Between the end of the eighth century and the first three decades of the ninth century, in a northern Frankish scriptorium – probably Saint-Denis – different gatherings were copied and/or collected in a single manuscript, now codex lat. 528 in Paris, Bibliothèque Nationale de France. The manuscript is actually composed by 181 leaves shared in twenty-two gatherings containing more than fifty texts and extended annotations on the top and bottom margins of the folios. At first sight, the miscellaneous structure of the manuscript seems to be ungovernable, but the codex can be divided in three main sections. The first section (gatherings 1-7) contains commentaries on the Pentateuch: Alcuinus, Quaestiones in Genesim, Exodum, Leviticum, Deuter., (P.L., 100); Eucherius, Instructiones. De Genesi (CSEL 31); Eugyppius, De singulis quaestionibus Augustini, (ed. Knoell, 1885). The second section (gatherings 8-17) collects a group of works all concerning grammatical and rhetoric matters, many by Bede De Schematibus Scripturae (P.L., 90), De arte metrica (P.L., 90), De divisione temporum (P.L. 90), Quaestiones in libros Regum (P.L., 91); a large series of lyrics by Paul the Deacon, a Greek grammar and some letters. Finally, the third section (gatherings 18-22) contains hagiographic texts circulating in monastic milieus: Vita Audoini (M.G.H., SS. RR. Merov., V), Breviarium Apostolorum cum Martyrologio hierolomit. (AA.SS. nov. II, I, II), Fulgentius, De Fide (P.L., 55), Augustinus Hipp., Excerpta. Before discussing the contents of the manuscript more precisely , it is important to give some information about dating and scriptorium. The manuscript, with the exception of three gatherings (11, 15 and 18) was probably copied in the scriptorium of Saint-Denis. Some specific calligraphic uses, which can be observed in the ms. BnF lat. 17371 (certainly a Saint-Denis product), are also identifiable in BnF lat. 528. Consequently, comparing the two mss., the copying of BnF 528 can be dated to the period in which Caroline minuscule writing was already in use but still in strict connection with the pre-Carolingian period. A more precise dating can be offered by some texts contained in the manuscript. Firm evidence is found in the lyrics by Paul the Deacon, which were composed in Aachen between the years 782-786. The Cyclus Decemnovennali (gathering 11) can safely be taken into consideration as terminus ante quem,. The gathering containing the text was bound later to the volume and reports the date 826. Therefore, the manuscript was for the grater part copied during Fardulfus’s abbacy (793-806) or immediately later. Considering the manuscript as a whole, the specific interest for rhetoric and grammatical topics is stricking. In particular, the presence of a letter of Charlemagne’s in which the Emperor exhorts clergy to literary studies is noteworthy (M.G.H., Epp. IV, Ep. Kar. Aevi, II, XXII, p. 532) and also the inclusion of the famous epistole of Jerome to Paula in which he explains the meanings of the Hebrew letters (Jerome, Ep. XXX). Other texts, for instance De arte metrica or De figuris in scriptura sacra, are devoted to rhetorical speculations, and also the Quaestiones in Genesim are copied here for their grammatical and rhetorical use rather than for edification purposes. It is namely a matter of fact that the first gatherings - containing biblical commentaries - show an anthological choice in which the passages are more connected to language and grammar than to religious edification. Finally, the same consideration can be made for Paul the Deacon’s lyrics, which were copied in that period in connection with grammatical works. If the texts are then analysed together with the marginal annotations, the apparently chaotic arrangement of the codex reveals its coherent structure. The annotations all over the codex are effectively, in a large amount of cases, strictly connected to the texts and unveil the presumable educational use of the manuscript. Four annotators' hands can be recognized. Some of them were particularly interested in rhetorical and grammatical aspects and this use probably was the original one for which the codex was conceived. The annotations concerning rhetorical subjects are inserted in the largest amount of cases as marginal glosses, and as single words. In a minor group of occurrences, the rhetorical glosses are more extended. the glosses concerning astronomical computing and alphabetical series are very different. In the first case, the glosses usually start on the verso side of the leaf and continue on the recto of the following one; it is therefore evident that the glossator wrote them when the manuscript was already bound. Anyway, it must be noticed that the copyist of astronomical glosses inserted the large part of them in coincidence with texts concerning astronomic computing, as in the 11th gathering. As far as the alphabetical series are concerned, a particular interest in alphabetical series and languages different from Latin must be pointed out. Glossators inserted different alphabets. A particular attention to Greek language and litterae is revealed by texts (grammar tì estìn doctus type) and by repeated Greek alphabets which were copied in the manuscript. Probably, this is due to the particular interest the scriptorium of Saint-Denis had for Greek, but it must be also remembered that the glossators filled margins of the codex with other alphabetical series: the Hebrew series, which is copied three times (ff. 78v-79r; 82v-83r) and other ones, for instance the Caldaic, Egyptian alphabets and the so-called Aethicus Ister alphabets (83v-86r). Moreover, two different glossators inserted Gothic letters and personal names in leaf 71v. In order to give a sense to our Paris miscellaneous codex, this proliferation of alphabetical series needs a particular attention. The presence of exotic alphabets was spreading in those years in the north-eastern part of France, and this codex is not an exception. Nevertheless, the specific grammatical and rhetorical characterization of the manuscript allows us to infer that the alphabets were copied there in order to complete the information imparted in the principal texts that were previously transcribed. The presence of marginal glosses, the alphabetical series, the average of margins covered by the computing of lunar phases, can reveal the compiler(s)’s intentions: to create a manuscript that could be used for educational purposes, in particular in the study of rhetoric, Greek, and astronomical computing. Probably, the compiler of the codex wanted to produce a manuscript that could presumably be used in a Carolingian schola monachorum, at an advanced stage of education, and therefore he condensed gatherings of different provenance or newly copied for the occasion in a single codex . The result was a new instrument to be used for learning activities. The third part of the manuscript, containing the vita Audoini and the Martyrologium, gives us supplementary information about the destination of the codex. Audoin was a saint particularly worshipped in Saint-Denis territories while the Martyrologium reveals that the manuscript, in the middle of ninth century, was probably used in the scriptorium of the abbey of Massay (department Cher, central France), which was re-founded some years earlier by the Visigoth Benedict of Aniane, to whom a specific interest in languages other from Latin can also be ascribed. Benedict of Aniane was also embarked on a monastic reform, which can easily explain why a educational manuscript created in a famous cultural centre like Saint-Denis was then used in a re-organizing monastery such as Massay.