Vigiliae Christianae 5 (1951) pp.193-203.




        It is now the second time that Mr. A.-P. van Schilfgaarde, the energetic archivist of the State of the province of Gueldre, has surprised me by sending me fragments of a Carolingian MS. Each time they were endpapers or chemises which had been employed to protect the bundles or the registers, pertaining to municipal or personal archives, in the keeping of the central depot. 1 Thanks to the comprehension of this benevolent colleague it is possible now for me to communicate a quite sensational find. I am all the more happy that, since this publication can appear in Vigiliae Christianae, everyone will understand that the communication of this fragment could not be better placed, since it concerns a vestige of a codex of Tertullian which actually dates from the beginning of the 9th century. It is a leaf cut into two of his treatise De spectaculis, one of the rarest works of this author.

        The baronial family of Pallandt, hereditary owner of the files of the house of Keppel, had the kindness to surrender to us on loan these two pieces of parchment; they will be placed at the disposal of paleographers and philologists in the cabinet of the manuscripts


1.   The first time it was a chemise for the accounts of the table of the poor (Provisorie) of Doesburg of 1558-' 70. Mr. A.-J. van of Ven, formerly archivist-assistant of the province of Gueldre, had found it while making an inspection and Prof J. - F Niermeyer of Amsterdam drew my attention to it. The fragment has been deposited on loan at the library of Leyden (B P. L 2507). It is part of a leaf of an Old Testament of the beginning of 10th century (P) of a very beautiful penmanship showing the influence of the school of Tours (here and there is to be found a semi-uncial of Tours). Undoubtedly this fully-realised writing must be assigned to a very significant milieu. Perhaps the perfect rhythm and the round forms prefigure the magnificent style of the Italian mss. of the eleventh and twelfth centuries.


of the Library of Leyden. The library is further very pleased especially to have obtained the right to renew this privilege each year before the first day of August. This small treasure will be there next to the famous large-folio manuscript of Tertullien, copied and illuminated for the library of Ferdinand I, king de Naples (1458-' 94). 2 This beautiful ms, acquired in 1659 for the sum of 125 fls., is certainly one of the jewels of the art of the Renaissance, but is however of only a relative interest for philology. It is different of our fragment: a remainder of the treatise De Spectaculis, coming from a ms of the 9th century, is a source of the first importance. The text of this treatise is based directly only upon one carolingian ms, the codex Agobardinus of Paris, 3 an ms offered by the Agobard bishop of Lyons (816-840) to the church of St Etienne of that city. Martin Mesnart made use of this ms around 1545, year of the editio princeps insofar as that relates to De Spectaculis, but he seems to have also used another ms, unfortunately lost since then. According to the methods of his time, he only accounted for the use of his sources in such way that it is very difficult to relate to any ms now.4 The text of Agobardinus, because it still exists, leads us to suspect the existence of another ms, but for lack of unquestionable evidence it has never been known to what degree the variants are really based on the text of another ms. One wondered whether it were necessary to sometimes allot those to the sagacity of the editor himself.

        The modern editors of De spectaculis have always generally preferred the text of Mesnart to that of the Paris codex.  Scholars will therefore be interested in  the fragment that we present to them here, because its text on several occasions approaches that of the editio princeps. It will be seen that the first edition is actually based on another old version of the text which deserves our full attention. I leave the critical study of the text of our fragment to others, more qualified, and I shall restrict myself to some remarks on its external aspect.


2. The traces of his blazon in the frontispice of the ms are still recognizable.
3. Paris, Bibl. Nat. ms. lat. 1622.
4. Cf. C. Schenkl  in his preface to the edition of  Cl. M. Victor (C.S.E.L. XVI), p. 339, 342-345. J.H.W.



        Here is the description: A single leaf cut into two by a bookbinder, but fortunately in such way that the text was not damaged. These two halves were employed formerly as endpapers of a book small-in-folio in which had been gathered copies of acts concerning vicariates of the manor of Keppel. It can still be seen how the needle of the bookbinder bored the parchment, leaving a small mitre. Originally the whole seems to have been wrapped with an border of very thick parchment, like the other registers of the 16th century of this small manorial chancellery. We will return to this later. First we will discuss our fragments themselves.

        The parchment is very thick and the concentrated ink is of perfect quality. The original dimension of the leaf seems to me to have remained almost intact (336 x c. 255). The justification (257 x 180 / 168) and the ruling (27 long lines) are in dry point, but the stylus did not reach our leaf directly: at least two sheets were ruled together. The vertical lines are doubled in accordance with the use which reserves a column for the large capital letters at the beginning of the chapters. One sees with which negligence this was done: at the top and bottom of the page the distance between the verticals (interior) shows a difference of 12 mm. Though the external margin is very large, 5 it is astonishing that there is no sign of the usual line of punctures for the use of the ruler. 6

        The script shows all the characteristics of the beginning of the 9th century. As the editor of V Chr. has been kind enough to allow us to reproduce the complete fragments as well as a sample in the original size, I estimate superfluous to give a detailed description of it. There is no point in underlining its characteristics


5.  Everything suggests that the source is a large scriptorium of a cathedral: large margins, imposing writing, higher quality of parchment (this one was very well prepared: it is almost impossible to distinguish between the flesh-side and hair-side). Cf. W. M. Lindsay, The (early) Mayence scriptorium - Paleographia Latina IV (1925), p. 15.
6. However I did find some of which I do not understand the purpose and I am not sure that all are original. Whatever it is, they show the angular or fissured form, which is significant in view of the provenance. [Cf. L. W. Jones, Pricking manuscripts : the instruments and their significance - Speculum XXI (1946), p. 389 sq.]


as the paleographer will gather them all from our facsimiles.7 Everyone will also be able to check me now, when I claim that this writing presents a great affinity with that of the scribes of the cathedral of Cologne at the time of the archbishop Hildebald (785-819). 8
      I will restrict myself to discussing some characteristics.

       Form of the letters: We can see that the a is written as a double c - the uncial form9, the capital N at the beginning of words, the ligatures re10, st, and NT (once) and the archaism ec (v°, 1.1). 11 With the long i of Item (v°, 1.26), the elongated i after l in seculj (v°, 1.21) and the e shown elongated after the last  r  of persecutores (v°, 1.26), I think that this underlines the essentials.
       Abbreviation: Apart from the nomina sacra, only the abbreviation for m is used.
       Punctuation: A very small point or a small feature for the short pause. A larger point and generally two small points and comma for the long pause. 12

       I will cite finally some characteristics of orthography.

    e = ae, oe reciprocally

    e = i (conflectemur, r°, 1.8; eregere, r°, 1.20; congemescentes, 
          v°, 1.25)

    e = o (tempera, , 1.5)

    a = i (dominiaci, v°, 1.26) 13

    t = d (aliut, aput, aliquit)

    u = b (liuacunculo, r°, 1.3).


7. One can observe the separation of the words which is very defective and in particular the bastes in the shape of bludgeons, those of h and d bent towards the left. Cf. L. W. Jones, (Script of Cologne from Hildebald to Hermann (Cambridge, Mass. i 932), pls. 32 et 38.
8.  Jones, pls. 22-66.The author distinguished more than 80 copyists at that time, which he called "middle Hildebaldian" (c. 795-c. 816).
9. Jones, pls. 221, 32 (upper part).
10. Jones, pls. 221, 32 (upper part).
11. A. Chroust, Monum. Palaeogr. II, 7, 1 (= Jones, pl. 221.), description.
12. Jones, pls. 44, 53, 57.
13. An insular characteristic. Cf. L. Traube, Vorles. u. Ahb. II (1911), p. 62.


       The text of the fragment

       Here are the variants: in the first column is the text of the Vienna edition, in the second that of the codex of Paris, finally in the last what our fragment gives us.14 When the text of our fragment is in conformity with the edition of Mesnart, the variants are printed in italics. Since this last is our starting point, I passed over all the variants of the Agobardinus which do not have any relationship with the text of our fragment. One will find this one in the edition of Vienna (C.S.E.L. 20), p. 26, cap. 27, 1. 22 gratissimis - p. 29, cap. 30, 1. 1 contra. The value of our text can be seen even at the beginning, since it supplies a gap in the Paris ms: in the Agobardinus lines 1-10 are missing.


V A Fr.
p. 26, 1.  22  illic B  - - - -   illa
23  perinde habe ac  - - - -  proinde habe ac si
24 lucunculo  - - - - liuacunculo
tanti  - - - - tangi
p. 27, 1. 2 tempera B  - - - - tempera
5 conflictamur B  - - - - conflectemur
8 delicatus B dilicatus delicatus
10 [quidam] quidem quidem
quieti et tranquillitati
dederunt in ea B
quieti in ea quieti et tranquillitati
 dederunt in ea
12  scaenam A scaenam scaenas
13 harenam  arenam harenas
14 debemus A debemus debebimus
est B et  est
15 apostoli AB apostoli apostolo
16 hic haec  hec; haec B
uotum ABmarg uotum ueitum; vitium B
17 putas A putas si putas
exigere [exige]re eregere; erigere B
18 ut tot et tales uoluptates ut putes tales a deo ut tot et tales uoluptates
a deo contributas . . . contributas . . . quam a domino con dei
quam dei patris B dei patris patris
20 quam ueritatis reuelatio quam quam ueritatis releuatio
quam B quam
25 quod calcas B quot calcas quod calcas


14.   V = Ed. of Vienna     A = cod. Agob.
        B = Ed. of 1445        Fr. = our fragment.


V A Fr.
26 deo AB deo d[e]o; deo B
27 haec uoluptates A haec uoluptates haec (hae B) uoluptates
28 gratuita AB gratuita grauita
circenses ludos A circenses ludos ludos circenses
p. 28, 1. 1 spatia [peracta] spatia - - - - ; spatia B
2 metas consummationis et has consumationes metas consummationis
exspecta B specta expecta
3 dei AB dei deus
4 martyrium A martyrium martyrii
5 est satis versuum est est satis sententiarum est sans versuum est
satis sententiarum B satis sententiarum
7 pugilatus pugilatus que praesto   pugillatus
8 parua et multa par sunt multa parua sed multa
impudicitiam B impudiciam inpudicitiam
9 caesam B caesas  caesam
a misericordia B misericordia a misericordia
10 tales sunt B tales talesunt
15 quae gloria B gloria quae gloria
21 tot spectans A tot spectans spectans tot et tales
22 ioue ipso A ioue ipso ipso ioue
ipsis suis testibus B sistestibus ipsis suis testibus
23 congemescentes A congemescentes congemescentes; con-
gemiscentes B
24 saeuierunt insultantes saeuferunt insultantibus seauierunt (saeuierunt
B) insultantibus

        Is it possible that we have here a leaf of Mesnart's ms? It must be said that we do not: thus, if he had found ueitum (p. 27, 1. 16), he must have corrected it to uotum. Now since he has printed vitium, clearly he must have found this word and that he did not correct it. He printed in margin votum (ms of Paris). One sees here how scrupulous Mesnart was. In addition this ueitum of our fragment is very interesting because now the passage of votum > ueitum > vitium appears clearer: perhaps the fault is understood most easily if one imagines a model written in uncial or semi-uncial writing with a malformed o which could have been misinterpreted as ci? 15

        There is possibly another indication of a very old model, written in large letters. Our fragment presents us with an omission:  . . .tales uoluptates a domino con- [tributas tibi satis


15. As for the second passage: a carolingian copyist not knowing ueitum would have changed it into vitium.



non habeas neque recognoscas? quid enim iucundius quam] dei patris et domini reconciliatio  . . .  (r°, 1. 21).  This omission - which it is not easy to justify, I acknowledge - would be understood most easily as a jump of one or two lines of the model.  However, the omitted words count 63 letters.  That would be almost too much for a single line of an ms written in carolingian minuscule, but it would be precisely enough for two lines of an ms written in uncial or semi-uncial:

................................ a domino (AB: deo) con
dei patris et domini .......................................

We have seen that it is unlikely that our fragment was a leaf of Mesnart's codex: though independent of this last, its text is very close to it. However this proves the existence in the 9th century of a family of mss. of Tertullien with a text differing considerably from that of Agobardinus. 

        It is known that five years after the edition of Mesnart, Sigmond Ghelen made a new edition of Tertullian. In his foreword the scholarly corrector of the printing works of Froben in Basle mentions "complures codices veteres e Gallicanis Germanisque bibliothecis conquisitos ". Undoubtedly, it is necessary to take care not to take literally such assertions of an editor of the 16th century, but Gelenius had the reputation of a very modest scholar and an enemy of posing and boasting.16 Certainly his very hazardous conjectures sometimes have been the subject of severe criticism, but does one have really the right to describe it as "vanus "?

        Undoubtedly with a find like ours it is necessary to be on ones guard: the ms, of which our fragment is a remainder could have been one of the witnesses of the edition of Basle of 1550.  One may compare the following passage:

        V (p. 27, 1. 23) : omnia illic seu fortia seu honesta seu sonora seu canora seu subtilia perinde habe ac stillicidia mellis de lucunculo uenenato nec tanti gulam facias uoluptatis quanti periculum per suauitatem.  Saginentur eiusmodi dulcibus conuiuae sui: et loca et


16. B. Röse, Sigmund Ghelen in J. S. Ersch and J. G. Gruber, Allgem. Encyklopädie der Wissensch. u. Künste LXVI (Leipzig 1857), pp. 269-275.


tempera et inuitator ipsorum est.  nostrae coenae, nostrae nuptiae nondum sunt, non possummus cura illis discumbere . . .
        B: Proinde . . . ac si .. . deliuacunculo uenenatu nec tangi . . . imitator . . . nondum sum, non possum . . .
        Gel. (1562)17proinde . . . ac si . . . de liuacunculo uenenato: nec tanti . . . et invitator . . . nondum sunt.  non possum  . . .
        Fr.:  proinde . . . ac si . . . deliuacunculo uenenato nec tangi . . . et inuitator . . . nondum sunt non possumus  . . .

It can be seen, that in this place the edition of Ghelen follows almost everywhere the redaction of our fragment. 

        Can our endpapers be regarded as having formed part of one of the mss. consulted by Ghelen to the profit of his edition? Is it physically possible that this leaf can come from it? The answer is subordinate to another question: is it possible to roughly date the binding of which these two pieces of parchment formed part? If our register had been bound long enough before the date of the edition of Gelenius, the relation between the latter and our fragment would be excluded. At the first glance, one would be tempted to conclude in this direction; everything points to an earlier date: the principal part of the register was written between 1537 and ' 30. From the very start Baron Jean de Pallandt, first lord of Keppel of this family (1530-' 65), himself also made use of this almost unused book, to note there jobs, financial arrangements with his subjects etc.  But after 1545 one finds no further indication of use. Despite all this the ms only seems to have been bound much later.  Here are the results that an examination of the files of Keppel gave me.  Mr. Van Schilfgaarde, a very great expert on his treasures and all that relates to the history of the manorial families of the old duchy of Gueldre, had the kindness to assist me with it. It is thus under the tutelage of this Mentor that I had the privilege to examine closely the other registers of the time of Jean de Pallandt. Even though the book was thus not used after 1545, it is very probable that it was bound only towards 1563. For on this date, two years before his death, the lord of Keppel seems to have definitively regulated the businesses of his little chancellery: he had made


17.  I have not been able to consult the edition of 1550.



a new register of holdings, at the same time as he had other mss bound. 

        The technique of these bindings is similar from every point of view and in all these mss. one finds endpapers which are simply the remains of old mss. of various epochs. The binder, a simple travelling craftsman or perhaps residing at Doesburg, a neighbouring small city18, seems me to have bought sometime a job lot of parchment made up of pieces from old mss. (the majority of German origin) of the 9th to the 14th century. Without any doubt our leaf comes from such a retail activity. I thus would by no means exclude that we have found here a leaf detached from one of the mss. consulted by Ghelen for his edition. 

        If, on the contrary, the codex of our fragment was unknown to the editors of the 16th century, we have the proof here that formerly there existed several mss. of Tertullian currently unknown. 

        Indeed, it is by no means impossible that the codex of our fragment was known in the 16th century. We know already that at that time an ms of Tertullian in the Library of the Cathedral of Cologne was known. It was Jones who first drew  my attention to this fact. In his  foreword (p. 4) he affirmed that formerly it contained an ms of Tertullian. It could be said that this is only an assertion which remains to be  proven, for all that is known is that towards 1579 Jacques Pamelius - the scholar of Bruges who prepared the Paris edition of 1579 - borrowed
an ms of the Apologeticus from the library of the Friars Minor of Cologne.19 However, a source from the cathedral is also possible, as even in the 13th century the Friars Minor regularly borrowed mss. from the cathedral.20 

        But there is more still: in consequence of the benevolence of Mr. E. Kuphal, director of the municipal archives of Cologne, I was able 


18.  See above. The fragment of Doesburg was used as a chemise for the accounts of 1558-' 70 of this city. However, precisely at the same epoch the remains are to be found at  Doesburg of another very old fragment of an ms. The technique of this bookbinder however, differs from that of the bookbinder of the lord of Keppel.
19.  P. Lehmann, Franciscus Modius als Handschriftenforscher - Qu.u. Unters. zur lat. Philol. des M.A. III, 1 (1908), p. 99 -100.
20.  Ph. Jaffé and G. Wattenbach, Eccl. metropol. Coloniensis codd. mss. (Berol. 1874), p. VIII.


to consult a well-known, but extremely rare, study by A. Decker, Die Hildebold' sche Manuskriptensammlung des Kölner Domes.21 Though out of date now and superceded by the study of P.Lehmann, Erzbishof Hildebald und die Dombibliotek von Köln,22 this former provided us at the same time with a supplemented and annotated edition of a catalogue of the books of the cathedral, written into 833. This catalogue was known still in 1634 by the historian Egide Gelenius,23 who unfortunately published only one appendix containing only the list of the  manuscripts on loan, followed by the names of the borrowers. In 1752, the date of the catalogue by the Jesuit Jos. Hartzheim,24 this  invaluable document had disappeared. Fortunately, in his introduction, the author does not fail to inform us that he consulted letters of Hittorpius, Calenius and Pamelius, of which he quotes passages, showing that this ancient catalogue was consulted by these scholars. Decker deserved well of the history of the carolingian era by finding a very old codex of the Liber Fulgentii Ferrandi diaconi ad Reginum comitem  (8th - 9th s.) which had been mislaid at the vicariate-general of the archbishop, and which contained at the same time this famous catalogue.  The value of this last, which mentions no fewer than 175 mss. of which 46 still exist, cannot be over-estimated and one regrets only that Jones, while writing his monograph, did not republish this catalogue with new annotations.25  Indeed, the edition of 1895 is very rare and in the future this list will have to serve as the basis  for all research on the anicent mss. of Cologne. 

        It is after having read what was published by Paul Lehmann about the Library of Cologne and of the scholars of the 16th century who  had visited it, that I was highly interested in this catalogue of 833, for my paleographic studies of the fragment in question


21. Festschrift der 43. Versamamlung deutscher Philologen u. Schulmänner . . (Bonn 1895), p. 213-251.
22. Zentralbl. f. Bibliotekswesen XXV (1908), p. 153-153.
23. One must take care not to confuse this Cologne historian with Sigmond Ghelen, the corrector of Froben.
24. Catalogus historicus criticus codd. mss. Bibl. Eccl. metrop. Coloniensis.
25. Once again this invaluable codex seems lost. Mr. B Bischoff of Munich, one of the greatest experts on German mss., had the kindness to inform me that since 1934 he has been attempting to find it, but unsuccessfully. Did Jones who discusses it too superficially, really ever see this ms with his own eyes?


led me towards scriptorium of Cologne at the beginning of the 9th century. I thus very much hoped to find there mentioned an ms of Tertullian. This was indeed the case. Decker transmitted to us a very clear and indubitable description, though, just like his predecessor of the 9th century, he did not recognize its identity:
De resurrectione mortuorum. lib. I
& de fide, libri II
De praescriptionibus hereticorum lib. I
de jejuniis adversum phisicos lib. I
de monogamia lib. I
de pudicitia lib. I
in une corpore sed auctorem ignoramus 26

It is a fact that De Spectaculis is not cited and the list appears to have been compiled with a lot of care, since other little treatises like De jejuniis and De monogamia are well and truly there, but it is always possible it was forgotten or that this work had lost its title in the ms.  However this mention of an ms of Tertullian at the beginning of the 9th century is very interesting, because in the first place it reveals that the works of Tertullian were copied at Cologne in the era of Hildebald.  Thus it corroborates my hypothesis: the Keppel fragment comes from an ms of Cologne cathedral, perhaps from the same ms cited in the catalogue of 833. Moreover, this detailed description has much value for the study of the history of the texts of Tertullian, because it informs us of the order of the treatises in this ms which is at the moment the  oldest known to us.  No doubt, Pamelius, by reading the titles of the list of 833, recognised an ms of Tertullien, but would hel have found it in
the library? I fear not, but it remains always possible that Sigmond Ghelen, his predecessor by about thirty years, consulted it. I thus hold that it is by no means excluded that we have found here a leaf detached from one of the mss. consulted by Ghelen to the profit of his edition.


26. Decker, nr. 98 (pp. 227 et 249).

Leiden, Universiteits-Bibliotheek.

G.I. LIEFTINCK, Un fragment de De Spectaculis de Tertullien provenant d'un manuscrit du neuvieme siecle, Vigiliae Christianae 5 (1951) pp 193-203.  © Brill Academic Publishers, 1951.  Reproduced by permission of the publisher.  All rights reserved.

This translation by Roger Pearse.

This page has been online since 27th October 2000.

Return to the Tertullian Project Original language About these pages