The temporal volume of a late-13th-century notated breviary is kept in Strahovská knihovna, the library of the Royal Canonry of Premonstratensians at Strahov in Prague under the shelf mark DE I 7.^[1]  This codex was classified by Janka Szendrei in 1984 as a medieval Hungarian liturgical manuscript that originated from the archiepiscopal see of Esztergom (Strigonium, Gran). It is the first part of a two-volume notated breviary exemplar of Esztergom Cathedral’s scriptorium. For a detailed description of the codex, referred to as Breviarium notatum Strigoniense (BNS), and for a selected bibliography, see CANTUS Database.^[2]  The temporal volume of Breviarium notatum Strigoniense is abbreviated as BNS I, while the sanctoral volume, once thought to be lost, as BNS II.

 

The discovery of the fragments of BNS II

In 2019, the sanctoral volume of BNS was identified in the Metropolitanska knjižnica Zagrebačke nadbiskupije (Metropolitan Library, Zagreb) in 258 smaller or larger fragments glued onto the covers of 129 early prints. Books bound with the leaves of BNS II hold two fragments: one on the front cover and another one on the back.

This particular type of book binding was made in Zagreb around 1692, under Bishop Aleksandar Mikulić who had a new library building erected. In that time, all the printed books of the Zagreb Cathedral library were rebound, and at least 22 codices were cut up and used for this work as raw material.^[3]  The binder(s) created a characteristic binding type that is called today Mikulić binding. (Formerly, it was referred to as Metropolitan binding or Metropolitanski uvez in the book-historical literature.)

Between 2019 and 2023, 14 other fragments of BNS II were identified on books that had been moved from Zagreb Cathedral library over the centuries and today are kept in different collections of Croatia and Hungary. Altogether 272 fragments are known of BNS II at present. There are four collections where its host volumes are found: the Metropolitan Library in Zagreb; the library of the Archiepiscopal Seminary in Zagreb; the library of the Franciscan convent in Zagreb; and ELTE University Library and Archives in Budapest.

The Budapest fragments were detached from the host volumes in the 1980s, so both sides of these leaves can be studied. The others today show only one side, the glued down sides being hidden.

The large folios of BNS II, around 460 × 320 mm, made them ideal bookbinding material in late-17th-century Zagreb. Even for the largest bookboards in the cathedral library, a single folio (2°) of the manuscript was sufficient, i.e. half of a bifolium. For medium-sized books, quarto (4°) and octavo (8°) fragments of BNS II were used, while the smallest volumes were covered with fragments of sextodecimos (16°).

On the basis of the fragments recovered, there is no thematic correlation between the books bound in the codex; books belonging to a given size range make up smaller or larger groups. Among them there are mostly theological works, but we can find Boccaccio’s Decameron^[4] and a famous handbook on witches^[5]  as well.

 

The reconstruction of the corpus of BNS II

Reconstructing BNS II from its fragments means answering two basic questions: (1) What is the original order of the fragments? In other words, how did the content on the fragments flow in the former codex? (2) Which fragments belong to the same folios? To answer these questions, a three-step iterative reconstruction method was needed.

The first step was the content analysis and grouping of fragments by feast, determining a „loose” fragment order. In certain cases, the result of the content analysis was hypothetical.

The second step was the comparison of the fragments with the sanctorale of the Breviarium Strigoniense, printed in 1484,^[6]  and determining the fragments’ order within the feasts as precisely as possible. The main difference between the printed breviary and BNS II is in the readings of matins, which are mostly abbreviated in the printed book. The gradual shortening of the length of legends, sermons and homilies is a well-known phenomenon from the Middle Ages, so this difference between the manuscript and the print can be considered natural. Because of this, unfortunately, there were several fragments for which no reference were provided in the printed breviary.

The third step was considering the working method of the bookbinder. He completed his task very systematically. First, he cut off a quire from the body of the codex already separated from its binding, unfolded it and started working with the innermost (or uppermost) bifolium at hand. When the binding was complete, the printed book’s two boards were covered with continuous or very closely situated liturgical material. After that the binder continued his work with the next bifolium of the quire. Its liturgical content, however, was no longer continuous, but symmetrically surrounded the previous bifolium, and so on. Obviously, the liturgical items on the book boards bound in the last bifolium of a quire fall furthest apart in the calendar. This systematic working method preserved the former structure of the manuscript and it helps to correct the content-based ordering of the fragments. Once the system of the host volumes had been checked according to the preliminary order of the fragments, the complete collation of the codex could be determined. That made it possible to determine the precise liturgical ordering and to refer to the fragments easily by their number in that order (Frs. 1–261). The steps of the iterative reconstruction can be illustrated with the following flowchart.

According to the revealed collation of the codex, BNS II consisted of 24 quires. As it is certain that five quires were quinions, the whole manuscript must have consisted largely of quires of this size, so that originally it would have comprised at most 10 × 24 = 240 folios.

The iterative reconstruction was also verified by examining the initials of BNS II. The hands of four illuminators were identified in BNS II, two of them also appear in BNS I. These four illuminators worked on certain quires of BNS II almost simultaneously. Their separate work thus confirmed the reconstructed physical structure of the manuscript, and therefore, the correctness of the aforementioned iterative method.

According to estimation, 165.25 folios of the at most 240-folio manuscript have been identified; that means 68.9% of the codex. At present, the glued down sides of the identified fragments cannot be studied, so we have access to about half of this surviving material, that is about 35% of the codex. These figures also mean that we currently know nothing about up to 74.75 folios of content (31.1%). It is therefore worth continuing to search for elusive fragments, even though a significant proportion of the material may have been destroyed.

 

The appendices of BNS II — FFCC and Appendix Zagrabiensis

BNS II had a special appendix, a quire of non-Hungarian origin (Frs. 262–269), that is a witness of the historia Sapientia aedificavit sibi domum for the feast of Corpus Christi. The leaves of this quaternion were preserved on folio-size (2°) printed books, so the loss of the written material is not too much. The host volumes are kept partly in the Metropolitan Library, Zagreb, and partly in the library of the Franciscan convent in Zagreb. This fragmentary quire is referred to as Fasciculus festi Corporis Christi, or FFCC in abbreviated form.

The provenance of FFCC was identified through a complex analysis. As a result, the original musical notation of the quire can be localized in the region of Central France, its surviving psalm differentiae are the same as those of the 12th-century antiphonal of the Benedictine monastery of Saint-Maur-des-Fossé.^[7]  The hymn settings in FFCC are echoed in early-14th-century sources of Central French origin, and finally the palaeographical analysis confirmed that it was copied by a Parisian hand. Thus, FFCC is a physical evidence of the ‘liturgical import’ of the new feast of Corpus Christi, and consequently, a written evidence of a direct or at least an indirect 13th-century cultural relationship between Esztergom and the region of Paris. It is the earliest known Corpus Christi libellus in the world and the earliest known notated witness of the historia Sapientia aedificavit, including the famous Corpus Christi hymns by Thomas Aquinas. On the other hand, FFCC was proved to be the textual exemplar of historia Sapientia aedificavit in BNS I (see fols. 213v–221r).

Three folio-size fragments (Frs. 270–272) are labelled as the ‘Zagreb appendix’, the Appendix Zagrabiensis of BNS II. In fact, these fragments are late additions to the intentionally-left-blank folios of the codex. These leaves reveal readings of the feasts of St. Demetrius and the Sanctification of Mary, added to BNS II in Zagreb.

 

A short summary of the history of BNS II

Having put the fragments of BNS II in order, it is possible to evaluate the liturgical and melodic history of the codex’s contents. Obviously, this analysis cannot be exhaustive, since today only about half of the textual and melodic material is accessible. Nevertheless, even the content to be studied on the visible pages of the fragments characterises the liturgical environment of the codex, and provides a unique insight into the literary and musical milieu that shaped the everyday life of Esztergom Cathedral in the second half of the 13th century. In the light of the known offices of BNS II, the codex can be dated between around 1285 and 1295. Consequently, BNS I may also have been copied in this period.

According to the most plausible hypothesis, BNS II was taken to Zagreb already in the first half of the 14th century, and served as a kind of ‘secondary literature’ for liturgists, and may have been kept on the shelf of the local scriptorium. Unfortunately, by the 17th century, the codex had completely lost its function, and the reasons for which it had come to Zagreb had become obscure. Around 1692, it was found by the bookbinder in charge of rebinding the cathedral’s printed books, and its fate was sealed.

 

The digital representation of BNS II and the future research

Given the current state of knowledge, the most integrated way to present BNS II is a visual reconstruction collated from the visible parts of the codex’s known fragments. It is available in digital form under the tab ‘Virtual codex’. On the right, a table shows the basic data of the chosen fragment. (A fragment becomes active when the user clicks on it with the left button of the mouse.) Besides the data table, the names and calendar dates of feasts are also to be found above or below the image of the fragment. The fragment’s number is included in brackets (Fr.) and the host volume’s shelf mark is provided next to it. Where it has been successfully determined that a fragment shows the verso or recto side, that too is included in the margin in red. The Budapest fragments are already detached from the host volumes (Frs. 15, 25, 206, 220), and therefore both sides of these folios appear in the reconstruction. Fragments presented in the edition on one page that did not once belong to the same folio are separated with a red line. In a few cases, the binder cut up a folio and glued it down so that now opposite sides are facing up, making the visible sides of these 4° fragments unmatchable. These cases are highlighted with crossing arrows by the red line. Symbols (listed below) also indicate whether a fragment belongs to the top or bottom half of what was once a folio. Faded columns indicate when the vertical positions of the fragments on the former folio cannot be determined precisely.

The fragment belongs to the upper part of a former folio.  The fragment belongs to the lower part of a former folio. The two fragments are probably part of the same folio, but were pasted onto the book covers with opposite sides.

 

 

As a result of the present work, the sanctoral part of the late-13th-century office of the archiepiscopal see of Esztergom can be studied in detail; a missing notated source of the most important liturgical tradition of medieval Hungary has been revealed. The search for the fragments of BNS II, the reconstruction of the codex and the analysis of its content are finished for a while in the certain knowledge that there is still work to be done in these areas. In principle, previously unknown books in Mikulić binding could turn up in any Croatian or Hungarian collections. On some of these volumes we can hope to find fragments of BNS II, witnesses of Esztergom’s bygone medieval liturgical and musical culture. This homepage, a virtual representation of BNS II will be continuously expanded with fragments that will hopefully be found in the future.

 

_________________________

 

A hardcover facsimile edition of BNS II

A printed visual reconstruction of BNS II with a detailed introductory essay in three languages (Hungarian, English, and Croatian) is published in 2025:

 

 

 

Order information:

HUN-REN Research Centre for the Humanities, Institute for Musicology

Email: zti.titkarsag@abtk.hu

Web: https://zti.hu/en/contacts