Fragments of the Liturgical-Musical Codex from the Archdiocesan Archive of Gniezno (Poland): Source Analysis and Provenance Hypotheses

: This paper discusses hitherto unidentified loose folios of a parchment liturgical and musical book held in the Archdiocesan Archive of Gniezno (Poland), containing the offertory and communion antiphons for the feasts De Trinitate and Corpus Christi. The author provides the codicological description of the leaves (analyzing Latin script, musical notation, ornamentation); identifies the time of their creation (15th century); indicates the type of the liturgical book to which they belong (graduale); seeks a melodic model for them and puts forward provenance hypotheses. He states that the melodics of the antiphons, although closest to the Cistercian tradition, are nevertheless variantly different from the melodic line preserved in foreign and Polish codices. It is possible to narrow down the dating of the leaves thanks to the type of Latin script, the calligraphic ornamentation of the initials and the spelling of certain letters.


Introduction
The collections of the Archdiocesan Archive of Gniezno, despite many years of systematic research, still hide unidentified examples of liturgical monody that remain a mystery to musicologists.This group of artefacts includes two loose parchment leaves (Figures 1-4), not belonging to any of the manuscripts kept in the archive, which have recently been found during the cataloguing of music documentation.All we know is that they were delivered there in 1970 by the archaeologist Dr Gabriela Mikołajczyk, then head of the Archaeological Museum in Pozna ń, Gniezno Branch.The scarcity of the analytical material undoubtedly poses the greatest difficulty to obtaining the most credible research results possible; however, it is not clear if the codex from which the surviving folios come still exists at all and if so, where it can be found.The aim is the introduction of existing parchment leaves into scientific circulation, identifying the type of book they originate from, and determining its dating and provenance.

Codicological Description
The text of the gradual was written on northern type parchment, tanned on both sides, called charta theutonica.The leaves measure 543 × 375 mm and the text area 360 × 240 mm.Both folios have their lower right corners beveled (30 mm).The book format of the codex should be described as in maior forma-forma regalis.The text is written in a single column, which was common in the 15th century.The manuscript has original Arabic foliation, with numbers written in black ink at mid-height of the outer margin.The first leaf is numbered 92, the second 96.Folio 96 recto has a provisional shelf mark Muz.Lit.41 in the upper right corner.Assuming that the codex from which they are derived consisted of regular quires (each of eight leaves stitched together), the surviving folios would belong to the 12th gathering (fol.91-98).The use of Arabic numerals, documented in Polish sources no sooner than in the second half of the 14th century, provides an important clue for dating the leaves.There is also a red Roman numeral LXXXXIII written twice, in the middle of the outer margins of folios 92v and 96v.The communio chants are marked in red with the rubric 'com'.
The liturgical text was carefully written in the 15th-century Gothic rotunda script, which is particularly evident in the way the letter 'o' is written.The writing tends towards a broad duct with a horizontal tendency.The Gothic breaking of stems and sharpening of curves also considerably relaxed compared with French or German Gothic.The letters are wider and have gently pointed curves and gently broken stems.The rotunda's features of the letters 'a' (low bowl and deeply drawn arch), 'g' (the upper part sometimes trapezoidal in shape), 'f', 'm', 'n', 'r' (legs truncated horizontally on the bottom line) (Semkowicz 2002,  pp.318-20) are reflected in the leaves under examination.The letter 'd' resembles the uncial script.
The scribe also had his own style.This can be seen in the way some letters are penned: the crossbar of the letter 't', when at the end of a word, has a delicate hook-shaped flourish on the right side; the letter 's' in ligature with 'b' follows the shape of the bowl of the letter 'b', forming an abbreviation 'bus' (omnibus, viventibus) and in the lower part also ends with a delicate flourish; the letter 'h' in its lower part has a delicate ornament (hunc); the letter 'm' (suam, calicem) resembles the shape of the numeral 3 with a thin line drawn upwards and rounded at the end.The shape of this letter appears in the 15th century (Semkowicz 2002,  p. 322), which is an essential lead as to the dating of the source under investigation.In addition to the simple minuscule 'r', the script also features a round 'r' which is a ligature of OR.The conjunction et is written in two ways: as a ligature and as a tachygraphic sign in the shape of the Arabic numeral 7. The Latin script tends to maintain an even structure of individual lines.Even, wide letters with gently sharpened arcs and gently broken shafts allow the script of the Gniezno leaves to be classified into the category of the so-called Italian Gothic, which is an important provenance clue.
The overall state of codex preservation should be described as good.The saturation of ink in both the liturgical and musical writing seems fine.The visible traces of neumes on the four-line staffs next to the original notation are the result of encaustic penetration from verso to recto and the other way round, and further confirm the quality of the writing material.Varying degrees of the soiling of the lower corners of the leaves and wrinkling of the parchment provide evidence of the manuscript's use.Both folios have been trimmed with a protective tape 0.5 cm wide to prevent tearing.

Musical Notation
In the leaves under study, the standard four-line handwritten notation system, employed until the late Middle Ages, can be found (Hughes 1982, p. 108).The graphic scheme is formed by five red four-line stave (spanning 35 mm) on a leaf with neumatic notation.The melodies were written in square notation (nota quadrata).At its core there is the desire to mark each note clearly and unambiguously, which conditioned fast reading of intervals (Szendrei 1999, pp.197-98) 1 .C-clef was used throughout.On fol.92, it is always located on the fourth line, while on fol.96, it changes its position depending on the ambitus of the melodies.The copyist therefore ensured that the tune did not go beyond the musical system.
The neumes represent a mature stage in the development of square notation, as evidenced by their paleographic features: the shapely form of the single squares showing no tendency to elongate or blur, and the regular form of compound neumes, such as climacus, with square and rhomboid notes maintaining the correct proportions.Due to the limited research material, it is not possible to determine other features, for example the presence or absence of liquescent neumes or quilisma, although the latter-as pointed out by medievalists-does not occur in medieval square notation as a result of less importance being attached to the Gregorian chant in the 12th and 13th centuries (Rampi and Lattanzi  1991, p. 346).
An equally important mark to identify the writer is the custos, located at the end of each line of music.Like the clefs and neumes, the symbol was made in black ink.It is written in the same style on both leaves and takes the form of a delicate rhombus with a long diagonal thin line pointing upwards, always in the same direction.It is very easy to read thanks to its rhomboidal form contrasting with the square shape of the neume.All the custodes were transcribed properly in the correct line or space, attesting to the copyist's accuracy and musical competence.An analogous form of this sign was also noted in the above-mentioned Italian codices.
In neumatic notation, there are bar lines-usually single bars, double appear only twice-crossing all the lines and separating longer stretches of melody.One explanation for the presence of these distinguishing marks may be the need to coordinate the flow of melodies and text.Bars, an important graphic and punctuation element, would divide the musical notation into smaller melodic units and thus make it easier to read the text and melody from a considerable distance in the monastic choir (Wiśniewski 2022, p. 124).It is possible that they were added later, when the book was already in use, as may be suggested, for example, by their very delicate duct.They could also have been related to the rhythmic interpretation of the melody (Augustyniak 2000, pp.121-22).The double bars appear to have indicated the incipit of the chant intoned by the cantor in Benedictus sit, and in Quotiescumque.This practice is confirmed by other books, inter alia the above-mentioned Italian codices, and by the 15th-century gradual from Maastricht (NL-Uu: Hs. 0415), in which the same fragment of the chant (Benedictus sit) is separated from the further part by a red bar crossing all the lines of the stave (fol.223r).Comparing it with music books used at present, it can be said per analogiam that the double bar fulfilled the role of the asteriscus.

Adornments 17
To break the monotony of the visual design of a medieval manuscript, different measures were taken with the intention of rendering it more attractive.The most common choice was use of ornamentation.The practice of supplementing the text with illuminations resulted from the most important needs of manuscript users, which include articulating the text, presenting the content of the text in a visual form for better understanding and memorization, and giving the text an ornate, solemn and beautiful character (Miodo ńska  1993, p. 103).In an attempt to improve the readability of the codex, among other things, different sizes of initial letters were used.Some fitted the linear system, others surpassed it, and some spanned several systems (Miazga 1977, p. 18).
The discovered leaves do not contain rich and elaborate ornamental elements.Their only decorative element is the calligraphic ornamentation of three initials of the heights of one stave and one verse, fol.92r: B(enedictus), dimensions 530 × 520 mm (letter frame 620 × 590 mm); fol.92v: B(enedicimus), dimensions 500 × 530 mm (letter frame 540 × 510 mm); fol.96r: Q(uotiescumque), dimensions 470 × 520 mm (letter frame 520 × 530 mm), drawn by the same illuminator.The initials are consistently decorated with very precise pen-flourishing with buds drawn in contrasting color: red decoration for blue letters and vice versa.The interweaving of these colors served above all practical purposes.By emphasizing the painterly side of the codex, the scribe enabled a better orientation within the book for those using it during the liturgy.The decoration of the initials is coherent, both in its overall conception and in the elaboration of the details; it is modest and limited to the block of the letter (without appendages).The single-color letter bodies with flowing contours are embellished with narrow vertical stripes and point-like thickening at the ends of the serifs and are inscribed in geometric quadrilateral filigree frames, the corners of which are filled with small quasi-masquerie forms.In the counters of the letters, divided into irregular fields by narrow bars, there are symmetrical arrangements of buds of different sizes; the incipit initial B(enedictus) has two upward and downward growing clusters with pointed buds, the incipit initial B(enedicimus)-diagonally arranged triple clusters with rounded buds, while the counter of the incipit initial Q(uotiescumque) is filled with a diagonal band with two rows of long-stemmed buds running opposite each other and single caterpillar forms along the curvature of the counter.Openwork patterns between the buds are opaquely filled in the color of pen-flourishing.
The frames of the letters are formed of simple thin double bars covered with minute pearls; each side of the frame is additionally decorated with a fine segmental arch with pearls on the outside and regular hatching on the inside, giving the effect of a flattened quatrefoil in the background, which, combined with the double contour around the curves of the letters, produces a multi-layered effect of the whole composition.The motif of pearls adorns the external lines of the serifs and tails extending beyond the frame of the letter.It is worth noting the repetition of given ornamental motifs.In addition, the letter 'u' in the word Quotiescumque has very delicate pen decoration.According to Prof. Monika Jakubek-Raczkowska, the adornments indicate reduced and very late forms of so-called caterpillars.Another typical feature is writing the letter immediately following the initial letter in majuscule script: BEneditus, BEnedicimus, QUotiescumque.

Source Identification
Despite the paucity of analytical material, it can certainly be stated that the fragments of the Gniezno codex come from the gradual.This is clearly evidenced by their liturgical and musical content, typical of the Mass liturgy (more specifically part of the Proprium de tempore) and intended for two Church feasts: De Trinitate and Corpus Christi.The Mass forms transcribed for them have only the offertory and communion antiphons, which vary in their degree of completeness: C o m m u n i o: Quotiescumque manducabitis panem hunc et cali (Figure 3) [verso] cem bibetis mortem domini annuntiabitis donec veniat itaque quicumque manducaverit panem vel biberit calicem [further text missing] (Figure 4).
The above antiphons were adopted for the same liturgical purposes by the Roman Gradual.In the case of the offertory antiphons, the absence of a psalm verse bears witness to their "modern" form.The addition of verses to offertory antiphons was historically linked to the rite of bringing the gifts to the altar, which ended with the celebrant giving the schola ad hoc a hand sign to end the singing.The process of the disappearance of the verses in offertories occurred between the 10th and 12th centuries (Maciejewski 1973, pp.226-27).The fact that there are only antiphons on the Gniezno folios indicates liquidation the cessation of the procession with gifts in the liturgy.

Melodies
The specificity of the musical content of liturgical codices primarily results from the fact that the way of singing was initially the domain of oral tradition.The oldest books with musical notation appeared in the Carolingian Empire in the mid-9th century.In spite of this, they were still not used to perform chants, but were only referred to in case of doubts related to the detailed course of a concrete melody.Books began to be used for singing after the introduction of diastematic notation, between the 11th and 13th centuries.Thus, singing from memory had its consequences in the shape of a melody.It resulted, in fact, in a large quantity of melodic variants of liturgical chants, often differing only in details (Kubieniec 2013, pp.119-21).This is confirmed by the melodic layer of the preserved antiphons from the Gniezno Codex differing in various ways from the commonly transmitted tradition in foreign manuscripts 18 .The stability of the modus cantandi of the antiphons: Benedictus sit Deus and Benedicimus Deum, however, can be observed in medieval Polish graduals of Cistercian provenance.They are generally consistent in their choice of melody with the Cistercian graduals PL-Wn 12496 IV 19 , PL-WRu I F 411 20 , PL-WRu I F 414 21 , PL-WRu I F 413 22 and PL-WRu I F 416 23 .
According to the list of non-psalmodic offertories published by R. Maloy, this antiphon is not recorded in Old-Roman sources, but it occurs in a few Frankish sources.Benedictus sit is a chant which belongs to the Frankish-instituted Solemnity of the Most Holy Trinity.In terms of the melodics, it is a contrafact of the offertory Constitues eos (Graduale Triplex, p. 434), probably a chant of non-Roman provenance, as it shows textual variants corresponding to the psalter used in Gaul in the pre-Carolingian period (Maloy 2010, p. 81).It would seem that Benedictus sit was composed by the Frankish for the needs of the new feast (Bernagiewicz 2013, p. 301).
In all records, the antiphon Benedictus sit Deus (Figure 5) was recorded in mode III.In the Gniezno record, compared to the Cistercian versions, there are slight differences in the melody, consisting in the lack of repetition of the sounds on the words Dei, sanctus, quoque, fecit, and in slight changes in the shaping of the melody on some syllables of the words: sit, Dei, quia, nobiscum, misericordiam.In the Gniezno record, the sounds accompanying the second syllable of the word mi-se-ri-cor-di-am are written one tone higher than in the other codices; in turn, on the word quia, the melody begins with the same sound in all codices (re), but the melody continues differently: in the Gniezno codex re-mi-mi-la, and in the Cistercian sources re-sol-sol-do.The discrepancies between the Gniezno version and the Cistercian codices are most visible in the word quia, the second syllable of nobiscum and the first two syllables of misericordiam.Although melodies may come from one source, they have followed different paths of development.
In all records, the antiphon Benedictus sit Deus (Figure 5) was recorded in mode III.In the Gniezno record, compared to the Cistercian versions, there are slight differences in the melody, consisting in the lack of repetition of the sounds on the words Dei, sanctus, quoque fecit, and in slight changes in the shaping of the melody on some syllables of the words sit, Dei, quia, nobiscum, misericordiam.In the Gniezno record, the sounds accompanying the second syllable of the word mi-se-ri-cor-di-am are written one tone higher than in the othe codices; in turn, on the word quia, the melody begins with the same sound in all codice (re), but the melody continues differently: in the Gniezno codex re-mi-mi-la, and in the Cistercian sources re-sol-sol-do.The discrepancies between the Gniezno version and the Cistercian codices are most visible in the word quia, the second syllable of nobiscum and the first two syllables of misericordiam.Although melodies may come from one source they have followed different paths of development.The melodics of Benedicimus Deum (Figure 6) are fully consistent with the Cistercian version.The only difference is that it-and one Cistercian melody-were transposed a fifth down in relation to the four Cistercian records.The scribes of G and PL-WRu IF 416 either made conscious transpositions or used some local pattern.The explanation for the transposition may perhaps be more the convenience of singers or notation.Transposition was also used to avoid the flat (Miazga 1977, p. 163).From the performers' point of view as long as interval relations within a melodic formula do not change, a chant remains the same in terms of modality.This is what happens in this case.Regardless of the transposi tion, the complete conformity of the Gniezno melody (15th century) with the Cistercian melody (13th and 14th century) indicates that the Gniezno scribe faithfully copied the Cis tercian variant.Moreover, the variants of the antiphons Corpus Christi differ from the melody hand down in the foreign codices 24 and the three Polish books consulted: PL-SAk 40 25 , PL-Kl R 1 26 and PL-Wn 12722V 27 .

Benedictus sit Deus
The preserved Gniezno fragment of the antiphon Sacerdotes Domini presents great affinity with the 14th-century Gradual from Wiślica (PL-Kl RL 1) (Figure 7).In relation this record, the Gniezno melody is less ornamented.The Sandomierz record (PL-SAK 4 and the Gradual from Tyniec (PL-Wn 12722V) have a clearly different shape of the m ody.The Gniezno version retains the same notation as the Sandomierz gradual in on four places, on the words: non, nomen and the last syllable of alleluia; while with the Gra ual from Tyniec, the melody is recorded identically on seven syllables (the initial fragme of the melody, the first syllable of pollunet, nomen, eius and the first syllable of alleluia).T differences in the course of the melody between individual records confirm the the about the individual character of local musical traditions.In this case, a stricter criterion that should be considered is the tonal qualificatio which is not easily modified in the transmission of the liturgical repertoire (Kubieni 2013, p. 137).In three sources (G, PL-SAK 40, PL-Kl RL 1) the antiphon is recorded in mo III, and in PL-Wn 12722 V in mode IV.Based on the CANTUS database, it is known th some sources classify the antiphon Sacerdotes Domini as mode I 28 .This means there is lack of correspondence between different transmissions of the same repertoire.Therefo it is impossible to establish permanent boundaries between musical traditions 29 .

Sacerdotes Domini
In all entries available in the CANTUS database, foreign and Polish (48 concordanc in total) 30 , the melody of the antiphon Quotienscumque manducabitis is recorded in mo VII.Therefore, the consistency of the choice of a tonal version of this chant is not region Much more interesting are some discrepancies in the melody of this antiphon betwe Polish sources (Figure 4).In the case of the word manducabitis (Figure 8), between t Gniezno message and other sources, the third and fourth syllables of the word are orn mented inversely: the Gniezno melody records the syllable -ca-with torculus and cliv Moreover, the variants of the antiphons Corpus Christi differ from the melody handed down in the foreign codices 24 and the three Polish books consulted: PL-SAk 40 25 , PL-Kl RL 1 26 and PL-Wn 12722V 27 .
The preserved Gniezno fragment of the antiphon Sacerdotes Domini presents greater affinity with the 14th-century Gradual from Wiślica (PL-Kl RL 1) (Figure 7).In relation to this record, the Gniezno melody is less ornamented.The Sandomierz record (PL-SAK 40) and the Gradual from Tyniec (PL-Wn 12722V) have a clearly different shape of the melody.The Gniezno version retains the same notation as the Sandomierz gradual in only four places, on the words: non, nomen and the last syllable of alleluia; while with the Gradual from Tyniec, the melody is recorded identically on seven syllables (the initial fragment of the melody, the first syllable of pollunet, nomen, eius and the first syllable of alleluia).The differences in the course of the melody between individual records confirm the thesis about the individual character of local musical traditions.Moreover, the variants of the antiphons Corpus Christi differ from the melody hande down in the foreign codices 24 and the three Polish books consulted: PL-SAk 40 25 , PL-Kl R 1 26 and PL-Wn 12722V 27 .
The preserved Gniezno fragment of the antiphon Sacerdotes Domini presents greate affinity with the 14th-century Gradual from Wiślica (PL-Kl RL 1) (Figure 7).In relation t this record, the Gniezno melody is less ornamented.The Sandomierz record (PL-SAK 40 and the Gradual from Tyniec (PL-Wn 12722V) have a clearly different shape of the me ody.The Gniezno version retains the same notation as the Sandomierz gradual in onl four places, on the words: non, nomen and the last syllable of alleluia; while with the Grad ual from Tyniec, the melody is recorded identically on seven syllables (the initial fragmen of the melody, the first syllable of pollunet, nomen, eius and the first syllable of alleluia).Th differences in the course of the melody between individual records confirm the thesi about the individual character of local musical traditions.In this case, a stricter criterion that should be considered is the tonal qualification which is not easily modified in the transmission of the liturgical repertoire (Kubienie 2013, p. 137).In three sources (G, PL-SAK 40, PL-Kl RL 1) the antiphon is recorded in mod III, and in PL-Wn 12722 V in mode IV.Based on the CANTUS database, it is known tha some sources classify the antiphon Sacerdotes Domini as mode I 28 .This means there is lack of correspondence between different transmissions of the same repertoire.Therefore it is impossible to establish permanent boundaries between musical traditions 29 .

Sacerdotes Domini
In all entries available in the CANTUS database, foreign and Polish (48 concordance in total) 30 , the melody of the antiphon Quotienscumque manducabitis is recorded in mod VII.Therefore, the consistency of the choice of a tonal version of this chant is not regiona Much more interesting are some discrepancies in the melody of this antiphon betwee Polish sources (Figure 4).In the case of the word manducabitis (Figure 8), between th Gniezno message and other sources, the third and fourth syllables of the word are orna mented inversely: the Gniezno melody records the syllable -ca-with torculus and clivi while the remaining codices use punctum in this place, and, unlike the Gniezno varian they ornament the melody on the syllable -bi-using clivis twice, while the Gniezno nota tion has a single note at this point.The place reveals an important difference betwee source G and the other books, which record this fragment of the melody in the same way In this case, a stricter criterion that should be considered is the tonal qualification, which is not easily modified in the transmission of the liturgical repertoire (Kubieniec 2013,  p. 137).In three sources (G, PL-SAK 40, PL-Kl RL 1) the antiphon is recorded in mode III, and in PL-Wn 12722 V in mode IV.Based on the CANTUS database, it is known that some sources classify the antiphon Sacerdotes Domini as mode I 28 .This means there is a lack of correspondence between different transmissions of the same repertoire.Therefore, it is impossible to establish permanent boundaries between musical traditions 29 .
In all entries available in the CANTUS database, foreign and Polish (48 concordances in total) 30 , the melody of the antiphon Quotienscumque manducabitis is recorded in mode VII.Therefore, the consistency of the choice of a tonal version of this chant is not regional.Much more interesting are some discrepancies in the melody of this antiphon between Polish sources (Figure 4).In the case of the word manducabitis (Figure 8), between the Gniezno message and other sources, the third and fourth syllables of the word are ornamented inversely: the Gniezno melody records the syllable -cawith torculus and clivis, while the remaining codices use punctum in this place, and, unlike the Gniezno variant, they ornament the melody on the syllable -biusing clivis twice, while the Gniezno notation has a single note at this point.The place reveals an important difference between source G and the other books, which record this fragment of the melody in the same way.
The Gradual from Wiślica (PL-Kl RL 1) shows the most discrepancies compared to the other sources, which are particularly visible in the words annuntiabitis donec veniat.The notation of the melodies, although consistent for all sources in 33 places (syllables) out of 56 in total, proves that they were not copied from any one archetype.What is noteworthy in the Gniezno record, in relation to all other variants of the melody, is the fact that in several places it substitutes two-and three-note neumes with single notes: the last syllable of calicem, the second syllable of bibetis, and in the case of the word Domini, its first syllable is additionally extended by one note (do), while other sources unanimously note pes.Such minor transformations of the Gniezno melody in relation to other versions, including analogous changes on the word manducabitis, clearly demonstrate the distinctiveness and uniqueness of the local cantus ecclesiasticus.The Gradual from Wiślica (PL-Kl RL 1) shows the most discrepancies compared to the other sources, which are particularly visible in the words annuntiabitis donec veniat The notation of the melodies, although consistent for all sources in 33 places (syllables out of 56 in total, proves that they were not copied from any one archetype.What is note worthy in the Gniezno record, in relation to all other variants of the melody, is the fac that in several places it substitutes two-and three-note neumes with single notes: the las syllable of calicem, the second syllable of bibetis, and in the case of the word Domini, its firs syllable is additionally extended by one note (do), while other sources unanimously note pes.Such minor transformations of the Gniezno melody in relation to other versions, in cluding analogous changes on the word manducabitis, clearly demonstrate the distinctive ness and uniqueness of the local cantus ecclesiasticus.
This demonstrates that the process of shaping and passing on the musical tradition was complex.It seems that both the written source from which the Gniezno version was copied and the oral tradition, confirming the individual character of these melodies, in fluenced the creation of the examined leaves.

Provenance Hypotheses
Dating and locating a manuscript is not an easy task.In the case of the leaves unde investigation, the fundamental difficulty lies in the very scarce research material, with no even indirect, clues to its origins.The information obtained about the type of Latin script palaeographical features, ornaments and liturgical content is helpful in formulating hy potheses on the dating and provenance.As the folios examined are written in the Gothic rotunda, they were compared with other Gniezno manuscripts representing this scrip This demonstrates that the process of shaping and passing on the musical tradition was complex.It seems that both the written source from which the Gniezno version was copied and the oral tradition, confirming the individual character of these melodies, influenced the creation of the examined leaves.

Provenance Hypotheses
Dating and locating a manuscript is not an easy task.In the case of the leaves under investigation, the fundamental difficulty lies in the very scarce research material, with no, even indirect, clues to its origins.The information obtained about the type of Latin script, palaeographical features, ornaments and liturgical content is helpful in formulating hypotheses on the dating and provenance.As the folios examined are written in the Gothic rotunda, they were compared with other Gniezno manuscripts representing this script (Rył  1976, p. 245)  31 .Unfortunately, no suitable model was identified for them in the local book collection.
A certain clue as to the provenance of the leaves may be the convent of Canons Regular existing in the Middle Ages in Trzemeszno near Gniezno, which dates to the 12th century (Dorszewski 2013, p. 98).Given the above-mentioned gradual NL-Uu: Hs. 0415, originally also belonging to the Canons Regular, and the characteristic graphic form of the flat sing observed both in that book and on the folio of the Gniezno gradual, one cannot rule out the monastic usage of the Gniezno leaves and their connection with the order of Canons Regular.It is possible that the unidentified loose gradual leaves stored in Gniezno may have belonged to the Canons of Trzemeszno, who needed various types of books for their worship.Such a hypothesis seems to be, at the present stage of research, only a suggestion regarding the provenance of leaves.If it is assumed that the book was indeed owned by the Canons Regular of Trzemeszno, it could be either an import or a work of the local scriptorium, which carried out particularly intense activity under Abbot Maciej of Kraków (1444-1481), during whose reign the convent was home to many scholars and distinguished writers (Dorszewski 2013, p. 138).Nevertheless, the codices brought to Trzemeszno at that time (Evangelia et lectiones and Expositiones psalmorum), now kept in the Archive of Gniezno, are dated to the 12th century and show no analogy with the examined leaves.
However, considering the type of the Latin script of the Gniezno leaves (Italian Gothic), the way of decorating the initial letters in the form of outlining the letters with vertical and horizontal lines, characteristic in this form mainly for Italian manuscripts (examples are given in the section Musical notation above), and the custos written in a rather specific way noted also in Italian books, Italian provenance of the leaves is most likely.Although determining a specific center or even region is difficult, because both leaves do not contain any special elements, such as miniatures, but the Latin script, musical notation, transparency and very large margins are characteristic of Italian manuscripts from the 14th and 15th centuries.According to Dominika Grabiec, only medieval Spanish and Portuguese manuscripts are similar to them, but these did not reach Poland 32 .
On the other hand, the mass form of the Solemnity of Corpus Christi proves exceptionally valuable for determining the dating of the aforesaid folios.The beginnings of this solemnity are attributed to St. Thomas Aquinas (1225-1274).It was first introduced in Liège (Belgium) in 1247 and instituted a feast for the entire Church in 1264 by Pope Urban IV.However, due to his death, the bull establishing the Solemnity of Corpus Christi was not promulgated.It was not until 1317 that the feast was inaugurated in the Universal Church by Pope John XXII (Zalewski 1973, p. 106).Taking these circumstances into consideration, it must be inferred that the Gniezno manuscript could not have been created before the 14th century.This means that the leaves must have been written after the feast had been introduced.The type of Latin script used (Gothic rotunda), the ornamentation of the calligraphic initial letters and the spelling of some letters clearly indicate the time of the leaves' creation was the 15th century.

Conclusions
The discussed gradual folios therefore constitute, due to their unique character, an interesting source material for determining, for example, the subtype of the local liturgical and musical tradition they represent.However, due to the limited research material, any conclusive statements on this issue at this point remain purely speculative.If the missing part of the codex from which the preserved leaves originate was found, this would offer completely new research perspectives, allowing the relationships between the manuscript and other sources to be understood and perhaps the place of its origin and use to be accurately identified.
Funding: This research received no external funding.
recto] O f f e r t o r i u m: Benedictus sit Deus pater unigenitusque dei filius sanctus quoque spiritus quia (Figure 1) [verso] fecit nobis cum misericordiam suam.C o m m u n i o: Benedicimus Deum caeli et coram omnibus viventibus con [further text missing] (Figure 2).Corpus Christi [fol.96 recto] O f f e r t o r i u m: [lack of beginning] suo et non polluent nomen eius alleluia.

Figure 5 .Figure 6 .
Figure 5. Antiphon Benedictus sit Deus.The melodics of Benedicimus Deum (Figure6) are fully consistent with the Cistercian version.The only difference is that it-and one Cistercian melody-were transposed a fifth down in relation to the four Cistercian records.The scribes of G and PL-WRu IF 416 either made conscious transpositions or used some local pattern.The explanation for the transposition may perhaps be more the convenience of singers or notation.Transposition was also used to avoid the flat(Miazga 1977, p. 163).From the performers' point of view, as long as interval relations within a melodic formula do not change, a chant remains the same in terms of modality.This is what happens in this case.Regardless of the transposition, the complete conformity of the Gniezno melody (15th century) with the Cistercian melody (13th and 14th century) indicates that the Gniezno scribe faithfully copied the Cistercian variant.