Laban’s seizing the image of Mahound and smashing it, which is related in «S» 2507, is omitted in «C», &c.

Looking back now to our investigation concerning the original of the «Sowdan», we sum up what results from it, in the following «resumé»:

Most probably the «Destruction de Rome» is the original of the first part of the «Sowdan». As to the second part, we are unable to identify it with any of the extant versions. The French «Fierabras», as edited by MM. Krœber and Servois, is not the original, but the differences between the two poems are not significant; apparently a version similar to the Hanover MS. may be thought to be the original.

The «Sowdan» is no translation, but a free reproduction of its originals; the author of the «Sowdan» following his sources only as far as concerns the course of the principal events, but going his own independent way in arranging the subject-matter as well as in many minor points.

The «Sowdan» differs from the poem of «Syr Ferumbras» in two principal points:

(1) In being an original work, not in the conception, but in the treatment of the subject-matter, whereas the Ashmole «Ferumbras» is little more than a mere translation. [p-xxxiv]

(2) In representing, in its first portion, the first part of the old «Balan» romance, whereas «Syr Ferumbras» contains only the second. But as that second part of the old «Balan» romance appears to be considerably modified and greatly amplified in the Ashmole «Ferumbras», so the first part of the «Sowdan» contains a likewise modified, but much shortened, narration of the first part of the old «Balan» poem, so that the «Sowdan» has arrived to become quite a different work from the original «Balan» or «Fierabras» romance, and that a reconstruction of the contents of that old poem would be impossible from the «Sowdan».

LANGUAGE AND SUMMARY OF GRAMMATICAL FORMS.

As regards the language of the «Sowdan», the first point is the dialect. Looking at the plurals of the present indicative in «-en» or «-n», we at once detect the Midland peculiarities of the poem. Thus we find, l. 1331, «gone» rhyming with «one», l. 1010, «goon» : «camalyon», l. 506, «gone» : «than», l. 1762, «lyven» : «gyfen», l. 1816, «byleven» : «even».

The verbal forms of the singular present indicative and of the second person sing. preterite of weak verbs lead us to assign this poem to an East-Midland writer. The 2nd and 3rd person singular present indicative end in «-est», «-eth»; and the 2nd person sing. preterite of weak verbs exhibits the inflection «-est»: l. 1202, «goist» : «moost»; 1314, 1715, «knowest»; 1344, «trowest»; 1154, «blowest»; 1153, «saiest»; 2292, «forgetist»; 560, «doist»; 1193, «doistowe»;—1093, «goth» : «wroth», 1609 : «loth», 1620 : «doth»; 1728, «sleith» : «deth»; 561, «sholdest»; 1244, «shuldist»; 603, «madist»; 563, «hadist»; 2219, «askapedist», &c.—Twice we find the 2nd person preterite without «-est» («made», «wroght»); but see the note to l. 2.

If, now, we examine the phonological and inflectional peculiarities of the «Sowdan», we find them thoroughly agreeing with those of other East-Midland works,[72] which still further confirms the supposition of the East-Midland origin of the poem. [p-xxxv]

«I» or «y», the descendants of original «u» (which in Old English [Anglo-Saxon] had already become «y» or «i» in consequence of «i-» mutation or «umlaut»)—are found rhyming with original «i»:—ll. 449, 881, «kyn» : «him», 2060 : «wynne»; 1657, «fille» : «stille»; 1973, «fire» : «desire», &c. It must, however, be noted that the rhyme «king» : «inne» (l. 372) or «king» : «thing» (ll. 173, 236) cannot be regarded as an East-Midland peculiarity, because «king», «drihten», «chikken», the «i» of which is a modification of original «u», are to be met with in all Middle-English dialects, as has been shown by Professor Zupitza in the «Anzeiger für deutsches Altertum», vol. vi. p. 6.

Old English short «a», which is liable to change into «o», appears in this poem—

(1) always as «o», before «n-» combinations («nd», «nt», «ng»):—531, «stronge» : «istonge»; 3166, «bronte» : «fonte»; 214, «amonge» : «longe», &c.

(2) as «a», before the single consonants «m» and «n»:—1120, «name» : «shame», 935 : «same», 1739 : «grame»; 785, 1773, «man» : «Lavan»; 3125, «came» : «Lavan» (cf. 2579, «Lavan» : «tane»); 2160, «came» : «dame», &c.—The fact that «com» (ll. 547, 1395, 3095, &c.) is used as well as «cam» as sing. preterite indic. need occasion no difficulty if we remember that the original short «a» (or «o») of «cam» (or «com») had already been lengthened into «ô» in the O.E. period.[73] «Came» and «come» as pret. sing. are employed indifferently in «Chaucer» as well as in the «Celestin» (ed. Horstmann, «Anglia», i. 56), which is known to have been composed in the East-Midland dialect.

«O» long, from O.E. «â», in our poem has that broad sound which is peculiar to the East-Midland dialect. We find it rhyming with—

(1) original «ô»:—1025, «wrothe» : «sothe»; 801, «goo» : «doo»; 60, «inowe» : «blowe»; 325, «so» : «ido», &c.

(2) unchangeable «a»:—257, «Aufricanes» : «stoones»; 506, «gon» : «than»; 2049, «agoon» : «Lavan», &c.

As many East-Midland works[74] the «Sowdan» has three forms for O.E. þâr:—«thare», «thore», «there», all of which are established by the rhyme:—1805, «thore» : «Egremoure» (cf. 2895, «Egremoure» : «tresoure», 1003, «Agremore» : «more»); 126, «thore» : «lore»; 430, «thare» : «sware»; [p-xxxvi] 2245, «there» : «chere», 2404 : «bere»; 2604, «there» : «were» (w[^æ]ron), 208 : «were» (werian), &c.

We likewise find «sore» and «sare»[75] (O.E. sâre):—1196, «sore» : «more»; 166, «sare» : «care»; 1377, «sore» : «thore».

The O.E. diphthongs «ea» and «eo» and the O.E. «ŷ» (mutated from «êa» or «êo») appear as «e» in this poem:—1595, «me» : «see», 632 : «fee», 1339 : «free», 405 : «be»; 1535, «depe» : «slepe»; 1011, 1523, «dere» : «here»; 963, «yere» : «vere», 1257 : «Olyvere»; 996, «nere» : «were»; 596, 1528, «nede» : «spede»; 1702, «eke» : «speke»; 1726, «leke» : «speke»; 184, 215, 1208, «shelde» : «felde»; 2530, «hevene» : «elevene», &c.

A brief summary of the grammatical inflexions employed in the poem will also give evidence of a great similarity with the forms used by other East-Midland writers, and will serve to show that the language of the «Sowdan» agrees closely with that of «Chaucer».

In the declension of substantives the only remnant of case-formation by means of inflexions is the ending used to form the Genitive Singular and the Plural.

The genitive singular of nouns ends in «es» (sometimes written «-is» or «ys») for all genders:—356, «develes»; 1209, «stedes»; 849, «worldis»; 1804, «worldes»; 3035, «dammes»; 1641, «nedes»; 1770, «shippes»; 1072, «faderis».

Substantives ending in «-s» in the nominative case, remain unchanged in the genitive case:—1214, 1287, «Ferumbras»; 2006, «Naymes»; 3207, «Charles»; 1639, 1350, «Floripas».—«Florip», l. 614, is the genitive case of «Floripe» or «Florip», l. 2027, 1571.

The nominative plural of all genders is formed by «-es» («-is», «-ys») or «-s»:—919, «knightes», 1947, 2276, «knightis»; 1384, «horses», 1401, «horsys»; 429, 2054, «gatis»; 192, «wordes»; 837, «swerdes»; 174, «hedes»; 2289, «ladies»; 3271, «soules»; 26, «bokes»; 606, «peres»; 297, «tours», &c. Examples of a plural case without «s» are seen in «thinge», l. 2, 1709:—O.E. «þing»; «honde», 987, O.E. «handa», as well as «hondes», 1412, 2568; «frende», 3212, O.E. «frŷnd», as well as «frendes», 1011, O.E. «frêondas». Other plurals which are equally easily explained by their O.E. forms are:—«eyen», 825, O.E. «êagan»; «shoone», 1381, O.E. «scêon»; «fete», 1403, O.E. «fêt», «fote», 1427, O.E. «fôtum», 2673, O.E. «fôta». [p-xxxvii]

To mark the difference between the definite and indefinite forms of adjectives is a difficult task; as the final «-e» had in most cases already become silent in the poet’s dialect, it seems probable that he no longer observed the distinction.

The pronouns are the same as in «Chaucer» and in other East-Midland poems:—«I», «me», «thou», «the»; «he», «hym»; «sche», «her» and «hir»; «it» and «hit» (cf. note to l. 41); «we», «us»; «ye», «you». The plural of the personal pronoun of the 3rd person is «thai» and «he» (cf. note to l. 2698) for the nominative case; «hem», and in some doubtful passages (see note to l. 88) «thaym» for the accusative case.

As in «Chaucer», the pronoun of the 2nd person is often joined to the verb:—«hastow» 1680, «maistow» 1826, «shaltow» 1669, «woltow» 1727, «wiltow» 1151, «artow» 1967, «kanstow» 2335, &c.

Possessive pronouns:—«myn» and «thyn» are used before vowels and before «h»; «my», «thy» before consonants. Only once, l. 90, «my» is placed before a vowel. «His», «hire» and «here»; «our», «your»; «here» and (twice, 623, 1244) «thair».

The demonstrative pronouns are «this», «these» or «thes»; «that».

The definite article «the» or «þe», is used for all cases singular and plural. But we find besides, the following examples of inflexion:—«tho», 2063, O.E. «þâ», and the accusative sing. «þon», 108. In l. 2052, «tho» means ‘them, those’ = Lat. eos. «Tha», l. 2639, seems to be a mistake of the scribe, it is perhaps miswritten for «þat» (day), cf. l. 619.

«Men», 115, 1351, and «me», 287, are used as indefinite pronouns. «Everyche», «every», «everychone» occur frequently. Note also «ichoon» 2774, «ilka» 2016; «thilke» 2644, «eche» 1865.

«That» or «þat», «who», «whome» are used as relative pronouns. The interrogative pronouns are «who» and «what».

«Verbs.» The plural imperative ends in «-eth» or «-th», which, however, we find frequently omitted, as in l. 194, «prove you», 2078 «proveth»; 2131 «sende», 167 «sendith»; «telle» 1977, «tellyth» 1625, &c.

The «-n» of the infinitive mood is often dropped, as in «Chaucer»:—274, 1588, «sene» : «bene»; 1124, «see» : «tre»; 658 : «cite»; 600, «be» : «cite»; 1225 : «contre»; 1411, «flee» : «cite»; 3065, «fleen» : «men»; 1282, «sloo» : «mo»; 792, «sloone» : «one», &c.

The final -(«e»)«n» of past participles of strong verbs is in most cases [p-xxxviii] dropped, as in «Chaucer»:—3176 «forlorne»: «borne», 32 «born», 3011 «wonne», 21 «wonnen», 2756 «comen» : «nomen», 155 «come», 2476 «holpe», 1362 «bygote», 1026 «blowe», &c.

Weak verbs form their past participles in «-ed», «-d», «-et», «-t», much as in «Chaucer»:—«lerned» 3042, «eyde» 1648, «toolde» 670, «bogt» 111, «delte» 526, «displaied» 133.

The prefix «i-» or «y-» occurs sometimes, «icome» 784, «come» 155, «istonge» 533, «itake» 49, «taken» 1430, &c.

The present participles end in «-inge» and «ande», as is often the case in East-Midland works:—2831 «prikande» : «comande», 435 «cryande», 924 «makande», 3225 «mornynge» : «kynge», 2399 «slepynge» : «honde», where evidently «slepande» is the true reading.

As in «Chaucer» the 2nd person preterite of strong verbs is sometimes formed by «-est» or «-ist», «letist» 2167; but we find also regular forms, as in «slough» 1259, where, however, the O.E. «e» («slôge») is already dropped.

The «-en» or «-n» of the preterite plural and of past participles is commonly dropped, «ronnen» 3007, «ronne» 2959, «took» 477, «tokene» 2621, «slough» 78, «sloughen» 401, «ido» 327: «so», &c.

The «-d» in the past participles and in the preterite of weak verbs is sometimes omitted, as often happens in East-Midland works. Thus we find «comforte» 2242 and «comforted» 312, «commaunde» 57 and «commaunded» 228, «graunte» 607, «liste» 1132, «list» 1966, «discumfite» 1464, &c. On the same analogy we find «light» 1125, 1189, and «lighted» 3109, «worth» 1203, and «worthed» 1163.

As regards the final «-e»’s, it may be remarked that the scribe has added many final «-e»’s, where the rules would not lead us to suspect them, and has often given a final «-e» to words which in other passages of the poem, although similarly used, have no «e»:—«note» 245, 274, «not» 255, 313; «howe» 19, «how» 275; «undere» 61, «under» 713; «bute» 247, «but» 8; «cooste» 202, «coost» 3062; «crafte» 424, «craft» 2335; «ashamede» 1295, «ashamed» 558, &c.

This is due either to carelessness on the part of the scribe, or perhaps to the fact that in the speech of the copyist the final «e»’s had already become altogether silent, so that finding many words ending in «-e» and not knowing its meaning, he considered it as a mere [p-xxxix] “ornament in writing” (Ellis, «Pronunciation», i. 338), and sometimes added, sometimes omitted it.

With respect to the composer of the «Sowdan» himself, there may be some doubt left whether in his speech the final «e» had become altogether silent, or was still pronounced occasionally. From the following instances it may be concluded with certainty that the poet very frequently did not sound the final «e»:—757 «boghtẹ» : «noght», 3154 «hat» : «fat», 961 «wrongẹ» : «distruccion», 556 «onlacẹ» : «was»; cf. also 1383, 1611, 2163; 2795 «spékẹ we of Ríchard», 2999 «fought», 2093, 859 «bringẹ», 9, 2547 «keptẹ», 834 «wentẹ», 142 «comẹ», 713 «wodẹ».

In other cases there is no certainty whether the final «e» is quite silent or must be slightly pronounced or slurred over, so as to form trisyllabic measures. It must be noted, however, that in supposing trisyllable measures in all these doubtful cases, the number of this kind of measure will increase to a great amount in the «Sowdan». Therefore I rather incline to think the final «e» silent also in the following instances:—2090 «défendẹ this place», 1201 «brékẹ both báke», 861 «cómẹ from ál», 2119 «askẹ consaile», 1597 «wólẹ these traítours», 1783 «whéns comẹ yé», 2317 «pássẹ that brígge», 1100 «rónnẹ bytwéne», 2997 «fóught so lónge», 175 «brokẹ nothinge», 1658 «béddẹ with ríght», 713 «grénẹ wodẹ síde», 571 «hómẹ to Rómẹ that nýght», 1610 «the fáls jailoúr feddẹ yoúr prisonére», 2152 «fáls traitóurs of Fránce», 921 «chárged the yónge with ál», 380 «aboútẹ midnýghte», 726 «sóne to hím», 160 «únneth not óne» [Chaucer still pronounces «unnethë»].

Nevertheless there seems to be some instances where the final «e» is to be sounded, as in ll. 298, 2790, 1332, 1619, 2740, 592, 2166, 2463, 1405, 2386, 895, 332, 91.

Final «en» also seems sometimes not to constitute a separate syllable:—1365 «waítẹṇ uppon mé», 459 «brékẹṇ our wállis», 45 «slépẹṇ with ópyne ýȝe», 485 «cómẹṇ by the cóst», 2313 «dídẹṇ it aboút», &c.

In all these cases «n» had very probably already fallen off in the speech of the poet, as the following examples lead us to suppose:—178 «wynne»: «him», 1582 «dye»: «biwry», 2309 «shewe» : «trewe», 2107 «slépe to lónge», 861 «cóme from ál», &c.

As regards the final «es» of nouns, the poet seems to have observed the same rules as those followed by Chaucer; viz. «es» is sounded when [p-xl] joined to monosyllabic stems; it does not increase the number of syllables (and therefore is often spelt-«s» instead of-«es»), when the stem has two or more syllables:—197, 277 «goddës», 665 «nailës», 445 «tentës», 2068 «tentïs», 174, 1799 «hedës», 2032, 2868 «swerdës», 2327 «wallës», 1209 «stedës», 1770 «shippës», 2702 «somers», 2687, 2591 «felowes», 2660 «felows», 2412 «maydyns», 647, 1597 «traytours», 2036 «orders», 45 «lovers», 2612, 3098 «develes», 1072 «faderis», 203, 862 «sowdons», 881 «sarsyns».

The final «es» of adverbs seems no longer to constitute a separate syllable:—2213 «hónged’ els bý», 2786 «éls had’ hé», 2109 «éllis I may sínge», 1525 «élles wol’ hé», 2061 «théns», 1783 «whens».

METRE AND VERSIFICATION.

The poem is composed in four-line stanzas. The arrangement of the rhyme is such that the 1st and 3rd lines rhyme together, and the 2nd and 4th together, which gives the following rhyme-formula: «a b a b». The rhyme-endings employed in one stanza do not occur again in the next following.

But it must be noticed that there seem to occur some instances of eight-line stanzas, one of which, beginning at l. 1587, is built on the model employed by «Chaucer». Others are arranged differently. Those beginning at ll. 1059 and 1219 show the rhyme-formula «a b a b a c a c», in that of l. 1411 the 2nd and 4th lines are rhymed together, and the 5th and 7th, whilst the 1st, 3rd, 6th, 8th, all end with the same rhyme. The formula for the stanzas beginning at ll. 807, 879, 1611 is «a b a b c b c b». In the stanza of l. 939 all the pair lines are rhymed together, and the odd ones also, which is the only instance in the poem of eight consecutive lines having only two rhyme-endings, as generally eight lines show four different rhyme-endings, and three only in the passages cited above. But the whole stanza of l. 939 seems not to be due to the author; he has very probably borrowed it from some other poem.[76]

Turning now our attention to the fact that the lines occurring between the Initials or Capital Letters, which are met with in some passages in the MS., are often divisible by eight, we might feel [p-xli] inclined to regard this as an additional reason for considering the stanza employed in the «Sowdan» as an eight-line one. Indeed, the portion from the Initial of l. 1679 to the next one of l. 1689 might be taken for one single stanza. The 24 lines from l. 575 (beginning with an Initial) to the next Initial in l. 598 might equally be considered as three stanzas, whilst there are 5 times 8 lines = 5 eight-line stanzas from the Initial of l. 2755 to the next Initial in l. 2795.

In all these instances the supposition of eight-line stanzas would suit the context, as is the case also with other passages. Thus in the following cases it might seem as though eight lines taken together were more closely connected and made better sense than four lines, «e. g.» ll. 583–598, 1703–1710, 1679–1686, 939–962, 1043–1050, 244 ss., 455 ss., 631 ss., 1059 ss.

But, on the other hand, it must be borne in mind that there are also a great many cases where, as regards the sense, four lines can be considered as an independent whole, when, «e. g.», the speech spoken by a person is contained in four lines, and the words of another person replying to the first follow in the next four lines. Very often also these next four lines contain only a part of the second person’s reply, so that the remainder of his reply falls into the following stanza. This ‘enjambement’ or continuation of the sense, and sometimes of the syntactical construction from one stanza to another, need not, of course, prevent us from admitting the supposition of eight-line stanzas; as, upon the whole, it is met with in all poems composed in stanzas, and as it is frequently used in «Le Morte Arthur» (Harleian MS. 2252, ed. Furnivall), which is written in eight-line stanzas; but as there is no instance known of an eight-line stanza containing four different rhyme-endings, which at this supposition it would be the case with the «Sowdan», the eight-line stanzas containing either three rhyme-endings, as in «Chaucer», or two, as in «Le Morte Arthur», and as in some passages of the «Sowdan» (ll. 1691, 1695, 1699, 1711, 1715), we find Initials placed after four lines, I believe a stanza of four alternately rhyming lines to be the one intended by the composer—a metre which, according to Guest, «History of Eng. Rhythms», ii. 317—‘must have been well known and familiar during the fifteenth century.’ The few eight-line stanzas quoted above, may [p-xlii] then be owing either to the inadvertence of the poet, who somewhat carelessly employed one of the two rhyme-endings of one stanza a third and fourth time in the following one, or, perhaps also, he intentionally retained that rhyme-ending, and he inserted eight-line stanzas amongst those of four verses as a mere matter of variation. It is perhaps not impossible that the retention of this rhyme-ending was not greatly felt.

As regards the rhymes themselves, they are both monosyllabic or masculine rhymes, and dissyllabic or feminine ones. Frequently they are used alternating with each other, as in the stanzas beginning with l. 2755.

Sometimes we find four feminine rhymes occurring in an unbroken succession, as in ll. 1263–66. But it must be noticed that the number of masculine rhymes is predominant. Thus the stanzas beginning with ll. 3047, 3063, 3123, 1123, 791, 1035, 1271, 1275, 2019, 1311, 1351, 1463, &c., contain only masculine rhyme-endings.

The rhymes are not always full and true; there occur many imperfect ones.

(1) A word in the singular number is often rhymed with a word in the plural number, which therefore has an additional «s» (or «es»):—797, «thinge» : «tidyngys»; 2647, «fyght» : «knyghtes»; 2087, «light» : «knightes»; 1455, «cosynes» : «kinge»; 2272, «laye» : «dayes»; 2395, 885, «Ogere» : «peres»; 2456, «alle» : «walles»; 2682, «nede» : «stedes»; 944, «mone» : «stoones»; cf. also 2376, «wile» : «beguiled». In l. 68, «poundis» : «dromonde»; the rhyme becomes perfect in reading «pounde», as in l. 2336, instead of «poundis».

(2) Single «n» is found rhyming with «n-» combinations.

α. «n» : «nd»—cf. 814, «ychoon» : «Mahounde»; 912, «pavilone» : «Mahounde»; 1201, «crowne» : «Mahounde». The rhyme, 162, «Rome» : «houne», may be explained in the same manner, for «houne» stands for «hounde», as it is spelt in ll. 237, 2377, 935, 1756.[77]

β. «n» : «ng»—cf. 2349, «Mapyne» : «endinge»; 86, «Apolyne» : «tithinge»; 370, «inne» : «kinge»; 1455, «cosynes» : «kinge»; 3249, «Genelyne» : «kinge»; 3171, «serpentyne» : «endinge»; 959, «distruccion» : «wronge». [p-xliii]

In 614, «love» : «vowe», the second rhyme «vowe» does not contain the consonant «v».

(3) Rhymes imperfect as concerns the consonants.

«m» : «n»—cf. 76, «Rome» : «one»; 1672, 364 : «done»; 2443, 366, «come» : «done»; 747, «some» : «soudone»; 1323, «came» : «than»; 1488, «came» : «ranne»; 2128, «tyme» : «pyne»; 177, «him» : «wynne»; 2375, «him» : «tene»; 447, 859, «him» : «kyn»; 2004, «hyme» : «skyne»; 2353, «him» : «inne».

«f» : «v»—cf. 341, «twelve» : «selve»; 415, «wife» : «alive»; 1762, «gyfene» : «lyvene»; 1912, «gife» : «lyve». But in all these cases the rhymes are really perfect, they seem only imperfect in consequence of the copyist writing indiscriminately «f» and «v». Thus the rhyme of l. 341 reappears in l. 1867, «self» : «twelf». In l. 2336 we find «ge»f«e», which is written «ge»v«e» in l. 198; «lefe», l. 764; «safe», l. 864, are spelt with «v» in ll. 1340, 1529, 2808.

«l» : «n»—cf. l. 363, «consaile» : «slayne». Quite similar is l. 1251, «felde» : «sende».

«p» : «k»—l. 820, «stoupe» : «stroke». A similar rhyme occurs in «Guy», l. 10903, «scapid» : «nakid».

«d» : «t»—l. 2868, «gyrde» : «sterte»; 1151, «plete» : «dede».

«d» : «p»—l. 283, «tyde» : «depe». But this rhyme is very probably owing to the scribe. For «depe» we ought to read «wide».

A single consonant rhymes with a double consonant. The only certain instance occurs in l. 311, «tyde» : «chidde». For in ll. 312, 317, «dele» : «welle», we might read «wele», as this word is frequently spelt in the poem; cf. ll. 385, 2618, 1173, 1651, &c. For «dedde» in l. 2980 («rede» : «dedde») we may substitute «dede», which occurs in l. 2510. The rhyme «glad» : «hadde», 2687, becomes perfect if we read «gladde», which is the usual spelling of the word in the poem; cf. ll. 439, 570, 918, &c. Besides, I believe «hadde» to be monosyllabic. «Ferre» : «nere» l. 1575; in l. 117 we find «fere».

The rhyme, l. 2654, «sloughe» : «drowe» can easily be restored in reading «slowe», which occurs frequently, as in ll. 2401, 2683, 304, 2208, &c. The rhyme «ane» : «shafe», 555, seems to be due to some clerical error.

(4) Rhymes imperfect as concerns the vowels.

«a» : «e»—2803, «gate» : «lete»; perhaps we are justified in reading «late», [p-xliv] cf. «Havelock», 328; l. 2752, «made» : «dede». The rhymes «thare» : «were», 1383; «bare» : «there», 671; «Agremare» : «there», 33, are really perfect ones, as we know the poet to have used «thare», «there», and «thore» indiscriminately; cf. ll. 208, 2604, 430, 1805, 1003; l. 1436, «ladde» : «nede»; 2365, «ladde» : «bedde», the author probably pronounced «ledde». For «lefte», l. 2335 : «craft», we may read «lafte», as is shown by l. 424, «lafte» : «crafte». In ll. 1781, 544, «tene» : «than», the rhyme will be improved by reading «then».

«a» : «o» (cf. p. xxxv)—504, «thane» : «gone»; 1143, 1079, «Rolande» : «honde»; 133, «sowdone» : «Lavan» (where we might read «sowdan», as in l. 1491); 627, «sowdane» : «towne»; 2527, 1684, «Roulande» : «londe».

«i»(«y») : «e». This rhyme also occurs in «Chaucer»; cf. Ellis, «Pron.» i. 272; see also «Guy», p. xiv.—l. 21419, «him» : «hem»; 1299, «dynte» : «lente»; 523, «strike» : «breke»; 1643, «mylde» : «shelde»; 1263, «togedere» : «thidere»; 1277, «wepenless» : «iwis»; 344, «shitte» : «mette»; 2538, «hende» : «wynde» (read «wende»), &c.; l. 82, «vilane» : «remedye» (read «vilanye», as in ll. 179, 2577); but 1015, «vilane» : «me», cf. «Guy», xi, ν—813, «sle» : «curtesye»; 895, «we» : «lye»; cf. Ellis, «Pron.», i. 271.

The monophthong «y» is rhymed with a diphthong, the second part of which is «y»:—l. 441, «Sarsynes» : «Romaynes»; 2761, «Apolyne» : «agayne»; 2105 : «slayne»; 2175 : «eyne»; 2280, «dye» : «waye» (cf. 1582); 589, «fyne» : «Bourgoyne».

«o»: «ou» («ow»).—l. 1023, «wrothe» : «southe» (which is written «sothe» in ll. 2014, 2024, 2246, 2719); 779, «fonde» : «grounde»; 260, «clarione» : «soune»; 879, «lione» : «crowne»; 2780, «malison» : «towne», &c. Cf. also 1264, «endured» : «covered».

«o»: «e».—463, «oost» : «best». The rhyme is restored in reading «rest» instead of «oost».

«o»: «i».—l. 966, «sonne» : «begynne».

«ue»: «ewe».—l. 2312, «vertue» : «fewe». But this rhyme cannot be objected to, as “final French «u» (as in «due») was diphthongized into «eu» in Chaucerian English.”[78]

Other irregularities are:—l. 112, «douȝte» : «rowte»; 1987, «use» : «house»; 1131, «thou» : «lough»; 1200, «moost» : «goist»; 1730, «dethe» : «sleith»; [p-xlv] 2136, «pas» : «grace»; 1611, «was» : «mace» (in which cases «e» is silent); 931, 1144, «peris» : «fiers».

A line or verse generally contains four accented syllables, separated from each other by one or by two unaccented syllables, so that there are some instances of trisyllabic feet, as in ll. 817, 834, 2035, 2301, 2791, 3020, 3073, 2313, &c. In ll. 692, 695, two accented syllables are put close together without being separated by an unaccented one, which is altogether wanting. In some passages we find lines of three accented syllables alternating with those of four accents, as in ll. 575–582, 763–770, 839–846, 871–878, 2287–2290, &c. But in most cases lines with four accents follow each other in an unbroken succession, as in ll. 1–372, 995–1010, 1026–1029, 1067–1107, 1147–1154, 1731–1734, &c.

A few instances of verses with more than four accented syllables are also to be met with in the «Sowdan». They are either due to the author and therefore intended, as in l. 37, where the poet almost literally imitates his original,[79] or they may be considered as due to some clerical error, in which case the metre generally can be restored by a slight emendation.

A verse has generally an iambic effect, that is to say, the first foot begins with an unaccented syllable, which is followed by an accented one. Frequently, however, the first accented syllable is preceded by two unaccented ones, as in ll. 41, 75, 127, 151, 367, 849, 1060, 1815, 1819, 2289, 2758, &c. There are some instances of the first foot consisting of a single (accented) syllable only, the unaccented one being altogether wanting, as in ll. 2120, 2288, 2374, 2394, &c.

DATE OF THE POEM AND NAME OF THE AUTHOR.

George Ellis attributes the present poem to the end of the fourteenth or beginning of the fifteenth century. “I think,” he says in his «Specimens of Early English Metrical Romances», ed. Halliwell, p. 380, “it would not be difficult to prove from internal evidence, that the present translation[80] cannot be earlier than the end of the fourteenth or beginning of the fifteenth century.” [p-xlvi]

Having seen from the summary of grammatical peculiarities that there is a great similarity between the language of Chaucer and that of the composer of this romance, we might be inclined to consider the latter as a contemporary of Chaucer. From some passages of the «Sowdan», which seem to contain allusions to Chaucerian poetry, we may conclude that the poet must have known the «Canterbury Tales». Thus ll. 42–46:—

“Whan kynde corage begynneth to pryke,
Whan ffrith and felde wexen gaye,
And every wight desirith his like,
Whan lovers slepen with opyn yȝe,
As Nightingales on grene tre” . . .

appear to be imitated from the «Prologue of the Canterbury Tales», ll. 10–12:—

“And smale fowles maken melodie,
That slepen al the night with open eye,
So priketh hem nature in her corages.”

Further on we remark in ll. 939–40:—

“O thow, rede Marȝ Armypotente,
That in the trende baye hase made þy trone.”

some traces of resemblance with the «Knight’s Tale», ll. 1123–26:—

“And downward on a hill under a bent,
There stood the tempul of Marȝ armypotent,
Wrought al of burned steel, of which thentre
Was long and streyt, and gastly for to see,”

which may still be compared with the first lines of the «Prologue of Queen Anelida and False Arcite»:—

“Thou ferse God of armes, Mars the rede,
That in thy frosty contre called Trace,
Within thy grisly temples ful of drede,
Honoured art as patroun of that place.”[81]

Now the «Prologue of the Canterbury Tales» and the «Knight’s Tale», being written in couplets, or lines arranged in pairs, were certainly composed after 1385,[82] or rather after 1389.[83] From the treatment of [p-xlvii] the final «e»’s, which, contrary to Chaucer’s usage, seem to have been silent in a great number of cases in the poet’s speech, we may further conclude that the «Sowdan» must be somewhat later than the «Canterbury Tales». Therefore the poet of the «Sowdan» cannot have been merely a later contemporary of Chaucer; I rather think it to be more probable that he must have lived some time after him. This would bring us to the beginning of the fifteenth century as the date of the romance.

As to the name and profession of the poet nothing is known, and we have no clue whatever from the poem.

MS. OF THE SOWDAN.

The present edition of the «Sowdan» is printed from the unique MS. of the late Sir Thomas Phillips, at Middle Hill, Worcestershire, which is now in the possession of the Rev. John E. A. Fenwick, Thurlestane House, Cheltenham. Sir Thomas Phillips purchased the MS. at Mr. Heber’s sale.[84] The oldest possessor’s name which we find noted, is on the reverse of the last leaf of the Manuscript, where is written, “This is John Eteyes (or Ebeye’s) boke, witnes by John Staff”—in a hand «circa temp.» Eliz. or Jac. I. By some notes made by former possessors on the first fly-leaf of the MS., and by the autograph names which we find there, we learn that Geo. Steevens bought the MS. “at Dr. Farmer’s Sale, Friday June 15, 1798, for 1: 10. 0.” On May 20th, 1800, it was “bought at the Sale of Geo. Stevens, for 3. 4. 6.” by “O. Grah^m Gilchrist.”

A transcript of the MS. made by Geo. Stevens had been presented by him to Mr. Douce. This copy was re-transcribed by Geo. Ellis, who, in 1811, published some extracts with an analysis of the romance in the «Specimens of Early English Metrical Romances».[85] The same copy has been followed by Halliwell, who in his «Dictionary of Arch. and Prov. W.», has several quotations[86] from the present romance, which he styles as “«MS. Douce», 175.” [p-xlviii]

The poem of the «Sowdan» was first printed by the Roxburghe Club in 1854.[87] The text of the present edition differs from that of the «editio princeps» in so far as punctuation is introduced, which is altogether disregarded by the MS. and the Roxburghe Club edition. In some passages words which have been written as one in the MS. are separated in the text; thus «a laye», l. 2694; «a ras», l. 645, are printed instead of «alaye», «aras». Sometimes also words written separately in the MS. are united by a hyphen, as «be-falle», 14; «i-wiss», 71; «i-sought», 725; «with-oute», 841; «a-bide», 818; «a-ferde», 1337, &c. These slight deviations from the MS., which are always indicated in the footnotes, seemed advisable on account of the great help they afford the reader in understanding the text. More important emendations and corrections of evident scribal blunders and other mistakes are given in the foot-notes, and will be found explained in the Notes.

The Index of Names will be useful to those who wish to compare the «Sowdan» with any other version of the romance.

The Glossarial Index contains besides the obsolete terms all those words the spelling or the signification of which essentially differs from that now accepted. Words which show only slight orthographical variations from their modern form have not been included, as the reader will have no difficulty in identifying them.

* * * * *

In conclusion I have the pleasant duty of acknowledging the invaluable assistance which Professor Zupitza at all times readily and freely gave me. My best thanks are also due to Mr. Furnivall and to Mr. Napier for their kind advice and suggestions, and to Mr. Herrtage for collating a transcript of the poem with the MS.

EMIL HAUSKNECHT.

«Berlin», «January», 1881.

FOOTNOTES:

[1] «Histoire Poét.», p. 133–4.

[2] Gautier, «Epopées», ii. 308.

[3] Cf. the French «Fierabras», l. 84; «Sir Ferumbras», l. 102; «Sowdone», l. 1067.

[4] Thus in «Scarron», Gigant, iii.

[5] Pantagruel, ii. chap. 1.

[6] See the most interesting account of this piece and its curious manner of representation in «Histoire Littéraire de la France», xvii. 720–21.

[7] Gautier, «Epopées», ii. p. 308; and «Histoire Poétique», p. 99.

[8] See «Huon de Bourdeaux», edd. Guessard and Grandmaison, p. xxxviii.

[9] See G. Nottebohm, «Thematisches Verzeichniss der im Druck erschienenen Werke von Franz Schubert». Wien, 1874.—Op. 76.

[10] Cf. besides, «Histoire Poétique», pp. 97, 143, 155, 214, 251; «Epopées françaises», ii. pp. 307–9; and the «Préface» of the French edition of «Fierabras».

[11] See also Mone, «Uebersicht der niederländischen Volksliteratur älterer Zeit». Tübingen, 1836. p. 56.

[12] Cf. Warton, «Hist. of Eng. Poetry», 1824, vol. i. pp. 147–8.

[13] It is worthy of notice that the account of the Fierabras romance as given by Barbour, may be considered, on the whole, as identical with the subject of the French «Fierabras» or the English «Syr Ferumbras», but not with the «Sowdan», as there is no mention made of the combat before Rome, nor any trace of what makes up the first part of the «Sowdan». But the spelling «Lawyn» for «Balan» agrees with the spelling of the same name in the «Sowdan». As to the relics mentioned in the passage above, they differ from all other versions.

[14] In the Sowdan the Bridgeward is called «Alagolofre»; cf. Index of Names.

[15] This MS. consisting of 71 parchment leaves in 4to, with coloured initials at the beginning of each rhyme-strophe, had formerly been in the possession “Majoris Monasterii congregationis Sancti Mauri,” at Paris. Having passed through many hands during the French Revolution, it finally came to the Library of Wallerstein.

[16] Der Roman von Ferabras, provenzalisch. Berlin, 1829.

[17] British Museum, MS. Reg. 15. E. vi.

[18] Cf. also the «Préface» of the French «Fierabras», p. iv.

[19] See «Leben und Werke der Troubadours», by Friedrich Diez, Zwickau, 1829, p. 613 note, and Berliner «Jahrbücher für wissenschaftliche Kritik», 1831.

[20] In a footnote to his «Histoire de la Poésie scandinave», p. 183, where he says:—“Le roman de Ferabras, publié à Berlin par M. Bekker, est . . . évidemment traduit du français, et en a conservé trop de formes et d’expressions pour avoir la moindre valeur grammaticale.”

[21] «Fierabras chanson de geste», edd. Krœber and Servois, in the collection of the «Anciens Poètes de la France».

[22] For a more detailed analysis, see «Histoire Poét.», p. 251, and cf. the account given of the old «Fierabras» or «Balan» romance by Philippe Mousket, ed. Reiffenberg, Bruxelles, vol. I. v. ll. 4664–4716, which runs as follows:—

4664 Puis fu Roume par force prise
et la gent destruite et ocise
et li apostoile ocis
Castiaus-Mireors ars et pris
4668 et toute la cité bruie.
li dus Garins et sa mesnie
entrerent en Castiel-Croisant,
quar Sarrasin, Turc et Persant
4672 amenerent trop grant compagne
et devers Surie et d’Espagne;
si furent crestien dolant,
et manderent tot maintenant
4676 soucours al bon roi Charlemainne
ki sa fieste en France demainne,
et li rois en cele besogne
lor tramist Guion de Bourgogne,
4680 ki nouviaus chevaliers estoit
et des jovenes enfans avoit
devant çou la couronne prise.
et soucoururent sans faintise
4684 lor bon roi en la tiere estrange
u il n’orent ni lin ni lange.
en France estoient revenu
et soujourné et bien péu,
4688 mais à cel soucours le tramist
li rois, ki moult s’entremist,—
et si tramist de Normendie
Ricart à la ciere hardie,
4692 si reprirent li Mireour:
et dus Garins vint à l’estour,
ki tint Pavie en quité
s’ot bien Castil-Croisant gardé,
4696 et Karles ot sa gent mandée,
si vinrent de mainte contrée,
quar il lor faisoit tant de biens,
qu ’à ses amis ne faloit riens.
4700 si trest vers Rome li bons rois
et fist as paiens moult d’anois.
dont se combati Oliviers
a Fierabras ki tant fu fiers;
4704 d’armes l’outra, si reconquist
les .ii. barius qu’à Rome prist,
si les gieta enmi le Toivre
por çou que plus n’en péust boivre;
4708 quar c’est bausmes ki fu remés
dont Ihesu Cris fu embausmés.
puis furent mort tot li paien
et mis en Roume crestiien,
4712 si ot autre apostoile fait
et Karles s’en revint à hait,
si gratia Dieu et St. Piere,
que recouvrée ot sa kaiere,
4716 soujourner vint dont à Parise . . .

[23] «Romania», ii. 1873, pp. 1–48.

[24] Cf. «Jahrbuch für romanische und englische Sprache und Literatur», edd. Lemcke, vol. xiii. p. 111.

[25] Printed in «Verhandlungen der 28sten Versammlung deutscher Philologen und Schulmänner in Leipzig». Leipzig, 1873, p. 209 «et seq.»

[26] Corresponding to ll. 1410 «et seq.» of the Ashmole «Ferumbras».

[27] Cf. «Sir Ferumbras», ll. 8192–3.

[28] Cf. also l. 2784 and «Sir Ferumbras», ll. 1860 and 2059.

[29] See above, p. xi, footnote, and «Histoire Poétique», p. 251.

[30] Cf. Grœber, «Verhandlungen», pp. 217–18.

[31] The following differences between the «Destruction» and the narration of Philippe Mousket are worthy of note:—

(i) the combat around Château-Miroir is described in a different manner in the two poems.

(ii) the scene of action, which at the end of the «Destruction» is transferred to Spain, remains, according to Philippe Mousket, in the neighbourhood of Rome for the whole time.

(iii) Guy of Burgundy and Richard of Normandy play a most important active part before Rome, according to Ph. Mousket, whereas in the «Destruction» this is not the case.

Now, as to the last two items, they must have been in the original such as they are related by Ph. Mousket. For only thus some obscure passages of «Fierabras», of which even the «Destruction» affords no explanation, are cleared up. Thus, «Fierabras», l. 1049,

“Près fu du far de Rome, ses a dedens jetés”—

which is in contradiction to the «Destruction», is explained by ll. 4705–6 of Mousket’s account (see above). Only Mousket relates that Floripas has seen Guy before Rome («Fierabras», l. 2240; Ashmole «Ferumbras», l. 1413), and that Richard took part at the combat there. Therefore the account as given by Ph. Mousket, agreeing with what must have been the contents of the old original, is based on a version older than the «Destruction», which exhibits significant differences.

These differences between Mousket and the «Destruction», as well as the fact that several references to preceding events contained in «Fierabras» remain unexplained by the «Destruction», were some of the reasons which led me in my «Dissertation», pp. 41–49, to consider the «Destruction» as a poem written by another author than that of the «Fierabras». In order to clear up the allusions to preceding events contained in the «Fierabras», the very beginning of which necessarily requires some explanatory account—a circumstance which also gave rise to the ‘episode’ of the Provençal version—the «Destruction» was composed as a kind of Introduction to the «Fierabras», whereby it happened that some allusions remained unexplained.

[32] For a description of this magnificent MS., see «Sir Ferumbras», p. vi, footnote.

[33] Cf. Warton, «Hist. of Eng. Poetry», ii. 197–8.

[34] Edited for the E. E. T. S. in 1879, by S. J. Herrtage, B.A.

[35] Cf. Gautier, «Epopées Françaises», i. 221.—“Rien n’est plus fréquent, dans la Chanson de Roland et dans nos poèmes les plus anciens, que la répétition double, triple et même quelquefois quadruple, de certains couplets. Cette répétition n’a pas lieu dans les mêmes termes, ni surtout avec les mêmes rimes. Tout au contraire, la même idée est reproduite en vers différents, munis d’assonances ou de rimes différentes.”

[36] The variations of this MS. are printed in the «Jahrbuch der roman. and engl. Sprachen», vol. ix. pp. 43 ss.

[37] This edition, although printed from the MS. «a», may be said to represent a group («w») of four MSS., called «a» «b» «c» «d» (see above xv). Another group («z») is formed by the MSS. «E» and «D». Both groups belong to the same type «y». Cf. Grœber, «Die handschriftlichen Gestaltungen der chanson de geste Fierabras», Leipzig, 1869, p. 27, where we find the following stemma:

«y»
╱╲
╱ ╲
╱ ╲
╱ ╲
«z» ╲
╱╲ ╲
╱ ╲ ╲
╱ ╲ ╲
«E» «D» «w»
╱╱╲╲
╱╱ ╲╲
╱╱ ╲╲
«d»«a» «b»«c»

[38] «Epopées Françaises», ii. 307, and «Cat. rais. des livr. de la bibl. d’Ambr. F. Didot», I, 361.

[39] Grœber, «Handschriftl. Gestaltungen», p. 6.

[40] «Jahrbuch», xiii. p. 111, and «Zeitschrift für romanische Philologie», iv. p. 164.

[41] “Die Vergleichung weniger aus allen Hss. bekannten Versen macht gewiss, dass «H» mit «D» und «E» aus der nämlichen Quelle «z» geflossen ist.” «Jahrbuch», xiii. 113.

[42] «Handschriftl. Gestalt.», p. 10.

[43] See the note to l. 5763 of «Sir Ferumbras», and cf. «Fierabras», 5955.

[44] The number of instances where «A» varies from «C’s» version might easily be increased. Thus we find «A» 340 differing from «C» 52/111 and from «F» 357; «A» 814 differing from «C» 79/3 and from «F» 1548; «A» 1616 differing from «C» 102/10 and from «F» 2424; «A» 1238 differing from «C» 92/5 and from «F» 2083; «A» 4652 differing from «C» 171/26 and from «F» 4900, &c.

[45] «Specimens of Early English Metrical Romances», ed. Halliwell, p. 379 «et seq.»

[46] «Histoire Poétique», p. 251; cf. also «Revue critique d’Histoire et de Littérature», ii. 1869, p. 121 «et seq.»

[47] Cf. Mr. Shelley’s Paper in Warton, «Hist. of Eng. Poetry», ii. 197–8.

[48] pp. 17 «et seq.»

[49] «Dissertation», p. 18.

[50] «Introduction to Sir Ferumbras», p. xiv.

[51] The French text will be found in the «Notes», which see.

[52] For these names, the «Index of Names» may be referred to.

[53] In some passages the «Destruction» shows also the spelling «Balan», but «Laban» is more common.

[54] See note to l. 1000.

[55] See note to l. 2842.

[56] «Dissertation», p. 20.

[57] See note to l. 1663.

[58] Cf. note to l. 1723.

[59] Mr. Herrtage, in his note to the Ashmol. MS., l. 259, reproduces—from the Roxburghe Club edition, «Introd.» p. vi.—the list of the twelve peers in the French version of the Grenville copy, 10531, which he erroneously takes for that of the «Sowdan».

[60] But there is one “Alorys þe erld of Brye,” mentioned in the Ashm. MS., ll. 935, 2842, 4076, &c.

[61] There is one «Templer» mentioned in the Ashm. MS., l. 2673. But he is not identical with «Tamper» of the «Sowdan», ll. 2641, 2667.

[62] Greek σινδων. Cf. «Dissertation», pp. 45–46.

[63] See note to l. 2535.

[64] There being only a small fragment printed of the Didot MS. («Epopées Fr.» ii. 307), a comparison of the «Sowdan» with this version is impossible at present. But as the Didot MS. belongs to the same group as «E», what results from a comparison of «S» with «E» may be assumed for the Didot MS.

[65] See «Zeitschrift für romanische Philologie», iv. pp. 164, 170.

[66] «Jahrbuch für romanische und englische Sprache und Literatur», xiii. p. 111.

[67] This example is not very striking, as the spelling «Ferumbras» may simply have been retained from the first part of the poem; see above, p. xxxi.

[68] «Syr Ferumbras», «Introduction», p. xiv, footnote.

[69] See «Handschriftliche Gestaltungen», p. 14, and «Dissert.», p. 29.

[70] «Histoire Poétique», p. 157.

[71] And to which only a few very insignificant additions were made by the author; see «Hist. Poét.», p. 99, bottom.

[72] See Morris’s Preface to «Genesis and Exodus», Skeat’s Introduction to «Havelock the Dane», and Mall’s edition of «Harrowing of Hell» (Breslau, 1871).

[73] See Sweet, «Anglia», iii. 152.

[74] Cf. Mall, «Harrowing of Hell», p. 18.

[75] Cf. Schipper, «Alexiuslegenden», 98/121.

[76] See note to l. 939.

[77] “This elision of a final «d» in such words as «hond», «lond», «sheld», «held», &c., is by no means uncommon in ancient poetry, and arises simply from pronunciation.”—Morris, «Specimens of Early English», 320/261.

[78] Cf. Mr. Nicol’s «Paper in the Academy» of June 23, 1877, vol. xi. p. 564, col. 1, and «Seventh Annual Address of the President to the Philol. Soc.», p. 2.

[79] See the note.

[80] Although l. 25 says that the story of the «Sowdan» “is written in Romance,” this cannot induce us to consider our poem as a mere translation. It is, on the contrary, a free reproduction of a French original.

[81] Cf. also Lindsay’s «History of Squyer Meldrum», l. 390:

“Like Mars the God Armypotent.”

[82] Cf. «Prioress’s Tale», ed. Skeat (Clarendon Press Series), p. xx; and Furnivall’s «Trial Forewords», p. 111.

[83] Cf. «Chaucer», ed. Morris, i. 205, footnote.

[84] «Bibliotheca Heberiana», Part xi. p. 162. MSS. Lot 1533.

[85] Ed. Halliwell, p. 379 «et seq.»

[86] For instances, see the following words:—«Atame», «alayned», «ameved», «assorte», «avente», «forcer», &c.

[87] London. Printed by William Nicol, Shakspere Press, MDCCCLIV.

[p-xlix]

ADDITIONS.

Since the «Introduction» was written, I have had an opportunity of seeing the Hanover MS. of the French «Fierabras». The kind offices of Professor Koner exerted on my behalf secured me the consent of the Administration of the Royal Hanoverian Library to have the MS. sent to Berlin, and their most generous permission to consult it freely in the Reading Room of the University Library.

Having now compared the «Sowdan» more closely with the Hanover MS., I must state that the final result arrived at in my investigation concerning the original of the «Sowdan» (cf. p. xxxii) is in no way altered.

As already stated above (p. xxxii), and as the subsequent examination and the passages of «H» quoted below will serve to confirm, the Hanover version is, generally speaking, the same as the printed version of the «Fierabras», differing only in slight variations of readings.

The names in which «S» differs from «F», but agrees with «H», are already spoken of on p. xxxi. But there are several others in the spelling of which «H» agrees with «F», but differs from «S». Thus we find «Balans» or «Balant» in «H» for «Laban» in «S»; «Guarin», «H», leaf 80, back, «F» 438 = «Generyse», «S» 1135; «Agolafres», «H», leaf 81 = «Alagolofer», «S» 2135; «Amiotte», «H» leaf 83, back = «Barrokk», «S» 2939, etc.

As to the subject-matter, there are no instances where «S», differing from «F», agrees with «H». In all points in which «S» differs from «F» we find it also differing from «H».

Thus the game of blowing a burning coal, in the description of [p-l] which «S» slightly differs from «F», is related in «H» with nearly the same words as in «F». As, besides the small fragment printed by Grœber in the «Jahrbuch», xiii, and some few remarks in the «Zeitschrift für rom. Phil.», nothing is known of the Hanover MS., the following passages printed here may serve to show how little «H» differs from «F». The game of the coal («S» 1996–2016, «F» 2907–2934) is thus described in «H», leaf 58:—

“Veillard, dist Lucafer, vo«us» ni savez juer,
Vo«us» ne savez en F«ra»nce le g«ra»nt charboun soffler.
Certes, ceo dist li dus, mais n’en oie soffler.
Et respont li payen: Mais te feray mostrer.
Ly payen vait le duc au g«ra»nt fowel mener.
Q«ua»nt Rollant l’ad veu, a Berard l’ad mostre
Ore p«or»res boue jeu ver «et» esgarder.
Dahait qui ne laira ly «et» Naimes juer.
Lucafer se beysa pur un tison combrer,
Trestote le plus ardant quil i poet trover,
Par tiel air soffla le fu qil li fist voler.
Puis ad dist a Names ‘Ore vo«us» cove«nt» soffler.’
Names prist le tison qui bien se sout aider,
Vers le payen s’en va pur le tison sofler,
Pur ceo le fist ly dus qa ly se volt meller,
Si suffla le tison qe le fist allumer,
Le barbe «et» le menton fist au payen bruler,
Tres p«ar»my le visaie en fist la flame virer,
Qe par un sule petite qe nel fist souuiler.
Q«ua»nt le voit ly payen, le sanc quida deueher.
Il jette a .ij. ses maines, qi le quide frapper,
Mais ly dus le ferry tres p«ar»my le costes,
Qe les oilz de la teste ly fist en fu voler.
Puys l’ad pris par le flank, s’il voit en le fu ruer.
Lichiers, dist dus Names, Dex te poet mal doner,
Tu me quidoies ore come fole cy trover.”

The distribution of the relics, in which «S» (cf. note to l. 3238) differs from «F» 6195 «et seq.» is related as follows in «H», leaf 100:—

“Au baron seint Dynis fu mult g«ra»nt l’assemble«e»
Au perron au londy fu la messe chante«e»,
Illok fu la corone p«ar»tie «et» dessevere«e»,
L’un moite fu a saint Dynis done«e»
Et un clow ansiem«ent», cest v«er»ite prove«e»,
De la Corone fu un p«ar»tie a Ais porte«e»,
A Compaign«e» est l’ensigne en l’eglise honore«e»,
Et les altres .ij. clowes a Orliens fu enveie«e»,
Maint p«re»sant fist Charls de France la loi«e»
Des saintisme reliqes, Jh«es»u de maiestes.
En l’onur de Deu est mainte eglise fonde«e»,
La feste de lendit fu pur iceo estore«e».
Jaiaz videront cens ne taille donee. [p-li]
Ne tardoit q«ue» .iiij. ans k’Espaign«e» fu gaste«e».
La fu la treison de Rollant p«or»pense«e»,
Qe Ganes le vendist a la gent diffaee,
Puys fu as chiuals sa chars destreine«e»,
Pinables en fu mortz de suz Lyons en la pre«e»,
La le vengea Terris au trenchant del espee,
Puys fu pendu armes par gule«e» pare«e»,
Toutz iours vegnent traitors a mal destine«e»
Ou aloignee ou apres ia ni aueront duree.
Charles voit a Orliens, la chancheon est fine«e»
Au deu vo«us» commande, tote j’ai ma chancon fine.
De cels romance est bone la fine «et» l’entre«e»,
Et en mileue «et» p«ar»tote qi b«ie»n l’ad escoute«e»
La beneiceon aez de Deu «et» del v«ir»gine honore. Amen.”

The miracle («F» 6101–6123)[88] of the glove, in which Charles had placed fragments of the thorns, remaining suspended in the air for over an hour, the description of which is omitted in the «Sowdan» (cf. «Dissert.», p. 29), is related as follows in «H», leaf 99:—