A New Witness to the Fayyūmic Version of First Corinthians (P.MorganLib. 265). Part I: Notes on Codicology, Language, Provenance and Date

Abstract This is the first part of a two-part article focused on a fragmentary parchment codex, whose three extant leaves, designated in Leo Depuydt's catalogue as P.MorganLib. 265, are housed at the Morgan Library & Museum in New York. These fragments bear witness to 1 Cor 2.12–3.18; 7.16–30; 15.3–30 in the ‘classical’ variety of Fayyūmic Coptic (dialect F5). Most of these verses have been hitherto unattested in Fayyūmic and thus allow us to attain better insight into the history and text of the Coptic Bible. In the first part of this article, I discuss the codicology of P.MorganLib. 265, its linguistic features, provenance and date.

anticlockwisefacing the rectos of frs. (c) and (d). Both paste-downs were then attached to the papyrus boards.
In 1912, the Morgan Phantoou codices were sent to the Vatican Library for restoration. There, the restorers separated the original binding of MICH.BM from its text block and its paste-downs from its covers. The manuscript, along with its covers and paste-downs, was photographed for the facsimile edition, published in 1922. 9 Fragments of P.MorganLib. 265 were separated from each other, inlaid in modern parchment and bound together with the text block of MICH.BM. The codex then received modern binding, while the original binding received a new housing and was assigned a new inventory numberviz. M.585A. 10 In the rebound codex M.585, the Fayyūmic fragments occupy fols. i-ii and 50-1. This position reflects the one they had when they were pasted together; thus, frs. In sum, P.MorganLib. 265 seems to have had a veritable afterlife. After it became worn out, this codex made its way to the binder's shop, where three of its leaves were reused as paste-downs of another codex, MICH.BM. This latter codex, along with many other manuscripts from the Monastery of the Archangel Michael, was eventually hidden in a stone cistern, where, centuries later, it was rediscovered by Egyptian farmers. Then, at the Vatican, the restorers removed the paste-downs from MICH.BM, separated the fragments of P.MorganLib. 265 from each other and rebound them with the rest of MICH.BM. It is in this state that the fragments are now available to researchers visiting the Morgan Library & Museum in New York.

Description of the Manuscript
Having discussed the fate of P.MorganLib. 265, I now turn to a description of the manuscript and its contents. All three extant leaves of P.MorganLib. 265 comprise portions of Paul's First Letter to the Corinthians in Fayyūmic. It seems certain that, when the codex was intact, it contained the complete text of this letter. It is also plausible that the codex contained more than one text. Although it is impossible to ascertain what these lost texts might have been, the most likely possibility is that the codex comprised Pauline letters. 12 The contents of each surviving page of P.MorganLib. 265  None of the extant margins seems to preserve any page numbers, which most probably indicates that the three surviving leaves were not paginated. 13 The text of the manuscript is divided into two columns. The longest columnsviz. col. ii of fol. (d + e) r and col. i of fol. (d + e) vcomprise forty-one lines; the shortest columnviz. col. i of fol. (c + a) vthirtyseven lines.
The text is written in unimodular uncials. Vertical strokes are upright and thick and horizontal strokes thin, with curved strokes alternating between thicker and thinner segments (e.g., the letter ⲟ is written in one stroke, whose left and right edges are thick and upper and lower edges thin). The letters ⲝ, ⲣ, ⲯ, ϣ, ϥ and ϯ extend below the baseline; the letter ⲫ breaks the bilinearism both at the top and at the bottom.
Two letters have shapes attested only in Fayyūmic manuscriptsviz. ϣ and ϭ. The letter ϣ seems to be written in three strokes, with a thick vertical stroke connected to a finned stroke at the top and a horizontal stroke at the bottom; the horizontal line stretches to the left then takes a sharp turn and extends all the way back to the right. The letter ϭ is usually written in two strokes, with a circle resembling the letter ⲟ and a diagonal stroke stretching from the lower edge of the circle to the right. However, in at least three instances, 14 the scribe did not give the letter ϭ this uniquely Fayyūmic shape; instead, in these cases, the oblique stroke joins the left edge of the circle and stretches to the right, reaching the bottom of the preceding line.
There are two other letters in P.MorganLib. 265 written in two different shapesviz. ⲁ and ⲙ. The letter ⲁ is usually made up of two thick oblique strokes connected at the top and joined by a thin oblique middle stroke at the bottom. However, occasionally the letter is shaped differently, with a thick vertical stroke joined to a curved stroke on the left and a thin horizontal stroke on the right; the horizontal stroke connects ⲁ with the following letter. As for the letter ⲙ, it seems to be always written in three strokes, usually with two thick vertical strokes connected by a curved middle stroke. Sometimes, however, ⲙ is written in three concave strokes, the right stroke extending along the baseline and connecting to the following letter.
When it appears near or at the end of the line, the letter ⲧ is sometimes enlarged, its vertical stroke being written above the adjacent letters. For the same reasoni.e. to accommodate more textwhen the combination ⲟⲩ occurs at the end of the line, the letter ⲟ is sometimes significantly reduced in size and written above ⲩ (see ⲛⲓⲉⲛⲧⲟⲗⲏⲟⲩ at 1 Cor 7.19 and ⲡⲟⲩ|ⲉⲓ at 1 Cor 7.20). Other letters written towards the end of the line can also be reduced in size; of those, ϥ is especially noteworthy (when it occurs at the end of the line, its vertical stroke is slightly elevated and its curved stroke very small, positioned above the bilinear space).
The new sections of text are marked with initials, which are often enlarged and written in ekthesis. The initials are always accompanied by paragraph marks, usually shaped as simple angular coronides (>). 15 Smaller textual units within paragraphs are separated from each other by means of middle dots. The middle dot also occurs on the last line of a paragraph, followed by a blank space. In one instance (at the end of verse 15.20), the middle dot is followed by a long horizontal stroke. The coronis, signalling the beginning of the following paragraph, is zeta-shaped, its lower branch growing into an ornament, with four spirals of identical size.
In at least six instances (ⲓⲙⲓ̇at 1 Cor 2.14, ϣⲱⲡⲓ̇at 1 Cor 7.12 and 15.21, ⲛⲁⲃⲓ̇at 1 Cor 15.17, ⲛⲓⲃⲓ̇at 1 Cor 15.19 and 15.28), the scribe placed a dot above the letter ⲓ. It is noteworthy that, in all these instances, the dotted ⲓ is word-final and preceded by a consonant; this phenomenon is also attested in P.Mich.inv. 158 (9) + P.MorganLib. 264. The function of the dot is obscure.

History of Research
The first to study P.MorganLib. 265 was Henri Hyvernat, the scholar responsible for the production of the facsimile edition of the Morgan codices, comprising fifty-six volumes. In the 'index tabularum' of vol. XXXVIII (comprising M.585), Hyvernat mistakenly wrote that each of the two paste-downs was made up of two fragments, somewhat imprecisely identifying their contents (e.g. saying that fr. (b) comprised 1 Cor 15.4-31this mistake was later repeated by Depuydt). After the publication of the facsimile, Hyvernat must have occasionally revisited the manuscript, as his copy of vol. XXXVIII housed at the Catholic University of America in Washington, DC contains numerous marks witnessing his attempts to trace the faded letters, identify the biblical verses and restore some of the lacunae. An entry on M.585 in his unpublished Catalogue of the Coptic Manuscripts in the Pierpont Morgan Library 16 includes a note on the 'fly leaves' (i.e. paste-downs), 17 which, however, merely repeats what Hyvernat had written in the 'index tabularum'.
Although only thirteen copies of Hyvernat's facsimile were made and although the photographs of P.MorganLib. 265 printed therein were taken before the paste-downs were dismantled (meaning that portions of text were inaccessible to the reader), this publication still allowed Coptologists to make use of the manuscript. Thus, the text of P.MorganLib. 265 is occasionally referenced in W. E. Crum's Coptic Dictionary. Unfortunately, Crum's references to P.MorganLib. 265 are often indistinguishable from his references to codex C, both of which he designates with the siglum 'F' (i.e. 'Fayyūmic & related dialects')even though the language of the latter manuscript is better defined as 'unevenly Fayyūmicised Crypto-Sahidic' 18and which happens to also contain 1 Cor 7.16-30 and 15.3-30. 19 That is, Crum never specifies when a passage derives from codex C and only occasionally marks his references to P.MorganLib. 265 with 'Mor 38'. Thus, in his entry for 'circumcise', the stative ⲥⲉⲃⲏⲟⲩⲧ and the construction ⲁⲓ ⲛⲁⲧⲥⲏⲃⲃⲓ ('be uncircumcised'), which both come from codex C, are marked '1 Cor 7.18 F'. 20 On the other hand, in his entry for 'have pity, mercy', the source of ϫⲓ ⲙⲡⲛⲉⲉⲓ ('get pity, take alms')marked '1 Cor 15.19 F'is P.MorganLib. 265. 21 The importance of P.MorganLib. 265 for Fayyūmic lexicography is further discussed in the following section (see pp. 14-15 below).
Later, P.MorganLib. 265 was briefly mentioned by Theodore C. Petersen in his unpublished monograph on Coptic bookbindings (written between 1929 and 1950) and by Paul E. Kahle in an appendix on the relationship between the Fayyūmic (fa) and the Bohairic (bo) version of the New Testament. 22 Finally, Depuydt wrote an entry on this manuscript in his catalogue, which, although quite short, needs several corrections. For example, 16 I consulted the hand-written copy housed at the Semitics/ICOR Library of the Catholic University of America. 17 Hyvernat also erroneously used the term 'fly leaf' instead of 'paste-down' in his Check List of Coptic Manuscripts in the Pierpont Morgan Library (New York: 'privately printed', 1919) 14 ('2 fly-leaves'). 18 W.-P. Funk, 'Formen und Funktionen des Interlokutiven Nominalsatzes in den koptischen Dialekten', Langues Orientales Anciennes. Philologie et Linguistique 3 (1991) 1-75, at 17 n. 38; cf. idem, 'Gedanken zu zwei faijumischen Fragmenten', Christianisme d'Égypte. Hommages à René-Georges Coquin (Cahiers de la Bibliothèque Copte 9; Paris: Peeters, 1995) 93-100, at 97 n. 13. 19 The designations 'codex A', 'codex B' and 'codex C' refer to three fragmentary parchment codices comprising the Book of Isaiah (codex A), the Tetraevangelion (codex B) and the Pauline Epistles (codex C), which were once deposited at the library of the White Monastery; of those, the first two belong to the F5 corpus. The extant fragments of the three codices are divided between the Institut français d'archéologie orientale in Cairo, the National Library of France in Paris, the Vatican Apostolic Library in Rome, the Austrian National Library in Vienna and the University of Oslo (042/3, unpublished). while he rightly observes that there are five fragments (rather than four), he fails to identify the contents of fr. (e) or notice that frs. (d) and (e) join together. The description of P.MorganLib. 265 provided in this article thus supersedes that by Depuydt.

Language
The consistent lambdacism of P.MorganLib. 265i.e. the substitution of /r/ with /l/ (e.g. ⲗⲱⲙⲓ, 'human', for S ⲣⲱⲙⲉ, B ⲣⲱⲙⲓ)immediately informs the reader that the manuscript is written in a Fayyūmic dialect. The two major varieties of Fayyūmic Coptic are F4 (Early Fayyūmic) and F5 (Classical Fayyūmic). 23 Although the dividing line between the two dialects is not as clear as we might wish it to be (i.e. F5 manuscripts often exhibit features characteristic of F4), the distinction still seems to be worth retaining. Wolf-Peter Funk has formulated several criteria for distinguishing between F4 and F5, 24 according to which P.MorganLib. 265 should be classified as a witness to the latter dialect: There are, however, also words spelled without vowel gemination whose form thus coincides with that in F4viz. ⲕⲉ-(prepersonal form of ⲕⲱ, 'to put'; cf. ⲕⲉⲥ, 'in order that'), ⲙⲏⲟⲩⲓ ('to think'), ⲟⲩⲉⲃ ('holy'), ⲟⲩⲉⲓ ('one') and ⲟⲩⲁⲉⲧ-('self'). It is worth noting, however, that, although other F5 manuscripts often geminate the stressed vowel in these words (thus ⲕⲉⲉ-, ⲙⲏⲏⲟⲩⲓ, ⲟⲩⲉⲉⲃ, ⲟⲩⲉⲉⲓ, and ⲟⲩⲁⲉⲉⲧ-), the F5 corpus is by no means uniform in this respect. As Funk notes, manuscripts that regularly employ forms like ⲛⲏand ⲛⲉⲙⲏyet exhibit certain inconsistency with regard to vowel gemination should still be classified as witnesses to F5. 26 -In F4, the indefinite pronoun used in negative contexts is ⲗⲁⲡϯ, while F5 either uses ⲗⲁⲡⲥ or oscillates between ⲗⲁⲡⲥ and ⲗⲁⲡϯ. In accordance with the F5 standard, P.MorganLib. 265 consistently employs ⲗⲁⲡⲥ. -For the relative perfect, F4 usually employs ⲉⲧⲁ-, ⲉⲧⲁ-, while, in the F5 corpus, we predominantly encounter ⲛⲧⲁ-, ⲛⲧⲁ-. P.MorganLib. 265 always employs ⲛⲧⲁ-, ⲛⲧⲁ-, except for the hitherto unattested form ⲉⲧⲁⲁ-, which appears to be reserved for instances where the relative perfect occurs in the 'glose' of a cleft sentence (see the discussion below). -In F4, the definite plural article is usually ⲛⲓ-; in F5, ⲛⲉ-. In P.MorganLib. 265, the form ⲛⲉoccurs only twice (1 Cor 15.3, 4), both times in the expression ⲕⲁⲧⲁ ⲛⲉⲅⲣⲁⲫⲏ ('according to the scriptures'). In all other instances, the manuscript has ⲛⲓ-. Just like the occasional absence of vowel gemination, this preference for ⲛⲓshould probably be interpreted in the light of the fact that the transition from the 'early' norm of Fayyūmic (F4) to the 'classical' one (F5) is likely to have been a long and gradual process.
In sum, despite some notable departures from what can reasonably be conceived of as the dialectal norm, the dialect of P.MorganLib. 265 is undoubtedly F5.
In what follows, I briefly discuss the orthographic and morphological features of the manuscript, its treatment of Greek loanwords and, finally, its significance for Coptic lexicography.
Overall, the orthography of P.MorganLib. 265 is quite consistent and compliant with what seems to be the standard Fayyūmic scribal practicefor example, the grapheme <i/y> is always represented by ⲓ. There are, however, several orthographic features that are worthy of note.
-The definite singular article ⲡ-, ⲧusually has the long form (viz. ⲡⲉ-, ⲧⲉ-) if the determined noun begins with two consonantshence ⲡⲉϩⲙⲁⲧ ('the grace') and ⲡⲉⲕⲗⲱⲙ ('the fire'). The nomen sacrum ⲡⲉⲡⲛ︦ ⲁ ('the Spirit') is similarly written with the long article. Yet the long form is never employed with the nomen sacrum ⲡⲭⲣ︦ ⲥ ('Christ'). This latter spelling, which in the F5 corpus competes with ⲡⲉⲭⲣ︦ ⲥ, is most certainly a reflection of the Bohairic practice, just like the abbreviations ⲫ︦ ϯ︦ for B ⲫⲛⲟⲩϯ, 'God', and ⲡϭ︦ ⲥ︦ for B *ⲡϭⲱⲓⲥ, 'the Lord' (F5 does not have consonant aspiration and thus the 'proper' F5 forms of these words would be ⲡⲛⲟⲩϯ and ⲡϫⲁⲓⲥ). 27 -The scribe often doubles the letter ⲛ when it is morphemic (serving as the linkage marker, direct object marker, negative scope definer etc.) and precedes vocalic onsetse.g. ⲛⲧϩⲏ ⲛⲛⲟⲩⲉⲓ ('like someone') in verse 7.25. Sometimes, morphemic ⲛ is not doublede.g. ⲛⲧϩⲏ ⲛⲟⲩϩⲟⲩϩⲏ ('like an untimely birth') in verse 15.8. Non-morphemic ⲛ, on the other hand, is never doublede.g. ϩⲓⲧⲉⲛ ⲟⲩⲗⲱⲙⲓ ('through a man') at 1 Cor 15.21 and ϩⲛ ⲁⲇⲁⲙ ('in Adam') at 1 Cor 15.22. -The scribe omits the non-morphemic initial syllabic nasal in the first-person singular form of the conjunctive, writing ⲧⲁrather than ⲛⲧⲁ-(1 Cor 7.25). In the F5 corpus, ⲛ is similarly omitted in codex MICH.AX (comprising the Investiture of Michael (cc 0488)), where ⲧⲁoccurs eleven times, 28 ⲛⲧⲁonly once. 29  conjunctive, writing ⲛⲧⲉⲛ ̣rather than ⲧⲉⲛ-(1 Cor 2.12). The latter spelling occurs in three manuscripts in the F5 corpusviz. MICH.AX, codex B and the still unpublished papyrus codex comprising the Martyrdom of Paphnutius (cc 0294) and the Life of Onnuphrius (cc 0254). 32 -In the expression ⲙⲙⲛ ϣϭⲁⲙ ⲙⲙⲁ-('to be unable'), the initial syllabic nasal of the preposition ⲙⲙⲁis omitted, resulting in the spellings ⲙⲙⲛ ϣϭⲁⲙⲙⲁϥ ('he is unable') at 1 Cor 2.14 and ⲙⲙⲛ ϣϭ̣ ⲁⲙⲙⲁⲧⲉⲛ ('you [pl.] are unable') at 1 Cor 3.2. However, in the one instance where this expression is used with the affirmative existentialviz. ⲟⲩⲁⲛ ϣϭⲁⲙ̣ ⲙⲙⲁⲕ ('you [sg.] are able') in verse 7.21the nasal is not omitted. Similarly, the negative existential is usually ⲙⲙⲛ (1 Cor 2.14; 3.2; 15.13) but is at least once spelled with the initial syllabic nasal omittedviz. ⲙⲛ (1 Cor 3.11). 33 The initial syllabic nasal is also omitted in the only occurrence of the phonologically reduced form of the possessive predicate ⲙⲙⲛⲧⲉ-, ⲙⲙⲛⲧⲏ-(derived from the negative existential ⲙⲙⲛ and the preposition ⲛⲧⲉ-, ⲛⲧⲏ-)viz. ⲙⲛϯ ('I do not have'), in verse 7.25. -The modal auxiliary of the conditional can be spelled either with or without the nasal sonorantviz. ϣⲁⲛ and ϣⲁ. The former spellingwhich is the norm in F5 34occurs in all five instances when the modal auxiliary is preceded by the pronominal subject (1 Cor 7.28 (bis); 15.24 (bis), 27). In the two instances where the subject is nominalviz.,1 Cor 3.4 and 15.28the morpheme is spelled ϣⲁ. 35 This pattern (ϣⲁ before nouns, ϣⲁⲛ before pronouns) seems to be the same in P.Mich.inv. 158 (9) + P.MorganLib. 264. -The form ⲥⲏⲃⲃⲓ ('to circumcise'; S ⲥⲃ̄ⲃⲉ, B ⲥⲉⲃⲓ) and its cognatesviz. the construct form ⲥⲏⲃⲃⲏⲧand the stative form ⲥⲏⲃ ̣ⲃⲏⲟⲩⲧdeserve a special comment. This form also occurs in another F5 manuscript extracted from the covers of a Phantoou codex viz. P.MorganLib. 267 (unpublished). However, yet another F5 manuscriptviz. P.Mich.inv. 158 (9) + P.MorganLib. 264always reads ⲥⲩⲃⲃⲓ, ⲥⲩⲃⲃⲏ-. These latter forms attest to the Fayyūmic scribal practice also reflected in ⲧⲩⲃⲧ ('fish') and ϣⲩⲃϣⲓ ('shield')namely, to represent the reduced vowel (i.e. schwa /ə/) with ⲩ in a closed stressed syllable, with /v/ 36 as its coda. 37 P.MorganLib. 265 and 267 thus bear witness to an alternative practice, in which the schwa followed by /v/ in a closed stressed syllable is represented by ⲏ. It is tempting to surmise that this difference in spellingviz. ⲥⲏⲃⲃⲓ (P.MorganLib. 265 47. 35 In the latter instance, since ϣⲁ is written at the end of the line, the missing ⲛ could have been represented with the supralinear stroke, which could have faded; however, since, a few lines above, the supralinear stroke after ϣⲁ (verse 15.27) is perfectly visible, I am inclined to think that, in verse 15.28, the manuscript reads ϣⲁ, not ϣⲁ⳯. 36 For a critique of the traditional phonological interpretation of ⲃ as /b/, see C. H. Reintges, Coptic Egyptian (Sahidic Dialect): A Learner's Grammar (Study Books of African Languages 15; Cologne: Köppe, 2018 2 ) 20-1. 37 The use of ⲩ in these words probably indicates that, due to its proximity to a labial consonant, the schwa comes to resemble /u/. 38 Admittedly, it seems impossible to ascertain where these three manuscripts were produced. However, the little that we do know about their provenance gives no reason to believe that any two of them could have been copied in the same scriptorium. Thus, unlike P.MorganLib. 265 and 267, P.Mich.inv. 158 (9) + P.MorganLib. 264 is not part of the Phantoou discovery. While codex MICH.BH (from whose covers P.MorganLib. 267 was extracted) -With regard to the schwa followed by a tautosyllabic ϩ, the scribes of the F5 manuscripts had three strategies: sometimes, the schwa is represented by ⲁ (e.g. ⲧⲱⲃⲁϩ, 'to entreat'), sometimes it has no graphic representation (ⲧⲱⲃϩ); in late manuscripts, the schwa is occasionally represented with a supralinear stroke (ⲧⲱⲃϩ̄). The scribe of P.MorganLib. 265 seems to be consistent in opting for the second approach, writing ⲗⲱⲕϩ ('to burn') at 1 Cor 3.15 and ⲱⲛϩ ('to live'/'life') at 1 Cor 15.19, 22. -At 1 Cor 15.9, the preposition whose usual spelling is ⲟⲩⲧⲉis spelled ⲟⲩⲇⲉ-. In the F5 corpus, this spelling likewise occurs in P.Mich.inv. 158 (9) + P.MorganLib. 264 (at Rom 2.15, although at Eph 3.8 this manuscript also has ⲟⲩⲧⲉ-). The form ⲟⲩⲇⲉis also found in codex C (at 1 Cor 15.9), in Sahidic manuscripts (including those produced in the Fayyūm) 39 and some of the first-millennium B5 manuscripts. 40 This spelling could perhaps be explained by the scribe drawing an analogy with the Greek loan words ⲟⲩⲧⲉ (οὔτε) and ⲟⲩⲇⲉ (οὐδέ), which were often confused with each other due to the lack of contrast between voiced /d/ and voiceless /t/ in Coptic. Of the morphological features of P.MorganLib. 265, the most remarkable is the relative perfect form ⲉⲧⲁⲁ-, which occurs twice in the extant text (1 Cor 2.12, 16). In the history of Coptic linguistics, the reasons behind the occasional doubling of the letter ⲁ in the F5 perfect conjugation base have been a matter of debate. While Ludwig Stern considered the pattern ⲁⲁ-ϥ-ⲥⲱⲧⲉⲙ to be a mere variant of the first perfect (ⲁ-ϥ-ⲥⲱⲧⲉⲙ), H. J. Polotsky argued that ⲁⲁ-ϥ-ⲥⲱⲧⲉⲙ is, in fact, a special form of the second perfect, an interpretation later endorsed by Walter C. Till and Georg Steindorff. 45 However, as Funk noted in 1995, this interpretation does not take into account those instances in the F5 corpus where the doubled ⲁ occurs in relative perfect. 46 This Gordian knot was severed in the latest instalment of Funk's F5 concordance, where two homonymous morphemes are distinguished, one occurring in the second perfect, 47 the other in the relative perfect. 48 P.MorganLib. 265 is thus an important new witness of the second morpheme. Its importance lies in the fact that, in this manuscript, the relative form occurs for the first time with the spelling ⲉⲧⲁⲁ-(rather than ⲛⲧⲁⲁ-). Moreover, P.MorganLib. 265 is the only known F5 manuscript where the relative perfect with the doubled ⲁ appears in the 'glose' of a cleft sentence. In fact, it seems that, in all other contexts, P.MorganLib. 265 employs the form ⲛⲧⲁ-, ⲛⲧⲁ-, which may indicate that, in this manuscript, the form ⲉⲧⲁⲁwas reserved exclusively for cleft sentences.
The following two morphological features of P.MorganLib. 265 are also worth mentioning: − The manuscript uses two sets of definite articlesviz. ⲡ(ⲉ)-, ⲧ(ⲉ)-, ⲛⲉand ⲡⲓ-, ϯ-, ⲛⲓ-. As I have already noted, in the plural the definite article is predominantly ⲛⲓ-, while ⲛⲉoccurs only twice, and only in the expression ⲕⲁⲧⲁ ⲛⲉⲅⲣⲁⲫⲏ ('according to the scriptures'). In the singular, the situation is reversed: the text normally employs ⲡ(ⲉ)and ⲧ(ⲉ)-, while ⲡⲓoccurs only five times (1 Cor 2.14; 7.16, 22 (bis); 15.5) and ϯonly once (1 Cor 7.28). As far as I can tell, the use of one of the two sets of articles is not conditioned syntactically, nor is there any discernible difference in meaning between them. 49 − The manuscript also attests to two competing forms for causative infinitiveviz. ⲧⲣⲉ-, ⲧⲣⲉand ⲧⲉ-, ⲧⲉ-. The pattern ⲧⲣⲉ-, ⲧⲣⲉoccurs six times: twice in the purpose clause construction ⲉ-ⲧⲣⲉ-(1 Cor 7.26; 15.9), twice in the negative causative imperative ⲙⲡⲉⲗⲧⲣⲉ-(1 Cor 3.18; 7.18) and twice conjugated (1 Cor 3.6, 7). The pattern ⲧⲉ-, ⲧⲉoccurs only once, in the negative causative imperative ⲙⲡⲉⲗⲧⲉ-(1 Cor 7.21). This datum seems to support Funk's observation that, in F5, the pattern ⲧⲉ-, ⲧⲉwas out-competed by ⲧⲣⲉ-, ⲧⲣⲉand that, as a rule, its use was restricted to the definite purpose clause ⲉ-ⲡ-ⲧⲉand the negative causative imperative ⲙⲡⲉⲗⲧⲉ-. 50 The treatment of Greek loanwords in P.MorganLib. 265 reflects the standard practices of the Fayyūmic scriptoria: that is, our scribe is consistent in representing all Greek epsilon-iota sequences with ⲓ (e.g. ⲓⲧⲉ for εἴτε) and all alpha-iota sequences with ⲉ (e.g. ⲙⲉⲗⲉⲥⲑⲉ for μέλεσθαι). Similarly, Greek verbs always occur in the infinitive and are always derived by means of the light verb ⲉⲗ-(e.g. ⲉⲗⲇⲟⲕⲓⲙⲁⲍⲓⲛ). The following features of the manuscript are worthy of note: -Although the scribe of P.MorganLib. 265 often misspells Greek wordsreplacing ω with ⲟ (e.g. ⲅⲛⲟⲙⲏ for γνώμη), ο with ⲱ (e.g. ⲱⲧⲓ for ὅτι), τ with ⲇ (e.g. ⲁⲣⲭⲓⲇⲉⲕⲧⲟⲛ for ἀρχιτέκτων) etc.he never uses the monogram ϯ for the tau-iota sequences in Greek loanwords (hence ⲉⲧⲓ, ⲡⲓⲥⲧⲓⲥ, ⲡⲛ︦ ⲁⲧⲓⲕⲟⲥ, ⲱⲧⲓ). -The verb ὑποτάσσεσθαι ('to be subject'), which occurs in P.MorganLib. 265 six times, is remarkable with respect to its spelling. As Funk observes, in Bohairic and Fayyūmic, Greek verbs ending with -τάσσειν and -τάσσεσθαι would normally be spelled -ⲧⲁⲍⲓⲛ or -ⲧⲁⲍⲉⲥⲑⲉ. 51 This observation is confirmed by P.MorganLib. 265, where, five out of six times, the verb is spelled ϩⲏⲡⲟⲧⲁⲍⲉⲥⲑⲉ, yet once it is spelled ϩⲏⲡⲟⲧⲁⲥⲉⲥⲑⲉ, which indicates that the scribe had difficulty distinguishing voiced /z/ from voiceless /s/. It is worth noting, however, that at least one manuscript that can be considered for inclusion into the F5 corpusviz. P.MorganLib. 268goes against this tendency, reading ϩⲩⲡⲟⲧⲁⲥⲥⲉⲥⲑⲉ (at 1 Cor 14.32, 34). However, given that the orthography of P.MorganLib. 268 is often at odds with the F5 standard for example, in this manuscript, the full (stressed) form of the possessive predicate is ⲟⲩⲁⲛⲧ-(rather than ⲟⲩⲁⲛⲧⲏ-)the form ϩⲩⲡⲟⲧⲁⲥⲥⲉⲥⲑⲉ should probably be explained by the scribe's insufficient acquaintance with the orthographic norm. 52 -Another medio-passive infinitive that occurs in P.MorganLib. 265 is μέλεσθαι ('to be an object of care'). In the Greek text of the New Testament, the verb μέλειν occurs only in the active voice. It is, therefore, rather remarkable that, at 1 Cor 7.21, P.MorganLib. 265 employs the medio-passive form; a similar 'predilection for the longer form' was recently detected by Funk in the 'Curzon Catena' (P.Lond.Copt. 2.249), one of the key witnesses to the medieval Bohairic (dialect B5) of the first millennium. 53 Interestingly, in the Bohairic New Testament (bo), the active voice of the Greek original is also occasionally rendered with ⲙⲉⲗⲉⲥⲑⲉsee e.g. Acts 18.17but not at 1 Cor 7.21, where all the witnesses use the active form. -The word ⲯⲩⲭⲏⲕⲟⲥ, which occurs in P.MorganLib. 265 at 1 Cor 2.14, also occurs in one other manuscript of the F5 corpusviz. P.Vindob. K 3280 + K 3921 + K 9311where it is similarly spelled with ⲏ. It is tempting to suggest that ⲯⲩⲭⲏⲕⲟⲥ was the standard F5 spelling of ψυχικός (rather than a mere case of iotacism). This suggestion receives support from the fact that, in the F5 corpus, some of the words formed with the prefix ἀρχare similarly spelled with ⲏ 54viz. ⲁⲣⲭⲏⲁⲅⲅⲉⲗⲟⲥ (in National Library of France, Copte 163 1 fol. 1 (comprising an otherwise unknown martyrdom) and MICH.AX), 55 ⲁⲣⲭⲏⲉⲡⲓⲥⲕⲟⲡⲟⲥ (in P.Carlsberg 300, comprising De incredulitate (cc 0013) and De providentia (cc 0012), both attributed to the fictitious Agathonicus of Tarsus), 56 ⲁⲣⲭⲏⲡⲗⲁⲥⲙⲁ (in MICH.AX) 57 and ⲁⲣⲭⲏⲥⲧⲣⲁⲇⲓⲕⲟⲥ (in MICH.AX). 58 -The F5 corpus is divided on the pluralisation of the noun ἐντολή. Two manuscriptsviz. codex B (at John 14.21) 59 and P.Mich.inv. 158 (9) + P.MorganLib. 264 (at Rom 13.9) employ the singular form (ⲉⲛⲧⲟⲗⲏ), specifying the number of the noun only by means of the article. Two other manuscriptsviz. P.Carlsberg 300 and Pushkin State Museum of Fine Arts, I.1.б.637also add the plural marker -ⲟⲩ. 60 P.MorganLib. 265, which uses the form ⲉⲛⲧⲟⲗⲏⲟⲩ (at 1 Cor 7.19), belongs to the second group. 61 -The F5 corpus is also divided on the spelling of the conjunction ἵνα. Some manuscripts always read ϩⲓⲛⲁ (e.g. codex B and National Library of France, Copte 163 1 fol. 1), others always ϩⲓⲛⲁⲛ (e.g. P.MorganLib. 259 and 261). 62 P.MorganLib. 265 seems to belong in the former group. It is worth noting, however, that, in both instances where ϩⲓⲛⲁ occurs in this manuscript (viz. 1 Cor 2.12; 15.28), it is followed by the initial ⲛ of the conjunctive. It is thus possible that the form ϩⲓⲛⲁ in these two cases is due to 'phonetic' haplography 63 and that, elsewhere in the manuscript, the