The Stowe Missal: Warner’s Introduction

The Stowe Missal: Warner’s Introduction

Stowe Missal (Irish Mass-book)

c. AD 800

George Warner’s introduction to his edition of the Stowe Missal. It describes the manuscript and its shrine, how it came to Stowe and then to Dublin, its scribes and the later hands that corrected it, its date and the monastery of Tallaght where it was written, and the sources of its prayers.

1 The text here printed is that of the oldest Mass-book of the early Irish Church known to have survived, and is intended to accompany the collotype facsimile of the MS, which has already been issued in a separate volume. Incongruous as it may seem that it should take its title from an English country seat, the Stowe Missal is so called, not with any reference to its origin, but merely from the fact that for a few years it was in the library at Stowe House in Buckinghamshire, formed early in the last century by George Grenville, first Marquess of Buckingham, who died in 1813, and Richard his successor, afterwards Duke of Buckingham and Chandos. Its existence there was first revealed in 1819, when a lengthy, but not altogether accurate, article was devoted to it in an Appendix to vol. i of the privately printed Catalogue of the Stowe MSS., by Dr. Charles O'Conor, and as no mention is made of it in the Catalogue itself, which is dated only a year before, it was probably a quite recent acquisition. Without explaining how it came to be at Stowe, O'Conor states that it was discovered somewhere on the Continent by "the late John Grace, Esquire, of Nenagh in Ireland, who was formerly an officer in the German service"; but unfortunately no particulars of the precise locality and circumstances of its discovery had been recorded, and nothing is known of its peregrinations before it was thus rescued from oblivion.

2 From a pedigree in Memorials of the family of Grace, by Sheffield Grace, privately printed in 1823, it appears that a John Grace, captain of carabineers in the Imperial service, died at the siege of Belgrade in 1789; and there was also a John Dowell Grace, captain of Würtemberg dragoons, who retired from active service in 1776 and died in Ireland in 1811. The latter is described in the Memorials as of Mantua House, co. Roscommon, and it is uncertain which of the two, if either, was the John Grace to whom O'Conor refers. There is little doubt, however, that the second Marquess of Buckingham obtained the MS. directly from some member of the Grace family, with which he was connected through his wife. As he was on friendly terms with Sheffield Grace, nephew of the John Grace who died at Belgrade, it perhaps came to him by gift or purchase from that quarter.

3 When the contents of Stowe House, then belonging to the second Duke of Buckingham and Chandos, were partially dispersed in 1849, the Missal was catalogued with the rest of the MSS. for public auction. Ultimately, however, the whole collection was sold privately to the fourth Earl of Ashburnham, and was removed to Ashburnham Place in Sussex, where it remained intact, and still distinguished as the Stowe Collection, until 1883. By a vote in Parliament it was then purchased from the fifth earl for the nation and was thereupon divided into two portions, the majority of the MSS., which were chiefly of historical interest, going to the British Museum, while those written in the Irish language or otherwise connected with Ireland were handed over to the Royal Irish Academy at Dublin. The Stowe Missal, with the metal case within which it is inclosed, of course went to Ireland, and under the press-mark D. II. 3 in the library of the Academy thus at length found a permanent resting-place in the country from which it originally came.

4 The full contents of the MS. comprise the following articles, all of which, except the last two, are in Latin:— 1. Extracts from the Gospel of St. John. f. 1. 2. The Ordinary and Canon of the Mass, followed by a few special Masses. f. 12. 3. The Order of Baptism, with the Communion of the Newly Baptized. f. 46v. 4. The Order of the Visitation of the Sick, with the Administration of Extreme Unction and Communion. f. 60. 5. A short treatise in Irish on the Mass. f. 65v. 6. Three short spells in Irish. f. 67v. 5 While it was still at Ashburnham Place the liturgical portions (artt. 2-4) were edited, for the first time, by the Rev. F. E. Warren, The Liturgy and Ritual of the Celtic Church, 1883, pp. 198-268; and in 1886, after its removal to Dublin, artt. 2 and 3 were again edited, together with the text and a translation of the Irish treatise on the Mass (art. 5), by the Rev. B. MacCarthy, D.D., in a dissertation "On the Stowe Missal," printed in the Transactions of the Royal Irish Academy, vol. xxvii, 1885, p. 185. A later edition of art. 5, with an independent translation, was included by Whitley Stokes and Strachan in their Thesaurus Palaeohibernicus, 1903, vol. ii, p. 252, where it is accompanied, pp. 250, 251, also with translations, by the Irish spells (art. 6) and the brief Irish rubrics interspersed in the liturgical text. In the present volume a literal transcript is given of the whole of these articles (2-6), the extensions of the abbreviated forms being printed in italic type and the additions to the original text in smaller type. A translation of the Irish text, borrowed by permission from the Thesaurus Palaeohibernicus, is added in an appendix. For reasons stated below, the extracts from St. John's Gospel (art. 1) are omitted. Reproductions of three pages, are, however, given in Plates VII-IX, and a collation of the text with that of the Codex Amiatinus will be found in a paper by the Rev. J. H. Bernard, D.D. (now Bishop of Ossory), "On the Stowe St. John," in the Transactions of the Royal Irish Academy, vol. xxx, pt. viii, 1893.

6 As will be seen from the facsimile, the volume is of small dimensions, containing in its present state sixty-seven leaves of vellum, which measure 5 5/8 inches in height and 4 1/2 inches in width. The vellum is normally rather thick and in some parts is much soiled and darkened by age and wear. The binding is of wooden boards covered with discoloured vellum and in addition round the three outer edges with strips of kid-skin ¾-inch wide stained red; its back is of modern dark brown leather. Owing to removals and insertions the present composition of the five gatherings of which the volume is made up is somewhat irregular. The first, which contains the extracts from St. John, consists of eleven leaves (ff. 1-11), but another has been torn away at the beginning, leaving only a narrow strip along the inner edge, on the verso side of which are slight remains of a border resembling that on f. 11v (see Plate IX). The second gathering, which now contains seventeen leaves (ff. 12-28), consisted originally of the ten leaves numbered 12, 13, 15-17, 20, 21, 26-28, together probably with two others between 17 and 20, now replaced by the present pair 18 and 19. Fol. 14 is an inserted narrow slip, 2⅞ inches in height at most, its blank inner edge projecting ⅞-inch between ff. 26 and 27; and the gathering of four leaves numbered 22-25 is also an insertion. The existing arrangement will be seen at a glance in the following diagram, where the dotted lines indicate the inserted leaves:—

7 The third gathering contains eighteen leaves (ff. 29-46), of which fourteen, viz., 29, 32-34, 37-46, are original, while, as shown below, two pairs, ff. 30, 31, and ff. 35, 36, are interpolated. The fourth gathering contains twelve leaves (ff. 47-58), all of which are original and regularly arranged; and the fifth is made up of the nine leaves numbered 59-67, together with a tenth, blank and unnumbered, which is pasted down to the cover. For the text of the Missal and of the other liturgical matter in its original form lines were ruled with a hard point on one side only of the vellum, two for each line of text, which was written with more or less regularity between them. Single or double vertical lines were also ruled on the left and right, but little attention was paid to the latter by the scribes, the text being as often as not carried beyond them. The number of lines of text on a page varies. Excluding, it must be understood, for the present all interpolated matter, on ff. 12-28v, 47-49, 50v-64, 65, there are thirteen lines; on ff. 29-46v there are fourteen; and on ff. 49v, 50, 64v there are twelve. These differences correspond to some extent with changes of hand. Thus the first hand (A) ends (on the last page of the gathering) at f. 28v; a second hand (A) writes the next gathering (ff. 29-46); and a third hand (A), with more ornamental initials, extends from f. 47 to f. 51v. After this the hands are more difficult to distinguish, but ff. 52, 52v were perhaps written by A, and ff. 53-64 by a fourth hand (A), while another (A) certainly begins on f. 64v and continues to the end (f. 65).

8 The character of the original script is Irish minuscule, with a tendency to be angular rather than rounded. In all the hands there is one exception to the prevailing minuscule forms of letters, the majuscule or capital R being almost invariably used. Three instances of a minuscule r, shaped like the Saxon letter, almost like n, may be seen on f. 33v, l. 1, deprecamur, l. 5, donare, and f. 65, last line, ordo. There is also a single instance of a majuscule N on f. 54, l. 2, fons, where N and long s are combined. With regard to the formation of the other letters a few points may here be noticed. Thus a is formed like u, sloping slightly backwards, with the tops connected by a fine transverse line. The second stroke, especially in hand A, is often higher than the other and sharply pointed, and the connecting line cuts it below the top. In a few cases (f. 12v, l. 10, f. 52v, l. 13) a following m is subscript.

9 The first stroke of b is bent, and is thickened at the top by an angular spur-like projection on the left. The latter feature also characterizes the other tall letters, d, h, l, and more or less all vertical strokes. c consists of a backward-sloping straight stroke turning up sharply at the bottom, with a separate curved top affixed to it, drawn from left to right. The bow of d closely resembles c, and the stroke on the right is nearly always vertical. It is very rarely bent round almost horizontally to the left (e.g., f. 38, l. 7, deum), as is the case with a capital (see f. 54v), but a sloping variant from the normal shape sometimes occurs, as on f. 15v, l. 11, mundo, f. 17, last line, deprecemur, etc., and more commonly in the smaller script of the rubrical directions (e.g., f. 58v). e is formed on the same principle as c, the rounded top and the cross-stroke being affixed separately. The top is frequently raised above the line of writing, the cross-stroke being utilized in making the following letter. At the end of a word, and occasionally also elsewhere, the cross-stroke is long, thick and wavy. The diphthong æ is commonly found, but ae, ϵ (e.g., f. 47v, l. 7, diuine) and e (f. 17, l. 12, catholice) also occur in its place.

10 f is composed of a vertical stroke carried only a little below the ruled line, with a curved or wavy line, from left to right, at the top and a cross-stroke low down, along the ruled line. The down-stroke of g springs from the centre of the horizontal top; it is only slightly curved to left and right, and usually ends in a small knob. i after m, n, u is often subscript, being attached to the final stroke of the preceding letter and curved round to the left, something like the bow of 3 (see f. 13, ll. 8, 12, f. 32v, l. 4, etc.). The same is the case after c and long s, where the i is attached to the curved top of the letter. The curve of p is generally rather pointed at the top and ends at the bottom in a knob. q is open at the top, with a curve above the first stroke (see f. 15v, l. 1, and passim).

11 s is of the long form, resembling f without the cross-stroke; but at the end of a word, instead of a convex top above the line, there is a small wavy stroke on the right of the vertical stroke, making the letter somewhat like r (see f. 39, end of lines, and for the same form in the middle of a word, f. 41v, l. 5, ipse). In one case (f. 45v, l. 4, esse) the form s is used. The cross-stroke of t extends well both to right and left of the main stroke and is generally somewhat heavy. As a ligature with e in et the letter is smaller and differently formed (f. 12v, l. 2, et; f. 13v, l. 2, eterne; f. 20, l. 2, habet). The characteristic forms of x, y, z all occur on f. 12v, l. 4, that of y being especially noticeable. The letters f, p, s, and more rarely r, are sometimes written with a flourish on the left, making an acute angle with the main stroke at the bottom (see ff. 12v, 16v). In some cases these forms represent initial capitals, but they are also used for minuscules, and in the case of s are even found at the end of a word (f. 41v, l. 6).

12 Most of the other initials of collects, etc., are merely the minuscule forms on a larger scale, but N in IN is used (ff. 42, 45, etc.). The form of d as an initial is invariably that in which the vertical stroke is bent back to the left horizontally; instead of an oval it sometimes has an oblong shape (ff. 39v, 41v, etc.), and the same is also the case with o (ff. 21, 45v). On ff. 13v, 47-51 there are more ornamental initials, with spirals, interlacing (ff. 47, 48), a human head (f. 50, see also f. 12v), and dogs' heads (ff. 13v, 48). Abbreviations are frequent and, besides the usual nomina sacra, include the following :— ā, aut. h̄, autem. b, bene. c̄s, cuius. c̄, cum. d̄cs, dicens. dī, dicit (f. 57v). d̄x̄, dixit (f. 62). ə, eius. ‡, enim. g°, ergo. ēē, esse. ‡, ʒ, est. 7, et (rare, f. 49, ll. 9, 10).

13 fr̄s, fratres. h̄, haec. h̄, hoc. h̄s, huius. h̄c, hunc. ig', igitur. ī, in. ī̄t, inter. mm, ms, meum, meam, meus. nōis, nōe (nō, f. 49), nōa, nominis, -mine, -mina. n̄, non. n̄r, nost, noster. n̄ra, nostra. nm, n̄rm, nostrum, nostram. n̄i, nostri. n̄rae, n̄ræ, n̄re, nostrae. n̄o, n̄ro, nostro. n̄rōm (f. 13v), nostrorum. nās (f. 13v), n̄s (f. 17), nostras. n̄ris, n̄rs (ff. 12v, 40v), nostris. n̄c, nunc. ōmes, ōms, omnes. ōmi, ōmis, omni, omnis. ōmia, ōma, omnia. ōmim, ōmm, omnium. ōnibus, ōbus, omnibus. ōnipot, ōnipōs, ōnipō, ōmpot, ōmpō, omnipotens. p̄, per. p̄, prae. p̄, pro. p̄i, pri-. p̄o, post. p̄t, praeter. prop̄t, p̄t, propter. q̄si, quasi. q̄:, q̄:, que. q̄i, qui. q̄:, q̄:·, q̄:·, quae. q̄, quod. q̄ (also quā), q̄ (f. 27v), quam. q̄, q̄:, q̄:· (f. 61), quem. q̄s, quis. q̄d, quid. q̄mo, quomodo. q̄m, quoniam. q̄ms, q̄su, q̄us, quaesumus. r̄l, reliqua. f, secundum (f. 20v).

14 s̄, sed. s̄r, supr, super. s̄t, sunt. t̄i, tibi. t̄e, tunc. l, vel. The following symbols for omitted syllables and letters also occur, viz. :— An inverted c for con (e.g., consubstancialem, f. 20, l. 13). A hooked horizontal stroke (¬) for final m and for m before a consonant in the middle of a word (e.g., ambire, f. 13) and in Amen (f. 21) even between two vowels; also for um in pacificum (f. 17). A short horizontal stroke, generally slightly concave, for n and -er (e.g., in, inter, uerbo, pater) and for -un in -runt (e.g., rogauerunt, f. 62). For -um in the genitive plural it is placed, not over r, but to the right of it (e.g., tuorum, f. 27). The same symbol is also used in many of the forms of abbreviation in the list above, as well as in abbreviations by simple suspension of all kinds. These are not uncommon, especially in words of frequent recurrence, as—

15 quesū for quesumus. orā and ō for oramus. p̄ (also p̄s) for presta. apos̄ for apostolos, -orum. mart̄ for martirum. uirḡ for uirginum. archanḡ princip̄ et potest, etc., for archangeli principes et potestates, etc. (f. 39v). respō, res̄, r̄ for respondet. ał for alleluia. sæcū sæc̄, sæ sæc̄ for sæcula sæculorum (ff. 48, 58). pā et fī for pater (-ris) et filius (-ii). leē for lectio. pr̄ iḡ for per ignem. vd̄ for vere dignum. A comma over t, slightly to the right, for -ur in -tur (e.g., largiuntur, f. 17). : ; 3 for us in -bus and elsewhere, as peccauimus (f. 12), uirtus (f. 58), opus (f. 61), etc. An acute accent is placed over monosyllables and frequently over -is of the dative and ablative plural. It is also occasionally found elsewhere, as sollemnitás (f. 38), circumdás (f. 40), confirmá (f. 44), isác, iacób (f. 45v), tribús (f. 62v); and even over a short syllable, as plebís (f. 38v).

16 Punctuation is marked in a variety of ways, by : : : ~ : : : : and, less often, by a full point, generally medial. The spelling is very irregular, but the irregularities are not systematic and the normal orthography in many cases is also found. The following list includes most of the peculiar forms, some of which may be merely clerical errors :— Consonants doubled : missericordiam (f. 13v, and so all derivatives of miser), æclessia (16v), uissiuiilium (visibilium, 20), quessumus (27), accussantes (44), cappillis (47), deffensa (48), tutellam (48v, 59v), sallientes (54), effussione (55), pressens (57v, 60), uissita (61v, 64v), inuassit (63), essurientem (64v). One of two similar consonants dropped : abiso (abysso, 12), emite (emitte, 12v), altissime (13, and so frequently in other superlatives), misa (13v, 38), imbiciles (imbecilles, 15v), aceptos (17v), remisionem (21, etc.), sufragia (38v), gutore (gutture, 47), acipientibus (48v), promisæ (49v), suplices (50), gehinam (gehennam, 50v), suplicii (50v), mitire (mittere, 52v), iusit (56v), acipio (58v), dimite (61v), acesserunt (61v).

17 a for e ... catacominis (catechumenis, 17, 52v), piatas (43), alimento (elemento, 47v), adsequandi (54v), faciam (faciem, 65). a for o ... manachorum (38v). a omitted ... ionne (Iohanne, 53). æ for a ... aeclesiae (abl., 27), lætentes (52v). æ for e ... iustæ (adv., 14), æclesia (16v, 17, etc.), præspiter, (29v, etc.), tribuæ (imper., 40), fontæ (45), sanctæ (voc., 47), æternæ (voc., 49v, 50, etc.), sempiternæ (voc., 54), frontæ (57, 58), saluaræ (59v), fratræ (60), æucharistiam (63v). b for m ... imbulamus (immolamus, 45). b for p ... babtismum (21, etc.), scabulas (52v), babtitzantes (53). b for v ... commobebuntur (62v), sperabimus (perf., 34). c for ch ... catacominis (catechuminis, 17, etc.), cresmate (chrismate, 57v). c for qu ... trancillitate (16v), corum (33v), scaloribus (57). d for t ... adque (27v, 50v), uelud (50v), debedis (58v), fortidudo (64v). d inserted after n concidunt (41). e for a ... emanere (emanare, 53).

18 e for æ ... querite (16), quesumus (passim), presta (passim). e for i... ... adstantes (13), respices (respicis, 16v), præspeteris (presbiteris, 16v), confeteor (20v), feceretis (feceritis, 28), cremine (37v, 55), preuelegia (38), marteria, (38, 40), emitatione (40), perfeciat (43v), refregeri (45v), abieciendos (47v, 52), lætentes (latentis, 52v), uirtutes (-is, 54v), affluentes (-is, 55), fontes (55), cresmate (57v), lineat (57v), uiuentes (-is, 61), homines (62v). e for ie ... maestas (42). e for ue ... pingesce (44), apinge (pingue, 44). ea for e ... fereabatur (54v). f for p ... adepti (16). f for ph ... orfanis (17), profetas (and other cases, 20v, 38, etc.), fantasma (54). g for c ... calige (15v), crugifixus (20v). h omitted ... tronum (13v), Pilippi (26v), ostium (hostium, 37v), teoricam (37v), immo (hymno, 41), triumpare (50), alitus (61); and see above, c for ch. h inserted ... baltheo (13), habundantius (41). i for e ... diabiso (de abysso, 12), ficisti (12v), urire (12v), promiriri (promereri, 13), accidentem (13), imbiciles (15v), redimisti (15v), perigrinantibus (17), meriamur (17, 63), discendit, etc. (20, 42v, etc.), sanguinim (20), sedit (20v), patiretur (27v), rimisionem (28), aspicire (28v), spicialiter (38v, 42v), proficire (40), difunctis (40v), sinciritatem (42), intigra (42), offerri (offerre, 42v), obteniant (42v), susteniant (44v), elimento (47v, 52, etc.), mistirii (47v), residiat (47v), flagillas (50), gehinam (50v), supplicis (-es, 51), benedictionis (-es, 52), cariat (52), mitire (mittere, 52v), apiriatur (58v), dispicias (58v), fidiliter (63), respondit (-et, 65).

19 i for ie ... maistatis (28v), mais[ta]tis (42v), maistate (62v). i for ii ... dilui (diluuii, 12), sacrificis (-iis, 16), pissimis (16v), fili (gen., 27v), insidís (37v), gaudi (38), gaudís (41v), iudici (-ii, 50v), misterís (54), uitís (55). i for u ... obtilit (28v), incolimitate (52v), uirtites (62v). ie for e ... diciens (28), agiens (28). ii for i ... aperii (imper., 49v). m for b ... neximus (13v). m for n ... imferis (28v), immo (hymno, 41). o for a ... euchoristia (63v). o for u ... tonica (13), catacominis (17, 52v), commonione (37v, 63v), gutore (gutture, 47), incolomitate (47v), commonicatio (63v), homano (64) b 2 p for b ... præspeteris (16v), præspiter (29v, etc.), opsecremus (38v). qu for c ... loqutus (20v), consequti (42v), loqui (loci, 52). t for d ... retdat (17v), tetrimentum (44v). tz for z ... babtitzantes (53). u for b ... coenouitarum (39v). u for o ... diabuli, etc. (16, etc.), idulorum (27v).

20 u for uu ... dilui (diluuii, 12), æqum (40v). u omitted ... cognoerunt (34), spiritale (38v), langoris (60). u inserted ... uirguinum (38). u for üs ... spiritu (gen., 39, 41, etc.). x for c ... resurrextionem (21). z for s ... zabaoth (13), zeth (32). zi for is ... elezion (eleison, 12v). Some small ornamental initials have already been mentioned (p. xiv). The only other decoration in the Missal is on the first page, which has a large rectangular initial P, together with a border inclosing the text on the other three sides. The colours employed in both initial and border are rather dingy red, yellow and brown. The initial is filled with a form of the keypattern ornament, familiar in the Book of Kells, the Lindisfarne Gospels, the Gospels of St. Chad, and other remains of early Celtic art, combined with lozenges and spirals. The border presents a striking analogy with that on the first page of St. Luke in the last two of the MSS. named. At the top, facing the initial, is the head of a cat, the hindquarters of which, with legs and tail, are at the lower extremity, the rest of the body being represented by horizontal and vertical panels containing a running pattern of swans with an edging of narrow bands. In the Lindisfarne Gospels the cat's head is at the lower extremity, and the treatment of the hindquarters here is more like that in St. Chad's Gospels, where the rectangular initial Q also contains key-pattern panels.

21 The Missal, however, is not in the state in which it was left by its original scribes, and the question of its date may be deferred for the present and be considered later in connexion with that of the subsequent additions. By far the most extensive and important of these are in the hand (a) of a scribe who gives his name, "Moelcaich scripsit," on f. 37. They are written, not only on the leaves which, as before mentioned, have been interpolated (ff. 14, 18, 22-25, 30, 31, 35, 36), but on others from which the original text has been, either in whole or in part, effectually erased (ff. 13, 15, 17v, 21, 21v, 26, 26v, 34 and 37), and also in the margins of ff. 16v, 28, 28v, and on f. 46v, which was originally left blank. As will be noticed at once in the facsimile, Moelcaich wrote a smaller, rounder and more facile minuscule hand than his predecessors, and, except when over an erasure, not between double ruled lines, but upon or pendent from a single one. There is also a considerable difference in the forms of some of the letters. Thus a is composed normally of two strokes only, which meet at the top and are not connected by a third transverse stroke. The exceptions to this are in the rather smaller minuscule rubrics, where the top is open (e.g., oblata, f. 21v), as is the case also in terrarum, f. 24, l. 11, and in fratribus, f. 25, l. 13.

22 In d the stroke on the right is most usually bent over to the left, but the vertical form, which largely predominates in the original hands, is also found (ff. 14v, 15). The difference in q and r is still more striking, for in Moelcaich's hand both these letters are of the ordinary minuscule type, the former being closed at the top and without the curve over the first stroke. Abbreviations are mostly the same as in the original text, with the addition of s^i for sicut (f. 13). The form q̄ is, however, used (f. 17v) instead of q:· for quae, and the horizontal stroke for -um in the genitive plural is placed over the preceding r and not to the right of it. The cases of noster are generally written in full; but when they are abbreviated, they follow the older system n̄i, n̄o, etc. (ff. 22v, 26, 46v, etc.), with one exception (f. 21), where n̄rae occurs. In its spelling of words Moelcaich's text is no less anomalous than the other. Thus it includes :—

23 aceptu, alis (aliis), altisimus, animamus (-abus), arcisire, audii (imper.). clementisime, commonionis. desederium, dicire, dirilinquisti, discende. effussione, eliuatur. feccis (faecis), filii (voc.), fructum (gen. plur.). incolomitate, inmotabilis, intigritate, iteneris. lintiamen. maestatis, meriar, misserere. ortodoxis, ossanna. prespeteris, pulluto. redetu, remisione. sarephin, spicialiter, spiciosus, sussum. uaris (variis), uitis (vitiis). Another later hand (b) is found in a few rubrics (ff. 12, 13, 13v, 15) and in two inserted collects, one at the foot of f. 15v and the other on f. 19, where the writer utilized a narrow strip of vellum which was left blank by Moelcaich and to which another piece was afterwards stitched. Although the two are very much alike in type, this hand may be distinguished from Moelcaich's by characteristic forms of t and u (see f. 13, l. 1). There is, however, apparently very little, if any, difference in date between the two, and both scribes were most probably contemporary members of the same religious community.

24 Some interlineations on ff. 13, 15v are in a smaller and rougher hand than either of the two last mentioned. They were either written by yet another corrector (c), or possibly by Moelcaich himself, the fact that he was writing under cramped conditions between the lines being accountable for any apparent difference of hand. Finally, there are two short Irish rubrics on ff. 51, 58, which are in a small, sloping hand, different from all the rest, but apparently not much later. With regard to the age, both of the MS. as it was first written and of the subsequent insertions, very different opinions have been expressed. For one reason or another the original text has been dated by Dr. Todd not later than the sixth century, by Dom S. Bäumer about 627-640, by Dr. MacCarthy in the second quarter of the seventh century, by Monsignor Duchesne in the eighth century, by Mr. F. E. Warren in the ninth century, and by Sir F. Kenyon, with some hesitation, at the beginning of the tenth century or end of the ninth. The diversity of opinion about Moelcaich's additions among the same authorities is not so great. As before, S. Bäumer is the most precise, dating them about 740-750, and Dr. MacCarthy also assigns them to the eighth century, but the rest all agree in the belief that they were inserted two centuries later.

25 It would materially assist us in solving the vexed question of date if the identity of Moelcaich could first be satisfactorily established. The name, which is explained to mean Calvus luscus, "the bald (tonsured), one-eyed man," does not appear to be an infrequent one. It was borne by a contemporary of St. Buite, who died in 521, and who, according to legend, miraculously restored his head to his body and later, as penalty for a misdeed, reversed the process; and the Annals of Ulster record the death of Maelcaich, son of Scannal, king of the Cruithni, in 665 (6) and of Maelcaech, son of Cuscrad Menn in 783 (4). Two others of the name appear in a table of the descendants of Ailill Ólom compiled by Dr. MacCarthy from the Book of Leinster. One of them is there described as son of Aed Bennan, king of Kerry, who died in 619, and his own

26 fault of his own, Dr. Todd's examination of the MS., which was then still at Ashburnham Place, was very hasty and superficial. death is approximately dated about 650. This is undoubtedly too early for the Moelcaich of the Stowe Missal, but Dr. MacCarthy confidently identifies the latter with the other Moelcaich in his table, who was son of Flann and presumably lived towards the middle of the eighth century, being fourth in descent from Aed, king of Cashel, who died in 601. The chief ground on which he bases this belief is the date of the latest saint in the Litany on f. 30, which is in Moelcaich's hand, and it is on this ground also that S. Bäumer dates his insertions about 740-750. On referring to the Litany in the printed text (p. 14), it will be observed that, although all the names are preceded as usual by Sancte or Sancta, some are in the vocative and others in the genitive case. This discrepancy rather suggests that the latter came originally from a different source and were ignorantly copied without the necessary change of inflexion. The following are the Irish saints with their dates, so far as they are known, and the days on which they are commemorated in the Irish Martyrologies:—

27 Patricii, Patrick, ob. 493. Ailbei, Ailbe, bishop of Emly, Sept. 12, Dec. 30 (O, T, G, D), ob. 534 or 542. Finnio, or Finnian, abbot of Clonard, co. Meath, Dec. 12 (O, G, D), ob. 549. Finnio, or Finnian, bishop of Movilla, co. Down, Sept. 10 (O, T, G, D), ob. 579. Ciarani, Ciarán, abbot of Clonmacnois, King's Co., Sept. 9 (O, T, G, D), ob. 549. Ciarani, Ciarán of Saiger (Seirkieran, King's Co.), Mar. 5 (O, T, G, D), ob. 560. Brendini, Brénann, abbot of Birr, King's Co., Nov. 29 (O, G, D), ob. 573. Brendini, Brénann, abbot of Clonfert, co. Galway, May 16 (O, T, G, D), ob. 577 or 583. Columba, Colombcille, abbot of Hí or Iona, June 9 (O, T, G, D), ob. 597. Colomba, Colomb, abbot of Terryglas, co. Tipperary, Dec. 13 (O, G, D), ob. 600. Comgilli, Comgall, abbot of Bangor, co. Down, May 10 (O, T, G, D), ob. 600.

28 Cainnichi, Cainnech, abbot of Aghabo, Queen's Co., Oct. 11 (O, T, G, D), ob. 599. Findbarri, Barr, bishop of Cork, Sept. 25 (O, T, G, D), ob. 622-3. Nessani, Nessan, deacon, of Mungret, co. Limerick, July 25 (O, T, G, D), ob. 551. Fachtni, Fachtnae, bishop of Roscarbery, co. Cork, Aug. 14 (O, T, G, D). Lugidi, Lugaid or Lugach, bishop of Cúil-Bennchoir, Oct. 6 (O, T, G, D). Lacteni, Lachtain of Achad-úr (Freshford, co. Kilkenny), Mar. 19 (O, T, G, D), ob. 622. Ruadani, Ruadán, abbot of Lothra or Lorrha, co. Tipperary, Apr. 15 (O, T, G, D), ob. 585. Carthegi, Carthach or Mochutu of Rathen, bishop of Lismore, May 14 (O, T, G, D), ob. 637. Coemgeni, Cóemgen, abbot of Glendalough, co. Wicklow, June 3 (O, T, G, D), ob. 618 or 622. Mochonne, Mochonna or Conna of Inis Pátraic (Inishpatrick, near Skerries, co. Dublin), or another, bishop of Lemchaill (Lowhill, Queen's Co.), both Jan. 13 (T, G, D, cf. O, p. 42). There were other saints of the same name, including an abbot of Derry, Apr. 12 (T, G), ob. 704; a bishop of Connor, May 15 (G), ob. 725; and Mochonna of Cluain Airdne, Sept. 30 (T, G, D), ob. 713.

29 Brigta, Brigit, abbess of Kildare, Feb. 1 (O, T, G, D), ob. 522. Ita, Ite, virgin, of Killeedy, co. Limerick, Jan. 15 (O, T, G, D), ob. 570 or 577. Scetha, Sciath or Scéthe, virgin, Jan. 1 (G, D), Sept. 6 (O, T, G, D). Sinecha, Sinech, virgin, Nov. 9 (O, G, D). Samdine, Samthann or Samdann, virgin, of Clonbroney, co. Longford, Dec. 19 (O, T, G, D), ob. 739. 30 The last saint invoked is thus also the latest in date; and, so far as this evidence goes, while it is clear that the interpolated Litany must at the earliest be later than 739, there is nothing in it to preclude the possibility of its having been written, as Dr. MacCarthy supposes, by Moelcaich, son of Flann, or between 740 and 750. There is, however, on f. 32, a list of saints to be commemorated, or what is known as the diptychs, which is in the hand of one of the original scribes and must therefore have been written before Moelcaich's interpolations, and it is necessary to ascertain whether it includes any saint posterior to Samthann. It begins with Biblical names from Abel to the Apostles and Evangelists, who are followed by the three early martyrs Stephen, Cornelius and Cyprian, and the two early hermits Paul and Anthony. Of the two remaining sections, the first comprises the names of as many as thirty-five bishops. All of them are insular except the first four, Martin, Gregory, Maximus, and Felix, who are identified by Dr. MacCarthy with Martin of Tours (ob. 397 or 401), Gregory of Tours (ob. 594), Maximus of Riez (ob. 462), and Felix of Nantes (ob. 584). Martin of Tours was extremely popular in Irish hagiography, and the identification in his case is no doubt correct, but neither Gregory of Tours nor Felix of Nantes is mentioned in any of the Irish Martyrologies.

31 The Gregory referred to is therefore more probably Pope Gregory the Great (ob. 604, Mar. 12, O, G), the "Grigoir Rúamae" of Oengus, who was held in the highest honour; and unless the connexion between Ireland and Brittany is sufficient to account for the Nantes bishop, Felix may also be a Pope, either Felix I. (ob. 274) or Felix II. (ob. 358), both of whom are in Gorman (May 30, July 29). Maximus of Riez points to the influence of Lerins, of which he was abbot. He is no doubt Gorman's "Maxim" entered on Nov. 27, which is his usual day. The insular names, which, like the rest, are all in the genitive case, are: 32 Patrici, Patrick, ob. 493. Patrici, Patrick (Sen-Phátric), of Glastonbury, Aug. 24 (O, T, G), ob. 457 or 461. Secundini, Sechnall, St. Patrick's nephew, Nov. 27 (O, G, D), ob. 448. Auxili, Auxilius or Usaille, bishop of Killossy or Killashee, co. Kildare, Aug. 27 (T, G, D), ob. 460. Isernini, Iserninus, companion of St. Patrick. Cerbani, Cerpán, bishop of Fert Cerpáin at Tara, ob. 504. Erci, Erc, bishop of Slane, Nov. 2 (O, G, D), ob. 512. Catheri, Cathar. A Cathar, as Dr. MacCarthy remarks, is included in the list of bishops (next to Iserninus) in the Book of Leinster, 365d, but he is otherwise unknown. Ibori, Ibar, bishop of Becc-Ériu or Beggery Island, co. Wexford, Apr. 23 (O, T, G, D), ob. 500. Ailbi, Ailbe. See above, p. xxiv. Conlai, Connla, bishop, May 10 (T, G, D), or another, bishop of Rúscach or Rooskagh, co. Westmeath, Dec. 30 (T, G, D).

33 Maicnissæ, Maccnisse, bishop of Connor, Sept. 3 (O, G, D), ob. 514; Mar. 1, May 1 (T). Moinenn, Móinenn, bishop of Clonfert, Mar. 1 (O, T, G, D), ob. 570. Senani, Senán, bishop of Inis Cathaig or Scattery Island, co. Clare, Mar. 1 (O, T, G, D), ob. 544. Finbarri, Barr. See above, p. xxv. Colmani, Colman, bishop of Cloyne, Nov. 24 (O, G, D), ob. 600; or Colman, bishop of Lindisfarne (661–664) and of Inisboffin, co. Mayo, Aug. 18 (O, T, G), ob. 676. Cuáni, Cua or Mo-Chua, of Cluain Dolcáin or Clondalkin, near Dublin, bishop, Aug. 6 (O, T, G, D). Aedach. This appears to be the true reading, and an Aedach or Aedhach is commemorated in G, D, on Apr. 9. He is not, however, called a bishop and Aedach is not a genitive form, as it should be here. Possibly the dot over h is a punctum delens applying also to c. Aeda would then be the genitive of Aed, and the saint may be identified with Aed mac Bricc, bishop of Cell-áir or Killair, co. Westmeath, Nov. 10 (O, G, D), ob. 588.

34 Laurenti, Laurentius, archbishop of Canterbury, ob. 619. Melléti, Mellitus, archbishop of Canterbury, ob. 624. Iusti, Justus, archbishop of Canterbury, ob. 627. Aedo, Aed, Aedoc or Maedoc, bishop of Ferns, Jan. 31 (O, T, G), ob. 624. Dagani, Dagán, the "Daganus episcopus" whose name appears in the letter of Laurentius, Mellitus and Justus to the Irish Church, Mar. 12 (T, G, D). Probably identical with Dagán of Inber Dóile or Ennereilly, co. Wicklow, Sept. 13 (O, T, G), ob. 639 or 642, though the latter is spoken of (O, p. 206) as a priest. Tigernich, Tigernach, bishop of Clúain Enis, or Clones, co. Monaghan, Apr. 4 (O, T, G, D), ob. 549. Muchti, Mochtae, bishop of Louth, Mar. 24, Aug. 9 (O, T, G), ob. 535. Ciannani, Cianán, bishop of Duleek, co. Meath, Nov. 24 (O, G, D), ob. 489. Buiti, Buite (Boetius), bishop of Monasterboice, co. Louth, Dec. 7, (O, G, D), ob. 521.

35 Eogeni, Eogan, bishop of Ardstraw, co. Tyrone, Aug. 23 (O, T, G, D), ob. 618. Declani, Declán, bishop of Ardmore, co. Waterford, July 24 (O, T, G, D), ob. circ. 540. 36 Carthuin or (possibly) Carthain. Perhaps (Aed) MacCarthainn, bishop of Cloghar, member of the household of St. Patrick, Mar. 24 (T) and, as Fer dá chrích, Aug. 15 (T, G, D, cf. O, pp. 68, 100, 186). There seems to be some confusion, however, between him and another (Aed) MacCarthainn, al. Fer dá chrích, abbot of Dairinis, ob. 747, described in a note to the Félire, Aug. 15 (p. 187), as a bishop and as "maternal uncle of Máel-rúain (see next name), Oengus' tutor, and from him Máel-rúain brought Fer dá chrích's bell, which is in Tallaght." In some verses quoted in the Martyrology of Donegal, Aug. 15, he is called Maelruain's teacher. The absence of "Mac" before "Carthuin," which would be a serious objection to the identification, may be more apparent than real. On reference to the facsimile it will be seen that "Carthuin" is at the bottom of the fourth and last column of f. 32v, on a line with, and close up to, "Colmani" at the bottom of the third column, and that (for the reason given on p. 16, note 1) the name above "Colmani" has been erased with the exception of "ni" at the end. There must have been some reason for leaving these two letters, and they were perhaps spared to serve as the end of "Colmani" in the line below, and what has there been read as "ni" is really "m," the abbreviated form of "Mac" or "Maic" before "Carthuin." Elsewhere in the case of MaicNissæ the "Maic" has been written in one line and "Nissæ" in the line below, but the scribe may have hesitated to write "Maic" at the bottom of one page and "Carthuin" at the top of the next and so took advantage of the erasure, which was made before he reached the end of the next column, to squeeze in the full name.

37 Maile-ruen. This name, which is the last in the series of bishops and is also the only one, with the doubtful exception of the two Patricks, which begins with a capital initial, is of crucial importance for fixing the date of the MS. Dr. MacCarthy takes Maile and Ruen, which are in different lines, to be two distinct names, identifying the first with Mel, bishop of Ardagh (ob. 488), while the other, which he regards as a phonetic form of Ruadain, genitive of Ruadán, is presumably Ruadán of Lorrha (ob. 585). There can, however, be very little doubt that, as in the cases of Melchi-sedech, Macha-beorum, Bartha-lomæ and Maic-nissæ, the single rather lengthy name Maileruen (gen.) has been divided between two lines so that it should not protrude into the next column on the right; and if this be so, it can hardly represent any other than the well-known Maelruain, bishop and founder of the monastery of Tallaght, near Dublin, who died in 792 (July 9, O, G).

38 In support of this conclusion I have the high authority of the Rev. C. Plummer, who kindly allows me to quote the following communication:— "As to Maile-ruen two things are in any case fairly certain: (a) Maile is the genitive of Máel = bald, tonsured, and (b) Ruen is phonetic writing for Ruadain, genitive of Ruadán, the d softened and disappearing between the vowels. The question whether we are to take this as one name or two depends on whether the genitive Ruen is governed by Maile or is an independent genitive ranking with the other names in the list, which are in the genitive. The former seems to me much the more probable. "Names formed of Máel with a following proper name in the genitive are very common, as you know, in Irish, Máel-Ísu, Máel-Eoin, Máel-Ciarain, etc., meaning the tonsured slave of Jesus, of John, of Ciaran, etc. When such names are declined, it is always the former part Máel which is declined, the following genitive being constant. Thus Maileruen would be the correct genitive of Máelruen.

39 "On the other hand Máel alone is hardly found as a separate name. If it does so occur, it would either be a hypocoristic shortening of one of the many names beginning with Máel, or a mere nickname, like in Máel," the bald one, applied to St. Iarlaithe. It occurs as a name of one of the Druids opposed to St. Patrick (Vita Trip., pp. 102–104), but as the other Druid is called Caplait (= Capitilauium), it is obvious that these are made-up names, the tonsured, the head-washed. Therefore the chances seem to me enormously in favour of taking Maileruen as a single name. "The identification of Maile (gen.) with Mél of Ardagh I regard as impossible. Only one MS. of the Félire of Oengus writes his name Moel (Feb. 6). The name is Mél, later Méal, showing that the l is not mouillé, and it seems to be indeclinable. The most decisive case is the genitive. We have Siur epscuip Mél," sister of bishop Mel (Vita Trip., p. 88), "a munigin epscuip Mél," in reliance on (lit. of) bishop Mel (Stokes, Three Homilies, p. 72 = Lebar Brecc facs., p. 64b).

40 "As to the identification of him, I think he can only be the great Tallaght saint. He must have been a well-known man to get into the Litany, and the only other Maelruain I have found in the Calendars is in the Lebar Brecc copy of the Félire at Dec. 13. The two Maelruains in the Four Masters, 880, 884, are probably too late." There are still thirteen more saints in the diptychs, who are classified as priests and with four exceptions may be identified with saints included in Moelcaich's Litany. The names are: Uinniaui, Finnio or Finnian, abbot of Clonard. See above, p. xxiv. Ciarani, Ciarán, abbot of Clonmacnois. See above, p. xxiv. Oengusso, Óenu or Oengus, successor of Ciarán, Jan. 20 (O, T). In G he is called a bishop. Endi, Endae, abbot of Áran, Mar. 21 (O, G), ob. 542. Gilde, Gildas Badonicus, Jan. 29 (O, T), ob. 570, or Gildas the Elder, ob. 512. T has a Gildas on Sept. 28, and G another on Nov. 4.

41 Brendini, Brénann Brendini, Brénann Cainnichi, Cainnech Columba, Colombcille Columba, Colomb Colmani, Colmán, abbot of Lynally, King's Co., Sept. 26, Oct. 3 (O, G), ob. 610; or abbot of Clonard, co. Meath, Feb. 8 (G), ob. 654. But there are twenty-one saints of the name in O, and many more in G. Comgelli, Comgall Coemgeni, Cóemgen See above, p. xxv. See above, p. xxv. 42 The result is that, excluding "Maile-ruen," all the saints commemorated in the above diptychs are of earlier date than Samthann, who is the latest in Moelcaich's Litany. But on the assumption that "Maile-ruen" is St. Maelruain of Tallaght, who died in 792, it is obvious that the original text of the Missal cannot have been written before the very end of the eighth century, and must have come into the hands of Moelcaich and have been revised by him at a later date still. Whoever he was, his identity with Moelcaich, son of Flann, is therefore impossible, unless, indeed, the words "Moelcaich scripsit" were copied from the same older source as the text. This, however, is very improbable, while on the other hand the fact that the Litany includes no saint later than the first half of the eighth century can be easily accounted for by its having been reproduced by Moelcaich from his archetype just as he found it without the addition of any names nearer his own time. Either on his own initiative or under direction, the particular scribe to whom we owe the diptychs or earlier list of saints appears to have acted differently.

43 As there is a very considerable interval of time between Maelruain and the latest of the other saints commemorated it is reasonable to infer that this list also represents a transcript with the single exception of Maelruain's name, the interpolation of which, in its proper place at the end of the bishops and with the unusual distinction of a large initial, marks it as that of a more recent saint who for some reason was the object of special veneration. Naturally his honour was greatest among the fraternity of his own foundation, and we thus get, in addition to an approximate date, a valuable clue to the place where the MS. was written. Its monastic origin is evident from the insertion in the canon (p. 10) of the words "et abbate nostro .n. episcopo," where the mention of the twofold dignity of abbot and bishop is also not without significance. Moreover, if it came from Tallaght, there is some slight evidence by which the posterior limit of its date may be brought within twenty years from Maelruain's death; for otherwise one might expect to find in it the name of another bishop and abbot of Tallaght besides that of the founder. The list of members of the community of Maelruain given in the Book of Leinster includes "Eochaid episcop Tamlachta," whose death is recorded in the Annals of Ulster under 811 (812), "Echaidh episcopus et ancorita princeps Tamlachta quieuit.

44 Like his predecessor, he was regarded after death as a saint, being commemorated on January 28, not only in the Martyrology of Tallaght itself, but also in those of Gorman and Donegal. The absence of his name from the list of commemorations, when Maelruain's was inserted, may therefore be taken to imply that it was completed while he was still living; and in that case his position as bishop and abbot makes it also probable that the original part of the Missal was written under his inspiration. 45 The hagiological evidence thus pointing to a date subsequent at least to 792 and possibly before 812, it remains to be seen how far there is palaeographical evidence to support or disprove it. The script of the various original hands is rather stiff and unformed and of an unusual type. So far, indeed, as I am aware, there is nothing quite like it in any other extant Irish MS., and it was possibly peculiar to Tallaght. From its general appearance I should not myself be inclined to give it a date before the end of the eighth century, and perhaps the nearest approach to it, though on the later side, is in a MS. of Priscian at St. Gall, No. 904, a page of which, showing two hands, is reproduced in Prof. W. M. Lindsay's Early Irish Minuscule Script, plate x. This MS. appears to have been written in Ireland, probably in 845 and is mainly by two scribes, Maelpatricc and Coirbbre, two others, Finguine and Donngus, also having a small share in it. Coirbbre's hand is seen in the first column of Prof. Lindsay's plate, and except as regards r, which is minuscule, the forms of letters resemble in many respects those in the Missal. The two most characteristic letters in the latter are a and q, and if Coirbbre's a is not precisely similar, this is the case in one of the other hands, viz., that of Donngus, where a, as Prof.

46 Lindsay describes it, "resembles u with a fine stroke connecting the two horns," just as it so often does in our MS. Coirbbre, however, does use the same open q, though with a less pronounced curve at the top of the first stroke. At the same time, this peculiar form of q is not a sign of late date, for it occurs also in eighth century MSS., as, for instance, in the Durham Cassiodorus, said to have been written by Bede, as well as in a Mercian charter of 793-4 and in the Book of Armagh dated in 808. The abbreviations in the two MSS. are of the usual Irish character and are practically identical; but one at least of the few instances in which they vary is significant, pointing to an earlier date for the Missal. The importance for dating purposes of the abbreviation-symbols of the cases of noster was shown by Traube in Nomina Sacra, and it is further illustrated and enlarged upon by Prof. Lindsay in his forthcoming Notae Latinae. The conclusion to which their investigations lead is that n̄i, n̄o, etc., do not survive later than about 815, by which time they were finally abandoned for the less ambiguous n̄ri, n̄ro, etc. Now in the St. Gall Priscian (845?), as in a Leyden Priscian (838) and a Carlsruhe Bede (836-848), which are also Irish MSS. that can be dated, n̄ri, etc., are exclusively found, whereas in the Stowe Missal we have both the n̄i and n̄ri forms.

47 The full number of the abbreviations of one case or another of noster is in fact sixty-two, of which, deducting instances of n̄r (noster, 3) and of mere suspensions such as nos̄ (4) and nostrōr, thirty-seven are examples of the two-letter symbol as against seventeen of the three-letter symbol. But, although the former are thus in a large majority, when they are examined it will be found that, with only two exceptions, viz., preces nas (f. 17) and patriarche n̄i (f. 29v), they all occur in the frequently repeated and, so to say, stereotyped formula dn̄i n̄i Iesu Christi and other cases of the same, and the probability is that the scribes of the Missal copied them mechanically from their archetype, while elsewhere, as a rule, they followed the more modern system. At any rate the double usage seems to mark a stage of transition such as might be expected shortly before or after 800, and in fact it also obtains in the Book of Armagh, the date of which is 808. There is a variation also in the abbreviation symbol for quae, which in the Missal takes the exclusively insular form q: while in the St. Gall Priscian it is q̄. Both these symbols, however, were in use by insular scribes concurrently, and they are found together in the same MS., as, for example, in the Schaffhausen Vita S. Adamnani as early as 713 and in the Book of Armagh very nearly a century later. As Prof.

48 Lindsay has shown, q̄ is really the older symbol of the two, being the ancient Nota for quae; but when in course of time quae came to be written que and in consequence was further erroneously represented by the que symbol q· or q:, insular scribes devised the modification q:· as a means of distinguishing between the relative and the conjunction and also perhaps, as Prof. Lindsay suggests, to avoid further confusion with q̄, which was their symbol for quem. This practice, however, did not continue indefinitely, but gradually fell into disuse, while q̄ reasserted itself, and there is at least a strong presumption that a MS. in which q:· is consistently employed, as in the original part of the Stowe Missal, is earlier than 850. 49 On palaeographical grounds therefore, no less than for the other reasons before stated, there seems to be sufficient warrant for dating the original text in the early part, or even within the first decade, of the ninth century. Nor is there any reason to believe that Moelcaich's and the other additions are, as has been supposed, so much as a century later. Although it differs considerably from those of the original scribes, Moelcaich's hand is nevertheless of quite as early a type. If his share in the MS. was not indisputably later than the bulk of it, but was in a separate volume altogether, there might even be some doubt which was really the older, and probably the difference in style is due rather to locality than to lapse of time. Taking the noster test, a comparison of the two scripts yields much the same result; for, although in most instances Moelcaich writes the possessive—even on f. 22, where it occurs nine times in ten lines—without any abbreviation, there are still six examples of the n̄i symbol and only one, n̄rae deuotionis (f. 21), of the n̄ri symbol, to which another, per d̄nm n̄rm (f. 140), in the hand of the second reviser (b) may be added. Here again, however, the shorter forms are all found in association with d̄ni, etc., and may have been copied mechanically. With regard to quae, on the contrary, there is a disagreement, for in place of q:· Moelcaich invariably writes q̄, which, so far as it has any significance, points to a later date.

50 While on the whole none of the inserted matter seems to be later than the first half of the ninth century, it is quite conceivable that the Missal was subjected to revision almost immediately after it was written. This is the more likely if, as I am inclined to believe, its primary object was to provide the monastery of Tallaght, which at the beginning of the ninth century was comparatively new and of growing importance, with an authoritative ritual. The original text would in that case be in the nature of a first draft derived from an archetype which, judging from the saints named in it (exclusive, of course, of Maelruain), may have been as early as the seventh century, while the completed MS. as we now have it represents the shape ultimately given to it after it had been revised with the aid of a second archetype dating from about the middle of the eighth century, and coming apparently from some other liturgical centre, possibly Armagh.

51 Further evidence as to date and origin, confirmatory or otherwise, may no doubt be drawn from the liturgical forms in the respective texts. The object, however, of the present volume is merely to give an edition of the text of the MS. in which its component parts are carefully distinguished, and also to fix its date so far as this can be done on other than liturgical grounds. Without a special knowledge of liturgiology, to which I make no pretence, it would be rash to enter upon other questions, and they must therefore be left for those who are more competent to deal with them adequately. This also applies to the contents of the Irish treatise on the Mass which is added at the end of the MS. But so far as its date depends on palaeographical considerations, this tract does not appear to me to be any later than the other insertions; on the contrary, the hand has a rather striking resemblance to that of the rubrics on ff. 57v-58v, which, although written in smaller characters than the text, are by one of the original scribes. Besides a general similarity we find in it the same majuscule r and the horned open q in addition to the ordinary form, which is also the commoner in the rubrics; moreover, as in the original text, the suprascript abbreviation-stroke for um in -rum is placed to the right of the preceding r, and not, as in Moelcaich's hand, above it.

52 The following, in addition to the usual nomina sacra symbols, are the abbreviations in the Latin words which occur in the tract: aecī, aecclesia. ciuī, ciuitatem. ə (a reversed c with cross-stroke, or, less correctly, the "eius" symbol as on p. xiv), contra. ÷, est. ʔ, et. fš, filius. li, hoc. ·i·, id est. ī, in. nō, nomen. p̄r, pat, pater. p̓, per (e.g., super). p̄, prae (e.g., deprecor). q̄n, quando. q̄, q̄3, que. s̄, sed. Most of the abbreviations in the Irish text are of the nature of mere suspensions, but they also include: q, the symbol for "quia," used for its Irish equivalent "ar." f̄, fō, for (upon). ·i·, idón (id est). ī, in (the, in). ʔ, ocus (and). The same syllable-symbols are mostly used both in Latin and Irish words. Those for er, um, ur and us, as well as the hooked horizontal stroke for omitted m, both final and medial, are the same as in the Missal (p. xvi); also two dots are placed over g and t for ra (e.g., Ir. grad, tra) and i over p for ri (e.g., Lat. prius). In some cases also Latin word-symbols (s̄, q̄, etc.) are used to express syllables in Irish words (e.g., issed, huisque = huisce).

53 As nothing in these criteria militates against an early date, there need be little doubt that the tract is nearly, if not actually, contemporary with the body of the MS.; and the linguistic evidence also seems to be in favour of this, for in alluding to the suggestion that this tract and the three short Irish spells on f. 67v are not earlier than the eleventh century the editors of the Thesaurus Palaeohibernicus remark that, if this be so, the language shows that the texts must have been copied from a much older original. With regard to the date of the rougher hands of the three spells, it is difficult to speak with any confidence; but my own impression of the whole MS. is that from first to last it contains nothing later than the ninth century. 54 Mention has been made above (p. ix) of the extracts from the Gospel of St. John which occupy the first gathering (ff. 1-11). These extracts are in a hand quite different from any of the other hands in the volume, and they clearly have nothing to do with the Missal and other liturgical matter. They have therefore not been included either in the facsimile or in the printed text; but they nevertheless claim some notice with respect to their character and probable date. Whether there was any particular reason for binding them up with the Missal it is impossible to say, nor is the date of the binding at all certain. They may conceivably have been prefixed out of reverence for the memory of some notable personage, such as St. Maelruain, to whom they had belonged. A note, however, to that effect might then have been expected, and possibly the only reason for their inclusion was the fact that the leaves on which they are written happened to have the same dimensions as those of the other MS. The portions of the Gospel comprised are: i-vi. 30; vii. 45-viii. 14; viii. 19-33, umquam; viii. 53, qui-59; xii. 9-39, credere; xvii. 11, Pater sancte-xviii. 1; xviii. 4-13; xviii. 15, discipulus-23; xix. 40-xx. 23; xx. 26-xxi. 6; xxi. 9-end. It is at once apparent that these are not passages purposely selected to serve as lessons or from any other intelligible motive, and some of them in fact break off abruptly in the middle of a sentence. The explanation, as Dr.

55 Bernard points out in his paper "On the Stowe St. John," seems to be that they were copied consecutively from a mutilated archetype of unknown, but probably early, date in which some leaves were lost, and others were torn or partially illegible. At the same time there is evidence which suggests that this defective Gospel of St. John is all that survives of a MS. which contained at least one other Gospel, and perhaps the whole four. It is curious that the rude miniature of St. John with his eagle, reproduced here in Plate IX, is not placed at the beginning of the Gospel, as is more usual, but at the end, on the back of the last page of text (f. 11v). The gathering, however, originally consisted of twelve leaves, the first of which, as already stated, has been torn away with the exception of a narrow strip along the inner edge, and as a small fragment of a border similar to that which incloses the figure of St. John is visible on the back of this strip, it seems probable that the missing page contained a miniature of St. Luke or one of the other Evangelists, which was also placed at the end of his Gospel.

56 With the text of St. John, or so much of it as the MS. includes, we are not here concerned, but the full collation given by Dr. Bernard proves it to be in his own words "Vulgate with traces, as is usual in Irish MSS., of old Latin 'mixture.'" In default of any other evidence its date can only be judged from the script, the first and last pages of which are reproduced in Plates VII, VIII. It begins neatly and carefully, but soon degenerates into a rather scratchy hand, becoming finally almost cursive, with the sinuous vertical stroke which Prof. Lindsay terms "the Irish (or rather insular) cursive form of i." This is specially observable in the colophon, the text of which will be found below, and which also shows the characteristic downward prolongation of the final stroke of m, n, u. Paragraphs begin on a new line with a plain capital initial set well out in the margin, the line being continued on any vacant space left in the line above at the end of the preceding paragraph. Apart from the miniature of St. John, there is no other decoration except on the first page (Plate VII), where the large ornamental capitals I N P are in black, the other colours in letters and border being red, yellow, and mauve. The key-pattern with lozenges in the N and in the border is almost identical with that in the initial on the first page of the Missal. Both the orthographical peculiarities and the abbreviations correspond generally with those in the Missal, but unfortunately the noster symbols do not occur, and for quae both q̄ and q:· are used, the latter, however, being less common.

57 The abbreviations are : a p̄, apud. ā, aut. h̄, autem. c̄s, cuius. c̄, c̄, cum. dī, dx̄, dcs, dr̄, dicit, dixit, dicens, dicitur. ɔ, eius. #, enim. g°, ergo. eē, esse. +, -, est. 7, et. hī, habet. h, h̄, h̄s̄, h̄c̄, haec, hoc, huius, hunc. hōis, hōem, hominis, -em. ī, in. m̄m, m̄s, meum, meus. m̄i, mihi. nō, nōe, nomen, nomine. n̄, non. ñc, nunc. ōme, ōms, ōma, ōmne, ōmnes, ōmnia. p̄, per. p°, post. p̄, prae. p̄, pro. ppus, proprius. propī, p̄t̄, propter. q̄n̄, quando. qṛe, quare. qsi, quasi. qʒ, que. q̄, q̄, q̄, qui, qua, quo. q̄, qʒ, quae. q̄, quem. q̄, quam. q̄, quod. q̄, quia. q̄mo, quomodo. q̄m, quoniam. q̄q̄, quoque. q̄t. quot. ṛdum, secundum. s̄, sed. s̄i, sicut. s̄n, sine. s̄t, sunt. supʳ, super. t̄n, tamen. t̄i, tibi. t̄s, trans. t̄c, tunc. ūn, unde. i, uel. u°, uero. The syllable-symbols are: an inverted c for con; a supra-script horizontal stroke for er (e.g., inter, noster, pater, propter, uerbum), for um in -rum (e.g., nostrum, uirum), and for un in -runt; i over p for ri (e.g., prior, primum); a comma over t to the right for ur (e.g., moriatur, turba); and ʒ for us in -bus, -mus, etc. Final m is represented by a hooked horizontal stroke.

58 The scribe gives his name under an Ogham form in the colophon, which is as follows: "Deo gratias ago. Amen. Finit. Amen. Rogo quicumque hunc librum legeris. ut memine From left to right the Ogham name reads Sonid, in the reverse way Dinos. Neither name is known elsewhere, but the former is the more probable, and it has been suggested that in the sentence which follows there is a play on the nearly homophonous Irish word sonaid, happy, prosperous, and the Latin sanus. The precise meaning of the epithet "peregrinus" is doubtful. Possibly the scribe was actually on a pilgrimage at the time, or he may have been, as Bishop Graves suggests, a foreigner who had come, like many others, to Ireland for the sake of religious instruction or, on the other hand, an Irish monk writing on the continent. 59 As regards the date of the Gospel, there has been almost as much difference of opinion as in the case of the Missal. Dr. Todd and Dr. MacCarthy make it seventh century, Dr. Bernard not earlier than the end of the eighth century or beginning of the ninth, L. Delisle not earlier than the ninth century, and S. Berger tenth century. As Dr. MacCarthy and others have remarked, the hand is to some extent similar in type to that of the scribe Dimma, who wrote the first three Gospels in the well-known Book of Dimma; but the age of that MS. is also uncertain, and it is the only one of those included in Prof. Lindsay's Early Irish Minuscule Script to which he has not assigned a date. As, however, it contains both the nī and nīr symbols, with a preference on Dimma's part for the latter, it is not likely to have been written much before or after 800, and the similar date which Dr. Bernard gives to the Stowe St. John is probably correct. Moreover, as it was in that case quite possibly written before 792, it may after all owe its preservation with the Missal to some connexion with St. Maelruain.

60 The importance attached to the volume by its former possessors is shown by the Cumdach, or casket, which was made to contain it, and which elsewhere than in Ireland would probably have taken the more usual form of an elaborate metal-work binding. It is one of five such characteristic early Irish book-shrines which still survive. The other four belong to the Gospels of St. Molaise (1001–1025), the Gospels of St. Patrick (11th century and circ. 1350), the Psalter of St. Columba (1084) and the Book of Dimma (1150), the date in each case being that of the Cumdach and not of the MS. itself; and besides these, three more notable MSS. are known from historical sources to have formerly been protected in a similar manner, viz., the Book of Durrow (877–914), the Book of Kells (before 1007) and the Book of Armagh (938). Like the rest, the Cumdach of the Stowe Missal is undoubtedly of later date than the book which it enshrined. This, however, is only what might be expected; for, unless a MS. was of exceptional splendour, such as the Book of Kells, or for some other special reason, it was not likely to be so distinguished until it had acquired, on account either of its contents or its associations, the sanctity of a relic by course of time.

61 as being the older of the two, it will be convenient to describe first, is a roughly squared oak board 7 inches long, 5¾ inches wide and ¼ inch thick, with the surface worked smooth by a chisel. Over this, attached by bronze nails, is a thin bronze plate with projecting edges, which are slightly bent up all round. The shallow depression thus formed measures 7½ inches by 6¼ inches. It is filled by a rectangular bronze frame, ⅝ inch wide throughout, bevelled at the inner edge and overlaid with a thin coating of silver, which in some places has peeled off; and two similar silver-coated plates of the same width intersect at right angles in the centre, forming a cross, the four limbs of which extend to the frame. The four sunk panels between the cross and frame are decorated with chequer-patterns of squares and triangles, the former in the first and fourth quarters, the latter in the other two. These designs are produced by punching through a thin bronze plate (or perhaps a separate plate for each panel), coated with silver, laid over the bronze plate first above mentioned and under the frame and cross; and, as the under metal is visible through the pierced work, a good effect is thus produced by the contrast between the silver and bronze.

62 The darker metal was also laid bare by punching a small square in the overlying silver at each corner of the frame and a small arc at each end of the cross, the eight recesses thus made being decorated with chiselled interlaced designs in slight relief. At a later time, evidently when the other face of the Cumdach was designed, the silvered plates forming the cross were rudely cut away at the point of intersection for a space measuring 2¾ inches by 2½ inches, in which a metal-work quatrefoil was inserted inclosing an oval setting for a crystal, now lost, corresponding with that on the other face, which will be described below. In its general features, including the bronze plates coated with silver, the frame and cross, and the panels of pierced work, this under face resembles one face of the Cumdach of the Book of Dimma. There, however, the panels are filled with zoomorphic interlaced designs and the central crystal is not a later addition.

63 The following inscriptions are cut in Irish characters round the frame and on the cross, and as there is no reason to doubt that they are contemporary, they enable us, as Dr. J. H. Todd has shown, to fix the date of the Cumdach within comparatively narrow limits. The four inscriptions on the frame are complete, viz.:— (a) At the top,

Bendacht De Ar Cech an Main As a Harilliuth·

64 i.e., "A blessing of God on every soul according to its desert"; (b) At the bottom,

Or Do Dondchad · Macc Briain · Do Rig Herend

65 i.e., "A prayer (orait or oroit) for Donchadh, son of Brian, for the king of Ireland"; (c) On the right,

+ Ocus · Do Maccraith · Hu D Ondchada · Do Rig Cassil

66 i.e., "and for MacRaith, descendant of Donchadh, for the king of Cashel"; (d) On the left,

+ Or Do Dunchad · Hu Taccain Do Muintir · Cluana · Do Rigni

67 i.e., "A prayer for Dunchad, descendant of Taccan, of the family of Cluain, who made it." Donchadh, son of the famous Brian Borumha, for whom prayer is asked in the second inscription, was joint king of Munster with his brother Tadhg until the latter was treacherously slain at his instigation by the tribesmen of Eli ua Cerbhaill, or O'Carroll, in 1023. Subsequently he appears to have claimed to be over-king of Ireland, but he was driven out of the country in 1064 and went on pilgrimage to Rome, where he died in the monastery of St. Stephen. The death of MacRaith, king of Cashel, for whom prayer is also asked, is recorded in 1052, the Annals of Tighernach styling him king of the Eoghanacht Caisil, who were a tribe settled about Cashel in Tipperary, and heir to the kingdom of Munster. From the fourth inscription we learn that the Cumdach was the work of Dunchad O'Taccan, or O'Tagan, of the family, or religious community, of Cluain, by which, no doubt, is meant the famous monastery of Clonmacnois in King's County, but beyond this nothing is recorded of him.

68 The two inscriptions on the cross are unfortunately both imperfect owing to the mutilation it has suffered in the centre, as above described. The letters remaining on the vertical plate are:—

+ Or Do . . . . . Main · H U Cat . . . [Las An]Dernad

69 i.e., "A prayer for . . . . . main, descendant of Cat . . . by (i.e., for) whom it was made." Dr. Todd and Miss Stokes read "nain" for "main" at the end of the first line, but the m is unmistakable. The t in "Cat" in the second line is followed by the remains of a vertical stroke, probably belonging to h; and the final stroke of an n before "dernad" is also visible. Altogether about ten letters appear to be missing in each line, or only seven, if, as is probable, the vertical plate was interrupted by the width of the horizontal plate crossing it. Any attempt to fill in the names must be purely guesswork, but there is no doubt as to the end of the inscription, which is the common phrase used in giving the name of the person for whom, or at whose cost, an object was made. In this case he was presumably some adherent of Donchadh mac Briain.

70 The inscription on the arms of the cross is still more hope-y mutilated, all that is left being

+ Oc[Us Do] . . . . . Ind Hu D . . . . . . . . Laig

71 i.e., "and for . . . . . ind, descendant of D . . . . . . laig." Dr. Todd reads "and" at the end of the first line, but the antepenultimate letter is clearly not a, which is formed quite differently. It is more like i preceded by c or possibly f, only the curved top of which remains, the sloping straight stroke beneath it being evidently a prolongation of the cut made by the mutilating chisel. In the second line the third letter has been read as t. Judging, however, from the roundness of the curve at the bottom, it is almost certainly a mutilated d. 72 As Donchadh mac Briain and MacRaith were both, no doubt, still living when the Cumdach was made, its date must thus lie between 1023 and 1052. It is not said that they had any share in the making of it, and whether their names were inscribed for any special reason other than merely a sentiment of loyalty on the part either of the unknown person for whom it was made or of the actual artificer, Dunchad O'Tagan, it is impossible to say. Reasons have been given above for the belief that the Missal originally belonged to the monastery of Tallaght, near Dublin, in Leinster. The Cumdach, however, has apparently a Munster provenance, and if it was expressly made for the purpose of enshrining the MS., it is not only evident that the latter was regarded two hundred years after it was written as a sacred relic, but it would seem that it was then no longer at Tallaght, but in Munster. Dr. Todd's theory, which has been accepted by others, was that, as the Cumdach was connected by its inscriptions with Munster, the Missal belonged from the first to some Munster church, which he supposed to be the monastery of Lothra, or Lorrha, in Tipperary, going so far, indeed, as to suggest that the book may have been the original Missal of St. Ruadan, the founder. As St. Ruadan died in 585, this is out of the question;

73 and not only is there really no evidence whatever of any early connexion of the Missal with Lorrha, but even the Cumdach is less likely to have come from there than from Clonmacnois, of which house Dunchad O'Tagan, who wrought it, calls himself a member. If the MS., on the other hand, was originally at Tallaght, it may have been carried off in 1026, when it is recorded in the Annals of Tighernach and of Ulster that Donchadh mac Briain invaded Leinster among other parts and exacted pledges or hostages. Books had the merit of being easily portable, and, especially when sanctity attached to them, they were regarded in Ireland as valuable property. How they were utilized in raids and the inter-tribal warfare which was so frequent is shown by a curious note in the MS. known as the Leabhar na h-Uidhri, the translation of which is as follows:— 74 "A prayer here for Aedh Ruadh, son of Niall Garbh O'Donnell, who forcibly recovered this book from the people of Connaught, and the Leabhar Gearr (or Short Book) along with it, after they had been out of our custody from the time of Cathal óg O'Conor to the time of Ruadri, son of Brian [O'Conor] . . . . And this is the way in which they were so taken: the Short Book in ransom of O'Docharta and the Leabhar na h'Uidhri in ransom of O'Donnell's ollamh of history, who was captured by Cathal and carried away as a pledge." These two MSS., it appears, were lost in 1359 and recovered in 1470. As I am informed by Mr. Robin Flower, Laud MS. Misc. 610 in the Bodleian Library (written in 1454) also contains a note which states that Thomas, Earl of Desmond, took the book in ransom for Edmund Butler. Both these are comparatively late instances, but it is by no means improbable that the Missal was similarly taken from Tallaght in pledge or ransom, or possibly as mere booty, during Donchadh's invasion of Leinster in 1026, and that it was enriched with a shrine by the person into whose hands it fell or who was entrusted with its custody. If this was so, the date of the shrine may be placed in that year or soon after, or approximately about 1030, so that it would be nearly contemporary with the more highly decorated shrine of the Gospels of St. Molaise, which was made between 1001 and 1025; and from the inscriptions on its other face it is evident that, together with the MS., it was still in Munster three centuries later.

75 But before dealing with these more recent additions, it will be well to describe the four sides of the shrine, where the whole of the metal-work appears to be of the same date as the older face. The wooden foundations of two of the adjoining sides, which measure respectively 7 inches and 5 3/8 inches in length, and are uniformly 1 1/4 inch wide and 1/4 inch thick, are also original; those upon the other two sides seem to have been renewed at a comparatively modern date. Except on the lower short side or bottom of the casket the bronze plates riveted to the wood have their edges bent upwards; their ends meet at the four corners, but are not connected together. In the centre of each of the short sides (Plates III, IV) a circular metal ornament of identical size and design is nailed to the underlying plate. The outline of the circle, which has a diameter of 1 3/8 inch, is composed of the long slim curved bodies and tails of two beasts, probably meant for lions, whose widely gaping jaws are fixed on either side of the head of a winged human figure, full-face, standing between them. The head of this figure is bullet-shaped, and the nose, if it ever existed, is quite worn away; the hair is indicated by incised lines, and there are also incised lines round the eyes and mouth.

76 The eyes themselves, as in the case of the two beasts also, are represented by tiny black or dark blue beads embedded in the metal. The figure is apparently half-naked, wearing no more than a kilted skirt reaching nearly to the feet, which are bare. In the middle of the breast a small square hole is sunk in the metal and lined with silver wire; in one case it is filled by a shining red stone (garnet?), in the other it is now empty. The arms, if they can be called so, are extended and terminate in what appear to be intended for wings, hanging down perpendicularly. In the upper part of them the metal is hollowed out so as to form a circular cavity, which is inlaid with dark blue cloisonné enamel round an inserted silver wire bent into a spiral. In one wing at the bottom of the casket (Plate IV) the enamel and spiral have perished. At the lower pointed end of the wings the feathers are represented by transverse lines. On the necks and flanks of the two beasts the metal is also cut away, except at the extreme edge, and the cavities are filled with similar blue enamel divided by zigzag and 8-shaped cloisons. Their forelegs rest on the shoulders of the central figure, while the hindlegs extend upwards along the inner edge of the wings. Their shoulders and thighs, through which pass the pins fastening the design to the plate below, are decorated with spiral chasing.

77 The whole design is worked out of a single disc of bronze, which is pierced through its entire thickness between the figures. Traces of gilding with which its surface was covered are still visible. 78 On the top of the casket (Plate III) out of four other small decorated plates which were originally attached to the plain metal backing only one is now left. Its upper and lower edges are straight, but its sides are concave, one of them fitting close up to the circular design above described. The plate is of bronze overlaid with silver, and is cut through so as to leave a spiral and interlaced pattern of metal. At the bottom of the casket (Plate IV) a similar pierced plate remains on each side of the central circle, one exhibiting a chequer pattern, the other, which is imperfect, a pattern of small circles connected by short intersecting diagonal bars. In all these cases a very thin sheet of gilt-foil separates the pierced plate from the bronze backing. The four semicircular end-pieces, which probably contained gilded figures in relief, like that in the centre, have unfortunately all perished.

79 Five small bronze plates, all being rectangular, were similarly attached by nails to the backing on both the longer sides of the casket (Plates V, VI), but one of the end-pieces is missing from each side. On the right of the casket (Plate V) the centre-piece, 1¾ inch in height and 1 6/12 inch in width, has a narrow beaded frame, within which is the figure of a bearded man, full-face, in high relief, between two beasts, who are leaping up at him. He appears to be without clothing, except a loin-cloth or short drawers, and his bare feet project over the frame. He holds in front of him with both hands a short sword, apparently sheathed, the pommel terminating in a semi-spherical knob. The two beasts have beads for eyes and striated bodies with spiral joints at the shoulders and thighs. They most resemble dogs in shape; but the design is probably a debased representation of Daniel in the lions' den. This was a motive common in early Christian art. It frequently occurs on sculptured stones in Ireland and Scotland; and it is possible that the two figures at the top and bottom of the Cumdach already described are another variation of it, though it is difficult to account for the wings. Above these beasts are two others, much smaller and more attenuated.

80 Their feet are on the sides of the frame and their heads and necks are bent backwards as if they were biting or licking the ears of the central figure, while the ends of their tails appear to be in the mouths of the larger beasts. Where the metal is cut through between the figures, the surface below is dimly seen to be covered with a pattern. The single end-piece that remains is 1 inch wide, and has a beaded frame similar to that in the centre, though the beading is less distinct. It contains the figure of a warrior in profile to the right, holding a spear and a round shield with large central boss. His head has a strange appearance, with long hair curling up at the ends. It strongly resembles that of a horseman, with spear and shield, on a sculptured stone on Inchbrayock at the mouth of the S. Esk. Possibly, however, he is wearing a large semi-classical helmet drawn down so as to cover the whole face, with orifices for the eyes, and surmounted by a crest and drooping plume. The figure has no visible clothing except round the loins. It is cut in the round out of the bronze, and the plain metal backing is seen beneath. Both this plate and that in the centre have evident signs of gilding.

81 The other two plates on the same side of the casket are of silvered bronze, and are perforated over their whole surface in patterns, consisting in one case (where there is a sheet of gilt-foil beneath) of spirals and interlaced work, as in Plate III, and in the other of small hollow squares connected by narrow vertical and horizontal bands bisecting their sides, so that the spaces between the squares are cruciform. A similar design is on the under face of the Cumdach of St. Columba's Psalter. 82 The same patterns, or very nearly so, are repeated on two of the plates on the left side of the casket (Plate VI). The former, however, is turned completely round, and in this position is strongly suggestive of a conventional tree with spiral and interlaced branches. Both plates on this side have gilt-foil beneath them. The plate in the centre contains four figures in high relief, but the metal is not, as elsewhere, pierced right through, nor is the design enclosed within a frame. Two of the figures, one on each side, are ecclesiastics vested in a dalmatic and chasuble, standing erect, full-face. Their chasubles are decorated with incised bars filled with gilding and with a border or orphrey of small ovals set side by side, and a similar border fringes the bottom of the dalmatic. The figure on the right holds a pastoral staff with a plain rounded crook, the other an oblong object widening slightly towards the lower end, with a round handle at the top and a small wedge-shaped projection at the bottom. Its surface, except at the edges, is covered with cross-hatching. From its shape it no doubt represents one of the portable bells common in the Irish Church, the projection at the bottom being the end of the clapper. Between the two ecclesiastics are two smaller figures, one above the other. That at the top is an angel with his wings folded in front of him;

83 the other is a human figure seated sideways on a low stool with his head turned to the front and his hands on a three-stringed harp. It is probably meant for David, introduced as a type of Christ, as in a panel under the Crucifixion on a sculptured stone at Monifieth. The eye-sockets in all four figures were filled with beads. The identity of the two large figures is doubtful; but a "pair of ecclesiastics, sometimes standing, sometimes enthroned, sometimes kneeling, with a bird holding a circular disc in its mouth between them, is a subject common to the early sculptured stones of both Scotland and Ireland. This may be another example, with an angel instead of a bird, unless, indeed, the winged figure is really meant for a bird. In the case of a cross at Nigg, in Ross-shire, Dr. Anderson takes the two ecclesiastics to be the hermits St. Paul and St. Anthony, and the bird would then be the raven which brought them a loaf in the desert. This is the more probable, as on the Ruthwell Cross, where the two (without the bird) are sharing a loaf, these names are given in an inscription. On this side of the casket also only one end-piece has been preserved. It represents within a beaded frame a stag hunted by two dogs, upon one of which he is trampling, while the other attacks him from behind.

84 Above the stag's neck is what seems to be the point of an arrow or spear shot or held by a hunter who is not included in the design. There are spirals on the shoulders and thighs of all three beasts, and the stag has serrated antlers and striated flanks. The design is made by piercing the bronze, and the surface is flat and not in relief like that of the centre-piece. In both cases the gilding in some parts is still adhering. 85 We come now to the other face, or front, of the Cumdach (Plate I), the decorative features of which obviously mark a later stage in its history. What became of the original decoration—whether it had perished or was ruthlessly destroyed, or whether after all what is now the under face was originally in the front and was removed to the back to make room for a new front—we cannot say. The wooden foundation of the later face and the bronze plate, or perhaps merely edging, laid immediately upon it, appear to be original, resembling those on the face already described; and the same is perhaps the case with another plate which covers the first up to a short distance from its upturned edges. Silver plates, 3/4 inch wide, are attached to the backing on all four edges so as to form a rectangular frame. They do not, however, meet at the corners, and those at the top and bottom are now imperfect. The inscriptions upon them will be given later on.

86 The centre is occupied by a highly ornate cross, the four extremities of which are soldered down to the plates of the frame. The four limbs of the cross are of bronze with a raised edging and middle line composed of a curb-link chain of fine gilt wire; and small pyramidal bosses, formed by narrow strips of gilt metal with a toothed upper edge turned round a central cone, are set upon the flat surface with other plain round-headed gilt studs between them. The ends of the vertical shaft of the cross are decorated with two oval rock-crystals. The one at the top is in a plain silver setting bent inwards, surrounded by two borders of twisted gilt wire. The setting of the other is deeper and has serrated teeth; round it is a low wall of silver, serrated along the top, with a frame beyond consisting of two rows of twisted gilt wire on each side of a flat surface decorated with minute gilt beads arranged in triangles of three and six alternately. At the end of the arm on the right, in a setting like that at the top, there is a round green chalcedony-like glass bead with a slightly concave surface, and on the left, in a similar setting, there is an opaque white circular object with a central hole drilled through it, perhaps for threading. Its nature is difficult to determine without removing it, but it is possibly a bead of coral.

87 In the centre of the cross is a large oval rock-crystal 1 7/8 inch long and 1 1/4 inch wide and rising 1/2 inch above the surface. It is polished in such a way as to leave a dividing ridge midway from top to bottom and it is backed (as is also the case with the two smaller crystals) with some substance which imparts a rosy tint to it when viewed from above. It is in a plain gilt setting with projecting clamps and is inclosed within a quatrefoil of metal and twisted wire, with the interior of each foil filled with cusped Gothic tracery over green enamel. In the angles outside the quatrefoil there are four stones in raised gilt settings, viz., a sapphire (? at the top and bottom and an almandine garnet on each side. The centre-piece which was inserted at the same time on the other face of the casket (Plate II) is of the same design, but does not appear to have had any gilding. All its five stones have perished, together with the enamel and most of the tracery in the quatrefoil. On the more modern face now being described thin silver-gilt plates are inserted in the four panels between the cross and the framing, the edges of which overlap them and help to keep them in place. A figure is rudely engraved on each of them on a background of toothed horizontal hatching, the incised lines being filled with niello.

88 The two panels on the left contain the Crucifixion and the Virgin, who is crowned, but without a nimbus. She holds the Child on her left arm and in the right hand a small round object, apparently either an orb or an apple. In the upper panel on the right is a saint, without nimbus, holding a book, probably meant for St. John, though the figure is more like a female. In the lower panel is a bishop with chasuble, mitre and pastoral staff with foliated head; his right hand (which is very much out of proportion) is raised in benediction and shows a large episcopal ring. 89 An approximate date for the later face can be ascertained from the inscriptions at the top and on the right of the frame, which read continuously, viz. :—

✠ Or Do P'Lib U [Cinneidig] — Do Rig · Urmu [Man Las] — Ar Cumdaiged · in Mindsa · 7 Do Aini · Dam — Nai ✠ Domnall · O Tolari · Docorig Misi

90 i.e., "A prayer for Philip O'Kennedy, for the king of Ormond, by whom this relic was covered, and for Aine his wife ✠ Domhnall O'Tolari decorated me." Of Domhnall O'Tolari, who names himself as the artificer, and who was probably responsible for the mutilation of the older work, nothing is known, but the Annals of the Four Masters record the death of Philip O'Kennedy, lord of Ormond, and Aine his wife, daughter of Macnamara, under the year 1381. The work must therefore have been executed before that date and may be placed approximately about 1375. The inscription on the left is :

✠ Or Do Gillaruadan U Macan · Don Comarba Lasar Cumdaiged

91 i.e., "A prayer for Gillaruadan O'Macan, the successor, by whom this was covered." Gillaruadan O'Macan, who thus appears to have borne a part of the cost of the later decoration or to have otherwise had some share in the credit of it, is also unknown. The name means "Servant of Ruadan," and it is possible therefore that, as Dr. Todd suggests, he was the comharb or successor of St. Ruadan as abbot of Lorrha, which was in Lower Ormond, in Munster. At the foot of the frame there are fragments of a fourth inscription, the lost portions of which on the right it is impossible to reconstruct. All that remains is : ✠ OR : DO (?) AE . . . . . . C : HU CEI . . . . . . The monogram which follows "Or[oit]" may be read as above, or perhaps "d'; and in the next line hu cei" represents the name "O'Cein," which belonged to a Munster family. This inscription seems to be in a different hand and is possibly somewhat later than the rest; and it will be observed that it includes some characters of Gothic type.

92 Although it may be assumed that O'Conor had sufficient reason for his statement that the Missal with its shrine was discovered on the Continent by John Grace towards the end of the eighteenth century, there is not the slightest evidence to show when and under what circumstances it left Ireland. If it were not for the above inscriptions on the later face of the shrine, it might have been thought that Donchadh mac Briain took it abroad with him when he went on his enforced pilgrimage in 1064; and it has in fact been suggested that it was among the gifts sent to the Irish monastery at Ratisbon by Tordelbach O'Brien, king of Munster, in 1130. As there is proof from the inscriptions that it still remained in Ireland in the second half of the fourteenth century, this supposition is no more admissible than the other, and unless further light comes from some unexpected quarter the history of the MS. from that time until it came into the possession of John Grace must remain an absolute blank.

93 In conclusion, it is my pleasing duty to express my cordial thanks to those who have assisted me in this edition of the Stowe Missal in various ways. I am much indebted for criticism and suggestions to the Rev. E. S. Dewick and the Rev. H. A. Wilson, and to Mr. J. A. Herbert for his careful transcript of the MS. On Irish palaeography I have had valuable aid from Professor W. M. Lindsay, of St. Andrews; on Irish hagiology and philology from the Rev. C. Plummer and my late colleague Mr. Robin Flower, who is cataloguing the Irish MSS. in the British Museum; and in connexion with the Cumdach from Mr. O. M. Dalton and Mr. Reginald Smith, of the British Museum, and the Rev. Dr. Lawlor, of Dublin. Finally, I have to thank the Council of the Henry Bradshaw Society for much indulgence in the long delay in completing the present volume.

94 G. F. W.