Fragments of Patristic and Other Ecclesiastical Literature in Norway

Fragments of Patristic and Other Ecclesiastical Literature in Norway from c. 1100 until the Fifteenth Century Espen Karlsen 1. Introduction The focus of research on surviving Latin manuscripts from medieval Norway has been mostly upon liturgical books. By contrast, patristic and other non-liturgical ecclesiastical books have been largely ignored. This article is intended to act as the first step towards filling this gap. It will discuss those manuscripts that contain patristic and post-patristic theological and devotional texts that once usually belonged to ecclesiastical institutions and the clergy but that did not have a liturgical function, i.e., library books.1 Also excluded from the discussion are ecclesiastical (canon) law manuscripts of which a small group of approximately seventy fragments are in the NRA.2 The term ‘patristic’ in this article refers to those authors known as the Fathers of the Church, going up until the time of Isidore of Seville (✝ 637).3 For convenience, the term ‘library books’ is used in what follows to refer to all the manuscripts under consideration. In what follows I will identify fragmentary texts, give an approximate date to the manuscripts and locate the place of copying, mostly using the evidence of the script. Earlier discussions of the holdings of such literature in medieval Norway have mostly been based on documentary evidence and the literary sources drawn upon by medieval authors.4 This article concerns the remains of copies of such literature in the NRA.5 What is left of them today is mostly leaves and smaller pieces used in the binding of account books from the sixteenth and seventeenth centuries (see Pettersen in this volume). In most cases, all that survives from each manuscript is one or a few fragments. Of the manuscripts discussed or described below, there are only two with several complete leaves or halfleaves surviving, including a copy of Isidore’s Etymologiae (‘Etymologies’) and a comprehensive commentary on the Psalms comprising a partly expanded and partly abbreviated version of Augustine’s Enarrationes in Psalmos­(‘Interpretations of the Psalms’). Most of the fragments of ecclesiastical library manuscripts are from authoritative and learned exegetical texts of the church fathers, in particular Augustine and Gregory the Great, as well as some later ecclesiastical authors. Also included among the patristic works is the sixth century * 1. The plates illustrating this paper are reproduced at same size unless stated otherwise. Johnsen (1908) published a survey based on surviving documentary evidence of what may be known about the content of ecclesiastical libraries during the late Middle Ages. Johnsen was aware that fragmentary manuscripts in the NRA and other institutions might provide further information about books in medieval Norway. 2. These fragments form a small group together with the fragments of Roman law, of which c. sixty fragments have been identified. The earliest known witness of canon law so far is an early version of Gratian’s Decretum copied by a very good scribe c. 1130–1160 (NRA lat. fragmenter 154). For Ivo of Chartres’ (c. 1040–1116) Panormia and an Anglo-Norman collections of decretals, see Gullick 2, nos. 82 and 85, in the present volume. Most of the Latin law fragments are collected in Eske (i.e., Box) 8 of the fragments collection in NRA. 3. It is possible that the Norman monk Robertus de Tumbalena’s commentary on Cantica canticorum, composed in the eleventh century, which is preserved in a fragmentary twelfth-century manuscript, was in this copy ascribed to Gregory the Great, as was frequently the case during the later Middle Ages, an attribution also given in the edition published in PL (vol. 79). These fragments will be discussed below. 4. For references, see below. 5. Sørensen (1910) is so far the only published presentation of some of this material of which I am aware. He quotes extensively from, and translates, some of the Latin texts, and identifies others. 216 espen karlsen Vitae­patrum (‘Lives of the fathers’, i.e., of the desert fathers), a non-exegetical work that is similar in content and compilation to later medieval collections of exempla. A text composed in Norway of a different nature is Brother Mauritius’ description of a journey from the west coast of Norway (probably Selje) to the Holy Land in the 1270s (Storm [1880] 1973, xxxxvii). Mauritius is the only Norwegian author found in the fragments in the NRA. Since the fragments of his work in the nineteenth century were regarded as more interesting than the liturgical manuscripts, they were located in the collection of Old Norse fragments.6 There are more surviving fragments of liturgical books than of other genres. We may assume that the priority during the first century or so of the establishment of the church and Latin learning in Norway was the provision of books for the liturgy. It seems likely that patristic and other library books only began to be imported during the first half of the twelfth century.7 This would correspond with the fact that, although some liturgical books had arrived in Norway by at least c. 1000 or early in the eleventh century,8 the composition of Latin literature in Norway only began in the second half of the twelfth century.9 Since the surviving ecclesiastical library books constitute so small a group compared with the liturgical manuscripts, the material does not allow for so many connections to be made between the fragments with regard to scribal identification and scriptoria. The bulk of the fragments of ecclesiastical library books fall into two main chronological groups, first, twelfth-century manuscripts and, secondly, manuscripts dating from the late thirteenth or early fourteenth centuries. A couple of late-medieval manuscripts contain versions of earlier texts that are either expanded or abbreviated, or both expanded and abbreviated. I mentioned above that library books mostly belonged to monastic houses and cathedral chapters.10 A bifolium from a late twelfth- or early thirteenth-century copy of Donatus’ (fourth century) widespread grammar Ars­minor, is remarkable in two ways. First, it was found in a stave church in the small settlement of Lom, and secondly, the copyist was exceptionally negligent of Latin. A copy of Ars­minor might (just like patristic and theological books) be expected to be connected with a centre. As Hohler commented: this sheet from Lom seems, alas! to show that Latin was taught widely, stupidly, and incompetently. For it is, as observed, nicely written; and the copyist has made a number of corrections, meaning he looked twice at his model to be sure he had copied it right. The ridiculous errors presumably come, therefore, from his model; which he must, therefore, have regarded as authoritative. His own knowledge of grammar must have been nonexistent, or he would have seen he was writing nonsense and changed things (Hohler 1978, 173).11 6. The same goes for some other Latin fragments that were regarded as particularly interesting in the nineteenth century. For instance a manuscript of Sallust’s Bellum Iugurthinum was included in the collection of Old Norse fragments (NRA norrøne fragmenter 93, 1–14). 7. Some scholars have argued that the Latin alphabet was used in written versions of laws in the Norse vernacular as early as the first half of the eleventh century (Eithun, Rindal & Ulset 1994, 9–12; Rindal 2002, 802), i.e., more than a century earlier than the oldest surviving vernacular material in the Latin alphabet. This presupposes the production of Latin books at an earlier date (Mortensen 2006). As there appear to be no remains of Latin manuscripts copied in Norway earlier than the second half of the eleventh century or c. 1050 as the earliest acceptable date (Karlsen 2003, 64–70), the production of vernacular books before this is highly unlikely. No argument so far has been produced to substantiate the claim that books in Old Norse were written in the first half of the eleventh century. For further criticism, see Mortensen’s extensive footnote 16 in Mortensen (2006, 255–256) with further reference to Helle (2002). 8. One of the earliest extant liturgical manuscripts almost certainly used in Norway is an Anglo-Saxon mass-lectionary (Lec-Mi 1) dating from the first half of the tenth century (Gjerløw 1957, who dated it in the second half of the century; Dumville (1991, 49) suggested an earlier date). One small fragment of this manuscript is now in the NRA, attached to an account book (NRA lat. fragmenter 201) and a larger piece of a leaf that also remained locally (now in NB, Ms.lat.fragm. 9). Other liturgical manuscripts dating from the first half of the eleventh century, mostly of English origin, also survive. 9. There is no evidence of Latin literature composed in Norway prior to 1150 with the possible exception of the account of miracles connected to the tomb of St Olav (Mortensen & Mundal 2003, 359; Mortensen 2006, 256). The earliest known liturgical texts relating to St Olav are found in English manuscripts from the eleventh century (Iversen 2000, 405–411). 10. Map 1 (p. 37) indicates such centres. 11. This fragment, subsequent to Hohler’s publication, was transferred from the NRA to the Museum of Cultural History at the University of Oslo (signature: C34738/B305/605, olim NRA Løsfunn Lom B 305), together with other fragments found at Lom. fragments of patristic and other ecclesiastical literature [karlsen 2] 217 The errors in the copy may also point to its being copied outside a large centre and possibly in a small place, such as Lom.12 Consequently, the possibility that the books may have been used in other contexts than in monastic houses and cathedral chapters cannot be entirely ruled out. Without such clear evidence of its provenance this copy of Ars­minor would probably have been assigned to a more important centre (Hohler 1978, 174). The few manuscripts to be discussed here must represent only a small part of what once existed in medieval Norway. The fact that there are no examples (with one exception) of Norwegian authors among the fragments underlines the degree of destruction of medieval manuscripts in Norway. As for the few such texts that have been preserved, they depend on very few textual witnesses surviving abroad and, in two cases, on post-medieval copies. Of the four major works in Latin that have been preserved, the Historia Norwegie, the above-mentioned work of Theodoricus Monachus, the Passio Olaui,13 and an anonymous account of a Danish and Norwegian expedition to the Promised Land, no Norwegian witnesses survive.14 Taking into consideration the fragile transmission history of Latin works from Norway, it is not unlikely that other works of the twelfth and thirteenth centuries may have been lost without leaving a trace.15 It is odd, however, that no copy of the works known to us is mentioned in medieval documents such as book lists etc. (cf., e.g., Johnsen 1908).16 In what follows, I shall briefly outline (in Section 2) the wider European context of book production and ecclesiastical libraries during the twelfth century, and provide a brief introduction to the Norwegian material on the basis of what has already been derived from study of the surviving liturgical manuscripts. Subsections 2.1 and 2.2 will examine the evidence for influence during the twelfth century from Germany and Lotharingia on the one hand, and France and England on the other. Section 3 provides an introduction to the patristic and theological manuscripts from the thirteenth and fourteenth centuries. Section 4 comprises a description of the fragments, organised alphabetically by author, and, within each author, not by text but in chronological order 12. 13. 14. The scribe of the fragment from Lom is not the only example among the Norwegian fragments of a scribe who did not understand what he wrote. The same goes for two leaves of a notated missal (late eleventh or early twelfth century, written in an English style of handwriting), in which word divisions are absurd (NRA lat. fragmenter 202 = Mi 3 in Gjerløw’s typewritten catalogue in the NRA; Hartzell 2006, no. 230). This indicates that the scribe did not understand what he copied (Gjerløw 1974, 77– 81; Karlsen 2003, 66). The missal may have been copied locally. The two leaves were never used in the binding of account books, but remained locally after the Reformation (Gjerløw 1974, 77– 79). However, there are not, to my knowledge, many instances among the Norwegian fragments of scribes who appear to have lacked any understanding of what they were copying. On the medieval literature on St Olav in Latin and Old Norse, see Mortensen & Mundal (2003). Theodoricus Monachus’ Historia­de­antiquitate­regum­Norwagiensium and the anonymous Historia­de­profectione­Danorum­in­Hierosolymam (late twelfth century) are both preserved in seventeenth-century manuscripts copied from a now lost medieval codex discovered in Lübeck in 1625 (Skovgaard-Petersen 2002; see the discussion in footnote 43 below). The Historia­Norwegie is only partially preserved in a codex copied in Scotland c. 1500 (Ekrem & Mortensen 2003, 28–31). The main witnesses to Passio Olaui were all copied outside Norway (for a list of manuscripts, see Jiroušková 2010, 238). The four major texts in Latin are thus only preserved abroad. Although these works are not represented among the fragments, it is likely that Norwegian-owned copies were lost in the century after the Reformation, and that either the parchment was reused for binding or other purposes such as fireworks (see Pettersen in this volume). It is also possible 15. 16. that the loss of copies of these texts began before the Reformation, due to the political situation prevailing under the unions of the fifteenth century, as a result of which there may have been little or no interest in Norway’s past among the political elite. The shorter texts concern inter alia the legends of the saints Hallvard and Sunniva and a text about the foundation of Lyse monastery outside of Bergen (published by Storm [1880] 1973; a more complete text of the Sunniva legend is published by Borgehammar 1997). The legends of Hallvard and Sunniva are preserved in liturgical fragments and the printed Breviarium Nidrosiense (Paris 1519). Ommundsen (2010c) surveys the four different known versions of the Hallvard legend. For a brief account of a journey to the Holy Land by Mauritius, see below. See Mortensen (2005) on the role of Latin in medieval Norway. As far as the Old Norse codices are concerned, it has been emphasised that the loss was greater in Norway than in Iceland. This disparity has been explained as the result of the different linguistic development in the two countries in the late Middle Ages (the linguistic change in Norway was more radical than in Iceland; see Rindal (2002, 804) with further reference to Eyvind Fjeld Halvorsen). It is clear that the scale of loss in Norway is just as great so as far as the Latin manuscripts are concerned, but this cannot be explained by linguistic change. The loss of medieval manuscripts in Norway was, therefore, considerable regardless of language. It should also be borne in mind that the number of Old Norse manuscript books that survive from medieval Norway is also low as well. About 130 Norwegian codices (complete as well as in fragments) produced before 1370 are still extant (see the introduction to Guðvarður Már Gunnlaugsson in this volume). 218 espen karlsen according to the date assigned to each fragment on the basis of its script. Section 5 includes further discussion of the list and a conclusion. 2. Patristic and ecclesiastical library books from the twelfth century The twelfth century was the last great age of monastic book production in Europe. Comprehensive collections of patristic and ecclesiastical texts became widespread, and included large exegetical commentaries on biblical texts in several volumes. A canon consisting of inter­alia Augustine’s De­Genesi­ad­litteram, In­Ioannis­evangelium­tractatus and Gregory the Great’s Moralia­in­Iob had for a long time been predominant on the Continent, and during the century after the Norman Conquest the same canon also became adopted in England (Ker 1960a, 8).17 The twelfth century was also a transitional period in the history of the development of handwriting, as the Caroline minuscule that had been written throughout most of western Europe as well as Norway until c. 1100, gradually changed its proportions and incorporated new features of style, developing into different varieties of script by the end of the century (Derolez 2003, 57–71). As we shall see, there is evidence in the 1100s of the presence of the same patristic texts in Norway as on the Continent and in England. As is the case for liturgical books, the twelfth century was an exceptionally rich period for the importation and production of ecclesiastical library books in Norway (Karlsen 2006, 20–21). The earliest physical evidence of such manuscripts in Norway dates from the late eleventh century or c. 1100. These fragments are the remains of manuscripts copied abroad. As with the liturgical manuscripts, it is hard to determine whether these books were brought to Norway a short time after they had been made or somewhat later (cf. Rankin in this volume). To judge from the surviving fragments I think that it the importation of such books may have begun early in the twelfth century, before the establishment of a Norwegian literature in Latin in the second half of the century (e.g. the Historia­Norwegie and the work of Theodoricus Monachus), 2.1.­Importation­of­ecclesiastical­library­books:­Germanic­influence There is not much material among the eleventh- and twelfth-century liturgical fragments in hands which display Germanic features that could bear witness to influence from the diocese of Hamburg-Bremen, to which Norway belonged until 1104. (The term Germanic is here used in its wider sense to refer not only to present-day Germany itself, but also adjacent areas such as Lotharingia and the Netherlands.) There are a few liturgical items from the eleventh century that are clearly Germanic or copied elsewhere but follow Germanic traditions of writing Caroline minuscule. Mi 100a (NRA lat. fragmenter (Unum. III) Tillegg til skapsakene pk. 2.-2)18 comes from an imported missal dating from the second half of the eleventh century, written by a competent scribe whose handwriting points to the Rhineland (plate 1). There is also a very unprofessionally written notated missal (Mi 17 = NRA lat. fragmenter 213, 1–4), the text hand of which reflects the influence of the same kind of handwriting.19 Its combination of a low quality Caroline minuscule of a Germanic variety with exceptionally low quality Anglo-Saxon neumes in­campo­aperto (according to Susan Rankin; see Karlsen 2003, 66– 67) suggests that this is probably one of the earliest known books to have been copied locally (plate 2).20 17. 18. 19. A recent study of changes in English book culture after the Norman Conquest is Thomson (2006). The provenance is a cadaster for the church in Tønsberg 1591/1631. The secondary provenance is account books for Fredrikstad, Moss and Åbygge in south-eastern Norway. 20. For this missal, see Karlsen (2003, 66–67 with pl. 9 on p. 79); Gjerløw (1974, 85–86 with pl. 6). Gjerløw dated it to the second half of the twelfth century, which is too late for the style of handwriting. The second half of the eleventh or the early twelfth century is a more likely date. fragments of patristic and other ecclesiastical literature [karlsen 2] 219 There is also a small group of missals written in hands that reflect Germanic or Lotharingian traditions of writing, namely Mi 10b–f (NRA lat. fragmenter 219 & 222). They date from the late eleventh or early twelfth century and have a provenance within the medieval diocese of Oslo. The scribe who wrote Mi 10c also copied Mi 10f and is also the second scribe of Mi 10e. Mi 10b and 10d clearly belong to the same family to judge from their script and the similarity of the light orange coloured ink used in rubrics and initials. They probably bear witness to an early local scriptorium (Karlsen 2003, 68 with pls. 13–14).21 Of Germanic high quality books there survives a fragment of a benedictional for a bishop in large format datable to somewhere between 1050 and 1150, but most probably from the eleventh rather than the twelfth century (NRA lat. fragmenter Unum. eske I Raj 192/1958; in Gjerløw’s typewritten catalogue it carries the designation Ben 2). Another manuscript that must have belonged to a bishop is a fragment of a pontifical in the NB (Oslo) dating from c. 1070 (NB Ms.lat.fragm. 16 = Pont 2 in Gjerløw’s typewritten catalogue in the NRA). These examples of Germanic influence all have a secondary provenance within the boundaries of the medieval diocese of Oslo, which is not surprising due to its geographical position. Moreover, three of the four Germanic manuscripts from the twelfth century to be discussed below also have a secondary provenance that points to the medieval diocese of Oslo. Among the earliest surviving patristic manuscripts are two manuscripts of Augustine, both of them written on a Germanic model, one of the Moralia­in­Iob of Gregory the Great, and one of the Verba­seniorum­(Sayings of the Elders),­a part of the Vitae patrum. Three of these are in folio format, as might be expected for copies of Augustine and Gregory the Great at this time (Ker 1960a, 40–41), whereas the Verba­seniorum manuscript is in quarto format.22 2.2.­Other­influences­in­the­twelfth­century:­France­and­England One of the oldest manuscripts of ecclesiastical literature is two fragments of a book in large format containing Hrabanus Maurus’ (780–856) Commentarii­in­Genesim. The script reflects a French or Germanic model dating from the late eleventh or early twelfth century. Another early witness of French influence is a copy of Rufinus’ Historia ecclesiastica. After c. 1150 there is further evidence of French influence, namely a copy of the commentary on the Psalms of Gilbert of Poitiers. There are also substantial remains of a copy of Isidore’s Etymologiae dating from the middle of the century or a bit later, which is in a French-looking hand. Among the liturgical material, English influence predominates in the eleventh century. This influence continued to remain strong in the twelfth century. The century after the Norman Conquest has been called the greatest period in the history of English book production, from which are preserved large numbers of books (Ker 1960a, 1). The standard of writing was also high ‘because the letter-forms of the script are essentially simple’ (Ker 1960a, 1). The books were accurately copied and competently written. From this golden age of English book production there survives the remains of one leaf of a large format copy of Augustine in a round English hand (see below). Further English or English-looking material is present among the fragments that date from the second half of the century 21. A minimal definition of scriptorium may be ‘a centre where (according to the surviving evidence) at least two scribes working in conjunction wrote significant amounts of the main text in at least two manuscripts’ (Thomson 2006, 23, footnote 19). This early local scriptorium seems to have involved more than two scribes. 22. The Verba­seniorum collection may have served a practical function as a collection of exempla for composing sermons. Thus a smaller format must have been practical. 220 espen karlsen 3. Patristic manuscripts from the thirteenth and fourteenth centuries There are few books from the first half of the thirteenth century. A copy of Liber­de­pastoribus­et ovibus by Hugo de Folieto appears to have been made locally (see Wifstrand Schiebe & Karlsen in the present volume). During the second half of the thirteenth century and the first half of the fourteenth there was a small renaissance in the production of manuscripts of patristic and other ecclesiastical authors. A couple of the manuscripts turn out to contain texts that are to be either expanded or abbreviated (or both simultaneously) versions of the original texts, or the versions we find in modern critical editions. Some patristic works, such as Gregory’s Moralia­in­Iob, were simply too unwieldy for some later audiences without some element of abbreviation and alterations. This happened not only, as Gorman (2001, 41) has suggested, because readers not easily any longer could master the Latinity of patristic writings created under the influence of classical culture, but also because of the expense of production, and, in particular, the cost of parchment.23 In the early fourteenth century four volumes of Roman law that belonged to the church of St Mary, probably in Oslo, were estimated to be worth a price of 42 mark­forngild, i.e., the price of a farm that could provide for a family (Karlsen 2005, 151–152).24 This means that a huge commentary on the Psalms from the second half of the thirteenth century, probably in several volumes (of which there are substantial remains), must have been an extremely expensive acquisition. 4. Alphabetical list of ecclesiastical literature arranged by author 4.1.­Introduction­ The list is arranged by Latin names of the authors, with the usual English form (when such a form exists) given within parentheses. De­obseruatione­monachorum by Athanasius’ is dealt with under the Vitae­patrum, as the fragments of the book in question contain both texts. Some items on the list are dealt with more extensively than others. There are fragments that still await identification; moreover, some fragments have come to my attention only recently. Some of the fragments discussed below were identified by Gjerløw in her records in the NRA; in a few instances (see below under Aurelius­Augustinus­and under Hugo­de­Folieto) her identifications have been revised. I have included reconstructions of some leaves which were cut into small pieces (cf., e.g., plate 6 of a large format copy of Augustine’s De­ciuitate­Dei). Reconstructing manuscript leaves from small pieces is a demanding and time-consuming task and, since such reconstruction is hard to describe in words, some of these are reproduced among the plates.25 Reconstruction is, in some cases, the only means of determining the actual size of the leaf. I have reluctantly chosen not to include fragments of Chalcidius’ translation and commentary on Plato’s Timaeus with text from Cicero’s Tusculanae disputationes26 (Lehmann 1937a), as they were removed from account books from Denmark.27 The medieval context for the reception of these texts was clearly christian. The fragments were once part of a manuscript copied in France in the first half of the twelfth century. These fragments, NRA lat. fragmenter Eske 45, II, 1–5, were published by Paul Lehmann (1937a) who regarded this commentary to be connected with 23. As there are some items like this among the Norwegian fragments, at least some Norwegians must have been able to master such works from the twelfth century on. 24. In documents from Norway from the thirteenth and fourteenth centuries the price of a psalter varies from three to seven cows (Karlsen 2005, 152). 25. When cataloguing fragments in the database in the NRA I developed a system of reconstruction that will enable the users to reconstruct the leaves (see footnote 33 below). 26. This manuscript was filed under Pseudo-Cicero in the records of Gjerløw in the NRA. 27. They were removed from account books from Vallø (1634), Vallø and Lellinge farms (1634–1635), Vallø (1636), and Vallø (sine anno) (Lehmann 1937a, 59). Lehmann suggested that the manuscript may have belonged to the monastery of Dalum. fragments of patristic and other ecclesiastical literature [karlsen 2] 221 the cathedral school of Chartres (1937a, 61). I have also excluded from my discussion a fragment of Petrus Comestor’s Historia­scholastica (NRA lat. fragmenter 60b-1), as it was removed from an account book from 1629 for Helsingør in Denmark. Two fragments of Magister Mathias’ Alphabetum distinccionum are discussed below despite the fact that they are the remains of a Swedish codex taken apart at Vadstena Castle in the years after 1595. They were removed from an account book for Akershus of 1621. The reason for including this manuscript is that the discussion throws light on the problematic issue of provenance. Further observations concerning post-medieval provenance are included in the list below under Aurelius­Augustinus d), Gregorius­Magnus a), Mauritius, Rufinus­and Robertus­de­Tumbalena. 4.2.­The­list­ Anonymus There are four fragments of the same leaf of a text of a lexicographical nature (NRA lat. fragmenter 1a: 1–4). They are reproduced on plate 3. The text is arranged in two columns, each 7 cm wide. With most of the upper margin missing the leaf now has a height of 28.5 cm and a width of more than 20 cm. The height of the writing space is 21 cm. This was once a large book. There are simple initials in red, green and yellow. Yellow initials are rare at this time and may be a local feature. The text is above top-line (Ker 1960b). The letter forms reflect a continental style of handwriting, and the character of the tick-and-point form of medial punctuation points to France (1r, right column, l. 15, reproduced on plate 3; 1v, left column, l. 26). The shaft of the d is bent (). Round s occurs, though not exclusively, in final position. Elsewhere tall s is preferred. Its descender turns to the left (∫) which may reveal the influence of early documentary script (Derolez 2003, 64). The ascenders of b, h, l have long serifs to the left and slightly to the right (cf. Derolez 2003, 79 on the phenomenon in Northern Textualis). Note also the final hairlines on r and e at word endings. The manuscript was probably copied in the fourth quarter of the twelfth century. The place of origin may have been France, but it is also possible that it was copied locally under continental influence or by a scribe trained on the continent. Some passages of the text were published and translated into Norwegian by Sørensen (1910, 33–35). The text contains lemmata that are interpreted in a manner that reflects christian teaching. The leaf includes (1r) Ocium, Quies, Litus, Pax­eterna, (1rv) <A>ridum, 1v Arena,­Sabulum­Litus, Portus,­Limus. I have not been able to find the text elsewhere, though I have not made a thorough systematic search. The purpose of the text was probably to support the composition of sermons. Aurelius Augustinus a) De­Genesi­ad­litteram The fragment with the designation NRA lat. fragmenter Eske 46, 35 is a large double leaf of which fol. 1v contains the beginning of Augustine’s De­Genesi­ad­litteram­(On Genesis word for word),­a work that has been called his ‘masterpiece of philosophical exegesis’ (Gorman 2002, 13).28 It was a huge work that comprises 501 printed pages in Zycha’s Vienna edition (1894). The first page of the fragment is blank with the exception of a note giving the name of the administrative district of the account book to which the fragment was attached, namely Bratsberg 28. This fragment is not included among the 195 manuscripts and fragments of manuscripts containing De­Genesi­ad­litteram known to Michael M. Gorman (2001, 1–5). As Zycha (1894, VI) observed regarding the number of manuscripts in the preface to his critical edition, ‘saeculum duodecimum et proxime sequentia libris manu scriptis abundant’ (‘the twelfth and the following centuries abounded with handwritten books’ sc. containing De­Genesi ad­litteram). 222 espen karlsen len 1586–1587.29 The format is large, as was usual for a manuscript of Augustine. The leaf measures 34.5 × 36.5 cm. The text is arranged in one column and measures 28.5 × 21 cm. There is one red initial (see pl. 4) and a rubric in the same red colour. The script, a late Caroline minuscule from around 1100 or a bit later, is Germanic in character, possibly from lower Rhineland. The letters lean slightly towards the right, a bit more so than the script in Mi 100a mentioned above. The following letter forms are continental: a, g, &. The tall s () is used throughout and extends in some cases slightly below the baseline, which may be an early feature. E­caudata (ę), i.e., e with a tail, which at this stage was frequently used for the classical diphthong ae (Bischoff 1990, 122), is used. The ligature ſt is regularly used. The combination re forms a ligature. The letter t connects with the preceding and following letter (e.g. nti, ate). The final stroke of r also connects with the following letter (ra, re). The rubric on fol. 1v (Liber­sancti­Augustini­super­Genesim­ad­literam) is a twelfth-century addition. The fact that the text is in one column in a manuscript of this size may be an indication of a date prior to 1100. As for the content, fol. 1r is blank, whereas fol. 1v contains the beginning of De­Genesi­ad­litteram (1,1–1,2 = Zycha 1894, 3, line 1 until p. 5, l. 27). Fol. 2rv contains 2,6–2,9 (= Zycha 1894, 40, l. 11 until p. 46, l. 12). This means that once there were probably three bifolia between fol. 1v and fol. 2r. b) In­Iohannis­evangelium­tractatus Two fragments are still attached to an account book from Akershus len, Hedmark and Østerdalen 1612, i.e., from the area north of Oslo (signature of the account book: RK Lensrsk. Akershus len, landskatt 1612, Hedmark og Østerdalen, 31.6, 27). This means that the book they once belonged to may have been used within the borders of the medieval diocese of Oslo. The two fragments are from the same leaf of a manuscript that contained Augustine’s tracts on the Gospel of St John (In­Iohannis­evangelium­tractatus), tract 99. The fragments both measure approximately 9 × 11 cm. The width of a column is 9.5 cm (cf. the lower fragment on plate 5 (incipit: equalis­est)). On both fragments there are thirteen lines of the text preserved. The manuscript had two columns as there is the trace of a second column on the lower fragment (see plate 5). On the verso side of lower fragment there is trace of a plain red initial. The script is a Caroline minuscule that is hard to date with precision. The manuscript is clearly Continental from either the second half of the eleventh century or first half of the twelfth. A cautious suggestion would be the early twelfth century, as there are some late features. The ampersand and the letter a have continental forms. E caudata occurs. The final stroke of e slopes upwards similar to a hairline (Derolez 2003, 61 with example 24). This may be an indication of a date after 1100. The tall  as well as f extending below baseline, however, may be an early feature (cf. Derolez 2003, 49–50 and 61). Major medial punctuation is by punctus­elevatus. The recto sides of the two fragments correspond to Willems 1990, 585, l. 38–47 aequalis­est . . . essentia,­ab­illo­sci<entia>; pp. 585–586 <cuius>libet temporis­uerba­. . . Numquid­duo­sunt, alius; the verso sides to p. 586 apparet­amborum­. . . eo­loco­clarum­est,­ubi­angelus; p. 587 <ta>libus quales nos sumus­. . . in­alium­sunt­differenda­sermonem). In­Iohannis­evangelium­tractatus is preserved in many copies, albeit many of them are incomplete (Willems 1990, VIII). If the codex contained a complete version of this text, it probably 29. Pettersen has observed that before 1610 large leaves or double leaves were probably used as wrappers on account books (Pettersen 2003, 50). Unfortunately many of the account books predating 1610 were discarded in the 1720s, but it is usually the case that the fragments preserved from early account books tend to be complete or almost complete leaves or double leaves (see Pettersen in this volume). fragments of patristic and other ecclesiastical literature [karlsen 2] 223 occupied a substantial part of it, if not the complete manuscript (the Latin text occupies 688 pages in Willems 1990). c) De­ciuitate­Dei These fragments are from a competently written English manuscript that contained Augustine’s De­ciuitate­Dei­(NRA lat. fragmenter 51, 1–7). The seven fragments are small and were part of the same large leaf which, reconstructed, measures approximately 40 × 31/32 cm and had approximately 43 lines in two columns (see plate 6 for the reconstructed leaf). The leaf contains text from the fourteenth book, chapters 13–17 (fol. 1r corresponds to Dombart & Kalb 1981, vol. 2, p. 33, l. 28–p. 37, l. 2; fol. 1v corresponds to Dombart & Kalb 1981, vol. 2, p. 37, l. 3–p. 40, l. 9), of this comprehensive work. Only slightly more than half of the text of the leaf has been preserved. The text is in a small, round hand that is typically English from the first half of the twelfth century. There is one plain green initial with no decoration marking the beginning of Chapter 15. There is frequent use of e caudata. The individual letter forms are typically English, e.g., the letter a has in many cases an ample curving head, especially at the beginning of words. Gullick (in Gullick 2, no. 6 in this volume) draws attention to a gospel book written at St Albans about the middle of the first half of the twelfth century where the hand of the main scribe is quite close to the scribe of this fragmentary leaf (see Mynors & Thomson 1993, pls. 3a, 3c, and 3d). d) An extended adaptation of Enarrationes­in­Psalmos The best documented manuscript now in Oslo that is relevant to this article is a huge commentary on the Psalms under the designations 34 (from account books from Sogn, i.e., Bergenhus), 53 (Stavanger), 54 (Stavanger), and 677 (Stavanger). The text is based upon Augustine’s Enarrationes in­Psalmos. Full signatures are listed in the leaf by leaf reconstruction below. There is evidence of eleven leaves. There are five leaves with their complete written space surviving (fols. 3, 4, 6, 7 and 9 in the reconstruction below). The writing space measures 27.5/28 × 19 cm and is divided in two columns 9 cm wide. This text was written on rather white parchment. The height of a leaf measures 36.5 cm, whereas the full width, to judge from the reconstruction must have been approximately 30 cm. The eighth quire contains comments on Ps. 103, which indicates that it must have been in a multi-volume copy in large format. A copy of Augustine (or an adaptation of Augustine of this kind) of this date would probably not have been imported. Considering its size this was probably not a private book, but belonged to an ecclesiastical institution.30 As I will argue below, it is likely that this institution was situated in or not very far from Stavanger in the late Middle Ages. The preserved parts of this huge book are written in textualis (see plate 7) by two or more different scribes whose hands are similar. The letter a is found with an open upper bow. This may indicate a date not later than c. 1300, as the upper bow tended to be closed from the fourteenth century on, but there are several instances with a closed upper bow as well in this manuscript. This a was the usual form in what Derolez has called Northern Textualis (Derolez 2003, 84).31 In the second half of the thirteenth century, and from c. 1300 it had replaced the a with an open upper bow completely (Derolez 2003, 84). Fusions (‘bitings’) are, as one would expect, frequent (e.g., e, be, o, o). There is a frequent bifurcation at the top of b, h, k, l (Derolez 2003, 79). Tall  is predominant except in final position, which would indicate a date at the latest in the first half of the fourteenth century. The ampersand is used instead of the Tironian nota. The quality of the initials is low. Punctuation is by punctus (.) and punctus elevatus () (Parkes 1992, 30. It is likely that books belonging to ecclesiastical institutions would have ended up as fragments on public account books. Parchment from privately owned books seems to have ended up elsewhere. 31. The adjective ‘northern’ in this context does not only refer to the Nordic countries but also to England, the Low Countries, much of France, Germany and Central and Eastern Europe. 224 espen karlsen 306). There is no flex punctuation.32 This might indicate a Benedictine or cathedral chapter origin rather than a Cistercian one. The book appears to have been professionally made probably from a full copy of the text. The script in itself does not reveal whether the book was copied in Norway or elsewhere. There is nothing in the script to contradict a Norwegian origin. This kind of book was frequently copied in England and on the continent in the twelfth century. The late date makes it more likely that it probably was copied locally in Norway or elsewhere in Scandinavia. The copy is, in fact, an expanded as well as in some parts an abbreviated version of Augustine’s Enarrationes­in­Psalmos which comprises c. 2250 printed pages in the edition in the Corpus­Christ­ianorum series. Long passages of the text of the fragments are more or less identical with the printed Enarrationes­in­Psalmos. The fragments from this large commentary also illustrate the problem of secondary provenance, as we shall see below. A full leaf by leaf reconstruction with the secondary provenance added is as follows: Fol. 1rv: Ps. 86 (54, 1; 54, 2; 54, 3; 54, 4).33 Secondary provenance: Stavanger 1614 Fol. 2r: Ps. 88, 24–29; fol. 2rv: Sermo II per vigilia sancti Cypriani (34, 6; 34, 10; 34, 11; 34, 7; 34, 8; 34, 9). Secondary provenance: Sogn 1636, Sogn Fol. 3rv: Ps. 103,22–Ps. 103, 26 (34, 5). Secondary provenance: (Jørgen Kaas, Sogn 1636) Fol. 4rv: Ps. 106, 13–Ps. Ps. 107 (beginning) (34, 2). Secondary provenance: (Jørgen Kaas, Sogn 1636) Fol. 5rv: Ps. 118, 44–Ps. 118, 67 (677, 2, 677, 1). Secondary provenance: Stavanger len 1606 Fol. 6rv: Ps. 134, 24–Ps. 134, 19 (34, 4). Secondary provenance: (Jørgen Kaas, Sogn 1636?) Fol. 7rv: Ps. 136, 21–22; Ps. 136, 21–22 (34, 3). Secondary provenance: (Jørgen Kaas, Sogn 1636?) Fol. 8r: Ps. 141, 7/8–Ps. 142, 1; fol. 8v: Ps. 142, 1–3 (677, 3; 677, 4). Secondary provenance: Stavanger len 1626 Fol. 9rv: Ps. 143, 11–Ps. 144, 4 (34, 3). Secondary provenance: (Jørgen Kaas, Sogn 1636?) Fol. 10rv: Ps. 146, 9–10 (53, 4). Secondary provenance: Stavanger len Fol. 11rv: Ps. 146–Ps. 147 (53, 1; 53, 3; 53, 5; 53, 6). Secondary provenance: Stavanger len 1615, Stavanger len In her records in the NRA Gjerløw identified one leaf of this manuscript as belonging to Augustine’s Enarrationes­in­Psalmos and made a note of some other fragments of the same manuscript which, on closer investigation, turned out to bear witness to a textual version rather different to the published version of the Enarrationes. All of these have signatures from account books from Bergenhus len. To these can now be added fragments removed from account books from Stavanger len. Gjerløw identified fols. 1, 10 and 11 as probably belonging to Peter Lombard’s commentary on the Psalms. She must have had some doubt about the identification, however, since 32. The occurrence of the flex in a manuscript from the first half of the twelfth century could indicate a Cistercian origin. Later in the century it was adopted by other monastic orders, e.g., the Praemonstratensians. 33. The fragments are listed according to their position within the reconstructed leaf, so the first mentioned is located at the top left and the next one just below that, and then from the left to the right. fragments of patristic and other ecclesiastical literature [karlsen 2] 225 she suggested that it could be an abbreviation of the commentary of Peter Lombard. However, the agreement between this text and that on the fragments is not close. It is also clear from the script, dimensions, parchment etc. that all the fragments belong to the same manuscript containing a commentary on the Psalms that only partially corresponds to Augustine’s Enarrationes­in Psalmos. Fragments from this large work were first used as binding material for account books from Stavanger len, mostly in smaller units between 1606 and 1626. As there is no evidence so far of fragments from this manuscript being used to bind material from other districts during this long period, it was probably kept in the chancery in Stavanger. Later, in 1636, fragments from the same manuscript were used to bind account books from Sogn with the name of Jørgen Kaas on the fragments. Sogn was administered from Bergenhus. However, there seems to be no evidence of a Jørgen Kaas in Sogn or Bergenhus that year.34 Another Jørgen Kaas (✝ 1619) was royal governor (or prefect, to use the Latin terminology of the sixteenth and seventeenth centuries) in Stavanger 1601–1616. Fragments from this manuscript were used during his administration there. Fragments from Stavanger contain commentary on Ps. 86 and on 146–147. The content of the material connected with Bergenhus indicates that it came partly from the same part of the manuscript as the fragments from Stavanger (i.e., from the commentary on Ps. 88 and 143). How was it that large parchment leaves from Stavanger ended up on account books in Bergenhus? Clearly there was contact between the two len. One particular period of such contact stands out (but there may of course have been others). As Bergenhus was one of the two main len at this time, it is reasonable that the settlement of Kaas’ administration after his retirement in 1616 was transferred to the nearest principal len, and the relevant documents were packed and sent from Stavanger to Bergenhus with Jørgen Kaas’ name on the wrapping material, i.e., on the parchment leaves. It is in this context worthy of notice that the five large fragments without loss of text were used on the account books from Sogn 163635 at a time when there seems to have been a lack of parchment, whereas the Stavanger fragments used to bind account books in the early 1600s are small. Birgitta There is a small group of manuscripts in the NRA related to the Swedish saint Birgitta (Bridget) (1303–1373). The fragments of her Revelaciones are certainly relevant to a survey of ecclesiastical literature. The other Birgittine texts are prayer books (Quattuor­oraciones), Horae, and the antiphoner of the Birgittine nuns called Cantus­sororum, i.e., texts that fall outside the scope of this article. A publication of the Birgittine fragments will appear in Gjerløw (forthcoming).36 Revelaciones There are three fragments of two leaves of a copy of Revelaciones written in cursive script of the fifteenth century (plate 8). Their designations are NRA lat. fragmenter 25 and NRA lat. fragmenter Eske 45, XXXIII, 1–2. These fragments were removed from account books from Tønsberg 1612 and Nedre Romerike 1618, i.e., within the borders of the medieval diocese of Oslo. The two leaves are by two different scribes, but their hands are similar. The leaves may have been part of the same complete manuscript of Revelaciones (Gjerløw, forthcoming), and they have the following content: 34. Dr Tor Weidling (NRA, Oslo) has kindly checked the account books to which the fragments from Sogn 1636 were attached. He can find no mention of Jørgen Kaas in the account books themselves. Dr Weidling has also confirmed that he can find no record of a Jørgen Kaas in Bergenhus in 1636 who might have had his name written on these fragments. 35. The exception in this respect among the Bergenhus fragments is fol. 2, which is reconstructed from six fragments (Sogn 1636). 36. There are also vernacular (Birgittine-Norwegian) extracts from the Revelations of St Birgitta that survive (see Seip 1956). 226 espen karlsen Fol. 1: Revelaciones 5 (Liber­Questionum), Interrogatio­sexta–Interrogatio­septima (= Bergh 1971, 110–112). Fol. 2: Revelaciones 7, cap. 18–19 (= Bergh 1967, 175–177).37 Gilbertus Porretanus (Gilbert of Poitiers) Two small strips are the remains of a leaf from a manuscript containing the commentary on the Psalms by Gilbert of Poitiers (c. 1080–1154) (NRA lat. fragmenter 50, 1–2) (see plate 9). I owe the identification of this as yet unprinted text to Michael Gullick, who collated this fragment with a manuscript in Cambridge (Corpus Christi College 67 fol. 63r). The written space of the two pieces is 15 × 1.5 cm. The script is very small, and conforms to French handwriting of the middle of the twelfth century. Gilbert was one of the most important scholars at the school of Chartres until he moved to Paris in 1147 (for a study, see Gross-Diaz 1995). These two fragments indicate that the text was known in Norway. This manuscript should be added to the fifty-one twelfth-century manuscripts of this work known to Gross-Diaz (1995, 27).38 However, slightly later Peter Lombard’s commentary on the Psalms became the ‘standard’ commentary by the late twelfth century and largely replaced the work of Gilbert. Gregorius Magnus (Gregory the Great) a) Moralia­in­Iob A large fragment of a single leaf (plate 10) is all that survives of a continental copy of the Moralia in­Iob (Moral expositions on Iob),­the monumental commentary of Pope Gregory the Great on Job that was frequently copied in the twelfth century (for a list of manuscripts, see the Prolegomena in Adriaen 1979–1985, vol. 1).39 The fragment carries the signature NRA lat. fragmenter Eske 52 [2] and was removed from a church register for the church in Brønnøy 1669—1677 (Ministerialbok for Brønnøy 1669–1677). This register was never sent to Copenhagen.40 A large format Moralia­in­Iob is more likely to have been used in a major centre than in a small place such as Brønnøy, so it is more probable that this copy of the Moralia was used in Trondheim itself. The large format is also typical many of books with Trondheim provenance.41 The text is from Book 11, cap. 13 (= Adriaen 1979–1985, 2: 619–620). The recto begins quam­timuit­in­culpa. The fragment measures 24 × 16 cm. The preserved part of the written area is 20.5 × 8.5 cm with 26 lines (see plate 11). Although this is a large fragment, a significant part of the text has been lost. As for the script, it is clearly continental. It has the Germanic form of punctus­elevatus (the so-called ‘tick-and-point’: ) (see, e.g., l. 8 and 12). There is also punctuation by the punctus­interrogativus. The letters a and g have continental forms. The letters o and e are narrow, typical of very late Caroline script (Derolez 2003, 58). There is an added hairline on the letter r; which is especially pronounced in line 21 in urgentur. In the twelfth century this is a Germanic feature. The script is related to, though not identical with, that of a bible manuscript in Brussels dated 1132–1135, included in Manuscrits­dates­conserves­en­Belgique 1, pl. 12–13 (Masai, Wittek & Brounts 1968). The probable date of the fragment is the first half of the twelfth century, perhaps towards the middle of the century. b) Dialogi Only one fragment has survived of a manuscript in two columns removed from an account book from Romsdalen 1605 (NRA lat. fragmenter 19). Trondheim is therefore a possible secondary 37. The fragment of Revelaciones 5 is cited by Sørensen (1910, 32– 33), but he did not identify it. 38. The manuscripts are listed by Häring (1978) with a supplement of Gross-Diaz (1995, 160–180). 39. In the Corpus Christianorum edition it comprises 1811 printed pages (Adriaen 1979–1985). 40. This account book is still in the SAT in Trondheim. It was rebound in 1902 and carries a note about the removed parchment fragment on the inside of the front board. See digitised version on http://www.digitalarkivet.no. 41. I am most grateful to Dr Tor Weidling in the NRA (Oslo) for discussing the provenance with me. fragments of patristic and other ecclesiastical literature [karlsen 2] 227 provenance for the manuscript. Only a small part of the text of the leaf has been preserved (see plate 11). It measures 11.5 × 9.5 cm. The written area is 7 × 7.5 cm. The column is 6 cm wide, the reconstructed width of the written space is 13 cm, and the lower margin is 4 cm. The final word of a rubric in red is preserved (eođē).The fragment contains a small part of book 1, chapter 4 (Pricoco & Simonetti 2005, 1: 32 and 36 = PL 77, cols. 168–169) of the Dialogi of Gregory the Great (fol. 1r (left column) <Cui­ille­respon>dit:­‘occasionem­queris­ut­non­debeas prestare­quod­peteris’ . . . a­cella­digressus­est.­Quo­discedente; Fol. 1v (right column) <de>moratur cum­eundem­uenerabilem­. . .­inquisitione­uir). To judge from the lengthy passage this was a large leaf. The large format is typical for books connected with Trondheim. The script is rounded, with English features, but perhaps Norwegian and dating from the late twelfth or early thirteenth century. c) Moralia­in­Iob Four fragments from account books from Jæren and Dalerne (1611) and Ryfylke (NRA lat. fragmenter 60) make up the upper part of a leaf (plate 12). The incomplete leaf measures *21.5 × 19 cm, with the lower part of the leaf missing. The text measures *15 × 19 cm and is arranged in two columns (column width 7.5 cm). 39–40 lines of the text on each side of the leaf are preserved. The script is compressed, and is a fully developed textualis (in Derolez’ terminology). The initials are plain and filled in with red. The ruling is in ink and the first text line is below the top ruled line, as was usual after c. 1250 (Ker 1960b; Derolez 2003, 39). The letter a has a closed upper bow. This a was usual in Northern textualis by the second half of the thirteenth century, and from c. 1300 it had replaced the a with an open upper bow completely (Derolez 2003, 84). Other typical features are bitings; d always has a sloping shaft (Derolez 2003, 87). The probable date of the manuscript is the second half of the thirteenth or the early fourteenth century. The content of this leaf is as follows: Fol 1r Moralia 23, 10, 19–23, 10, 19 (= Adriaen 1979–1985, 1689, l. 30–1690, l. 62; 1690, l. 6 (from the bottom)–1691, l. 41). Fol. 1v Moralia 23, 12, 22–23, 12, 23; lacuna; 23, 12, 23–24 (= Adriaen 1979–1985, 1692, l. 5–1693, l. 41; lacuna; 1693, l. 51–1694, l. 86). Hrabanus Maurus Commentarii­in­Genesim There are two fragments (NRA lat. fragmenter 59, 1–2) from the same leaf of Hrabanus Maurus’ (780–856) Commentarii­in­Genesim 2, chapters 21–22 (1r = PL 107, cols. 551d–554a panes­eo­quod sine­ferment . . . quare­Saram­redargu<at>; 1v = cols. 553a–554b homines­putabat­. . .­omnia­nuda et­aperta). The two pieces fit together (see plate 13). They were removed from an account book for 1624 from Jæren, close to Stavanger. The leaf was cut vertically. The preserved part measures approximately 25 × 10 cm with a written space of 22 × 8.5 cm and 33 lines (the bottom few lines are lost). The leaf once formed part of what must have been a large format book. The script is on a continental, probably French or Germanic model, and probably dates from the late eleventh or possibly early twelfth century (cf., e.g., the form of the letter a and of the ampersand). The script is comparable to that of a psalter copied in 1105 at the abbey of St Martin at Tournai (Thomson 1969, pl. 1). The letter forms are still caroline, but rather compressed, perhaps reflecting the relative informality of the hand rather than a later twelfth-century date. The Tironian nota (without a cross-stroke) and the ampersand are both employed for et. There is fre- 228 espen karlsen quent use of e caudata. A late feature is the closed lower lobe of g. Tall  is used throughout, whereas the round s only occurs as a capital. The descender of q turns to the left. Straight d is preferred, whereas uncial d is found only once (cf. Derolez 2003, 49 and 60). Punctuation is by punctus and punctus­elevatus. Hugo de Folieto It was once assumed that the works of Hugo de Folieto (Hugues de Fouilly; c. 1100–1172/74) did not reach Scandinavia (Van den Abeele 2003, 268). Fragments from two different manuscripts indicate otherwise.42 a) De­pastoribus­et­ovibus This work is represented by a leaf from the first half of the thirteenth century, probably copied locally by a good scribe. For further discussion of this fragment, see the contribution of Wifstrand Schiebe & Karlsen in this volume. b) De­claustro­animae Two fragments of a copy of this text with the designation NRA lat. fragmenter 8, 1–2, were filed under Iosephus Latinus43 with a question mark in the records of Lilli Gjerløw. They were removed from an account book from Trondheim 1613. This is also the kind of provenance to be expected for this kind of book. The fragments are part of the same leaf with two columns 6.5 cm wide (see reproduction of the verso side on plate 14). The leaf was cut horizontally in the middle, with the loss of two or three lines, and there are twenty-eight lines remaining. There is also loss of text at the left of one column and the right of the other column. The reconstructed writing space is c. 14 cm wide and probably c. 20/21 cm high. The leaf contains De­claustro­animae, Book 4, from the beginning of Chapter 18 and continuing until Chapter 19 (right column on the verso side; see plate 14) corresponding to PL 176, cols. 1156c–1158b. The manuscript dates from the late thirteenth century or the early fourteenth. It was copied by a poor scribe and has a poor initial in red on the verso side with blue decoration. The bow of the letter a is closed, a form of a occurring in the thirteenth century and predominant after 1300 in book script (Derolez 2003, 84). It fuses with preceding d and is connected to preceding c. Biting occurs but is not frequent. Tall minuscule  is used throughout (round s occurs as capital S). Capital J occurs in initial position; minuscule j in the combination ij (hijs). There is a trace of rubrication in red. Punctuation is by punctus. It is hard to locate the place of copying of this manuscript on the basis of script alone as the scribe was not very expert. Therefore, whereas the text itself may have been typical of the resources of a major centre, the book may have been copied in a small one by an individual for his own use. 42. For further evidence of Hugo de Folieto in Scandinavia, see footnote 16 in Wifstrand Schiebe & Karlsen in this volume. 43. No fragments of the Latin translation of the Antiquitates­Iudaicae of Flavius Josephus (37–95) have been identified among the Norwegian fragments as yet, despite the fact that this was a frequently read text in the twelfth and thirteenth centuries (Guenée 1980, 317–318). It is interesting to note that two historical works composed in Norway, the small history of Theodoricus Monachus and the Historia­de­profectione­Danorum­in­Hierosolymam (see footnote 14 above) were transmitted together with the Bellum­Iudaicum of Josephus (Skovgaard-Petersen 2002, 110). The content of the now lost codex in folio format discovered by Bernhard Caspar Kirchmann c. 1620, was probably copied in a Norwegian or Danish scriptorium to judge from its entire content (SkovgaardPetersen 2002, 111–112). Twelfth-century fragments of both the Antiquitates Iudaicae (from four different codices) and the Bellum Iudaicum (from one codex) have been used in the binding of Danish account books and other archival material (tingbog) (Mejer & Raasted 1973, 371–372). Michael Gullick tells me that Stockholm, SRA Fr. 10069 is a fragment of a copy of the Antiquitates Iudaicae probably written in Sweden s. xii/xiii. fragments of patristic and other ecclesiastical literature [karlsen 2] 229 Ioannes Cassianus De­institutis­coenobiorum John Cassian (c. 360–c. 435), an important figure of early monasticism, is represented by a fragment of De­institutis­coenobiorum­(On the regulation of monasteries) with the designation NRA lat. fragmenter 30, 1 (plate 15). This fragment was removed from an account book of 1634 for Nordfjord len north of Bergen where the account book was probably bound. The opening word of the recto is <r>edundancia. The recto side contains a passage from Book 2 (Book 2, 10, 2– 2,11,2 (= Guy 2001, 74–76); whereas the verso side contains Book 3, 3, 4–3, 3, 6 (= Guy 2001, 96–98). The original dimensions were therefore probably very large or the text must have been heavily abbreviated. The fragment itself measures 8.0 × 17.0 cm and was probably part of a large leaf with text in two columns, although only one column is now visible. The surviving written space measures 8.0 × 10.0 cm. It was ruled in ink. The script is compressed and the letters tend to fuse (e.g., the fusion of cco on the third line and of exc on the same line). The letter a still has an open bow and may indicate a date before 1300. It was probably written in the second half of the thirteenth century or, at the latest, early in the fourteenth. Isidorus Hispalensis Isidore of Seville (✝ 637) is represented by two manuscripts, both of which contain his Etymolog­iae (or Origines), a comprehensive christian encyclopaedia that became a model for later encyclopaedias and one of the most copied texts during the Middle Ages. Some 1000 manuscripts of this text still exist (Guzman 1996, 703). a) Etymologiae Thirty-one fragments from fourteen leaves are the remains of a twelfth-century manuscript, probably from the middle of the century (NRA lat. fragmenter 2, 1–8; 1, 1–11; 3, 1–13) (plate 16). The content falls between Book 3, 5, 6 and Book 20, 6, 2 (a reconstruction follows below). All of the signatures indicate that this manuscript was probably taken apart and reused for binding purposes in Trondheim (see Pettersen in this volume on this manuscript). The account books to which the fragments were once attached are from Romsdalen between the years 1625– 1627, from Trondheim between the years 1623–1624, from Fosen between 1623–1624, from Herjedalen 1626 and 1638, from Strinda 1624, and Reinskloster 1625. The manuscript was probably dismembered and used for binding purposes in the medieval archbishops’ palace which was the residence of the royal governor after the Reformation.44 The manuscript is likely to have been the property of the cathedral chapter or a monastic community in or close to Trondheim. The size of the leaves is 24 × 15 cm. The writing space is 19 × 11 cm. The pages have forty lines in two columns that are each five cm wide. This is a rather small size for a book of this kind and date. As the script is rather compressed, the book may have been a one-volume copy of this huge work. The initials are rustic and in red and green. The rubrics are red. As for the script, the letter a is clearly not characteristic of English script. It has a flat head and is of continental kind. There is a dot above ẏ. The Tironian nota for et is preferred to the ampersand. The general appearance of the script is more French-like than Germanic-like. The manuscript may be reconstructed as follows (signatures are found in parentheses): Fol 1rv Etym. 3.71.7–3.31.19 (2, 1; 2, 2) Fol. 2rv Etym. 5.26.10–5.27.25 with lacunae (2, 3;2, 4; 2, 5; 2, 6) 44. See Pettersen in this volume for a detailed discussion of the provenance of this manuscript. 230 espen karlsen Fol. 3rv Etym. 5.3.1–5.3.4 (2, 7; 2, 8) Fol. 4rv Etym. 6.2.8–6.2.33 (1, 1; 1, 2; 1, 3; 1, 4) Fol. 5rv Etym. 6.2.37–6.4.3 with lacunae­(1, 5; 1, 6) Fol. 6rv Etym. 6.17.23–6.18.2 with lacuna (1, 7) Fol. 7rv Etym. 8.6.20–8.9.3 (1, 8; 1, 9) Fol. 8rv Etym. 9.2.39–9.2.65 (1, 10; 1, 11) Fol. 9rv Etym. 11.1.61–11.1.70 (3, 1; 3, 2) Fol. 10rv Etym. 11.1.114–11.1.136 (3, 3; 3, 4) Fol. 11rv Etym. 13.22.1–13.22.4; Incipiunt­capitula­XIIImi; the verso is blank (1, 12) Fol. 12rv Etym. 16.10.1–16.13.2 with lacunae (3, 5; 3, 6) Fol. 13 Etym. 16.13.5–16.13.2 (3, 7; 3, 8; 3, 9; 3, 10) Fol. 14 Etym. 20.3.15–20.6.2 (1, 13) b) Etymologiae There also survive two small fragments from one leaf of another twelfth-century manuscript of Isidore (NRA lat. fragmenter 113b, 1–2; reproduced on plate 17). They were used to bind an account book from Strinda and Selbu for the year 1624. This also makes it likely that this account book was bound in the residence of the royal governor in Trondheim. The two fragments fit together and jointly measure approximately 12 × 6 cm. The written space is 12 × 3.5 cm. There are sixteen lines. Most of the leaf is lost. The preserved text is from Etymologiae 17, 6, 26–17, 7, 1 (the left column on the recto side) and 17, 7, 17–17, 7, 18 (the right column on the verso side). The manuscript probably had two columns and was ruled in hard point. In the printed edition by Lindsay (1911, vol. 2) the text from Etymologiae 17, 6, 26–17, 7, 18 occupies more than four printed pages.45 The letter a with an ample curving head occurs, especially at the beginning of words. This is typical of English script. The letter y has a dot (ẏ). There is biting between the letters d and e. The g is closely resembles the Arabic numeral 8. The degree of compression suggests a date late in the twelfth century, and the script is typically English of that date. Lietbertus de Insulis (Letbert of Saint-Ruf ) Flores­Psalmorum Three fragments survive of two consecutive leaves (NRA lat. fragmenter 37, 1–3) that contained a commentary on the Psalms of David, Psalm 49 and the beginning of Psalm 50.46 This commentary is ascribed in PL 21 to Rufinus of Aquileia, but since 1914 has been correctly attributed to Letbert (Augustinian abbot of Saint-Ruf in Avignon from 1100 until his death in 1110: Wilmart 1914–1919; Gross-Diaz 1996, 117–118). His commentary is known as the Flores­Psalmorum. A list of twenty-six manuscripts is supplied by Stegmüller (1950–1980: 3, no. 5395), and there are several fragments of a copy now divided between Stockholm and Helsinki that suggest this commentary was known elsewhere in Scandinavia (Lehtinen 2005, 123, no. 17). Letbert’s Flores psalmorum is also documented in an abbreviated version by a bifolium from the late twelfth century among the fragments at the UBB in Bergen (MS 1549, 4).47 It is from a continental manuscript of the late twelfth century. The provenance is unfortunately unknown. The three fragments discussed here were removed from account books from Hardanger len 1647 in western Norway. Fr. 37, 1–2, are conjoint and make up the upper part of fol. 1, whereas Fr. 37, 3, is an almost complete leaf of which a part of the left margin is cut (plate 18). The script 45. The edition has no pagination. 46. In PL 21 is included an edition of Letbert’s commentary on the first seventy-five Psalms among the writings of Rufinus (under the title In­ Davidis­ LXXV­ Psalmos­ commentarius), but some manuscripts contain his commentary on Psalms 1–150. 47. The fragment is reproduced and commented by Åslaug Ommundsen at the website of the UBB (http://gandalf.aksis.uib.no/ mpf/). fragments of patristic and other ecclesiastical literature [karlsen 2] 231 is English-like from the late twelfth century. The writing space measures approximately 20 × 12 cm. There are two initials in blue and red. In Gullick’s opinion these fragments were almost certainly written locally (see Gullick 2 in this volume, no. 75 among the manuscripts and fragments with English features). The surviving passages of text are as follows: Fol. 1rv: Ps. 49, 11–49, 14; lacuna; Ps. 149, 15–149, 18 (PL 21, cols. 846b–847b) cols. 846d–847b (eruam­te. Cum­fueris–Nonne­ipso­placito). Fol. 2: Ps. 49, 20–50, 3 (PL 21, cols. 847c–848d) Mathias Ouidi (Övidsson) Magister Mathias (c. 1300–1350) of Linköping in Sweden was the confessor of St Birgitta of Sweden and the author of at least six works48, some of which were also widely read outside Scandinavia. Alphabetum distinccionum (Concordancie­super­totam­Bibliam) Two fragments (NRA lat. fragmenter 43, 1–2) in the NRA of the Alphabetum distinccionum, also known as Concordancie­super­totam­Bibliam,­were identified by Lehmann in the 1930s. They were removed from an account book for Akershus of 1621. The account book was most likely bound at Akershus castle, now in central Oslo. Lehmann identified these fragments as parts of a manuscript of which there are considerable remains in Stockholm (Lehmann 1937b; 153) and elsewhere in Sweden, Helsinki, and London (Piltz 1995, 140). Around 180 leaves are known today (Abukhanfusa 2004; see the inside of the softcover where there is a good reproduction of a complete leaf in Stockholm). They probably constitute the remains of a copy of c. 700 leaves in three volumes that belonged to Vadstena monastery. It was dismembered in the sixteenth century.49 All the known leaves were used in the binding of account books (Piltz 1995, 139–140). The two small Oslo fragments, probably once belonging to the same leaf (plate 19), contain the lemmata (fol. 1r) Can<. . .>, Cau<sa>; (1v) Cecitas and have references to the Bible between the lines. They measure 6 × 9 cm. The script is a cursive dating probably from the late fourteenth or early fifteenth century. An early feature in the script is the form of g that consists of two lobes, whereas the interlinear references have a later cursive g (Derolez 2003, 146). In initial position the letter a sometimes occurs in a slightly archaic form in two compartments. 50 As the lemmata are early in the alphabet, these two fragments must have belonged to the first volume brought to Vadstena castle in 1595 (see footnote 49). How did these fragments from Vadstena end up as binding material in Akershus Castle? Most likely the leaf they once were part of was used as wrapper on material sent from Sweden to Akershus len. It is not surprising that there was probably direct contact between Swedish authorities and the bordering main len in Norway. The parchment leaf from Sweden was then probably reused once more in Norway51 and cut into smaller pieces, as was a usual practice in Norway at the time (see Pettersen in this volume).52 48. For his life and works, see Piltz (1985–1987); Piltz 1986, 137–160. Further works have subsequently appeared in print. 49. The second and third volumes seem to have been brought from Vadstena to the royal palace in 1543, whereas the first volume was brought to Vadstena castle in 1595. 50. I am indebted to Dr Nils Dverstorp (Oslo) for discussing the palaeography of these two fragments with me. 51. I have suggested a similar explanation regarding the fragments of a multi-volume copy of an adaptation of Augustine’s Enarrationes in­Psalmos (see under Aurelius Augustinus d) above). 52. Another manuscript used in the binding of account books in Norway and Sweden is a missal dating from the eleventh century. There are eighteen fragments of Stockholm, Mi 1 (= Oslo, Mi 7) in Swedish collections and four in Oslo (Hartzell 2006, no. 340; Gullick 2 in this volume). The rather smaller number of fragments in Norway (NRA lat. fragmenter 204b, 5–6 + 205, 1–2) may have ended up on account books in Norway in a similar manner as I have suggested for the two fragments of Alphabetum­distinccionum. The Swedish fragments of this manuscript are mainly connected with the areas close to Akershus len. The fragments in Norway are from Trondheim len 1622 and from a small len close to Akershus len,­Eiker 1620. The manuscript was probably first taken apart in Sweden. Some leaves were probably used to bind material sent from Swedish authorities to the administration in Trondheim and at Akershus where the leaves were used once more for binding purposes. Most account books were usually sent directly from Eiker len to Copenhagen, but these fragments were probably removed from an account book recording taxes which was sent via Akershus. I thank Dr Tor Weidling for enlightening me on matters concerning public account books in this period. 232 espen karlsen Mauritius Itinerarium­in­terram­sanctam The only text composed by a Norwegian author that belongs to the category dealt with in this article is a description of a journey to the Holy Land by ship from western Norway (probably Selje) in the 1270s by Brother Mauritius, who spoke of himself as a poor brother (pauperem fratrem; Storm [1880] 1973, 168) and who has been identified by Storm with a Franciscan.53 It was published under the title Itinerarium­in­terram­sanctam by Gustav Storm in 1880 (reprinted 1973, 165–168, with introduction on pp. xxxxvii–xxxxix).54 Six fragments from five leaves are preserved under the designation NRA norrøne fragmenter 92,55 1–6. They were foliated in the sixteenth century in Arabic numerals: 135, 138, 140, 144 and 145. The content of the manuscript is as follows: Fols. 135 and 138: The marriage treaty in Roxburgh 25 July 1281 copied from a vidisse by the Bishop of Bergen and three abbots from 1283. Fols. 139–144v: Itinerarium­in­terram­sanctam, with fol. 139 and 141–142 lost Fols. 144v–145: A certified copy (so-called vidisse) from 1286 of a papal bull concerning tithe issued by Honorius IV (published in DN 1, nos. 75 and 78). The itinerary comprised only six leaves, half of which are preserved. The writing space measures 14.5/15 × 10 cm and has one column with 32 lines (see plate 20). There are no decorated or coloured initials. The script differs from the other manuscripts discussed in this article in being a less formal book hand. The letter a has two compartments. The ascenders and descenders are extended. The ascenders are in many cases without loops, which they tend to have in cursive.56 Round s and  extend below baseline. As tall  in final position disappeared in the first half of the fourteenth century, at least in textualis (Derolez 2003, 92), c. 1350 may be the terminus­ante­quem for this scribe. The manuscript was probably written not very long after 1286, the date of the latest document. The letter g has an extensive lobe below the baseline (for a reproduction of an example in the Norse vernacular from Nidaros 1303, see reproduction in Simensen 2002, 68). Punctuation is by punctus. The script itself, as far as I can see, does not suggest a particular place of writing. It is the content and the fact that the manuscript is so clearly linked to Bergen (see below) that leaves little doubt that the manuscript was copied in Norway, most probably in Bergen. Gustav Storm identified the hand that paginated the manuscript with arabic numerals as belonging to the Norwegian sixteenth-century humanist and protestant bishop in Bergen, Absalon Pederssøn Beyer (c. 1528–1575) (Storm [1880] 1973, xxxxvii–xxxxviii). The diary of Absalon Beyer also shows that he knew the lost leaves of Mauritius’ itinerary (Storm [1880] 1973, xxxxvii). Since the content suggests that the manuscript was probably copied in Bergen and it was paginated by bishop (or ‘superintendant’ according to the terminology of the sixteenth century) Absalon of Bergen, it was probably taken apart in Bergen, as the preserved fragments were used to bind account books for areas administered from Bergen. It is very probable that the codex was a register of correspondence and other useful texts that belonged to the diocese of Bergen (Storm [1880] 1973, xxxxvii), which remained the property of the diocese for a while after the Reforma53. It was probably first mentioned in print by Riant (1865, 72–73; 357–358) who used a transcription made for him by the Norwegian medievalist P. A. Munch (1810–1863). 54. Mauritius’ Latin text with a brief introduction, notes and an Italian translation is reprinted after Storm by De Sandoli (1978–1984, vol. 4, 85–93). 55. The designation according to Storm’s edition is Old Norse fragment no. 29, probably a misprint for no. 92. 56. In some vernacular documents of the second half of the thirteenth century we find some ascenders with loops and some without loops in the same documents (see Seip 1954, nos. 17 and 18). These two documents dating from Oslo 1279 and Trøndelag c. 1280, however, were not written in the same kind of script as the one in the Mauritius manuscript. fragments of patristic and other ecclesiastical literature [karlsen 2] 233 tion as it was paginated by Absalon. This is an exceptionally clear example of a connection between account books and the origin of the manuscript from which leaves were taken to bind them. Petrus Comestor Historia­scholastica Historia­scholastica by Peter Comestor’s (c. 1100–1187) became widely disseminated after 1180 and is represented by one fragment of a codex from the late twelfth century (NRA lat. fragmenter 16, 1), reproduced on plate 21. The provenance is unknown. The leaf has 50 lines and is 31 cm tall with column width 7.5 cm. Rubrication is in red. There are initials in red and blue with simple decorations. The script is small and compressed with bitings, and with some odd features. Final  extends well below baseline. The letter f is very long also in initial position. The Tironian nota for et is not crossed. The preserved text is from Historia­scholastica­in­Actus­apostolorum chapter 11–15 (PL 198, cols. 1652b–1654c). Another manuscript of the Historia­Scholastica (NRA lat. fragmenter 7, 1–2) is mentioned in Gullick 2, no. 69, dating from the twelfth/thirteenth century. Petrus Lombardus Peter Lombard (1095/1100–1160), a theologian from northern Italy who became bishop of Paris, is represented by his commentary on the Psalms and by his Glossed Pauline Epistles. a) Collectanea­in­Paulum One leaf survives of Peter Lombard’s commentary on the letters of St Paul (NRA lat. fragmenter 36, 1). It was attached to an account book from 1605 for Lista len in Southern Norway. It is reproduced on plate 22. The leaf measures 32 × 16 cm. The right column is cut vertically with a small loss of text. The biblical text on the recto begins linguis­suis. Fol. 1rv contains the commentary on Ad Romanos 3, 13–19; Fol. 1v Ad­Romanos 13, 20; Fol. 1v Ad­Romanos 3, 21–22. The text corresponds to PL 191, cols. 1357a–1358a. It was copied towards the end of the twelfth century and the fragment is commented upon in Gullick 2, no. 72, in the present volume. b) Collectanea­in­Paulum Two small fragments (NRA lat. fragmenter 40, 1–2) survive from the same leaf of another copy of Peter Lombard’s commentary on the Pauline Epistles (fol. 1r 2 Cor. 8, 16–20; fol. 1v 2 Cor. 9, 2 = PL 192, cols. 59d–60b). Their provenance is an account book for Eker and Brunla len 1631.The two fragments are conjoint and are in two columns (plate 23). The written space is 8 × 15 cm. They are described in the section ‘Manuscripts and manuscript fragments made in England’ in Gullick 2 (no. 5) in this volume, and date from the late twelfth or early thirteenth century. c) Commentarius­in­Psalmos Peter Lombard wrote a commentary on the Psalms that achieved greater success than those of Letbert (or Lietbert) of Lille (or Saint-Ruf) (✝ 1110) and Gilbert of Poitiers (c. 1180–1154) whose commentaries also are found among the fragments in the NRA. The two fragments with the designation NRA lat. fragmenter 47, 1–2 (Romsdal len in northwestern Norway 1623), are pieces of the same leaf (plate 24). The manuscript measures 9 × 24 cm and has two columns (column width 8 cm). It was copied in the first half of the thirteenth century. Fol. 1r contains the commentary on Ps. 148, 14 and Ps. 149, 8 (PL 191, cols 1287a–1287b) and fol. 1v Ps. 148, 14 and Ps. 149, 4 (PL 191, cols. 1287a–1287b). 234 espen karlsen Robertus de Tumbalena In­Cantica­canticorum In approximately the same format as the oldest Isidore manuscript mentioned above is an incomplete bifolium (NRA lat. fragmenter 52, 1–2) of a commentary on Canticum­canticorum, chapter 7 (PL 79, cols. 533–534 and cols. 537–538).57 This commentary has frequently been attributed to Gregory the Great and was published under Gregory’s name by Migne in PL 79, which remains the only printed edition. Whether in the copy of which these two fragments are the remains the text was likewise attributed to Gregory or to the Norman Benedictine Robertus de Tumbalena (✝ 1078) to whom it is attributed today, must remain uncertain. As Robertus compiled his text from the commentary of Gregory, it is no wonder that it frequently became attributed to the latter. The best preserved of the two fragments measures 11 × 16 cm. The manuscript had two columns, each column 5 cm wide. A reproduction is found on plate 25. Sixteen lines of text are preserved, probably a little than the half of the leaf. One of the two fragments has as its secondary provenance an account book for Stavanger 1623, i.e., in the south western part of Norway. The other fragment was removed from a book of court records (bytingsprotokoll) from 1619 from Stavanger. There is a note on the fragment of the old protocol number in NRA for this kind of court records (Rettsprotokoll­1376). This book of court records was never sent to Copenhagen, and in the nineteenth century it was transferred from Stavanger to the NRA from where it was returned back to the SAS, i.e., the local State Archives in Stavanger, in 1914 after removal of the fragments. The script is English in appearance (see, for example, the form of a and g), though of a later twelfth-century date than the English copy of Augustine, De­ciuitate­Dei mentioned above. The letter a is of an English kind. The same goes for the letter g. There are several instances of the tailed e (ę) which confirm a date before 1200. The punctus­flexus (), a punctuation sign that indicated a minor medial pause (Parkes 1992, 195 and 197), is found in this manuscript and indicates a monastic origin. It belonged to a system of positurae58 that came mostly to form part of monastic culture, and especially, though not exclusively, books used in the liturgy (Parkes 1992, 38). The punctus­flexus suggests a Cistercian origin (Parkes 1992, 195). Rufinus Historia­ecclesiastica Among the fragments dating from the twelfth century are two fragments of Rufinus’ (344/345– 410) Historia ecclesiastica­(NRA lat. fragmenter 130, 1–2). This is the only church history of late Antiquity among the fragments in the NRA to be identified so far. Rufinus, who specialised in translating Greek Christian texts into Latin, translated Eusebius’ (c. 263–339) Church­History into Latin in A.D. 401 and he continued the work to A.D. 395 by adding two books. He also introduced numerous alterations to Eusebius’ text. In the twelfth century this work of Rufinus 57. In PL it is published under the title Super­Cantica­canticorum­expositio. 58. The positurae form a system of symbols that were introduced in the second half of the eight century to indicate the end of a sent­entia and minor and major medial pauses. It reaches its peak with the rise of monastic movements in the eleventh and twelfth cent- ury, in particular in books connected with the Cistercians and the Carthusians (Parkes 1992, 35–40), and, as Patricia Stirnemann added in a paper given in Bergen in November 2009, the Praemonstratensians. The first known instances of punctus­flexus occur in tenth century manuscripts (Parkes 1992, 36). fragments of patristic and other ecclesiastical literature [karlsen 2] 235 was frequently copied.59 The two fragments were removed from account books of 1622 for Hedemarken, a fogderi (in Akershus len) north of Oslo, which means that the manuscript is likely to have been used somewhere within the diocese of Oslo before the Reformation. There is a note on the two fragments indicating that they earlier were used on an older account book (see plate 26 Oc­tiill­same­aars­d. . . Igienn­Anno­1606; approximately: ‘And for the same year’s . . . again in the year 1606’). The fragments were once conjoint and formed the last ten lines of a two column manuscript of about 30 lines. They both measure 11.5 × 10 cm and were part of a manuscript in two columns. The column width is 9.5/10 cm. The preserved text is from Book 5, 18: 60 Fol 1r, left column: <arch>is­publicis­apud­ephesum­.­.­.­Jtem­post­pauca­de­ip<sis> (Hist.­eccl. 5, 18 = Schwarz & Mommsen 1999, 477, lines 24–30). Fol. 1r, right column: <s>piritui­qui­in­ipsa­loquebatur­. . .­ex­quibus­omnibus­uehementer­[Hist.­eccl. 5, 18 = Schwarz & Mommsen 1999, 477, lines 24–30; fol. 1v, left column: <lit>eris­inherebant­Verum­de­his­satis­. . . qui­tamen postea­in­errores­deuolutus (Hist.­eccl. 5, 18 = Schwarz & Mommsen 1999, 480, lines 13–19). Fol. 1v, right column <hęc> inquit­dogmata­florine­que­asser<is> . . . dum­adhuc­puer­essem­in­asia­apud (Hist.­eccl. 5,20 = Schwarz & Mommsen 1999, 483, lines 15–21). There is no doubt that the manuscript was written by a scribe trained on the Continent or under continental influence (cf. e.g. the letter a and the ampersand that are clearly of a Continental kind). The script has clear angular features, especially evident in the final minims of m and n. E caudata (ę) is frequent. J is used in initial position (e.g., Jtem); minuscule j occurs in final position (hij). The ligature ſt occurs. The y has a dot above (ẏ). Punctuation is by punctus and punctus­elevatus. Though I cannot find a very close match among the plates in the twelve volumes of the catalogue of dated manuscripts from France (Samaran & Marichal 1959–1984), there are examples that are not too dissimilar. A date in the first half of the twelfth century, perhaps towards the middle, is likely. Vitae patrum The texts included in the widespread Vitae patrum are divided into ten books in the edition published by Migne (PL 73–74). Among the fragmentary manuscripts to be discussed here are the Verba seniorum (The Words of the Elders), i.e., Book 5 of the Vitae patrum. 61 The Verba seniorum is the title of a collection of short anecdotes from the earliest monastic communities in Egypt, Palestine and Syria. They were translated into Latin from the Greek during the sixth and seventh centuries. These anecdotes were highly popular in the Middle Ages and were eagerly studied, 59. Another important church history was not to appear in Latin until the sixth century when Cassiodorus (c. 490–583) translated and compiled Historia­ tripartita from Greek sources (Vessey 2004, 325), a work that was to become the most widespread ecclesiastical history in the Middle Ages. Rufinus’ Historia­ecclesiastica was also one of the most influential historical works during the Middle Ages, along with inter­alia Historia­tripartita, Orosius’ (c. 385-c. 420) Historia­contra­paganos, the two works of Josephus in Latin translation (Guenée 1980, 301–303). Historia­tripartita (or a text based upon it) was known to Theodoricus Monachus in his passage on the fall of Julian in chapter 8) (Karlsen & Vatsend 2003, 239–255). Johnsen (1939, 53) had earlier suggested Hugo (i.e., Richard) of St Victor’s Excerptiones­ priores as the source, but the text of Theodoricus contains information at this point that is not mentioned by Richard of St Victor, whereas the information given by Theodoricus closely follows Historia­tripartita or a text based upon it. Excerptiones­priores­should therefore not longer be considered a possible source (as does, e.g., Foerster 2009, 110 with footnote 223). The anonymous author of Historia Norwegie may have known Orosius (Ekrem & Mortensen 2003, 28). Rufinus, Orosius and Historia­tripartita were part of the christian historical works that Cassiodorus regarded as indispensable in his Institutiones, and they were frequently copied during the Middle Ages. 60. Rufinus’ Latin version accompanies the critical edition of Eusebius’ Church History in Greek, whereas the text of Rufinus itself has not been published in a critical edition. 61. The third, sixth and seventh book of Vitae patrum are also published in PL under this title. The title was assigned to it by Heribert Rosweyde (1569–1629) in his printed edition of Vitae patrum (Antwerp 1615). Verba seniorum is Rosweyde’s translation of the Greek Apophtegmata patrum. The easiest available edition today is the one of Migne in PL 73 (cols. 855–988) which is merely a reprint of the text published by Rosweyde. There is as yet no critical edition of Vitae­patrum. 236 espen karlsen copied and excerpted.62 The Verba­seniorum­collection was an important precursor to the exempla tradition (Palmer 1996, 585).63 The three fragmentary manuscripts to be discussed below all contain text from Book 5 of the Vitae­patrum. a) Vitae­patrum (Verba­seniorum) An early example of a patristic manuscript preserved from medieval Norway is a bifolium of the Verba seniorum (NRA lat. fragmenter 18). Fol. 2r is reproduced on plate 27. A note on fol. 2v of the double leaf records that the bifolium was attached to an account book from c. 1540 from Akershus, the castle that was then just outside Oslo.64 The text on fol. 1 has 41 lines in one column and measures 20 × 16 cm. A line or two is lost at the top of the leaf, whereas the margin below was cut just below the last line. The second leaf has preserved forty lines. Approximately 3 cm of the text has been lost from the right margin of fol. 2, probably because the double leaf was larger than the account book. The script is a small and compressed late Caroline minuscule. The feet of minims and ascenders all curve to the right, a phenomenon typical of twelfth-century script (Derolez 2003, 58). The final stroke of e slopes upwards similar to a hairline (Derolez 2003, 61 with example 24). The a has a flat head, the ampersand has a continental form and the g is straight-backed. These features, together with the general aspect of the script, point to the Continent rather than to England as the place of origin. The form of majuscule A may be a Germanic feature. The tailed e occurs occasionally for the classical diphthong ae, though not frequently (e.g., twice on fol. 1r). The diphthong is usually realised by an e without a tail. There is variation in the choice of letter forms. The tall s (), predominant in Caroline script, is used as well as the round uncial s (cf. Derolez 2003, 51 and 64). The tall  as well as f tend to extend slightly below baseline, which may be an early feature (cf. Derolez 2003, 49–50 and 61). J is used initially (Jpse;­Jam) (cf. Derolez 2003, 50). There are two forms of the letter d. One form is found with a straight shaft and another with the shaft sloping towards the left (ꝺ), a form that was more usual in pre-caroline scripts (Derolez 2003, 49), disappeared in Caroline minuscule and reappears in late caroline and early post-caroline script. A majuscule variant of this d is also found on this fragment (similar to majuscule d on pl. 129 in Hoffmann (1986). The ampersand is in every instance preferred to the Tironian et­which does not occur at all. The ligature of o and r is used throughout consistently. Punctuation includes punctus­elevatus () and punctus interrogativus. The general aspect of the script is more Germanic than French. It is probably of a later date than the manuscript containing Augustine’s tracts on the Gospel of St John discussed above (see Aurelius­Augustinus c)). As for the content, it is from Vitae­patrum, Book 5, i.e., the Verba­seniorum. It is a rather more selective version compared with the text published by Migne (and Heribert Rosweyde before him65). Many episodes are left out and in those included are few major deviations from the printed edition. The two leaves do not have consecutive text. Fol. 1 contains five episodes between Vitae patrum 5, 7, 24 and 5, 8, 4 (= PL 73, cols. 898c–906a), whereas fol. 2 contains nine episodes from Vitae­patrum­5, 11, 26 (PL 73, col. 937a) until 5, 13, 13 (PL 73, cols. 946a–946c). It is hard to decide 62. Formulas, such as Legitur­ in­ uitis­ patrum, occur frequently in medieval collections of exempla and in sermons. 63. In early manuscripts the collection was called Adhortationes­pat­rum (Chadwick 1958, 35). The easiest available edition today is the one of Migne (PL 73) which is merely a reprint of the text published by Rosweyde (1615). There is as yet no critical edition of Vitae­patrum. In the following I refer to Migne’s edition by column. 64. See footnote 29 above. The fact that an account book from c. 1540 is bound with a double leaf supports Pettersen’s suggestion that complete leaves and double leaves were more frequently used for binding account books in the sixteenth century. 65. Migne’s edition which is the one that is easiest available to modern scholars is a reprint of the edition of Heribert Rosweyde (1569–1629) from Antwerpen 1615. There is as yet no critical edition of Vitae patrum. The Old Norse translation of Vitae­patrum, Book 5 (Tveitane 1968, 5) is likewise more selective than the text printed by Rosweyde. The anecdotes on this bifolium are also from the same book (5). fragments of patristic and other ecclesiastical literature [karlsen 2] 237 whether there were once one or several bifolia between the two preserved leaves, given the selective nature of the text (the text on fol. 1v ends on column 906a in the printed edition, whereas fol. 2r continues with the text on column 937a (. . . <l>uctus­cum­autem­aliquis­uenerit­ad­me). The content is as follows: Fol. 1r: . . . potui­cognoscere­nisi­ante­dies­quattuor­.­.­.­sed­ueni­et­uende­omnia­&­fac­secundum­uolun<tatem> (Vitae patr. 5, 7, 24 = PL 73, cols. 898c–899c). Fol. 1v: lacuna <in> conspec<tu> lacuna eroges­ea­pauperibus­et­egenis­.­.­. egredi­cellam­suam­(Vitae­patr. 5, 7, 24 = PL 73, cols. 899c–900a). Frater­quidam­dixit­abbati­arsenio­.­.­. uincebat­eas­(Vitae­patr. 5, 7, 27 = PL 73, col. 900b–900c). Senex­quidam­sedebat­in­eremo­.­.­. ab­aqua­illa­(Vitae­patr. 5, 7, 31 = PL 73, cols. 900d–901a). Eulogius­quidam­nomine­.­.­. senex­cum­gaudio­quod­habuit (Vitae­patr. 5, 8, 4 = PL 73, cols. 905d–906a) Fol. 2r . . . <Quando­in­cella­sum­l>uctus­cum­autem­aliquis­uenerit­ad­me­.­.­.­inuenit­eam­(Vitae­patr. 5, 11, 26 = PL 73, col. 937a). Di<xit> senex:­Sicut­nemo­potest­ledere­.­.­. miseram­animam­nostram­in­ignominie­passionem­(Vitae­patr. 5, 11, 42 = PL 73, col. 939a). Dixit­abbas­cassianus­quia­venimus­.­.­. tunc­ex­sua­potestate­ieiunant (Vitae­patr. 5, 13, 2 = PL 73, cols. 943d– 944a). Dicebant­de­quodam­sene­in <Syria> qui­iuxta­uiam­eremi­.­.­. et­gratias­egerunt­deo­semper­mirabilia <faci>enti (Vitae­patr. 5, 13, 9 = PL 73, col. 945b–c). Monachus­quidam­erat­habens­fratrem­secularem­pauperculum­.­.­.­Tercio­ueniens­attulit­multas­expensas­et­uinum et­pisces.­Qu<od cum­uidisset­frater­eius­.­.­. > (Vitae­patr. 5, 13, 13 = PL 73, col. 946a–946c). Fol. 2v: <Et­ille­dixit> non­domine­quoniam­quando­accipiebam­aliquid­a­te­.­.­. multiplicabitur­<labo>r­eius (Vitae patr. 5, 13, 13 = PL 73, col. 946c–d). Senex quidam­cum­alio­uno­fratre­habebat­communem­uitam­.­.­. Atque­ita­cognoscens­fidem­<et­u>irtutem­senis, glorificavit­Deum (Vitae­patr. 5, 13, 15 = PL 73, col. 947a–947c). Dicebant­de­abbate­Siluano­quod­habuerit­in­Scythi­discipulum­nomine­Marcum (Vitae­patr. 5, 14, 5 = PL 73, col. 948d).66 Huius­mater­aliquando <uenit­ut> uideret­eum­.­.­.­Et­consolatus­est­eam­et­ita­discedere­fecit­(Vitae patr. 5, 14, 6 = PL 73 col. 948a–948c). Venit­quidam­ad­abbatem­Sisoi­Thebeum­uolens­fieri­monachus.­Et­interrogavit­eum­senex­si­quid <haberet­in­saeculo> (Vitae­patr. 5, 14, 8 = PL 73, col. 949d). b) Verba­seniorum Two fragments survive of a manuscript in two columns written in a pointed hand (NRA lat. fragmenter 378, 1–2), for a reproduction of which see plate 28. The provenance is Eiker and Brunla len in southeastern Norway from the year 1632. They both measure 8 × 20 cm. The fragments are parts of the same leaf and contain text from the Verba­seniorum (Vitae­patrum 5, 11, 13–18), reprinted after Rosweyde by Migne (fol. 1r = PL 73, cols. 934c–934d; fol. 1v = cols. 935a–935c). The hand reflects a continental model (cf., e.g., the ampersand and the letter a). In one of the fragments there is a passage not included in PL in the left column: <ar>bore­magna et­attendenti­. . . non­prevaleo­aduersus. These lines form part of a passage not printed by Migne that Chadwick added to his translation (Chadwick 1958, 133) and which he probably supplied from the pre-Caroline manuscripts he mentions in his introduction (Chadwick 1958, 36).67 The manuscript mixes late features (such as forked ascenders on b, d, h, etc.) with rather caroline ones. The ascender of d is always straight and there are also rather few abbreviations. The letters, however, are rather compressed, and there are a few instances of biting. 66. This version lacks most of Chapter 5 (in PL 73, col. 948d et­hic fuerit­magnæ­obedientiæ­.­ . ­ .­ et­nos­diligimus­quoniam­et­Deus­diligit eum). 67. This clearly demonstrates the need for a proper critical edition of the Verba­seniorum. The absence of one is probably explained by the huge number of manuscripts. 238 espen karlsen c) Athanasius/Vitae­patrum Two large successive leaves from the same codex (NRA lat. fragmenter 24, 1–2) contain Athanasius, De­observationibus­monachorum (the end; PL 103, 670c–672b) and the beginning of Vitae patrum Book 5 (= Verba­seniorum), 1, 1–8 (PL 73, cols. 855a–856a). Their provenance is unknown. Fol. 1r is reproduced on plate 29. Text in red on fol. 2r marks the transition between Exhortatio­Athanasii (as it is entitled in the manuscript) and the Hortationes­sanctorum­patrum translated by Jerome (according to this manuscript), i.e., Verba­seniorum. The latter text is introduced by a large red initial. The written space is complete and measures 28.5 × 20 cm, the text being in two columns. The book was probably copied in the late twelfth century, but the script is old-fashioned. A striking feature is that the letters are not compressed and are widely spaced. There is a low frequency of abbreviations (9, the nasal stroke). The scribe seems to have made an effort to write clearly and with special care, perhaps following the conventions of his exemplar which may also have had few abbreviations. There is a striking similarity with the lexicographical text mentioned above (see under Anonymus) which has approximately the same written space but with smaller, compressed letters, narrower spacing and more frequent use of abbreviations. Tall s () is used almost without exception (there is one instance of round uncial capital s in the left column, line 8, castitatem­corporis). The ſt ligature is used whenever this letter combination occurs. The form of the ampersand (which is more common here than the Tironian nota which is also used for et) and the letter g are continental in appearance. The shaft of the d is bent (). The letter j is found in initial and final position several times (Jpsa,­Jtaque and in­filijs,­datij,­Judicij). In this context, j was probably used to avoid confusion: in this script the combination ii could easily have been mistaken for u, so that the scribe may have preferred ij for clarity (cf. Derolez 2003, 59). The descenders of p and q turn to the left. The letter g has an extensive lobe below the baseline which may be a late feature. The few abbreviations and the broad spacing may reflect the appearance of an older exemplar of the late eleventh century or early twelfth century. The script seems based on a caroline exemplar that the scribe took great care to imitate. The exemplar probably also had two columns. This manuscript may have been copied locally. It is rather the general appearance of the script more than specific letter forms, and the dependence the scribe reveals in copying features from an older exemplar that suggest a local origin.68 5. Concluding remarks In this article I have focused on fragments of manuscripts that are likely to have been connected with larger centres, i.e., monastic houses and cathedral chapters; though one should not discount the possibility that such literature might have been found elsewhere.69 The importing of ecclesiastical library books seems to have begun in the very late eleventh century or early twelfth. This accords well with the fact that the second half of the twelfth century saw the composition of non-liturgical texts in Norway, partly in connection with the newly erected archbishopric at Nidaros. The early books, copied before 1150, were all apparently imported. The copies themselves may have arrived later in Norway,70 but one may assume with confidence that such literature must have been available in Norway in the first half of the twelfth century in view of the fact that the dawn of a locally composed literature in Latin in the second half of the century presupposes an earlier importation of the literature dealt with here. It is also evident that ecclesiastical library books were copied locally in the second half of the century. 68. For some of the same reasons I have suggested a local origin for NRA lat. fragmenter 38, 1, a leaf of Hugo de Folieto’s Liber­de pastoribus­et­ouibus (see Wifstrand Schiebe & Karlsen in this volume). 69. See the introduction to this article on the fragment of Donatus found at Lom. 70. See Gullick 2 and Rankin in this volume on the date at which the imported books may have arrived in Norway. fragments of patristic and other ecclesiastical literature [karlsen 2] 239 As for the imported library books in the first half of the twelfth century and into the second half, there is a clear predominance of Germanic and French material, whereas the English contribution appears rather weak in the first half of the century. In this respect, the library books differ from the liturgical fragments among which English material for a long time had been predominant. The fragments offer few surprises as to which texts are found. Among the learned exegetical texts copied early are Augustine’s De­Genesi­ad­litteram, Gregory’s Moralia­in­Iob and Dialogi; Hrabanus Maurus’ Commentarii­in­Genesim, Isidore’s Etymologiae, Peter Comestor’s Historia scholastica, Rufinus’ Historia­ecclesiastica, and the Vitae­patrum (i.e., the part known as Verba­seniorum). There are three different commentaries on the Psalms from the twelfth century, i.e., the ones of Gilbert of Poitiers, Letbert of Saint-Ruf and Peter Lombard. A leaf from a text of a lexicographical nature from the late twelfth century is as yet unidentified. It is discussed under the heading Anonymus on the list above. It contains Christian definitions of words. It is reproduced in plate 3, the inclusion of which may lead to its identification. Palaeographically it is is in a French-looking hand. Despite what was once argued, the works of Hugo de Folieto reached Scandinavia (see the entry for Hugo de Folieto above and the contribution by Wifstrand Schiebe & Karlsen in the present volume).71 As for the later material discussed here, dating from the century after 1250, i.e., the following entries on the list: Aurelius Augustinus c), Gregorius Magnus c), Hugo de Folieto b), Ioannes Cassianus, the script itself does not provide clues regarding their place of production, but their dates suggest that they are more likely to have been copied locally in Scandinavia. The twelfth century was the great century for the copying of the large commentaries of Augustine and Gregory the Great throughout Europe, and the centres where such texts were needed would probably have acquired their copies before the late thirteenth century.72 For obvious reasons the work by Brother Mauritius was certainly copied locally, and he is the only Norwegian author relevant in the present context. It is striking that two later manuscripts (c. 1250–1350) contain textual versions that are either abbreviated or heavily expanded (Aurelius Augustinus c), and Ioannes Cassianus). In the thirteenth and fourteenth centuries patristic texts were integrated into Old Norse literary culture. It is interesting that there is a correlation between texts documented among the fragments and Latin sources used in the Old Norse rendering of the Old Testament called Stjórn. This translation was supplied with copious commentaries to be used at the royal court under King Håkon V Magnusson (1299–1319). Augustine’s De­Genesi­ad­litteram and De­civitate­Dei were known to the compilers of Stjórn, as well as Peter Comestor’s Historia­scholastica, the Dialogi of Gregory the Great, and Isidore’s Etymologiae. Petrus Comestor’s Historia­scholastica, and Isidore’s Etymologiae play a particularly central rôle (Astås 1987, 459–467 and 473–478; Astås 2009, vol. 2, 1305–1324). Sections of Vitae­patrum Book 5 (Verba­seniorum) were selectively translated into Old Norse (edited by Unger 1877, vol. 2, 489–671 and studied by Tveitane 1968). The selection of episodes was either made by the translator or was already present in his exemplar (Tveitane 1968, 9). In this respect the Old Norse translation resembles the oldest Vitae­patrum manuscript discussed above.73 71. Among the fragments of Norwegian provenance in Stockholm discussed by Gunilla Björkvall in the present volume are two fragments of Possidius’ Vita­Augustini (SRA NoFr. 42–43), a text composed before A.D. 439. No more fragments of this manuscript have as yet been discovered in the NRA, contrary to what has been the case with some other fragments of Norwegian provenance in Stockholm. 72. 73. Though the nature of the parchment gives no conclusive evidence in itself, it is worthy of notice that the parchment of some of these late books corresponds to the description of Scandinavian parchment given in Gullick 1 in this volume. For a concordance of episodes between the Old Norse version and Vitae­partum­Book 5 in PL, see Tveitane (1968, 10–11). 240 espen karlsen Gregory the Great had a direct impact on Old Norse literature (Wolf 2001; Wellendorf 2009, 144–157). His Dialogi were translated into Old Norse and a sermon of his found its way into the Old Norse Homily Book (Haugen & Ommundsen 2010). It appears that patristic and theological manuscripts constitute a small group indeed among the Norwegian fragments. Above I have drawn on fragments from twenty-nine library books, one of which was dismembered in Sweden (Mathias Ouidi). In comparison, there were fragments of forty-nine codices of this kind of literature among the thousand fragments in the DRA that formed the basis of the report published by Raasted in 1960 after fifteen years of cataloguing74 (Raasted 1960, 146).75 There is still more material of this kind to be identified in the NRA, but it is likely to remain a small group compared to the number of such manuscripts in the DRA and the SRA. The most likely explanation is that the destruction of codices in Norway was more radical than in Denmark and Sweden unless one supposes that such books were far more rarer in Norway than in other parts of Scandinavia. I have commented upon the provenance of manuscripts, e.g., when a manuscript is divided between two len­concerning Aurelius Augustinus d) or between Norway and Sweden (Mathias Ouidi, with footnote 52 on a missal divided between Norwegian and Swedish institutions) and why fragments of a book used in medieval Sweden are found on Norwegian account books. I have been able to connect the large leaf of Gregorius Magnus a) to an account book in Trondheim and identified one fragment of Robertus de Tumbalena that remained locally. Appendix There are few complete manuscripts preserved today of patristic authors or theological texts known to have been used in Norway before the Reformation (see Karlsen 1 in this volume volume). The NB in Oslo holds among its approximately sixty-five medieval manuscripts in Latin seven of this kind that arrived in Norway after the Reformation. As they are virtually unknown to scholarship I give a list of them: Ms.8º 1757: Julianus Archiepiscopus Toletanus (✝ c. 690): Prognosticon­futuri­seculi. Thirteenth-century. Ms.8º 1761: Petrus Comestor: Historia­scolastica. C. 1300–1350? Ms.8º 2823: Hugo Argentinensis: Compendium­theologicae­veritatis. Germany, fourteenth-century. Ms.4º 1195: Ioannes Cassianus: X­Collationes­sanctorum­patrum. Belgium, twelfth-century. The manuscript once belonged to the abbey of Le Parc in Belgium. Ms.4º 1792: Petrus Lombardus: Libri­sententiarum IV. Ms.fol. 729: Lactantius Firmianus: Divinarum­institutionum­libri­septem. Italy, ‘Anno 1410’. Ms.fol. 2576: Hieronymus: Epistolae. Italy, second half of fifteenth century. In addition there is a glossed book with a Pontigny provenance of Ezra, Tobit and Esther (Ms.fol. 3570) (saec. xii/xiii) with interlinear and marginal commentaries by Bede and Hrabanus Maurus.76 74. In the classification of Raasted TH AA refers to Theologiae­Auctores. 75. This is also striking when comparing with the material dealt with by Lehtinen (2005, 124–128) in Stockholm and Helsinki. 76. The following manuscripts listed above are mentioned by Kristeller (1989, 391–392): Ms.8º 1757, Ms.4º 1195, Ms.fol. 729, and Ms.fol. 2576. karlsen 2 pl. 1 Plate 1. NRA lat. fragmenter (unum. III) Tillegg til skapsakene pk. 2.–2, fol. 1v. Missal (or mass lectionary) (Mi 100a) 241 242 karlsen pl. 2 Plate 2. NRA lat. fragmenter 213, 1, fol. 3r. Missal (Mi 17) Plate 3 (slightly reduced). NRA lat. fragmenter 1a, 1–4, fol. 1v. An unidentified text of a lexicographical nature Plate 4 (reduced). NRA lat. fragmenter Eske 46, 35, fol. 1v. Aurelius Augustinus, De­Genesi­ad­litteram karlsen 2 pl. 5 Plate 5.­In­situ RK Lensrsk. Akershus len, landskatt 1612, Hedmark og Østerdalen, 31.6, 27,­fol. 1r. Aurelius Augustinus, In­Iohannis­evangelium­tractatus 245 Plate 6 (reduced). NRA lat. fragmenter 51, 1–7, fol. 1r. Aurelius Augustinus, De­ciuitate­Dei Plate 7 (reduced). NRA lat. fragmenter 34, 3,­fol. 7v. An adaptation of Aurelius Augustinus, Enarrationes­in­Psalmos 248 karlsen pl. 8 Plate 8 (reduced). NRA lat. fragmenter Eske 45, XXXIII, 1–2,­fol. 2v. Birgitta, Revelaciones karlsen 2 pl. 9 Plate 9. NRA lat. fragmenter 50, 1–2, fol. 1rv. Gilbertus Porretanus 249 250 karlsen pl. 10 Plate 10. NRA lat. fragmenter eske (52) [2], fol. 1r. Gregorius Magnus, Moralia­in­Iob Plate 11. NRA lat. fragmenter 19, 1, fol. 1r (above) and fol. 1v (below). Gregorius Magnus, Dialogi­ 252 karlsen pl. 12 Plate 12. NRA lat. fragmenter 60, 1–4v. Gregorius Magnus, Moralia­in­Iob Plate 13 (opposite). NRA lat. fragmenter 59, 1–2, fol. 1v. Hrabanus Maurus, Commentarii­in­Genesim 254 karlsen pl. 14 Plate 14. NRA lat. fragmenter 8, 1–2, fol. 1v. Hugo de Folieto, De­claustro­animae karlsen 2 pl. 15 Plate 15. NRA lat. fragmenter 30: 1, fol. 1r (above) and fol. 1v (below). Ioannes Cassianus, De­institutis­coenobiorum 255 256 karlsen pl. 16 Plate 16. NRA lat. fragmenter 1, 8–9, fol. 7v. Isidorus Hispalensis Etymologiae (or Origines) karlsen 2 pl. 17 Plate 17. NRA lat. fragmenter 113b: 1–2, fol. 1rv. Isidorus Hispalensis Etymologiae (or Origines) 257 karlsen 2 pl. 19 Plate 19. NRA lat fragmenter 43, 1–2,­fol. 1r (left) and fol. 1v (right). Magister Mathias, Alphabetum­distinccionum Plate 18 (opposite). NRA lat. fragmenter 37, 3 verso, fol.2v. Lietbertus de Insulis, Flores­Psalmorum­ 259 260 karlsen pl. 20 Plate 20. NRA norr. fragmenter 92, 3 + 92, 5,­fol. 3r. Mauritius, Itinerarium­in­terram­sanctam Plate 21 (reduced). NRA lat fragmenter 16, 1, fol. 1r. Petrus Comestor, Historia­ecclesiastica Plate 22 (reduced). NRA lat. fragmenter 36, 1, fol. 1r. Petrus Lombardus, Collectanea­in Paulum karlsen 2 pl. 23 Plate 23. NRA lat. fragmenter 40, 1–2,­fol. 1r. Petrus Lombardus, Collectanea­in­Paulum 263 264 karlsen pl. 24 Plate 24. NRA lat. fragmenter 47, 1–2, fol. 1v. Petrus Lombardus, Commentarius in­Psalmos karlsen 2 pl. 25 Plate 25. NRA lat. fragmenter 52, 1v. Robertus de Tumbalena, In­Cantica­canticorum 265 266 karlsen pl. 26 Plate 26. NRA lat. fragmenter 130, 1–2, fol. 1r. Rufinus, Historia­ecclesiastica karlsen 2 pl. 27 Plate 27. NRA lat. fragmenter 18, fol. 1v. Vitae­patrum (Verba­seniorum) 267 268 karlsen pl. 28 Plate 28. NRA lat. fragmenter 378, 1–2, fol. 1r. Vitae­patrum (Verba­seniorum) Plate 29 (reduced). NRA lat. fragmenter 24, 1,­fol. 1r. Athanasius, De­observationibus­monachorum