QUELLEN UND FORSCHUNGEN ZUR SPRACH- UND CULTURGESCHICHTE DER GERMANISCHEN VÖLKER. HERAUSGEGEBEN VON ALOIS BRANDL, ERNST MARTIN, ERICH SCHMIDT. Lxxvm. PLACE NAMES IN THE ENGLISH BEDE AND THE LOCALISATION OF THE M8S. STR ASSBURG. KARL J. TRÜB SEK. 1896. P L A CE N A M ES IN THE ENGLISH BEDE AND THE LOCALISATION OF THE MSS. BY THOMAS MILLER. STRASSBURG. KARL J. TRÜBNER. 1896. O Otto's Hof-Buchdruckerei in Darmstadt. INTRODUCTION. This examination of English Place Names in Bede's Church History was undertaken with a definite purpose, which has prescribed very narrow limits. It is an attempt to ascer- tain from the spelling in the Mss the range of local know- ledge possessed by the scribes. It was accordingly necessary to settle from contemporary records the form of the word locally in use and, where needful, the date of the use. For the object in view the necessary documents were, besides the Latin and Anglo-Saxon Versions of Bede, only the A.S. Chronicle, the charters, and the few names scattered in Oro- sius and iElfric. The modern names have been added, some freshly identified, but most according to the view generally accepted. It would have led too far afield to have included more of the literature bearing on the subject. As it is, the bulk has outgrown the compass of an introduction to the second volume of the edition of the Old English Bede, for which the paper was originally planned. It is now published separately, as it may be of interest to have presented at once the materials offered in the documents mentioned above. Every scholar has felt the difficulty of assigning date and place to Anglo-Saxon Mss. "Where possible, the effort should be made to localise Mss. on other than mere dialectical grounds, as an Ms. once localised becomes the starting point for fresh dialectical investigation. It is assumed that a scribe will be likely to give cor- rectly the names of familiar localities, and to trip over those QF. LXXVIII. 1 — 2 — less well-known, or to assimilate their forms to those in his own neighbourhood. It is also taken for granted, and is abundantly demonstrated, that the ancient names will be seen shrinking with age and tending to their present shape. False etymologies have perverted the shape, and literary ingenuity has exhausted itself in variations, as in the case of Ely and Malmesbury. Some of Bede's derivations are clearly wrong, or their immediate application dubious, as in the case of Streoneshalh, Heofonfeld, Selsey. Where it is necessary for purposes of localisation, an attempt has been made to map out the distribution of certain forms, like stede in connection with Medeshamstede, and ea, eg, ig, the names for 'isle' or 'eyot', which latter is treated at some length under Ceortes eig, and has been been separated from the now lost ge (perhaps surviving in yeoman) in course of the discussion on Lsestinge, Lindissi, 'Surrey' etc. The existing Mss of the Old English Bede are five in number, two only being complete B. and Ca., two others T. and 0. being defective at the beginning and end, and T. having also lost some intermediate leaves. The fifth survives only in fragments, and is known as C. since the time of Wheelocke and Smith, who consulted it before the disastrous fire of 1731. All are fully described in the Introduction to the first volume of the edition published by the Early English Text Society in 1890. For the purpose of comparison the Latin names, as written in the oldest Mss of Bede, have been given in each case first. These Mss are noticed in the Introduction just mentioned, but are fully described by Sweet, The Oldest English Texts 1885, from which work the various forms have been copied. M. was written about 737, I and II later in the same century; but M. is Northumbrian, the others come from localities further south. Another Ms. N., of the same century, is rarely cited., as it was written on the Continent and is not an authority at first hand. M. is always cited first, and without addition; the variations in other Mss are marked by their letter. The Mss. of the A. S. Chronicle aid in settling questions — 3 — of time and place. Ms. A. would be invaluable in deter- mining the currency of forms, if a competent palaeographer would undertake the task of settling the succession and dates of the hands. He would be aided by the copy made in the 11th c. Ms. G., which editors have not fully utilised, and in a less degree by Ms B. I have assumed that the hands in A. are contemporary from Alfred onwards, and that the Mss. have their localities correctly assigned, as in the list at the end of this introduction. Ms G. is however not a Canterbury copy made by a Kent scribe. Later on it is shown that the variations of a and o before nasals determine locality roughly. Now a is Kentish, and for o of Ms. A. the Canterbury B. writes a. There are only rare exceptions as 661, 897, 910, 918, 919, 975 and a few others. But G. while freely altering the Old West-Saxon spelling, especially as regards ie, gives o the preponderance up to 891 (A's first hand), and in the 10th century occasionally writes o where A. has a. This Ms. was in the same volume as Ms. C. of Bede. But 0. E. B. C. throughout gives o a great preponderance. The difference may be in part due to C's original, as undoubtedly G. varies with A., but the tendency in both is pro tanto a proof of their scribes not having come from South of the Thames line. But A. S. C. B. may fairly be assigned to Canterbury, at least on this score. The other Mss. of the A. S. C. are, with the exception of F., sometimes in agreement with A. as against B., but a prevails. In the Charters I have felt obliged to accept the dicta of the editors. A thorough examination of the dialect in the Chartularies, and a precise indication as to the appearance and disappearance of forms will in the end enable us to sift out the later copies. At present "late West Saxon" is a vague elastic term. Later on I shall speak of one or two criteria of date. In the Introduction to 0. E. Bede vol. I, it was shown that Mss. T. and B. must have sprung from one original, and the other three C. 0. Ca. from a second. But the rela- tionship between T. and B. is far more remote, than is the case in the second group, where C. and 0. are independent, l* — 4 — but closely akin, while Ca. is a transcription of 0., as Ms. G. is of Ms. A. But the grouping by spelling gives very different results. T. and B. stand at the different ends of a series, in which T. and C. come first. Date is of some influence here, but much more locality. In the following summary, details are omitted, as they are fully given later under the place names. In the spelling of 'Britain', B. is steadily Southern, the others Midland; but as to 'Briton', while T. is steadily archaic and presumably Midland, B. O. Ca. are later and Southern, C. in this latter case is defective. The word 'Angle' is by T. C. spelt on-, by B. an-; 0. has o : a = 1:4; but Ca. 1:9. Here date may partly influence Ca. The word 'Northumbrian in the A. S. C. appears early with Nordan-, later with Nord-, as first member. NorCt- is Alfredian and Southern. T. C. 0. Ca. maintain NorSan-, B. has NorS-. 'Mercian'. In A. S. C. A. the first hand (old West Saxon) has Mierce; then in the time of the 'lady of Mercia' (died 918), we find Merce, but from 922 on Myrce, which is the Southern form in all Mss. G. however occasionally writes Merce for Mierce of A. — a proof of Midland affinities. In 0. E. B., T. hais always e, B. y, C. has e : y = 11:1; 0. 23 : 19 (e often erased), but Ca. has the proportion reduced to 6 : 54. The conclusion is that T. is consistently Midland, B. not less steadily Southern. C. has more affinity for the Mid- lands than the South, 0. is rather more Southern, Ca. still more. But then Ca. is the latest of the Mss. 'London'. — Under this word it is shown that the trans- lator did not use the combination current in Alfred's day, as exhibited in the A. S. C. and in charters. The Archetype is followed by all Mss. Besides this B. makes a strange blunder, converting London into an 'island city': see under Thames'. Nor was the neighbourhood better known, see under Tilaburh, in Bercingum and Pente.