² See Antiquity, xiv (1940), 64Google Scholar.

³ [Birch, W. de G.] [Cartularium Saxonicum (1883-1893)], 1, nos. 4, 5 and 6Google Scholar: see the Appendix to the writer's ‘English and Gallic Minsters’ in the Trans, of the R. Hist. Soc, 1941Google Scholar. The best MS. source of these three charters is Trin. Hall MS. 1, which has Thomas of Elmham's facsimiles of the charters in a largely archaic writing. Through the kindness of Mr C. W. Crawley in gaining the permission of the College to have Trin. Hall MS. 1 photographed, the writer has been able to obtain further advice about this script: she should, in the Trans, of the R. Hist. Soc., have described the old letters imitated, not as ‘Merovingian’, but as cursive or half-cursive. ‘Merovingian'implies the Frankish form of cursive, the script used on papyrus, whereas the forms of Trin. Hall MS. I approach to a later or half-cursive form, such as that preserved in some ninth-century Anglo-Saxon MSS. on parchment. See infra p. 107, n. 11.

⁴ The donation would have been verbal: the charter, in diploma form, a record made ‘to provide testimony more permanent than that of mortal witnesses’: see Prof. Galbraith's, ‘Monastic Foundation Charters’, in Camb. Hist. Journ., iv, 206Google Scholar.

⁵ Dr R. Lane Poole wrote in 1917: ‘Anglo-Saxon Charters…unquestionably derived their form and structure from Italian models’, ‘Imperial Influences [on Papal Documents’, Proc. of Brit. Acad., VIII, 239]: but he was there dealing with the grants of the last quarter of the seventh century, belonging roughly to the age of Theodore, and passing over as spurious or questionable the early seventh-century Kentish grants. Moreover, when Dr Lane Poole wrote the sentence quoted, he could not consider the derivation of the early Kentish charters from the Frankish collection of Marculf, or the documents behind Marculf, as possible, because he believed Marculf's Formulate to be of a much later date. He accepted Zeumer's dating of Marculf in his edition in the M[onumenta] G[ermaniae] H[istorica], Leges, V, 1886Google Scholar, and wrote (Imperial Influences, 241): ‘This book is believed to have been compiled at about the year 700: it is more likely a little later.‘Zeumer's dating theory has, however, been disproved by Levillain's, M. Léon notable ‘Le Formulaire de Marculf et la critique moderne’, in Bib. de I'École des Charles, LXXXIV (1923), 21–91CrossRefGoogle Scholar. M. Levillain showed that Marculf was a monk of the Paris region, who dedicated his. compilation to the bishop Landri of Paris who granted a privilege to Saint Denis in 654: that Marculf used older charters as forms: and that he had at his disposition, besides the archives of his own monastery, those of the cathedral of Paris and the private collection of Landri himself, which included some cartae pagenses then current in the Paris region and documents from the referendariat of the kings of Paris; Landri, it would seem, had been referendary before he was bishop ( Levillain, , op. cit., 90–1)Google Scholar. The long dispute about Marculf's date was caused, as in the case of other collections, because the nucleus of the original editor received later additions. The altered dating of Marculf encourages the writer to differ from the weighty judgment of Dr Lane Poole in 1917. It affects the fact that he traced no connexion between the opening phrase: Regnante domino nostro jesu Christo in the very early form, Marculf, ii, 17 (see Deanesly, , ‘English and Gallic Minsters’, 67–9)Google Scholar and in Æthelberht's charter to Rochester (Birch, 1, 3), though he wrote of the phrase with obvious reference to this charter: ‘It is found in Anglo-Saxon charters of the eighth century and perhaps very much earlier.’ The altered dating affects too the fact, for knowledge of which and for much other help the writer is indebted to Prof. Galbraith, that Dr Lane Poole looked for the prototypes of Anglo-Saxon charter forms rather in Italian collections, such as the Regesta di Farfa (ed. Georgi, and Balzani, , 1878)Google Scholar, than in Marculf's collection. Nevertheless, the forms of donation in Marculf's book ii, the cartae pagenses, have many striking points of resemblance to the Anglo-Saxon diploma, ii, a formula for an important donation for a monastery or xenodochium, has a long preface, with all the usual scriptural texts exhorting to almsgiving, and a calling on posterity to observe the grant: Obsecro clementissimis regibus tarn praesentibus quamfuturis et omnibus episcopis busque potestatibus ac primatis, etc.: the donor's epistola donationis is not addressed to a recipient, but to so wide a circle. The grant has a penal clause beginning Si quis, involving penalties at the final judgment, and it is confirmed, not by the signum recognitionis of a notary, but by the stipulatio of the donor, or donor and witnesses. It is, in short, a diploma, as are forms ii, 2, 3, 4, 7, and 8. Till M. Levillain established the date of Marculf, there appeared no reason for comparing the cartae pagenses with the Anglo-Saxon diploma. The writer hopes to pursue this subject elsewhere.

⁶ The chief anachronisms in Birch, I, 5 are the use of the annus incarnationis for dating, the phrase Notum sit omnibus, etc., the sentences of subscription, and the place-names (for which see infra p. 107, n. 10). All may be due to the ninth-century copyist. The annus incarna-tionis appears first in English charters when, after 669, the use of the Roman Easter was being advocated by archbishop Theodore and made possible by the passing round of Dionysius Exiguus's Easter tables, which used this dating (see Migne, , Patrologia Latina, LXVII, 495)Google Scholar. It was popularized by Bede.

The phrase Notum sit omnibus tarn presentibus quam posteris is proper to the diploma making known to all that a legal act has been performed, rather than to an epistola addressed to a recipient by name; it does not therefore occur in early papal bulls, which were addressed to individuals, with the exception of letters of commendation and safe-conduct (e.g. M.G.H., Epist. III, 207 (of 630-655), 257 (of 718), 266 (of 722) and 270 (of 723)). The phrase Notum sit omnibus presentibus et futuris is regularly used in the letters of the emperor Louis the Pious (ed. Zeumer, in M.G.H., pp. 288Google Scholar, no. 1; 289, no. 3; 291, nos. 5, 6; 293, no. 9; 294, nos. 10, 11; 295, no. 13; and cf. pp. 296, 297, 299, 300, 304, 305, 312). The address has, however, a long Merovingian ancestry of phrases that occur frequently in the prefaces of Frankish diplomas and royal writs, one of which may have been modernized by the ninth-century scribe. The bishops of Gaul wrote to pope Leo in 450: Omnibus etenim Gallicanis regionibus notum est sed nee sacrosanctae ecclesiae Romanae habetur incognitum quod, etc. The normal Merovingian phrase is Omnibus non habetur incognitum qualiter, as in Marculf's, Formulare, 80Google Scholar, no. 9; 93, no. 29; 96, no. 35, etc. In the Carolingian additions to Marculf's Formulare is found: Rex omnibus agentibus vel cunctis fidelibus nostris, etc., 120, no. 20. The Formulae Turonenses, in the same volume, has the phrase slightly changed: Dum et omnibus habeturpercognitum qualiter, 148, no. 24; 151, no. 28; 152, no. 29; 155, no. 34, etc.; while the Cartae Senonicae has Omnibus non habetur incognitum, id. 206, no. 50; 217, no. 12; 232, nos. 11, 13, etc.

The phrase Notum sit, etc. in Birch, I, 5 is probably a modernization of the Merovingian Omnibus non sit incognitum. Since early Anglo-Saxon grants approach in form the continental private charter, and not the Merovingian royal writ, it is natural that this universal form of address scarcely occurs among them: though in an early grant issued by a referendary it may well have occurred. From the ninth and tenth centuries, however, the influence of the Frankish imperial formula is shown on Anglo-Saxon charters: cf. Birch, 1, 383, of 825; 1, 409, of 833; II, 595, of 901 (praesentibus, absentibus, posteris cunctis); II, 608, of 904 (omnibus…notum ac manifestum constai); II, 736, 761, 778, 787; in, 860, of 948 (universis…non solum praesentibus sed etiam futuris…quocirca notum esse volo omnibus); III, 872, of 948 (uni-versis…tarn praesentibus quam posteris); III, 885, of 949 (rex omnibus meis notum volo esse fidelibus); in, 996, 1001, 1004, 1030, 1044, 1050, 1057, 1059, 1264.

The sentences of subscription in Birch, 1, 5 would appear to be not earlier than the ninth century and the work of a copyist. They are more varied and elaborate than the sentences on Clovis II's privilege of 654, which run subscripsi, iussus subscripsi or consentiens subscripsi, with the name, rank, office and signum recognitions of the witness. Those who study the reproduction of this grant in [P.] Lauer and [C] Samaran, [Les diplomes originaux des Mérovingiens (Paris, 1908), pi. 6Google Scholar, ] will agree that no later scribe, faced with the task of copying such a papyrus grant, is likely to have tried to reproduce the casual, irregular spacing of the signatures at the foot of the document, although such a scribe normally copied the entries roughly from left to right, getting the signatory in the middle at the foot of the grant in the centre of the list of witnesses, now arranged in one or two columns. This is the case with the referendary's signature in Birch, I, 5. As to the modernizing of the sentences of subscription, or the supplying of sentences where none existed, no scruple would have been felt, because even in Anglo-Saxon original grants the scribe, as far as can be told from surviving originals, both wrote the name and the sentence and made the signum; we have no extant Anglo-Saxon original contemporary with the Frankish grant of 654, where the signatures are largely in the signatories'own, differing hands. The early forms in Marculf provide for stipulation, or the touching of the signum by the witness, but when the deed was the record of a verbal act made later, even stipulation could not have taken place. The sentences of subscription are slightly more elaborate than those of archbishop Wulfred's grant of 811 ([ Bond, E. A.,] Brit. Mus. Facsimiles], 11, 11)Google Scholar, where, though the grant is apparently an original, the sentences, names and signa of the witnesses are all in the same scribe's hand.

⁷ See frontispiece for Elmham's ‘facsimile’ of this charter in Trin. Hall MS. 1, f. 22, and James, M. R.' Descriptive Catalogue of the MS.S. in the Library of Trinity Hall, 2Google Scholar. The charter also occurs in a thirteenth-century register of St Augustine's, Cotton MS. Claudius D. X, f. 8 (for which, and other registers of St Augustine's, see Holtzmann's, W. Papsturkunden in England, 1, i (1930), 82Google Scholar, etc.); in Harl. MS. 358, f. 47b, as. one of a collection of charters and papers made in the late sixteenth century; and in Harl. MS. 686, f. 26, a sixteenth-century copy of Trin. Hall MS. 1. The latter MS., with its facsimile of an archaic hand, appears therefore to be the best authority; it has also the charter Birch, 1, 4, in a similar hand, but without the list of witnesses, and the charter id. I, 6, to be rejected as fictitious, with the list of witnesses. Birch added the list of witnesses to his no. 1, 4, which purports to be Æthel-berht's first grant to St Augustine's, and which ends in the Trin. Hall MS. at indictione viii, without any sentence of subscription or witnesses; the association of the witnesses with this charter had apparently been made before Edward II granted an inspeximus and confirmation for it, dated 18 July 1326: see Cal. of Charter Rolls, III, 492. In the case of two charters, both of which must originally have been written on papyrus, it is not possible to reject the possibility that the list of witnesses survived as a mere fragment, or in a deed in very damaged condition, from either; but the fact that the list is so associated in the best MS. authority with Birch, 1, 5 and not with Birch, 1, 4: that Birch, 1, 4 does not occur in Claudius D. X or in Thomas Sprott's Chronica (of c. 1272) as used by W. Thome (see Hist. Ang. Scriptores, ed. Twysden, R., col. 1761)Google Scholar: and that, of the words of donation, Birch, 1, 5 appears to use an earlier form (tradidi deo per antistitem would appear earlier than dabo et concedo in honore sancti Petri): points to the association of the witnesses with Birch, 1, 5 rather than with Birch, I, 4.

As to Birch, I, 6: the reasons for regarding this charter as fictitious, the work perhaps of an eleventh- or twelfth-century scribe, are as follows: (1) The uncials in which it is written, like those of the bulla plumbea of Augustine in Trin. Hall MS. 1, bespeak a scribe who regarded the uncial script as very old and suitable for his purpose, without realizing that in the case of both grant and bulla the originals must have been written in cursive on papyrus. Though Canterbury had several very old MSS. in uncials, they were all parchment codices, solemn treatises. The uncial hand developed on parchment and is rarely found on early papyri: see Chatelain's, Uncialis Scriptura Codicum LatinorumGoogle Scholar, and Lowe, E. A., Codices Latini Antiquiores, pt. 1, 1934)Google Scholar. (2) The scribe, ignorant of the fact that a referendary was always a layman, a super-notary, described Angemund the referendary in the text as Angemundum presbyterum. (3) The deed has no information which could not have been obtained from other sources. It notably includes a description of the gifts sent by Gregory I to Æthelberht, referred to in Bede's, Hist. Ec, i. 32Google Scholar and ii. 10, and it is attested by personages suggested by a reading of the Hist. Ec, namely, the bishops Mellitus and Justus and the priest Lawrence, as well as by the list of witnesses annexed from Birch, I, 5, or its prototype.

To the list of witnesses Claudius D. X, f. 9 added, as final signatory, Aldhunus. The name i s sufficiently common to render identification difficult, or rather, the accounting for the name n i the list difficult. Its occurrence among the lay signatories rules out any intention to connect the list with the Aldhunus who was abbot of St Augustine's from 748 to 760 (see, for Aldhunus, , Hist. Ang. Scriptores (ed. Twysden, R., 1642)Google Scholar, vol. n, col. 1772, and Monast. 1, 120). It is possible, however, that the Aldhunus referred to is the layman mentioned in archbishop Æthelheard's grant of 805 (Birch, 1, 319) as having in the past given land to Christchurch: he is called homo bonus, qui in hoc regali villa illustris dvitatis praefectus fuit. He had an official position, and he might have signed what Giry calls an acte récrit: an appennis or pancarta (see Manuel, 1894, 12–15)Google Scholar. The name does not appear in either list of witnesses in Trin. Hall MS. 1.

⁸ The form Pinca, as an Anglo-Saxon or Frankish proper name, presents great difficulties. On the other hand, the ranks and offices of court officers among Franks were, in the seventh century, written in contracted form, with three or four letters only of the word, and it is not difficult to suppose that an Anglo-Saxon copyist, ignorant of the office of pincerna, might write the contracted PINCA as a proper name. Cf. the contractions in the ranks of the signatories to Clovis IPs grant of 654 in Lauer and Samaran, pi. 6 and 6 bis, especially Viro illustri Aigulfo COM(iti) PAL(acii).

⁹ For a list of Frankish referendaries, as well as a description of the office and the use of referendaries by other courts, see Breslau, H., Handbuch der Urkundenlehre fur Deutschland und Italien (Leipzig, 1912), 264–72Google Scholar.

¹⁰ See ‘Kentish Place Names’, Wallenberg, J. K., in Uppsala Universitets Ársskrift, 1931, 5Google Scholar.

¹¹ See Giry, , Manuel, I, 12Google Scholar, and Bernard's, A. de Manuel de Diplomatique, I, 186ff.Google Scholar, for reference to which I am indebted to Dr C. R. Cheney. The co-existence of archaic letter forms and later interpolations in the same grant is not impossible; Giry, II, 740, writes: ‘Ce qui est surtoiit difficile, e'est de distinguer un original d'une copie figurée contemporaine ou de peu postéieure, susceptible par consequent de contenir quelques interpolations.’ Æthelberht's grant had a very long history, and the fact of its at least partial survival, rather than its not surviving intact, is remarkable. It must have been a papyrus deed (see infra), for papyrus at the date was universally used for loose deeds, and for some books: parchment, as far more expensive, was only used for gospel books and solemn treatises. Papyrus survived often in imperfect form, or in mere fragments, whereas parchment, if it survives at all, survives whole. It is conjectured then that at least the witnesses'names at the foot of the papyrus (for they would seem to fit into no other period and milieu than those of Æthelberht's court), and in all probability words connecting the deed with St Augustine's (for in what other circumstances would the addition of a list with a referendary be thought to add authenticity to the text?), were copied from papyrus to parchment, a process that modifies the letter-forms; that this transfer may or may not have been made on incorporation into an early register. Whether there was an intermediate scribe or not, some scribe writing not earlier than the mid-ninth century copied the deed, using some half-cursive forms, already then archaic.

As to the script of Elmham's copie figurée: Mr Neil Ker has very kindly examined the photograph of Birch, 1, 5 for me, and writes: ‘I think the kind of script which is being imitated is that seen in Bond's, E. A. Brit. Mus. Facsimiles, i, 5Google Scholar and 11, 11. The latter is dated in duitate dorouernia A.D. 811 and is, I take it, a contemporary copy. The former refers to s. viii inc. but was not copied much if at all before 800.’ This document (Brit. Mus. Facs., 1, 5) is a letter of between 693 and 731 addressed to archbishop Berhtwald; it must have been originally written at the time when parchment had not long superseded papyrus for diplomatic usage, and the script, even as copied shortly before 800 (according to Mr Ker's dating), is of a late cursive, or half-cursive character. The a is sometimes open, sometimes closed; e and r have lost their excessive height as cursive letters, but et, g, s, r, and f are cursive in origin, and i is sometimes tacked on (cf. those in Elmham's script). These may, in Brit. Mus. Facs., I, 5, be merely natural developments of the cursive script; but in some cases there was a conscious copying of outmoded forms. Mr Neil Ker writes: ‘One does find script imitations of eighth and ninth century documents made in the s. x and s. xi (?), for example, Brit. Mus. Facs., 11, 35; ii. 36 (of “862”), and, I think, Brit. Mus. Facs., 11, 3 (“778”). And one finds in the twelfth-century Winchester cartulary, Brit. Mus. Addit. MS. 15350, imitations of tenth-century writing where the intention is the same as in Thomas of Elmham: to confirm the. authority of the copy by showing that what was being copied really had ancient letter forms.” For the archaic forms in Elmham's script, see infra p. 109, n. 14.

¹² See Deanesly, , ‘English and Gallic Minsters’, 26Google Scholar.

¹³ The papal scrinium not only went on using papyrus, but it continued to use a development of the cursive hand as long as papyrus itself was used for bulls, and even when, as Giry says, n, 669, it must have been quite illegible to most of the faithful. Cf. Poole, Lane, Imperial Influences, 245Google Scholar, on the curial hand in the tenth century: ‘the documents were still written in the curial handwriting, an artificial and highly intricate development of ancient cursive.’ Under pope John XIII, 965–972, the Datum began to be written in the Caroline minuscule: ‘the distinction in handwriting between the document itself and the Datum was retained till the curial hand itself silently died out in the twelfth century.’

¹⁴ See Liebermann's Notes on the Textus Roffensis, reprinted from Arch. Cantiana, 1898Google Scholar. As to the transmission of the half-cursive hand which Elmham copied: it may be observed that the rapid or slow supersession of cursive forms was connected with the supply of papyrus. It was obtainable in Italy, and the notaries of the papal scrinium used papyrus and the cursive hand till the eleventh century. Among the Franks, it was in general use till the last quarter of the seventh century (see Giry, , Manuel, 404)Google Scholar, and the cursive hand was transferred (with the necessary modifications) to parchment: it was only superseded by the Carolingian hand, one proper to parchment, towards the end of the eighth century. Loose papyrus was, apparently, never obtainable in Ireland, though papyrus codices must have been; we hear of the Irish using waxed tablets and papyrus books together in their schools at Tuain Drecain in 638 (see Prof. Macneill's study of Cenn Faelad and Irish learning at the date in Studies, an Irish Quarterly Review, 11 (1922), p. 17)Google Scholar; the use of waxed tablets suggests knowledge of the cursive script. The Irish, apparently, evolved their half-uncial hand, a hand written with a straight-cut, nibbed (reed or quill) pen, early; there are, however, traces of cursive influence in it. The Anglo-Saxons occupy a position intermediate between the Irish and the Franks: they must have had more cursive books or deeds than the Irish, and less than the Franks, because they sometimes used a half-cursive hand on parchment, after adopting, for the most part, Irish letter forms.

As to the particular cursive forms in Elmham's script, which are, of course, not uninfluenced by his own hand, they include the open a, used regularly, the et, the r, the s, and the ligatures with i tacked on. The f and g are also archaic and cursive in origin, but indistinguishable from those of the Irish hand, here also influenced by the cursive (see Steffens's, Paldographie Lateinische, plates 19, 20, 21, 22, 23, 24)Google Scholar. The open a is found in the Ravenna papyrus of 572 and notably in the Merovingian script (see id., pis. 22 and 24, of about the same date, and pi. 25 for the open a in seventh-century Merovingian and Italian cursive hands). It crops up also however (not regularly or frequently, but sporadically) in eighth- and ninth-century MSS. written both in Mercia, (Brit. Mus. Facs., l, 10Google Scholar) and Kent, as do other cursive letters. Elmham's curiously written et has a cursive ancestor and is found in the Ravenna deed (id., pi. 19, and the half-cursives of plates 20, 21, 22 and 23); the different form which these two ligatured letters take in the Irish-Anglo-Saxon half-uncial of pi. 24 is clearly due to the impracticability of making the long, sloping stroke of the e with the straight-cut Irish nib. Elmham's r had a cursive ancestor and is markedly not Irish (see id., plates 19, 20, 21, 23, 24); this is true also of his s (see id., plates 19, 20, 21, 22, 23, 24).

¹⁵ It is of interest to note that the facsimile grants of Thomas of Elmham are in no way similar to these forged Merovingian grants at Saint-Denis, dealt with by Levillain, M. in the Bib. de I'École des Chartes, LXXXVI (1926), 255–7Google Scholar. He describes these as pseudo-originaux, written on the backs of authentic ancient acts: ‘ils sont, tous les huit, d'une écriture contre-faite, qui s'efforce d'imiter l'écriture mérovingienne des chartes pour le diplôme de Dagobert i et la charte de Landri, et l'écriture curiale pour les autres [papal bulls].’ That is, these forged Merovingian documents are copies of a homogeneous seventh-century script, whose falsity M. Levillain, among others, detects, not through the use of obviously later non-Merovingian forms, but the careless formation of certain Merovingian ones. ‘Cette écriture déguisée trahit la main d'un seul et měme scriptor, qui par distraction ou par lassitude, laisse quelquefois sa plume tracer quelques lettres cursives qui le montrent accoutumé a se servir de la minuscule Caroline.'After discussing the difficult question (257) of whether the monks of Saint-Denis intended to pass off these documents as originals, or to present them simply as ancient copies, M. Levillain decides that the facts that they were tendentious in character: were copied on the backs of genuine papyrus documents, in a ‘more or less happy’ transcription of the archaic script of certain genuine originals at hand: were, in the case of the forged papal bulls, furnished with seals from genuine bulls of the pope in question: that all these facts show that the intention was to pass them off as originals. ‘Nos pieces d'archives sont done bien des pseudo-originaux et non des copies figurees a preventions anciennes.’ Elmham's documents, on the other hand, do appear to be copiesfigurees;they are not tendentious in character, being merely land grants, and their script and contents are not homogeneous, but contain elements of very early and of much later date.

¹⁶ Dr Previté-Orton kindly informs me that vetustas laudabilis would have this sense, of approval after examination.

¹⁷ See the writer's ‘Canterbury and Paris in the reign of Æthelberht’, History, XXVI (1941), 103Google Scholar.

¹⁸ It is an earlier form of the name Aunemundus: cf. the bishop who signed the privilege of Clovis II of 654 as Aunemundus peccator subscripsi (Lauer, and Samaran, , pi. 6)Google Scholar. This s apparently the Aunemundus or Dalfinus, archbishop of the Gauls, with whom Wilfrid stayed at Lyons: see Plummer, C., Hist. Eccles. Gentis Anglorum, 2 vols. (1896), 1, 182, 324Google Scholar; 11, 321.

¹⁹ So Grimm, 1854 ed. of the Wörterbuch: the section with Graf in the new ed. is not out. The derivation has been disputed and the word, written graphic, has been taken to mean notarius; but the equation of grafio with comes in early docs, renders the Germanic derivation preferable. See Brinckmeier, E., Glossarium diplomaticum, 2 vols. (1856-1863): GrafGoogle Scholar.

²⁰ See Deanesly, , ‘Canterbury and Paris in the reign of Æthelberht’, History, XXVI (1941), 99Google Scholar.