THE BRITISH MUSEUM (Add. MS. 9067.) Mr. Anscombe, in the May number of Anglia, has drawn attention to the transcript of the Exeter Book made for the British Museum some time prior to the end of February 1832. By directing attention to this neglected, though not unimportant MS., Mr. Anscombe has done a service to students of Old English. He has also raised questions in connection with the MS. which fortunately admit of definite and conclusive answers. His arguments are these: (1) He believes Kemble's text of Widsith to have been taken, not from the Exeter Book itself, but from this British Museum transcript. (2) He thinks that Thorpe derived his text of Widsith, not from the Exeter Book, but from Kemble, and he continues: (3) "Both English and German scholars have kept on copying Thorpe and Kemble, and interpolating here and emending there, until the accepted printed text has ceased to be authoritative." The conclusion then is that the "textus receptus" of Widsith depends upon Thorpe, and that as Thorpe depended upon Kemble, and Kemble upon a transcript, it lacks authority. As to (1). That Kemble used the British Museum transcript is not a question of surmise, but of fact. Kemble writes: "My text is given from an accurate and collated copy of that codex [the Exeter Book] now in the British Museum." (Preface to Beowulf, xxvI.) As to (2). Mr. Anscombe asks, "Was Thorpe's text of Widsith drawn from the original document, or from the diplomatic transcript made at the British Museum, or from Kemble's imprint?" He instances two supposed errors which he believes Thorpe to have copied from Kemble, and concludes, "These two reasons would appear to justify the opinion that Thorpe relied upon Kemble for the text of 'Widsith"". But this is definitely disproved by the statement of Kemble himself, who continues: "[my text of the Traveller's Song] has moreover been most obligingly compared with his transcript from the original at Exeter by my friend B. Thorpe, Esq." So far, then, from Thorpe having made his transcript from Kemble's text, Kemble's text itself had been compared with Thorpe's transcript. 1) This comparison would account for any mistake which we find common to both Kemble and Thorpe: such mistake must have been taken by Kemble from Thorpe, not, as Mr. Anscombe assumes, by Thorpe from Kemble. One such common mistake there undoubtedly is: viz. 1. 125 Th. Kem., waron for MS. wæran. The two instances of common mistakes alleged by Mr. Anscombe are, however, not to the point, since reference to the Exeter Book shows that the reading of Thorpe and Kemble is accurate in both these cases. 2) Even if we had not Kemble's assertion, it could be proved that Thorpe did not rely upon Kemble for his text, for he has corrected Kemble in several places, and the correction is of such a character as could not have been made without reference to the MS., e. g. in 1. 2, where Kemble erroneously wrote three times for the MS. p: in 1. 94, where Kemble wrote hleodrihtne for the MS. hleodryhtne: in 1. 141, where Kemble wrote scaced for the MS. scæceð. 1) Thorpe did not publish this transcript till nine years after Kemble's edition appeared. แ 2) I think here there must be a misprint in Mr. Anscombe's article. He argues that both Thorpe and Kemble agree in the error of reading cafere as Casere". But how else can it be read? The other common error alleged is Wistlawudu. I return to this later. (Des verfs. vermutung ist richtig: cafepe ist fehlgesetzt für cafepe, wie A's. MS. zeigt. E.) As to (3): the assertion that "both English and German. scholars have kept on copying Thorpe and Kemble". This was true from 1838 to 1861.1) But in the winter of 1870-1 Schipper collated Thorpe's text of the Exeter Book with the original, and published his results in Germania, XIX (1874). Schipper corrected Thorpe's errors (which were not numerous in his text of Widsith) even in matters of the most minute detail. Wülcker collated the MS. for his Kleinere Dichtungen (1882) and managed to improve upon Schipper. Wülcker's text had, however, two slight inaccuracies of its own: as these were corrected in the text finally given in Wülcker's edition of Grein (1883), it seems quite likely that Wülcker had examined the MS. yet once again. The work of the late Prof. Wülcker certainly needs no commendation from me: but I may be permitted to remark that, having had occasion, for an edition of the poem which I am preparing, twice to compare Wülcker's transcript of Widsith with the MS. at Exeter, I have found it exceedingly accurate. Möller, Kluge, Holthausen and Sedgefield have all based their texts of Widsith upon Grein-Wülcker. So far, then, from the accepted printed text having ceased to be authoritative, it is based upon a double revision, first by Schipper and then by Wülcker, of the text of Thorpe, which was itself, as Schipper admitted, by no means inaccurate: and so far from Thorpe having copied from Kemble, we have Kemble's express statement that Thorpe copied "from the original at Exeter". Mr. Anscombe's problem "it would appear that it needs to be proved that Thorpe's text of "Widsith" was neither derived ultimately nor directly from the British Museum copy of the Codex" is then solved, by Kemble's own statement. It follows, from Mr. Anscombe having underestimated Thorpe, and overlooked the collations of Schipper and Wülcker, that, when he censures modern editors for what he supposes 1) Guest (1838) Leo (1838) Ettmüller (Scôpes vidsith, 1839) and Schaldemose (1847) copied Kemble: Ebeling (1847) copied Ettmüller: Klipstein (1849) copied Ebeling and Thorpe: Ettmüller (Scôpas and Bôceras, 1850) Grein (1857) and Rieger (1861) copied Thorpe and Kemble. to be departures from the original MS., it is in many cases really the editor who is adhering to the MS., and Mr. Anscombe who wishes to depart from it. For example, he blames Dr. Sedgefield for printing Hringweald, which should be, he says, "Hringwald: the e was written and expuncted." Now Schipper had already pointed this out (Germ. XIX, 333). And Wülcker in the Bibliothek had corrected Schipper's correction: hringweald, Sch. doch ist der punkt ein fleck, nicht beabsichtigt. Another departure from the MS. text for which Mr. Anscombe blames Dr. Sedgefield, is ongend, (1. 35) which should, he says, be ongean. But, on the testimony of the Brit. Museum transcript, of Madden, of Thorpe, of Schipper, and of Wülcker, the Exeter Book does read ongend. I, too, have twice verified from my own collation that the Exeter Book reads ongend. Against this we have only Mr. Anscombe's unsupported assertion that it is ongean: and Mr. Anscombe does not claim to have seen the Exeter Book himself. Yet again, Mr. Anscombe suggests that in reading Wistlawudu in 1. 131 Dr. Sedgefield is departing from the MS. "The Museum transcript has Wistlā wudu (i. e. Wistlan wudu). The mark above the a is differently formed from the mark of long quantity, and can only indicate that nor m is to be supplied." Now, there certainly is, both in the Exeter Book and in the British Museum transcript, a slight mark over the a like an imperfectly formed mark of length. But this is quite unlike the hooked, heavy, horizontal stroke which signifies n or m. The reading Wistlawudu, not Wistlan wudu, has accordingly been adopted by Kemble, working from the transcript, by Thorpe from the original, and has been approved by Schipper and by Wülcker; and Dr. Sedgefield, in common with every other editor of Widsith, naturally adopts it. He could do nothing else, for the stroke could not possibly be regarded as the n or m contraction. Besides, this contraction is not used in the Exeter Book for n, which, except in the contraction poñ for bonne, or in Latin words, is always written in full: the mark in the Exeter Book is invariably used as an abbreviation for m, not n. Mr. Anscombe's accusation against editors such as Wülcker, Kluge, Holthausen and Sedgefield of having "kept on copying Thorpe and Kemble, and interpolating here and emending there, until the printed text has ceased to be authoritative" is therefore based upon a misunderstanding. There is not, and there has not been since 1883, any doubt as to what is the MS. reading of any word in Widsith. So far from editors having drawn, as Mr. Anscombe thinks, from the transcript when they should have been collating the original, they have gone to the original, and ignored the transcript. And it is for this reason that Mr. Anscombe has done a service in drawing attention to the British Museum MS. For though Thorpe's edition of the Exeter Book (1842) was, in the main, accurate, he took very little trouble to record the exact readings of the MS. in places where it was fragmentary: (e. g. in portions of the Ruin, the Husband's Message, and the Riddles.) Such passages he occasionally omitted altogether. Conjectural emendations were made by various editors and scholars with the object of filling these gaps. The later collations of Schipper, and still more of Wülcker, who carefully recorded every letter and fragment of a letter which was visible, showed that many of these conjectures were merely wasted ingenuity. For the odd fragments of words which Wülcker found were sufficient to prove that the reading could never have been what Leo, Ettmüller, Dietrich or Grein had conjectured. But a comparison of the Museum MS. with Wülcker's collations will, I think, show that, when the Museum transcript was made, more was to be read in the Exeter Book than Wülcker was able to see when he made his examination half a century later. Wülcker was, in fact, aware of this, and for two poems in his second volume he has used the "Londoner abschrift der Hs. von Exeter" (II. pp. 217, 280). But the most mutilated poems from the Exeter Book are given in the first and third volumes: |