‘A battlefield of books’: this is how Solomon Schechter described the mass of tangled and damaged manuscript debris when he entered the Genizah chamber of the Ben Ezra synagogue in Fustat (Old Cairo) in 1896 (fig. 2.1). This windowless room, together with similar caches in other synagogues and in the cemetery Basatin in Cairo, yielded over 350,000 fragments of manuscripts, kept today in more than seventy collections worldwide. Most of the fragments date from the Fatimid and Ayyubid periods: more than ninety-five percent come from books while the rest are fragments of legal documents, letters, and other pragmatic writings. They were preserved thanks to the long-standing Jewish tradition of disposing of old writings with particular respect, founded on the belief that Hebrew texts containing the name of God are sacred: rather than being destroyed or thrown away, worn out books and documents—both holy and trivial—were instead placed in dedicated space, a Genizah, to decay naturally without human intervention. This massive necropolis of discarded writings offers us unprecedented knowledge of Jewish life in medieval Egypt in general and of Jewish book history in particular. Thousands of fragments are witnesses to the centrality of Hebrew books in liturgy, in professional activities, and in private life, as well as offering a mine of information about how these books were made and read: their materials, forms, and formats.
Particularly interesting, in this respect, are recent discoveries at the Genizah that attest to the unexpected importance of vertical scrolls, or rotuli, in the book culture of Oriental Jews. Indeed, as a result of a systematic search in various Cairo Genizah collections—a collaboration between Gideon Bohak of the University of Tel Aviv and myself—nearly 500 fragments of books written in rotulus form have been found. Judging from their palaeographical features, they were written in Egypt between the tenth and thirteenth centuries. It is likely that most of the rotuli were produced in Fustat and discarded in the local Genizah. However, as we shall see, some of the rotuli were produced in smaller Egyptian towns. It is unclear why the writings from the provinces were discarded in the Fustat Genizah, but their conservation is an important source for the study of reading and book-making practices outside of the Egyptian capital. In this chapter, I will focus on one fragment of a literary rotulus—now Cambridge University Library Taylor-Schechter (henceforth TS) C 1.67 (figs 2.2–2.3)—which was discovered in the Cairo Genizah and brought to Cambridge by Solomon Schechter, Rabbi and reader in Rabbinics at Cambridge University. A detailed palaeographical analysis traced its origin to the small Delta town of Damira. After a brief presentation of the corpus of rotuli in Hebrew script from the Cairo Genizah, I will turn to focus in some detail on the physical description, palaeography, and dating of this rotulus.
The Geniza Rotuli
The rotulus form has been used in traditions of Jewish book making since antiquity. Although no ancient rotuli have been preserved, the Mishnah and the Talmud both mention takhrikh (תכריך), a ‘roll’ or ‘wrapper’. This term usually refers to the practice of attaching together vertically three or more legal documents to facilitate their archiving. However, there is some evidence that the vertical scroll form was also used to copy literary or liturgical texts; the Talmud Yerushalmi mentions a takhrikh berakhot (תכריך ברכות), ‘a roll of blessings’. Yet despite these references, Hebrew books in rotuli form have been largely disregarded by book historians and codicologists, who instead tend to focus their attention on more traditional horizontal Bible scrolls and codices. Few would have suspected that preserved rotuli fragments would number so many: until now most known rotuli have systematically been dated before the year 1000, conceived of simply as a transitional hinge between the scroll and codex. Yet the recent discovery of hundreds of rotuli in the Cairo Genizah shows not only that this ‘third form’ of the Hebrew book was much more common in Oriental Jewish communities than previously believed, but also that use of the format extended well into the thirteenth century and even later.
The survival of this ancient book form in the community at Fustat is less surprising when we consider that this form was in fact relatively common in medieval Egyptian society; also among Christians, Muslims, and Samaritans. Indeed, Greek and Samaritan prayers on rotulus are preserved, as are rotuli with excerpts from the Koran in Arabic, probably used for magical and apotropaic purposes. Particular to the Jewish rotuli from the Cairo Genizah, however, is their specific function and proficient, professional readership. An analysis of their materiality quickly shows that the overwhelming majority of these Jewish books on rotuli were low-cost copies, often user-produced and destined for a personal devotion, individual study, or as a professional vademecum. It is likely that personal notebooks, megillot setarim (מגילות סתרים), literally ‘personal or concealed rolls’, mentioned in the Geonic literature were such rotuli, although the term megillah usually refers to horizontal scrolls. Various factors suggest the economic concerns of the readers: including the use of lesser-quality, often reused writing materials; a lack of decorative features; a high density of the text formed from the small size of its characters; minimal left-hand margins; and reduced interlinear spaces.
Their extremely varied contents, too, shed light on the potentially broad appeal of the fragments. More than 55% of their identified texts contain liturgy. Only a few include standard prayers (for instance TS H 10.310, TS 20.57, TS 6H 8. 3, and TS 13 H 1. 4), while the majority contain liturgical poems or piyyutim (פיוטים) (TS H 8. 43). Very few fragments contain passages directly drawn from the Bible: a few fragments of the Psalms exist (TS AS 43. 23), part of a prayer anthology rather than a Biblical manuscript as such, as well as various passages of haftarah (Bodl. MS Heb. b 18. 23; JTS, ENA 3974. 3). A few known rotuli with passages from the Pentateuch (e.g. TS AS 7.2), seem to have either been used as a copying exercise or were a copy of a short portion of the text, rather than that of the entire Biblical book. Secular poetry is attested, for instance, in fragments of work by Judah ha-Levi (TS 13 J 24. 13). A small corpus concerns science and materia medica (TS 20. 150, TS NS 90. 47), while the Genizah also preserves the earliest attested manuscript of the Sefer Yeṣirah (Book of Creation) (TS 32. 5, TS K 21. 56, TS K 12. 813), the version used by the tenth-century exegete Sa‘adyah Gaon for his much-renowned commentary. Magic and astrology also feature in the rotuli fragments (Bodl. MS Heb. a 3. 31), as do passages of Hekhalot literature, a mystical body of writings detailing chariot-bound ascents to heaven (e.g. Bodl. MS Heb. a 3. 25a).
Another relatively large group of the fragments contains scholars’ books. They include Biblical translations and commentaries, for instance Sa‘adyah Gaon’s Arabic paraphrasis of the Bible, the Tafsīr (TS Ar 1a. 140); lists of Biblical variants and textual difficulties known as the Masora (Bodl. MS Heb. a 3. 30); and lexicographical works, such as Sa‘adyah’s list of seventy words attested only once in the entire Bible (hapax legomena) (TS Ar. 53. 9). Likewise, the rotulus seems to have been a favoured book form for students of Jewish legal tradition. Several of the Genizah rotuli contain tractates of the Babylonian Talmud (Bodl. MS Heb. e 52 (R)), the Mishna (TS F 2(1) 167), legal compendia such as Halakhot Gedolot (TS F 5. 151, TS NS 329. 1020), Sheiltot of Rabbi Aha de Shabha, and commentaries or glossaries used to facilitate the study of the Babylonian Talmud (Bodl. MS Heb. b 12. 33, TS G 2. 20).
Lastly, the Cairo rotuli contain several copies of the so-called ‘late midrashim’, which seem to be anthologies of earlier Rabbinic texts and quotations, such as Pirqa de-Rabbenu ha-Qadosh (MS Bodl. Heb. a 2 fol. 24; TS H 7. 21; TS K 21. 85; TS K 21. 94; LG Talm. II. 95) or a composition similar to the work published under the title The Pearl of Rav Meir (Margenita de Rav Meir) or The Pearl of the House of Rav (Margenita de-bei Rav) (TS C 1. 67). In addition to this rotulus, I was able to identify yet another fragment in the Cairo Genizah containing passages of the Pearl, a fragment on a horizontal scroll dated c.1000. The Pearl is a short, homily-like ethical essay whose main subject is the punishment for bad actions and the failure to follow God’s commandments in this world and in the world to come. The Genizah fragments are the earliest witnesses to the heritage of this text.
The Damira Fragment
In the following sections, I will focus on the physical characteristics and palaeographical dating of the Genizah rotulus TS C 1. 67. A detailed study of its text is beyond the scope of this materially-oriented essay, however it is important to stress that, like most medieval midrashic and homiletic compilations, The Pearl is an example of a non-authoritative and ‘open’ text whose versions vary a great deal from one manuscript to another. Individual manuscripts present important differences of wording and intertwine passages found in other identifiable works within their unique texts. The rotulus TS C 1. 67, for example, contains a passage (lines 1–4) that appears somewhat closer in content to a collection of midrashic homilies printed under the title of Pesiqta ḥadta (the passage concerning Yom Kippurim) than it does to the corresponding passage of the printed version of the Pearl. However, despite such differences, the core of the rotulus text and its order in the context of TS C 1.67 does correspond to the later printed edition of the Pearl.
TS C 1. 67 is written with black-fading-to-grey carbon ink, on inferior quality, grey, thick Egyptian paper with clearly visible rag fibres. The fragment contains forty-two lines and the text is written in one continuous block on the recto. Paper is the writing material of some sixty percent of the rotuli from the Cairo Genizah studied thus far, and the preserved fragment measures 32.5 x 11.5 cm. It was composed of at least three sheets of paper (kollemata) glued together vertically before the text was written, as evidenced by the written line overlapping two of these sheets in line 3 of the fragment. Only a small part of the upper sheet is preserved but the full width of the rotulus is generally complete, with the end of the line preserved in most cases. As for its length, the rotulus is damaged, torn off at the beginning and at the end. When compared with the printed edition, the preserved portion corresponds to about one third of the text. It is therefore likely that the complete rotulus was originally about one metre long.
It seems that this was an optimal size of a paper rotulus. We know only three Genizah rotuli on paper whose length has been fully preserved: a section of the Sheiltot of Rabbi Aha de Shabha reconstructed from seven fragments joined together, measuring 120 cm; a copy of liturgical poems on the verso of a reused letter in Arabic by a Muslim official, CUL Add. 3336, measuring 150.5 cm; and a Babylonian Talmud, Tractate Beiṣah, Bodl. MS Heb. e 52 (R), measuring 158.5 cm (six paper sheets). Indeed, the Beiṣah roll appears to represent a maximum length for a paper rotulus. This version was in fact copied across three rotuli which together formed the same codicological unit, effectively presenting the text in three ‘tomes’. Fragments of the other two rolls of this same codicological unit have been found too, suggesting that the division of the tractate into three portions, copied on three separate rotuli, was judged the largest practicable solution for accommodating this long text. We must remember that unlike rotuli made of parchment, some of which reached up to three metres, paper was far less resilient: too long a rotulus would be easy to damage and tear. Still, a relatively short rotulus like TS C1. 67—about one metre long and also very narrow—could be easily rolled and unrolled when held in the hand. This was a perfect format for a small, inexpensive, light, and portable book intended for personal reading.
The verso of the rotulus in its present state is blank except for a note in Arabic and Hebrew, containing a magical formula for protection of the book against worms:
كبج كج ملعلون عبرانى من كتاب אל חסיד
kabij kaj mal‘ūn ‘ibrānī min kitāb al-ḥasid 
[litt. Kabij kaj is cursed, in Hebrew, from the Book of the Pious]
There are, however, some traces of Hebrew letters at the top of the fragment corresponding to the end of the upper sheet, just above the place where the two parts were glued together. They may suggest that the upper part of the verso, now lost, originally contained the end of the text. As is usually the case with the Genizah rotuli, the scribe did not make rulings to guide the lines of their text; the rotuli are, after all, informal books. This is why the fragments’ lines are not always regularly spaced, their writing often sinking in the middle of the rotulus and lifting again towards the end. However, here the scribe has taken particular care to justify the fragment’s long block of writing. The right-hand margin is narrow, only around 1 cm, but straight nonetheless. The margin on the left, though, is not even: most frequently, rotuli scribes tend to run lines of text right up to the paper’s edge, but sometimes when the writing is either too short or too long for the line, care is taken to reduce calligraphic interference. Here, to avoid large blank spaces, the scribe has created fillers either by extending the upper horizontal strokes of the lines’ final letters (for instance lines 19, 22, 24, 28, 31) by elongating the letter’s basis (see the nun at the end of line 32), or by using a horizontal line, sometimes a mark of an abbreviation, as a space filler (for instance in lines 18, 21, 25, 41, 42). To accommodate the ends of lines that are too long, the scribe chose to write the last word in slightly smaller characters, above the line, with a slant up to the right (lines 17, 26, 37, 38, 39). Besides this textual consideration at a linear lever, there are no other graphic indications of separate sections of paragraphs of the midrash in the fragment, nor any punctuation marks for that matter. The text runs as a regular block of uninterrupted short lines. However, the fact that the lines are short and relatively generously spaced, and given their low density and careful, clear handwriting, the scroll is not uncomfortable to read.
The script is Oriental of the Egyptian sub-type and belongs to the non-square register, similar to that used in legal documents and other less formal books. It also contains several cursive features, with characters measuring around 3 x 3 mm. The pertinent features of the script are consistent with Egyptian manuscripts of the first half of the thirteenth century (see Appendix), and both this date and location can be confirmed and further specified by the identification of the scribe of our rotulus as the scribe of a legal document in Arabic in Hebrew script, TS NS J 2 (figs 2.4–2.5). A systematic handwriting analysis leaves no doubt that the scribe of this text is the same as our rotulus, TS C 1. 67. His name in the related legal document is slightly damaged but can be read as Moshe ben Mevorakh. The document, published in 1971 by Shlomo Dov Goitein, records donations to the community chest (heqdesh) by several members of a family in exchange for the honour of their youngest member, Ibrahim (Abraham), to be chosen to read in the synagogue, intoning the scroll of Esther during the celebration of the festival of Purim in front of the assembled congregation. The father of Ibrahim, Abū al-Fakhr ben Abū al-Faraj, also offers in return to relinquish a reimbursement claim for the cost of transport by beasts, perhaps horses, which the community had hired from him for the trip of a prominent visitor. This was a member of an aristocratic family of Mosul, the Nasi (or ‘Prince’ of Davidic descent) Joshiah, son of Jesse ben Solomon, who had travelled to Ashmūn and al-Maḥalla al-Kubra in Lower Egypt. Most importantly for our rotulus, this legal record of the donation and settlement contains a precise date and place of writing: ‘in the second third of the month of Adar of the year 1555’ of the Seleucid era, that is in February 1243, in a town of Small Damira (Damira ha-qetanah). The town of Damira, situated on the al-Maḥalla canal rather than on the Nile proper as stated in the document, is mentioned in a number of Genizah documents, and according to the twelfth-century traveler Benjamin of Tudela it had a large Jewish community of some 700 individuals. As pointed out by Goitein, like other provincial towns in Egypt, Damira was also home of scholars and teachers.
TS C 1. 67 is thus provided with a context of production. But, equally importantly, the precise dating of this rotulus proves that the roll format was still in use for small, portable copies of literary texts in the thirteenth century, both in Fustat and across various Jewish settlements in Egypt. The contents of the Cairo Genizah now rest deep in library vaults, a corpus of minute fragments scattered across multiple institutions worldwide. But cases like this show there are still many codicological puzzles held within them, able to shed light both on the small, personal world of Jewish Egypt and the ongoing presence of the continuous page.
The handwriting’s specific features include: particularly rounded bases of the letters lamed, final mem, kaph, nun, and pe; concave upper horizontal lines of beth, daleth, and final mem; the reduction of strokes and a rounded execution as one movement of the upper part and the right-hand downstroke of the letters he, ḥeth, kaph, final mem, pe, final pe, resh and tav; and relatively long ascender of the lamed and descenders of nun, final pe and qoph. There is some slight shading or difference between horizontal strokes written with the full width of the calamus’ nib, and thin vertical and oblique strokes. The shapes of the pairs of letters, which can be similar in some script-types or handwritings, are different in this manuscript: beth versus kaph, gimel versus nun, daleth versus resh, he versus ḥeth, final mem versus samekh. The only confusion concerns one of the allographs of the letter aleph and a ligature of nun followed by a vav, both adopting an N-shape.
line 25: שאינו
The most salient morphological features include:
Gimel traced with two straight lines that cross at the level of the baseline; with the right-hand downstroke straight and almost perpendicular to the baseline, going below the meeting point. The downstroke points sharply upwards and does not contain an additional ‘roof’. The left-hand stroke is long and parallel to the baseline.
Zayin is wedge-shaped and its head is placed to the right of the downstroke
Final nun, with a wavy descender and a head placed to its right.
Shin whose ‘middle’ short stroke is attached to the extremity of the left arm. This left-hand part is sometimes detached from the right-hand stroke, which forms the letter’s basis.
The features particularly emphasised in the handwriting of the scribe of our rotulus include a tendency to allography of aleph and lamed:
a/ b/ Two allographs of aleph: one, kappa-shaped, with a characteristically long upper stroke of the right-hand part of the letter; and an N-shaped aleph, whose right-hand stroke also goes frequently above the line of writing.
a/ b/ c/ d/ עליו Several allographs of lamed: ranging from two strokes superimposed vertically (a), through the forms with the body of the letter forming a rounded base (b, c); to a rounded form written as one continuous movement forming a closed oval (d, here, the following yod is written decoratively inside the loop of the lamed).
Judith Olszowy-Schlanger FBA (PhD Cambridge) is Professor of Hebrew Palaeography and Manuscript Studies at the École Pratiques des Hautes Études, PSL, Paris, President of the Oxford Centre for Hebrew and Jewish Studies and Fellow of Corpus Chrisiti College, Oxford. Her research interest include, Cairo Genizah Studies, medieval Hebrew palaeography, codicology and diplomatics as well as intellectual contacts between Jews and their non-Jewish neighbours.
 Judith Olszowy-Schlanger, ‘The Third Form of the Hebrew book: Rotuli from the Cairo Genizah’, Report of the Oxford Centre for Hebrew and Jewish Studies (2010–2011), pp. 90–91. More fragments of rotuli still await identification. It is not always easy to see immediately that a small fragment belonged to a rotulus as only two rotuli in the corpus are complete with all their sheets still attached (Bodl. MS Heb. e 52 (R): BT, one third of the tractate Beiṣah, and CUL Add. 3336, a literary text), and it is not always easy to identify a fragment of a book as belonging to a rotulus. Such a claim can be made when the fragment in question is composed of two or more pieces of writing material still attached together; when it contains traces of stitching; when the fragment displays proportions which are incompatible with a codex; or when it contains a blank verso or a verso covered with the writing by the same hand, but laid upside down in respect to the recto. However, in the case of very small fragments, their identification as a part of a rotulus is only possible when other parts of the same codicological unit have been previously identified as such.
 A large proportion of the Genizah rotuli fragments are kept in the Taylor-Schechter collection in the Cambridge University Library, but smaller corpora have been located in the Bodleian Genizah Collection, the British Library, the John Rylands Library in Manchester, the Jewish Theological Seminary in New York, the Kaufman Collection in Budapest, and the Alliance Israélite Universelle in Paris.
 See Mishna, Baba Meṣia 1:8; BT Baba Meṣia 20b: takhrikh shel shetarot.
 See Shabbat 79b.
 For probably the earliest mention of the codex among the Jews, see Arieh Leib Schlossberg (ed.), Sefer Halakhot Pesukot o Halakhot Re’u, attributed to Rav Yehudai Gaon (in Hebrew) (Versailles: Defus Serf, 1886), p. 11: ומגלה כתובה במצחף אין אדם יוצא בה ידי חובתו שכתב ונכתב בספר ומצחף אינו ספר, (concerning the ritual reading of the scroll of Esther): ‘If man reads from a codex (maṣḥaf), he does not fulfill his religious obligation, because it is written “from the scroll (sefer)”, and the codex is not a scroll’. See Nahum M. Sarna, The Pentateuch: Early Spanish Manuscripts (Codex Hilleli) from the Collection of the Jewish Theological Seminary, New York, 2 vols. (Jerusalem: Makor Publishing, 1974), Introduction, n. 20; Mordechai Glatzer, ‘The Aleppo Codex: Codicological and Palaeographical Aspects’ (in Hebrew), Sefunot 4 (1989): pp. 260–261; Irven M. Resnick, ‘The Codex in Early Jewish and Christian Communities’, Journal of Religious History 17:1 (1992): pp. 1–17.
 For instance, Richard C. Steiner, A Biblical Translation in the Making: The Evolution and Impact of Saadia Gaon’s Tafsir (Harvard: Harvard University Press, 2011), p. 46, 85, 93.
 BL Or 5036.1, a rotulus on parchment in Samaritan. See Abraham Tal, Samaritan Targum of the Pentateuch (in Hebrew), vol. 2 (Tel Aviv: Tel Aviv University, 1981), pp. 105–109; TS 16.321, a Greek rotulus in Greek minuscule script. On medieval Arabic rotuli, see Solange Ory, ‘Un Nouveau Type de Mushaf : Inventaire des Corans en Rouleaux de Provenance Damascaine, Conservés à Istanbul’, Revue des Études Islamiques 33 (1965): pp. 87–149.
 On megillot setarim in a responsum of Hai Gaon, see Robert Brody, The Geonim of Babylonia and the Shaping of Medieval Jewish Culture (New Haven and London: Yale University Press, 1998), p. 171.
 Judith Olszowy-Schlanger, ‘Cheap Books in Medieval Egypt: Rotuli from the Cairo Genizah’, Intellectual History of the Islamicate World 4 (2016): pp. 82–101.
 Dalia Wolfson, ‘Medieval Chapbooks: Early Collecting of Rabbi Judah ha-Levi’s Poetry: T-S 13J24.13 and T-S K25.138’, Fragment of the Month (February 2015), accessed 20 April 2019.
 For the most recent edition, see Peter A. Hayman, Sefer Yeṣira. Edition, Translation and Text-critical Commentary (Tübingen: Mohr Siebeck, 2004), pp. 59–195.
 See Gideon Bohak, ‘The Magical Rotuli from the Cairo Genizah’, in Gideon Bohak, Yuval Harari, and Shaul Shaked (eds), Continuity and Innovation in the Magical Tradition, Jerusalem Studies in Religion and Culture, (Leiden and Boston: Brill, 2011), pp. 321–340.
 See Alexander Tal, ‘Between Talmud and Abridgment: A Geniza Scroll of BT Betzah’ (in Hebrew), Ginzei Qedem 7 (2011): pp. 75–144.
 For a rotulus of Sheiltot reconstructed from seven fragments and copied by Ephraim ben Shemaryah, see Ronni Shweka, Marina Rustow, and Judith Olszowy-Schlanger, ‘The She’iltot, Recycling Manuscripts and Efrayim b. Shamarya’, Fragment of the Month (October 2011), accessed 20 April 2019.
 The Cambridge fragment has been published by B. M. Lewin, ‘Explanation of the Words of the Talmud from an Early Source’ (in Hebrew), Ginze Kedem (1934): pp. 167–177. See Judith Olszowy-Schlanger, ‘Glossary of Difficult Words in the Babylonian Talmud (Seder Mo‘ed) on a Rotulus’, in George. J. Brooke and Renata Smithuis (eds.), Jewish Education from Antiquity to the Middle Ages: Studies in Honour of Philip S. Alexander (Leiden and London: Brill, 2017), pp. 296–323.
 Judith Olszowy-Schlanger, ‘Un Rotulus du Midrash Pirqa de-Rabbenu ha-Qadosh de la Geniza du Caire’, Annuaire de l’EPHE, 2012-2013 (145e année) (Paris: EPHE, 2014): pp. 26–40; Anna Busa, ‘The Rotuli Corpus of the Medieval Midrash Pirqa de-Rabbenu ha-Qadosh’, Fragment of the Month (July 2017), accessed 20 April 2019.
 The work was edited from early printed editions (alongside two hundred other ‘minor’ midrashim and ethical works) in two volumes by Judah D. Eisenstein, Ozar Midrashim (New York: Bibliotheca Midraschica, 1915), p. 355, and by Adolph Jellinek, Bet ha-Midrasch, Sammlung kleiner Midraschim und vermischler Abhandlungen aus der ältern jüdischen Literatur, vol. 2 (Wien: Brüder Winter, 1877), pp. 120–122. The early printed editions were the Reshit Ḥokhmah of the Safed kabbalist Elias ben Moses de Vidas (Amsterdam edition, 1737; editio princeps, Venice, 1579). This midrash was known and parts of it included into a fourteenth-century Ashkenazi anthology, Sefer ha-Zikhronot compiled and copied by Elazar ben Asher ha-Levi (Oxford, Bodleian Library, MS Heb. D 11), edited by Eli Yassif, The Book of Memory, that is The Chronicles of Jerahme’el (in Hebrew) (Tel Aviv: Tel Aviv University Press, 2001), pp. 93–94. In addition, Moses Gaster, Chronicles of Jerahmeel or the Hebrew Bible Historiale (London: The Royal Asiatic society, 1899), p. lxiv, mentions a Geniza manuscript of the ‘Pearl’ that he had in his possession, n° 289. I have not been able so far to identify this Geniza fragment in either the British Library or John Rylands Geniza Collection in Manchester, both stemming from Gaster’s collection.
 TS 16. 282, see Judith Olszowy-Schlanger, ‘The Anatomy of Non-biblical Scrolls from the Cairo Geniza’, in Irina Wandrey (ed.), Jewish Manuscript Cultures: New Perspectives (Berlin: De Gruyter, 2017), pp. 49–88, n° IV.
 Eisenstein, Ozar Midrashim, p. 497.
 Oxford, Bodlian Library, MS Heb. e 52 (R) was the last in the unit, containing the last third of the text of the tractate and bearing the number three (the Hebrew letter gimel) in its upper margin.
‘Kabij kaj’ derives from a name of a plant, and was used in Arabic books as a formula of protection from worms. I thank Professor Ronny Vollandt for his help with the interpretation of the formula. See Adam Gacek, ‘The Use of ‘Kabīkaj’ in Arabic manuscripts’, Manuscripts of the Middle East 1 (1986): pp. 49–53.
 This script register or style is often referred to as semi-square or semi-cursive.
 Shlomo D. Goitein, ‘Side Lights on Jewish Education from the Cairo Genizah’, Gratz College Anniversary Volume 83 (1971), p. 105 ff. See also Moshe Gil, Documents of the Jewish Pious Foundations from the Cairo Geniza (Leiden and London: Brill, 1976), p. 26.
 For this family of nesi’im, see Moshe Gil, Jews in Islamic Countries in the Middle Ages (Leiden and London: Brill, 2004), pp. 441–442.
 See Norman Golb, ‘The Topography of the Jews of Medieval Egypt, VI: Places of Settlement of the Jews of Medieval Egypt’, Journal of Near Eastern Studies 33 (1974): p. 126.
 S. D. Goitein, A Mediterranean Society: The Jewish Communities of the World as Portrayed in the Documents of the Cairo Genizah, 6 vols (Los Angeles and London: University of California Press, 1967–1986), vol. 1, p. 54, 404.