CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

The Manuscripts of Xenophon's 
Symposium 

John Cirignano 

KCENT SCHOLARSHIP has produced a clearer picture of the 
relationships among manuscripts of certain opuscula of 
Xenophon. 1 These shorter works have been shown to 

belong to two distinct families: one headed by Vaticanus gr. 
1335 (A), the other by a lost hyparchetype, Cl>. In this paper I 
examine and interrelate manuscripts containing Xenophon's 
Symposium, a work found only in the manuscripts of the 
second family. The twenty-three manuscripts described here 
contain a total of twenty-five Symp. texts, seven of which are 
incomplete. 2 Three of the four groups of the <l> family of Symp. 

1 Agesilaus: R. WIECZOREK, Xenophon 's Agesilaus: A Collation, Stemma, 
and Critical Text (diss. University of Iowa 1975 [Ann Arbor 1975: hereafter 
'Wieczorek']); Hiero: J. K. DEULING and]. CiRIGNANO, "A Reappraisal of the 
Later ABS Family Tradition of Xenophon's [fiero," Scriptorium 44 (1990: 
'Deuling and Cirignano') 54-68; D. O. Haltinner and E. A. Schmoll, "The 
Older Manuscripts of Xenophon's Hiero, " RHText 10 (1980) 231-36; L. 
LEVERENZ, "The Descendants of Laurentianus 80.13 in Xenophon's Hiero," 
StIt 7 (1989: 'Leverenz') 12-23; Memorabilia: M. BANDINI, "Osservazioni sulla 
storia del testo dei Memorabili di Senofonte in eta umanistica," StCIOr (1988: 
'Bandini, "Osservazioni''') 271-92, and "I Memorabili di Senofante fra il 
Bessarione, Isidoro di Kiev, e Pier Vettori," BollClass 12 (1991: 'Bandini, 
"Mem. "') 83-92; lIipparchicus: D. F. JACKSON, "A New Look at the 
Manuscripts of Xenophon's Hipparchicus," CQ N.S. 40 (1990: 'Jackson') 
176-86; Athenorum Politeia: G. SERRA, "La tradizione manoscritta della 
Costituzione degli Ateniesi dello Pseudo-Senofonte," AttiPadua 91 (1978-79: 
'Serra') 77-117; Poroi: D. F. JACKSON, "The Manuscripts of Xenophon's Paroi," 
Stlt 8 (1990: 'Jackson, "Poroi"') 166-79; Cynegeticus: E. A. SCHMOLl., 
"Xenophon's De Venatione: A Collation, Stemma, and Critical Text 
(diss.Univ. of Iowa 1982 [Ann Arbor 1982: 'Schmoll, De Ven.'j), and "The 
Fragmentary Manuscripts of Xenophon's Cynegeticus," SyllClass 1 (1989: 
'Schmoll, SyliClass') 21-25. 

2 Vaticanus Urbinas gr. 95 (V) has two short excerpts in addition to a com­
plete Symp. Vindobonensis phil. gr. 37 (H) contains two complete Symp. 
texts. I know of three other extant Symp. manuscripts, which I have, 
unfortunately, been unable to study. From catalogue information, it is known 
that one, Monacensis 494, contains only an excerpt of Symp.; another, Mutin-

187 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

188 MANUSCRIPTS OF XENOPHON'S SYMPOSIUM 

manuscripts-M, 'Y, and A.-are already familiar from work 
done on other opuscula. 3 The fourth group-the a-unknown 
previously, forms the largest group of Symp. manuscripts, and 
is usually joined to only one other work of Xenophon, most 
often Oeconomicus. An outstanding witness of this new group 
is a Symp. manuscript that has not previously been consulted: 
Vaticanus Urbinas gr. 95 (V); codex V and the a group are here 
described and evaluated in some detail. 

The M Group 

The M group of Symp. manuscripts dates to the fourteenth 
and fifteenth centuries and includes the following: M arcianus 
gr. 511 (M), Marcianus gr. 369 (0), Laurentianus 55.22 (L), and 
Marcianus gr. 368 (N). Three Venetian manuscripts (MON) 
once belonged to Cardinal Bessarion (1403-72). Devoted to his 
former countrymen, Bessarion in 1468 willed that his library be 
sent to Venice, where many Greek refugees were relocating. M 
and N were among the manuscripts shipped from Rome to 
Venice in 1469; 0, which was not written until 1470, arrived in 
Venice sometime after the death of Bessarion in 1472. There are 
conjunctive errors that firmly establish the M group: 

1.8 alt. ~EV om. MOLN 2.3 ne; l1~tV EVE'YKaL] MOLN 2.22 
lCaJ-l1t't0J-lEVll] lCaA,u1t't0J-lEVll MOL N 3.5 aA,A,a cru a~] aA,A,a 
a~ MOLN 4.14 KA,£lVta 'to. Qv'ta] 'to. Qv'ta KA,£LVla MOLN 
4.2511011 om. MOLN 4.iO ad a~] ad (bv MOLN 4.52 cru O£ 
oft] cru OE MOLN 5.2 'tOlVUV crE £q>ll] 'tOlVUV £q>ll OE MOLN 5.6 
npoooEXE08aL] 1tPOOOEXE'tUL MOL N 6.9 YEtote;) YE 'tOL 'tote; 
MOLN 8.15 craq>&e; Kat a1tO'tEA,Et'taL] Kat a1tO'tEA,Et'taL 
oaq>&e; MOLN 8.29 aya8ate; om. MOLN. 

ensis III D-IO (129), is a fifteenth-century manuscript that, judging from its 
contents, is probably a member of the a family; the third, Marcianus gr. 513, 
is a fourteenth-century manuscript and also probably an a derivative. 

J See especially the articles on Cynegeticus, Hiero, Hipparchicus, and Poroi 
cited supra n.1. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

JOHN CIRIGNANO 189 

The head of this group is M arcianus 511 ( M), a paper codex 408 
folios in length. Although previous editors have disagreed about 
M's date, it can on the basis of its script be placed in the first half 
of the fourteenth century.4 

Marcianus 511 (M) contains the following works of Xenophon 
(ff. 141-325v): Cyropaedia, Anabasis, Agesilaus, Hiero, Memor­
abilia, Hipparchicus, Peri Hippikes, Lacedaemoniorum politeia, 
Athenorum politeia, Poroi, Oeconomicus, Symposium (ff. 
311 v-318), and Cynegeticus. M 's first three opuscula­
Agesilaus, Hiero, and probably Memorabilia-all derive from 
A m bros ian us E 11 info (5).5 But because the Ambrosian 
manuscript does not contain the other shorter works of 
Xenophon that appear in M, the copyist was forced to go 
elsewhere for the remaining Xenophon opuscula. 

The three manuscripts OLN are easily shown to be indepen­
dent derivatives of M. Each contains the same opuscula of 
Xenophon in the same order as does M, except that M arcianus 
368 lacks Hiero. Also, each of the three has in Symp.-i n 
addition to the M group's readings-unique separative errors. 

Marcianus gr. 369 (0), according to Mioni, is a "very 
beautiful"6 parchment codex of 280 folios. He tells us that the 
beginning of each fascicle of this manuscript is colored in gold, 
blue, green, and red, and that folio 1 includes the shield of 
Bessarion skillfully reproduced in these same colors. On the 
front guard leaf Bessarion's ownership is spelled out in Greek 
and Latin. The last folio contains a subscription, written in red 
ink, that dates 0 to 1470 and identifies as the copyist George 
Tzangaropoulos, who at the time was working in Rome for 
Bessarion. 7 

4 For a list of articles containing descriptions of the M group see Jackson, 
182 n.23. For catalogued descriptions of the three Venetian manuscripts see E. 
M ION!, Bibliothecae Divi M arci Venetiarum codices graeci manuscripti II 
(Rome 1985: hereafter 'Mioni'). In addition to describing M, Mioni lists in his 
bibliography some earlier editors and the different dates given in M. Of 
further interest in M is a subscription on folio 398v, which is written in a very 
free hand different from the other hands in the manuscript; it incorrectly 
assigns the year 1166 to M. 

5 For a description of codex S see Wieczorek 13. Also, M's Cyropaedia may 
derive from S, but not its Anabasis, which is missing from S. There is some 
disagreement whether Memorabilia derives from S: see Bandini," Mem." 90j 
Jackson lSlf. 

6 Mioni 129: "cod icem pulcherrimum." 
7 Tzangaropoulos wrote Marcianus gr. 380 for Bessarion in Rome 16 August 

1469: Mioni 137. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

190 MANUSCRIPTS OF XENOPHON'S SYMPOSIUM 

Editors agree that the works of Xenophon and Arrian (ff. 
1-190v) in Marcianus 369 (0) derive directly from M.8 In Symp. 
(ff. 122-28) 0 has the following unique errors that confirm that 
it is not the source for extant manuscripts: 

1.2 tv nEtpatEl] tv nnpulC!> 0 1.4 oi)v] oi)v dm:v 0 1.4 alt. E i 
om. 0 1.8 ~U(JlAl1(OV 'n] ~UOlAlKOV 'nvu 0 1.9 npooclYE'tUl] 
npooclyncn cis 02.22 KUt CPO~EpOnEPOV] EV CPO~EpOnEpov O. 

Laurentianus 55.22 (L) is a fifteenth-century paper codex of 
237 folios housed in the BibJioteca Medicea Laurenziana in 
Florence. 9 L seems to have been written by the scribe who also 
wrote Laurentianus 69.12 and 69.18; 10 the three codices together 
form a complete collection of Xenophon's works. The close­
ness of L to M has been proven for a number of Xenophontic 
opuscula. ll It is no surprise then that Symp. (ff. 198-218v) in L 
also derives from M. The following list includes unique L 
readings that prove L's independence from 0 and N. 

2.5 Ut)'t'!> om. L 3.6 £8vos om. L 4.32 'to'tE oE] 'tOOE L 4.50 EV] 
tnt L 4.53 UnOK'tElVat uu'tov] U7tOK'tElV uu'tOv L 6.7 'tl om. L 
6.7 cP&.:;] oucp&S L 9.2 dOElOlV ds] etOElOlV 7tpOS L. 

Marcianus gr. 368 (N) is a paper codex of 184 folios. Two of at 
least four hands at work in this manuscript have been identified 
(Serra 81) as Demetrius Trivolis (ff. 1-98v, Is4v line 6-1ssr line 
10) and Cardinal Bessarion (ff. 99-1s4v line 5, 155 line 11-184v). 
Codex N has long been regarded as an apograph of M, except 

8 P. Cerocchi, "Prolegomena ad Xenophontis Hipparchicum, " Stlt 6 (1898) 
498; G. Pierleoni, "De fontibus quibus utimur in Xenophontis Cynegetico 
recensendo," StIt 6 (1898) 68; both name scholars who testify to this 
derivation. The concluding work of the manuscript, five books of Polybius' 
Historia (ff. 191-208v), seems to have been copied from Marcianus 371. J. M. 
Moore lists scholars who testify to this derivation: The Manuscript Tradition 
of Polybius (Cambridge 1965) 25f. 

9 Bandini argues (" Mem." 84ff) that Laurentianus 55.22 (and Marcianus 369) 
in Mem. was a later copy of Marcianus 511 than Marcianus 368, thus dating 
codex L to the second half of the fifteenth century. 

lQ See E. L. De Stefani, "Due codici delle Elleniche di Senofante," SLIt 5 
(1897) 106 n.t. 

11 Agesilaus: Wieczorek 18; Hiero: Haltinner and Schmoll (supra n.l) 234; 
Memorabilia: W. W. Baker, "Some of the Less Known Manuscripts of 
Xenophon's Memorabilia," TA PA 43 (1912) 149; Bandini, "M em." 84; 
Hipparchicus: Jackson 182; Peri Hippikes: V. Tomassini, "Prolegeomena ad 
Xenophontis De re equestri," Stlt 10 (1902) 109; Athenorum politeia: Serra 81; 
Poroi: Jackson, "Poroi" 176-79; Cynegeticus: Schmoll, De Ven. 3. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

JOHN CIRIGNANO 191 

for its H ellenica that is absent from M. 12 Recently, N's 
immediate dependence on M for Agesilaus has been questioned 
(Wieczorek 22, 38). A collation of Symp. in N also casts doubt 
on N's alleged direct descent from M. Codex N's Symp. has at 
least seven omissions of considerable length that cannot be 
assigned to homeoteleuton. These omissions do not correspond 
to lines in M, which averages approximately eighty-three spaces 
per line. A manuscript whose lines are shorter, between fifty 
and sixty spaces per line, should be N's exemplar. It is also 
possible, however, that these are not line omissions, but the 
result of unusual carelessnessY 

Unique N readings include: 

1.6 post /..EYEw, vuv bE ov'tU om. N 2.10 post 
XU/..E1tOO't(i"Cll, <.ht ... XP1loEo8Ut om. N 2.17 ci/../..u ... S11'tEtV 
om. N 2.20 "COtS cOllOtS om. N 2.25 'tu 'trov tv rfl] 'trov tv yft 
N 3.6 post E1tlO'tCXV'tCXt, ou ... /..E/..118E om. N 4.6 ~J(' ... 

1tEtpiio8m om. N 4.40 OUbEV ... 'tCXlltEUOllat om. N 7.37 OS ... 
\'O'tUIlEVOS om. N. 

The y Group 

The 'Y group is comprised of Laurentianus conventi suppressi 
110 (C), Parisinus gr. 2955 (P), and Perusinus B34 (K). This 
family is especially interesting for its connections with two 
famous Renaissance humanists, Guarino Veronese (1374-1460) 
and Vittorino da Feltre (1378-1446). Symp. occurs only in 
fragmentary form in the 'Y group. 14 The same portion of Symp. :;. 
(8.28 1tA.dovoC; to the end) is found in CPK, a fact that provides 

12 Baker (supra n.11) 148; Cerocchi (supra n.8) 478; Tomrnasini (surpa n.l1) 
110. 

13 Bandini, "' Mem." 86, argues for a date in the 1450s for Marcianus 368; D. 
Harlfinger, Die Textgeschichte der pseudo-aristotelischen Schrift DEpt 'A'tOI-WlV 
fpujljlwv (Amsterdam 1971) 249 n.3, seems to think that Trivolis was still in 
the East as late as 1465. Codex N appears in Bessarion's 1468 inventory and so 
was written in Rome before that date. See H. Omont, RevBiblioth 4 (1894) 
129-87. 

14 Fragmentary or incompleted works are marked with an asterisk (*). 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

192 MANUSCRIPTS OF XENOPHON'S SYMPOSIUM 

primary evidence for the 'Y group. 15 The following readings also 
establish their relationship: 

8.30 'tolno b' cx~ ... EibwS om. CPK 8.32 uAlCqlO'tCX'tQV] 
UAXtJ1W'tOV ca.c.pa,c·K 8.33 'to{nous] 'tolno CPK 8.36 1ct(HEUO"at] 
1ttO"'tEUEtV CPK 8.37 fV€(3CXAOV] UV€(3cxAoV CPK 8.43 'til> 0v'tt] 
'til> bOlCEtV CP K 9.5 ou O"lCW1t'tOV'tCXS] O"lCW1t'tOV'tcxS C 
O"lC01t'tOV'tcxS K 11 Tj O"lCW1t'tOV'tcxS P. 

Laurentianus conventi suppressi 110 (C) is an early fifteenth­
century parchment manuscript of 150 folios in quinternions 
containing the following works of Xenophon: M em., Cyn., 
Hipp., Peri Hipp., Lac. Pol., Ath. Pol. \ Poroi':-, Symp. 'c 

(ff.l09-111). C is apparently a companion codex to Laurentianus 
conventi suppressi 112 (J), which contains Gee., Cyrop., Anab., 
and Hiera-works that C lacks. 16 A second hand appears after 
the Xenophontic material in C for one folio (f. 113); otherwise, 
C was written by a single scribe. Antonio Corbinelli (d. 1425), a 
patron of Guarino Guarini in Florence, owned both C and J, 
which went into the Badia Fiorentina at his death, later being 
acquired by the Biblioteca Medicea Laurenziana. 

The main scribe in C gives Symp. 'c the title: fragmentum 1t£pi 
cruJ.l1tocriou ~£vo<p&V'to<; PTJ'tOpo<;. Codex C deviates from other 
manuscripts in its group by placing Symp. 'c at the end, following 
Poroi':- rather than after Gec., perhaps because of its fragmen­
tary state Qackson 185). 

Parisinus 2955 (P) is a composite paper codex of 253 folios, 
now housed in the Bibliotheque Nationale in Paris. 17 Codex P, 
another fifteenth-century manuscript, contains the following 
opuscula of Xenophon (ff. 68-115): Oee. 'c (up to 1.23 "tov 
AOl1tOV.), Hiero, Lac. Po!., Ath. Pol.'c (up to 1.16), Poroi':- (from 
5.4), Symp. ,.t (ff. 108-110v), Peri Hipp. 'c (4.5-12.15). The second 

15 Also characteristic of the y group-including two other manuscripts, 
Laurentianus 55.21 (T) and Vaticanus gr. 1619 (W)-is an omission that 
begins at Ath. Pol. 1.16 (q>tAot IlclAtO'!(l ~oav 'A811va tillV... ) and extends into 
the next opusculum Poroi 5.4 ( ... xat oOq>to'Wt xat q>tA.OOO<jlot ... ). This omis­
sion, probably the result of a lost gathering in y, is passed on to each offspring 
unnoticed by the copyists. As a result, the two incomplete opuscula appear as 
one continuous work in each of the y manuscripts. See] ackson 179; Serra 82f. 

16 Laurentianus conventi suppressi 112 was likely the manuscript mentioned 
in a letter sent to Guarino by Isidore Ruthenus who was also J's scribe. See 
Deuling and Cirignano 54f. 

17 H. Omont, Inventaire sommaire des manuscrits grecs de la Bibliotheque 
Nationale I (Paris 1886-98). 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

JOHN CIRIGNANO 193 

hand (ff. 147-226) of at least four hands at work in this 
manuscript (not the scribe of Xenophon's works) has been 
identified (Deuling and Cirignano 60) as George Gregoropoulos 
of Candia, active 1465-1501. The Symp. fragment in P follows 
Poroi'~ as in C and has the same title: mopt O"UIl1tocrtOU 'tou 
~£vo<p&v'tO~ Pl)'tOpO~.18 Codex P in Symp.'~ has all of C's errors 
plus one additional insignificant error. 19 

Perusinus B34 (K) is a parchment codex of 276 folios written 
in two columns in quinternions. Gerard of Patras, K's scribe, 
was active in Mantua in 1431 and was one of a group who 
worked for Vittorino da Feltre-the famous pedagogue of 
Manuta (d. 1446).20 Codex K contains only works of Xenophon: 
Mem., Gec. '~, Symp.'~ (ff. 59-60v), Cyn. Cyrop., Anab., Hipp., 
Hiero, Peri Hipp., Lac. Pol., Ath. Pol. "\ Poroi'~. An incomplete 
Gec. ending at 19.9 'to KA:ilIlU) is joined immediately to the 
Symp. fragment that follows mid-sentence without a break. The 
conflation of the two works was noticed and a retroactive title 
was written at the conclusion of the Symp. fragment and the 
beginning of Cyn. (60v): LYMnOLIONKYNHrETIKOL. 

There are separative errors common to C and P that show 
that K does not derive from either of them: 

8.29 Ulnas om. CP 8.30 OE 't'] 'taOE C"P 8.31 ano8uvov'tl] 
ano8uvov'tos Cpa.c. ano8uvouv'to pc an08uvovn pin margo 8.32 
'ta VUV KuAa] 'ta VUV 'ta KUAa CP" 8.34 anOH:Awcrl] 
anE'tArocrl Ca.".p 8.36 'til> 'tlJv] 'trov 'tlJV CP 8.38 aVEXE'tUl] 
aVEcrXE'tUl CP 8.39 fPfUVll'tEOV OE Kui] EPfUVll'tEOV Kui CP 
9.1 &.v8pomos om. CP 9.5 aAAlJAouS] &'AAouS CPt aAAlJAoUS 
pm. 

Neither does C nor P derive from K: 

8.30 &'AAo8i nou] &.1.1.0 OlJnou K 8.33 d yE] d OE K 8.33 
8uu!lucr'tu] 'ta 8uu!lucr'tu K 8.35 faV KUt] faV yap Ked K 
8.38 KUt 'tov om. K 8.39 !lEV crOll !lEV'tOl K 8.40 a~lOnpfnEcr­
'tu'tov] a~lOnpEnEcr'tfpoV K 8.42 oi !lEV] oi !lEV yap K 9.2 

18 In the margin is the abbreviation "FRAG" for "fragmentum." Also 
underlining the title in P is an olive branch--Dne of many seen throughout the 
first scribe's portion of the manuscript (ff. 1-146). 

19 8.42 au !lE] au !lEV ( sic) P. By itself this reading is not sufficient to prove P's 
derivation from C. It has been shown, however, that P derives from C in 
Poroi* (Jackson, "Poroi, » for whom Laurentianus cowlJenti suppressi 110=L 
and Parisinus 2955=K). C then is probably also P's source for Symp. *. 

20 A. Diller, "The Greek Codices of Palla Strozzi and Guarino Guarini," 
JWarb 24 (1980) 320. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

194 MANUSCRIPTS OF XENOPHON'S SYMPOSIUM 

u1to1te1t(Ol(w~] u1to1te1t't(Ol(w~ K 9.5 ov't~] o,h(O~ K 9.6 qnAe"t 
a{ytov] <plA£t a{nllv K 9.7 C£1tllAauvov] U1tllAaucrov K. 

The production of the y family is thought to have occurred in 
the following way. After four years of teaching in Florence 
(1410-14), Guarino moved to Venice, where he purchased y, a 
manuscript containing works of Xenophon that at the time were 
not available in Florence.21 In 1417 Amrogio Traversari (d. 1439), 
a humanist monk working for Corbinelli in Florence, wrote to 
Guarino asking that he bring him these "rarer works."22 
Eventually, a copy of y was made (~) and sent to Florence. The 
arrival of ~ allowed Corbinelli to have the opuscula not found in 
Laurentianus conventi suppressi 112 a )-Mem., Cyn., Hipp., 
Peri Hipp., Lac. Pol., Ath. Pol. \ Poroi", and Symp. "'-copied 
into codex C. Before ~ was sent to France, Guarino made a 
copy (x) that he took with him to Verona where he settled after 
marrying in December 1418. Codex 1t was the source of K 
whose scribe, Gerard of Patras, was working in Manua in 1431 
for Vittorino da Feltre. 

The A. Group 

The A. group includes the following manuscripts: Laurentianus 
80.13 (E), Parisinus gr. 1643 (A), Leidensis Vulc. 2 (X), Parisinus 
gr. 1645 (B), British Museum add. ms. 5110 ( b), Vindobonensis 
phil. gr. 37 (H 1), and Vindobonensis hist. gr. 95 (G). For the 
existence of the A. group in Symp. there is little evidence;23 the 
group has been established for other opuscula of Xenophon that 
appear in these manuscripts. 24 Furthermore, the similarity of 
contents among these manuscripts is good evidence in support 

21 On Guarino and his Greek manuscripts see Diller (supra n.20) 317-21. 
22 See the stemma at the end of this article. Bandini, "Osservazioni" 175ff, 

discusses this sequence of events and also quotes from Traversari's letter. For a 
more detailed discussion of this history as it pertains to Symp. see my X eno­
phon's Symposium: A Collation, Stemma, and Text History (diss.University of 
Iowa 199) [Ann Arbor 1993]) 12-19. 

2J Only one A. reading exits: 2.11 £()o~£vl £()£l~£V GEXBbH1; codex A, a 
descendant of E, has the correct reading EOO~£V. The shortage of A readings in 
Symp. indicates that A's Symp. was written very carefully, and so had no 
significant errors to pass on. 

14 For Hipp., which appears in five of the above A manuscripts, see Jackson 
177; for Cyn. *, which appears in four of these manuscripts, see Schmoll, 
SyllClass 21-25. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

JOHN CIRIGNANO 195 

of the group's existence. All seven manuscripts contain Lac. pol. 
followed immediately by Mem., a combination not found in the 
other groups. Also five manuscripts-EAbHl G-have the same 
unique sequence of Xenophontic opuscula, beginning with 
Hipp.25 

Of the seven A. manuscripts, five derive from Laurentian us 
80.13 (E) in Symp.:26 

2.4 1l0X9llV] 1l0<JXllV EAXBhHl 'tOlV IlUP01tllArov EAXBhHl 
2.6 ropT]<Jn] £UpT]<JEtS EAXBhHl 2.22 'tOU1tl<JS£v] 'tOUIl1tp0<JS£v 
EAXBbHl 4.1 'touS uv9poo1tous] 'trov UVSPO)1tllV EAXBhHl 
4.39 OOl(£l] OOKOU<JtV EAXB hHl 4.41 8cx<Jtcp] 9cxullcx<Jtcp 
EAXBbHl 4.62 1tpocx"{op£u<Jcxv'tcx EBH1: 1tpocx"{o"{£uocxv'tcx b: 
1tpocx"{op£uov'tcx At: 1tpOCX"{ll"{EUOCXV'tCX A in margo et rell. 9.6 
<ptA£l cxu'tov] <ptA£l CXU'tT]V EXBhHl q>tA£l cxu'tov A. 

Laurentianus 80.13 (E) is a composite manuscript 189 folios in 
length, written in four hands and bound in quaternions. 27 Codex 
E contains the following works of Xenophon: Ath. pol., Paroi, 
Hipp., Hiero, Peri Hipp., Lac. pol., Mem., Dec., Symp. (ff. 
132-44), and Cyn.'~ (up to 2.1 1tat8dlJ.w·W.). Folios 1 to 18 are 
paper and were written by one hand from the fifteenth century 
(Serra 81£); the rest of the manuscript is parchment and dates to 
the early fourteenth century. The number of each work­
beginning with Hipp.-and also the number of folios dedicated 
to it appear at the top of each recto; the folios are numbered 
consecutively within each work at the bottom of each recto 
(Leverenz 13). 

Parisinus gr. 1643 (A) is a fifteenth-century paper manuscript 
205+ folios in length and gathered in quaternions. Codex A 
contains these works of Xenophon: Hipp., Hiero, Peri Hipp., 
Lac. po!', M em., Dec., Symp. (ff. 151 v-170). 28 The opuscula of 

25 The sequence followed is: Hipp., Hiera, Peri Hipp., Lac. pol., Mem., Oee., 
Symp., Cyn. * The exceptions are that codex G lacks I fiero, and A is without 
Cyn.* 

26 The ordering of the descendants of Laurentianus 80.13 has been worked 
out for Hiero by Leverenz; since codices AbHI are nearly identical in ordering 
and content, her findings are sure to hold for Symp. as well; the details then of 
the relationships among the E derivatives in Symp. will not be repeated here. 
Manuscripts X and B, however, which lack Hiera, are shown to be E 
derivatives. 

27 A. M. Bandini, Catalagus codicum graecorum bibliothecae Laurentianae 
(Florence 1768) III 202ff; for a more recent description see Leverenz 12£. 

28 For a general description see OmOnt (supra n.17) I 115; Leverenz 13f. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

196 MANUSCRIPTS OF XENOPHON'S SYMPOSIUM 

Xenophon in A display the same running log as in E, giving the 
number as well as the total folios for each work at the top of 
each recto; at the bottom, folios are numbered consecutively 
within each work. A single scribe wrote the manuscript with 
corrections added by a different hand above the lines and in the 
margins. Before correction, Symp. is mistitled throughout its 
text as Lac. pol. On the back of the first guard leaf is a note in 
Latin (dated 1785) that harshly criticizes Michael Apostoles (d. 
1486) for ruining the text of Mem.: "Hie Michael Apostolius 
textum fere totum Xenophontis laniavit, mutilavit, pessum 
dedit. Etsi nonnullas antiquas lectiones exhibeat hic codex, quas 
alibi frustra desidcraveris, pessimus dici potest; ut ipse pronun­
tio, collato integro opere (l1tOJlvTlJlovEuJHltrov.» 29 

Leidensis Vulc. 2 (X) is a fifteenth-century paper codex 164 
folios in length. 3o Codex X is a composite of two fifteenth­
century manuscripts: 3! the first part (ff. 1-141) was written by 
George Hermonymus of Sparta and contains Lac. pol., M em., 
Gee., and Symp. (ff. 73-83v). At the top of folio 143 another 
hand has written: MOaX01tOAOU tauta dvat imOAaJlPavro-per­
haps referring to the author of the marginalia of Philostratus 
Maior's Iconum appearing on folios 143-163. A prefatory folio 
gives the contents and the names of two previous owners: "fuit 
liber Pauli Aemilii Paris(iensis); nunc Ioa(n)nis Arcerii Theodor­
(eti).» The date 1565 also appears on this sheet.32 

Parisinus gr. 1645 (B) is a fifteenth-century paper codex 168 
folios in length. Codex B contains three works of Xcnophon: 
Symp. (ff.I-33), Lac. po!., and M em. There is no logging system 
as in E; the gatherings and the folios within each gathering arc 
numbered at the bottom of each recto. George Hermonymus 
wrote the text of B as well as its scanty marginalia. 

Codices X and B have already been shown to contain codex 
E's group readings. Codex B displays all the errors of X and 

29 The note does not specify whether Apostoles was the scribe or editor of 
the text. The appearance of Michael Apostoles' hand in Parisinus gr. 1643 was 
rejected most recently by Bandini, "Osservazioni" 287. 

30 For a general description see P. C. Molhuysen, Bibliothecae Universitatis 
Leidensis codices manuscripti, I. Codices Vulcaniani (Lciden 1910) 1 f. 

31 The composite nature of codex X is most easily demonstrated by the 
physical difference in size between each part (given by Molhuysen [supra n.30J 
1 f): part one measures 265 x175 rnm.; part two is 235x135 mm. 

32 Also on this same sheet it is suggested that folios 143-164 were written by 
Macharios (sic) Calliergis-an unknown and perhaps fictional scribe. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

JOHN CIRIGNANO 197 

many of its own; B then derives from X. The following is only a 
selection of unique B readings: 

2.17 OoAtX,OOPO!lOt ... oi. om. B 3.12 ot'to~ ye !lilv ecp" 'tt~ om. B 
4.10 E7t(XUcr(X'to] 1teptE1t<:tuo<X'to B 4.28 rocr1tep] rocr1tep KUt B 
4.58 'tl O£ AOyOt] AoYOt B 4.62 ecp" cre] ecp"crev B 6.9 aAA el1tEp 
yE 'to'i~ 1tUOl. rep. B 8.21 EV 'to'i~ acppOOtcrtOt~ om. B 8.42 ro 
I:roKp<:t'te~ om. B 9.2 K<:t'tE'tES,,] ICU'tES" B. 

British Museum additional 5110 (b) is a fifteenth-century 
paper codex 312 folios in length written in four hands. 33 The first 
hand begins the manuscript with H ellenica (ff.1-84v); the 
second hand continues with Hipp., Hiero, Peri Hipp., Lac. po!., 
M em., Oec., Symp. (ff. 152-159v), and Cyn. ", following the 
same sequence as in Laurentianus 80.13 (E). The works of 
Xenophon, beginning with Hipp., are numbered only in the title 
at the beginning of each work (Leverenz 13 n.2). Codex b is a 
composite of at least two manuscripts: Hellenica was written in 
the early fifteenth-century, the opuscula of Xenophon in the 
late fifteenth. 34 

Vindohonensis phil. gr. 37 ( HI, H2) is a paper codex 292+III 
folios in length, written in five hands. The first half dates to the 
sixteenth century; the second half (ff. 165-292) to the end of the 
fifteenth. 35 This codex is a composite of at least four different 
manuscripts, three of which contain works of Xenophon 
(Leverenz 15 n.6). It is unique in having two copies of Symp. 
and Peri Hipp. Vindobonensis phi!. 37 contains Hipp., Hiero, 
Peri Hipp., Lac. pol., written by the same scribe; a second hand 
wrote Mem., Dec., and Symp. (ff. 112v-124v=H 1; Cyn., Peri 
Hipp., and Symp. (ff. 148v-163=H 2) were written by another 
scribe; George Gregoropoulos of Candia appears to have 
written Epictetus' Encheiridion and a commentary on this work 
by Simplicius (ff. 165v-245v). 

33 M. Richard, Inventaire des manuscrits grecs du British Museum I (Paris 
1952) 3; codex b has been described more recently by Leverenz 21 f and 
Schmoll, SyllClass 22. 

J.4 Hellenica in codex b was once one with British Museum Egerton 2625, 
which originally included Hellenica and Thucydides. See D. F. Jackson, 
"Varia Palaeographica," C] 65 (1969) 1 Iff. 

35 H. HUNGER, Katalog der griechischen H andschriften der osterreichischen 
N ationalbibliothek I (Vienna 1961: hereafter 'Hunger') 162, lists the water­
marks for Vindobonensis phil. gr. 37, three of which date to the late fifteenth 
and three to the early sixteen century: Verona 1518; Vicenza 1501; Udinc 1503; 
Treviso 1483; Udine 1485, 1479. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

198 MANUSCRIPTS OF XENOPHON'S SYMPOSIUM 

Vindobonensis hist. gr. 95 (G) is a paper manuscript 338 folios 
in length, which dates to the second half of the fifteenth 
century.36 The scribe remains unknown and there are no visible 
quire signatures. Codex G contains Anab., Hipp., Peri hipp., 
Lac. pol., Mem., Gee., Symp. (ff. 303-323), and Cyn.:~ (up to 2.1 
ro<pEA.do:~). 

Codex G lacks E's conjunctive errors and is therefore not 
derived from it. The following is only a selection of G separative 
readings:3? 

1.5 ad] ad 0.0 G 1.14 YEAoiov n EUSUe;] EUSUe; YEAOiov 'tt G 
3.5 <Ju~~tyvU'tat] ~lyvU'tat G 4.10 ~axoU~Eea] <Ju~~axou~ESa 
G amXTrle; OlK11V] amx't1lv OlK11e; G 4.18 Of ;,OU ... OEOt'tO om. 
G 4.44 'tt~&~at] n~& Kat G 5.8 'Vllcpous] \jI~ G 8.4 £p~e;] 
<>pq.e; G 9.6 aAA"Aoov om. G. 

The a Group 

The ex group, the largest of Symp. manuscripts, includes the 
following: Vaticanus Urbinas gr. 95 (V), Laurentianus 85.9 (D), 
Vindobonensis phil. gr. 109 (F), Laurentianus 55.19 (2), 
Ambrosianus A 157 sup. (Q), Ambrosianus E 119 sup. (R), 
Vindobonensis phil. gr. 37 (H 2), and Monacensis 495 (g). Two 
additional manuscripts with sizeable excerpts from Symp. may 
be included in this group: Laurentianus Acquisti e Doni 37 (q) 
and Bernensis 690 (r).38 A few readings establish the group: 

2.3 ~€v'tOt] ~EV n a 3.2 et1tav] et1tova 3.9 EuXapi'tcp] 
Euxapt<J'tcp a 3.13 <> o· d1tEV] <> o· EhEv on a 4.36 o{51:oo] 
ou'troc; aD a 8.4 mxpEXE] mxpa<JXE VDFZQ R mxpE<JXE H2. 

Unlike the other groups that contain a more extensive selection 
of Xenophon's works, the a group features Symp. and usually 
only one or two others, most often Gec., suggesting that the 
exemplar a had only Gec. and Symp. 

Although Urbinas 95 (V) has the above ex readings, it is 
without other significant errors and is therefore closest to Cl> of 

36 All watermarks date to the second half of the fifteenth century: Hunger 
103. 

37 The relationship between Laurentianus 80.13 and Vindobonensis hist. gr. 
95 has already been established for Cyn. >+ (Schmoll, SyllClass 22f), Hipp. 
(Jackson 178f), and Mem. (Bandini, ·Osservazioni" 290). 

38 Neither manuscript has previously been consulted as a Symp. codex. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

JOHN CIRIGNANO 199 

the a group. The rest of the family helps to establish the lost 
gemellus of V, codex ~, which contains the following errors: 

2.12 QUXt] QU DFZQRH2 3.7 £<1'11 i1 LrolCpa'tE~) i1 LrolCpa'tE~ 
£<1'11 DFZQRH2 4.12 OE~atf.L11V av] av OESatf.L11V 
DFZQRH2Q 4.35 1tEPl't'tEUQv'ta 'tll~] 1tEPl't'tEUQv'ta 'tex. 't1l~ 
DFZQRH2 4.56 d1tav] d1tQV DFZQRH2 6.4 'EPf.L0YEVES] i1 
'Epf.L6YEVE~ FZQRH2. 

Vaticanus Urbinas 95 (V) is a paper manuscript 333 folios in 
length, in quinternions, and is bound in two volumes: folios 
1-151 and 152-333. Folios 1-80 were originally a separtate codex 
and contain mostly works and excerpts from Xenophon: Gee. 
(ff. 1-31r) is followed by Symp. (ff. 31r-48v); the rest of folio 48v 
is filled with sententiae accompanied by illustrations from the 
Bible that were clearly added to an already existing blank space; 
the next seven folios contain a first series of excerpts from 
M em., followed by a complete H iero (ff. 57-66v), more 
excerpts from M em., six lines from Gee. followed by two 
SJmp. excerpts (71 v-72): the first runs from Kal 6 IWlCpaLT1~ 
dm:v (2.9) to £bO~EV EiPTla8al (2.11), the second from bnjl& (2.23) 
to Eydpnv (3.2) and is followed by more excerpts from a ee. 
(72v); the last eight folios of this section contain various 
apophthegms, sententiae, and excerpts. 

Codex V continues with more sententiae, epitomes, orations, 
homilies, epigrams, etc. 39 Much of the last section of V-folios 
239-333-was penned by John Eugenicus, an admirer of Pletho 
and ardent anti-unionist. 40 There is an interesting note, written 
by Eugenicus in the top margin of folio 239, which introduces a 
work on the Sabbath by John of Damascus; it indicates that the 
present book was copied from a "hagioretic" book on the holy 
mount of Athos at the time of the synod in Italy, "where we 

39 Folios 236-38 are blank. For a complete listing of the varied contents of V, 
see C. Stonajolo, Codices Urbinates graeci Bibliothecae Vaticanae (Rome 1895) 
139-47. 

40 Except for ff. 259-63 and 314-16. This according to G. Mercati, "Sopra 
alcuni autografi di Giovanni Eugenico," Opere minori IV (==Studi e testi 79 
[Vatican 1937]) 61 n.5. Canan mostly agrees with Mercati, but doubts whether 
ff. 239-57 and 306-11 are in his hand; he also asserts that only the marginalia 
of ff. 259-63v are written by Eugenicus. Also, according to Canart, only the 
interlinear phrases-not the text-for ff. 314-17v are in Eugenicus' hand: P. 
Canart, "Scribes grecs de la Renaissance: additions et corrections aux 
repertoires de Vogel-Gardthausen et de Patrinelis," Scriptorium 17 (1963) 76. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

200 MANUSCRIPTS OF XENOPHON'S SYMPOSIUM 

sailed from the big city-with no good reason."41 Assuming that 
this note refers to the Council of Ferrara-Florence, this section 
of V was written in late 1437 or early 1438. 42 Of interest also are 
the autograph orations of John Eugenicus found in the last 
section of V (ff. 265-305, 326-33), and a red monocondyl on 
folio 263v that reads: 'Iwuvvll<;; VO'tUPLO<;; 6 EUYEVlK6<;;.43 The first 
part of codex V-folios 1-235-dates to the fourteenth century, 
and except for an excerpt from Theognis (ff. 81r-83r) is in onc 
hand. 44 

Vaticanus Urbinas 95 contains no significant separative errors 
in Symp. and is probably the oldest of the extant Symp. 
manuscripts in the a family.45 It has been used previously in 
establishing a text of Xenophon. The manuscript's shabby 
appearance as well as its complicated contents must have 

41 tv hEpq> apXUWUltq> ~l~Aiq> 0 8wu xapm n£pLCH.o/;£tUl illllV i:e; o£upo, 
£UpiOl(£'tUl il £mypa<fl~ Pll'troe; ou't(J)()'iv' 'Ic.oavvou 'tun£\vou 1l0VUXOU np£(J~u'tEPou 
L\ulluox:iJvou A.6yo<; de; 'to aylOv oa~~u'tov. £Otl(£V oily dVUl 't<!> ayiq> ouyxpovov 
tl(£lVO. 'to DE nupov 1l£'tE)'pa<fl1l £l( ~t~Aiou aYlOp£t'tll(OU tv U\)'t<!> 'tip ayiq> opo 
'tou "Aero' ';Vll(U l(u'ta 'tr,v tv 'hUAi~ OUVODOV ano 'tile; llE)'aAll<; nON:ro<;, 00<; Ill, 
W<jl£AA£V, i:~£1tAEOIl£V. 

42 While en route to Florence, Eugenicus apparently took advantage of a 
stop at Mt Athos to obtain a copy of this work on the Sabbath by St John of 
Damascus. There is some doubt whether the work (ff. 239-57) is in Eugenicus' 
hand: Can art (supra nAO) 76. Assuming that John did not at that time bring 
codex V with him to Italy, this section was inserted into V at some later time. 
Perceiving the Council to be a failure, John left Ferrara soon after the 
procedings began in September 1438. His return voyage from Venice was 
delayed by a shipwreck that landed him in Ancona. While in Ancona, 
Eugenicus wrote an oration in thanksgiving for his survival, which is dated 
11-22 May 1439 and is found in Paris. 2075 (ff. 244-81 v); an excerpt from this 
oration-£uxupl0't~plOe; de; ... 8w'tol(oV £nt 'tU anUAAaYU 'tou tv euMoon 
nll(pOU euva'tou--appears on ff. 329-30 of codex V. 

43 Mercati (supra nAO) 61 nA. See also D. C. Young, "A Codicological 
Inventory of Theognis Manuscripts with Some Remarks on Janus Lascaris's 
Contamination and the Aldine Editio Princeps," Scriptorium 7 (1953) 7. There 
are marginal notes (esp. ff. 328-29) that mention Eugenicus. Also, the 
watermark on folio 324--cheval 3567 (Venice 1433) or 3568 (Venice 1434) fits 
well with the dates of John Eugenicus. 

HAs Mercati notes (supra nAO: 61 nnAf), a monocondyl on f. 8lv names 
Thomas Gorianites who may be mentioned at the Acta of 1318 (F. Miklosich 
and J. Miiller, edd., Acta Patriarchatus Constantinopolitani 1315-1402 e 
codicibus manuscriptis Bibliothecae Palatinae Vindobonensis [Vienna 1860J I 
86). 

45 V's lack of separative errors may indicate that it is a direct descendant of 
codex a. A potential rival for oldest member of the a family is Marcianus 513, 
a manuscript I have as yet not seen. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

JOHN CIRIGNANO 201 

contributed to its neglect. 46 Although codex V contains three 
complete works of Xenophon, Dec., Hiero, and Symp., the 
greater part of the manuscript consists of M em. fragments and 
other excerpts that are of little use to editors of complete texts. 47 

Moreover, the most important figure in recent times for studies 
of Xenophontic manuscripts, Karl Schenkl, never mentions 
Urbinas 95 (V).48 

Codex V probably came west in the 1460s soon after the death 
of John Eugenicus and entered the library of Federigo da 
Montefeltro, founder of the most famous of the princely 
libraries of the Italian Renaissance. Although Duke Federigo had 
willed that his library's manuscripts should remain in Urbino, 
Pope Alexander VII in 1658 transferred the manuscripts to the 
Vatican for safekeeping. Urbinas 95 remains at the Vatican and is 
one of the 1,767 manuscripts comprising the U rbino collec­
tion.49 

Codex V is the gemellus of codex ~, which is the lost source 
of two other lost manuscripts: codices ~ and C. The following 
errors estabish the lost codex ~-parent of Laurentianus 85.9 
(D) and Vindobonensis phil. gr. 109 (F): 

2.11 de; o{)v] de; !lEV o{)v DF 2.12 o{)'tco om. DF 3.4 1to'tEpav] 
1tou:pov DF 3.8 1tOAATtV] 1tOAAlJV £<1'11 DF 4.37 arav om. D F 
4.47 'tl u: ... ahouv'tat 'toue; 8wue; om. DF 5.6 eCf, £u8ue;] £u8ue; 
eCf, DF 8.9 eon 1m\. 't11~] eon 't11~ DF Kat vaol dot] dot Kat 
vaol DF 8.37 d5011AOV] a011Aov Dpa,c'rt 9.3 OU1tco] 0151co DF. 

Laurentianus 85.9 (D) is a parchment codex 434 folios in length 
and gathered in quaternions. In addition to Dec. and Symp., 

46 I have seen that the first half especially of Urhinas 95-where the three 
complete works of Xenophon occur-has a brownish color and is splattered 
with sizeable waterspots. 

47 For example, Young's inventory of Theognis manuscripts omits Urhinas 
95 with only 276 lines of Theognis as an independent but not valuable witness; 
he refers (supra n.43: 7) to this manuscript as a "scrap-book." 

48 Schenkl contributed three Xenophon studies to the SBBerlin series: the 
first (1869) on Anah., the second on Mem. without mentioning the fragments 
in Urbinas 95, and the third (83 [1876] 141-68) treating Oee., Symp., and Apol. 
again without menton of Urhinas 95. 

49 On the library at Urbino see especially C. H. Clough, "The Library of the 
Dukes of Urbino," Lihrarium. Revue de La Societe des bibliophiles suisses 9 
(1966)=id., The Duchy of Urbino in the Renaissance (London 1981) ch. VI; D. 
Robathan, "Libraries of the Italian Renaissance," in J. W. Thompson, ed., The 
Medieval Library (New York 1957) 536-44. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

202 MANUSCRIPTS OF XENOPHON'S SYMPOSIUM 

codex D contains the complete works of Plato. 50 Sicherl states 
that the works of Plato in D were copied from Laurentianus 
59.1 ca 1462 at the request of Marsilio Ficino, who received 59.1 
from Cosimo de' Medici. 51 Although Sicherl gives no evidence 
for Ficino's connection to 85.9, he is probably right in placing D 
in the fifteenth century. 52 

The text of codex D was written by a simple scribe who, 
starting at folio 28, at the top of every recto, has written the 
author, title, and number of folios that each dialogue occupies. 
At the bottom of every recto, each folio is numbered seq uen­
tially within each dialogue. 53 Another hand is apparent now and 
then in D's margins. The works of Xenophon, Oee. (f. 419) and 
Symp. (429v), follow the Plato section after orations by Aristides 
and Libanius, and conclude the manuscript. The editio princeps 
Symp., printed by the Giunta press in 1516, seems to be derived 
from codex D.54 

Vindobonensis phil. gr. 109 (F) is a paper codex 324+II folios 
in length, in quaternions, and contains ten works of Plato that, 
except for M enex., follow the sequence in D. Of the works of 
Xenophon, codex F contains only Symp. (f. 219).55 The hand 
that executed the text of F resembles that of George Chrysococ­
ces, grandfather and namesake of the scribe of Z and Q; 56 there 
are Latin as well as Greek marginalia, some of which appear to 

50 For a complete description of D's contents see Bandini (supra n.27) III 
257-66. 

51 M. Sicherl, "Platonismus und Texttiberlieferung," in O. Harlfinger, ed., 
Griechische Kodikologie und Textuberlieferung (Darmstadt 1980) 554. 
Laurentianus 59.1 is the oldest extant manuscript containing the complete 
works of Plato, but lacks works of Xenophon and is therefore not the source 
of 85.9: G. Boter, The Textual Tradition of Plato's Republic (Leiden 1989) 32f. 

52 M. Schanz, "Mittheilungen tiber platonische Handschriften," 11 ermes 10 
(1876) 174, shows that the text of 85.9 (D) contains readings from the margins 
of Laurentianus 59.1 and concludes that 85.9 must be younger than 59.1, 
contrary to the dating of A. M. Bandini who dates 59.1 to the fourteenth and 
85.9 to the thirteenth century. 

53 Laurentianus 80.13 (E) has a similar logging system, but instead of giving 
author and title, the top of each recto indicates the number of each work: 
A.6yo~ u'. Myo~ W. etc. 

54 The Symp. of the 1516 Giunta edition contains all of the OF (0) readings 
and all but one (at 5.40) of D's readings. 

55 For a more complete description of F see Hunger 217f. 
56 A sample of the elder Chrysococces' hand appears in A. Turyn, Codices 

graeci Vaticani saeculis XIII et XIV scripti (The Vatican 1964) pI. 110. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

JOHN CIRIGNANO 203 

be by George Gemistus Pletho. 57 There are also clear indications 
that Symp. was inserted in F sometime after the codex was first 
compiled. In the pinax, Symposium is written in an erasure by a 
hand different from that of the Plato contents. Also, the 
ordering of quire signatures is interrupted by a qumternion 
numbered a' on folio 119 where Symp. begins. 

Vindobonensis phil. 109 is usually assigned to the fourteenth 
century.58 All of the watermarks in F-except for one from the 
late thirteenth century-date to the fourteenth. 59 An inscription 
on folio 2 identifies John Sambucus (d. 1584) as an owner of 
codex F, which he purchased for five ducats. Hunger suggests 
(217) that F may have been used for the printing of Stephanus' 
1578 edition of Plato. 

Laurentianus 85.9 (D) is a carefully copied manuscript and 
contains only a few readings to show that it is not the parent of F 
in Symp.: 2.18 EV (Jl(tCf] EaV (Jl(tCf D 4.5 1tpoayopE1H:tV] 
TtP0(JCXYOPEUElV D 4.25 bil] i1bTj D 4.50 fTC' a\:l1:n <pPOVOtTj] 
cppovoiTj E1t' atytn D 4.56 £lEv] olov D. ' 

Nor is Vindobonensis phil. 109 (F) the parent of D, as this 
sample of errors in F shows: 

2.3 a 1.1.:1'1 flEv avSpl. aAAll OE yuv(wd] aAAll flEv yuvuud 
aAAll bE uVbpt F 2.6 KCd. aAAo] aAAo F 3.7 'hj..la~ TCoa:lv] TCOtElV 
'hj..la~ F 4.24 oUbUj..lOU om. F 4.25 aTCAll<J'tOV] a1tt<J'tov F 4.37 
Ex mv] EX m F 4.4 2 <JKoTCouV'ta~] U<JKOUV'tU~ Fa.c. 4.58 't { bE 'tn ... 
TCavu j..lEV o?Jv om. F 6.1 ECPll add. F post ExOt<; &v 6.2 uAAeX] 
aAA' 0 F 8.29 'to{nou<;] uu'tou<; F 8.33 \jIoyou 'tEl \jIEynv 'tt F 9.2 
'ttS] 'tE F. 

57 Noted are marginalia especially on folios 108, 208, 209, 215, compared 
with Pletho's hand in Marcianus gr. 406 (given in D. I-Iarlfinger, Specimina 
griechischer Kopisten der Renaissance, I: Griechen des 15. Jahrhunderts 
[Berlin 1974] pI. 1). The scribe of codex F also appears in A mbrosianus I 121, 
which is catalogued as a fourteenth-century manuscript and contains Cyrop. 
and A nab.: A. MARTINI and D. BASSI, Catalogus codicum graecorum 
Bibliothecae Ambrosianae (Milan 1906: hereafter 'Martini and Bassi') I 565. 

58 Hunger 217; L. A. Post, The Vatican Plato and its Relations (Middletown 
1934) 91; R. S. Brumbaugh and R. Wells, The Plato Manuscripts, A New Index 
(New Haven 1968) 14. According to Bater (supra n.51: 60, 142), Bk 10 of 
Plato's Republic in F derives from Laurentianus 85.9, a fifteenth-century 
manuscript. A date in the fifteenth century, however, would conflict with the 
dates of the watermarks in F. 

59 The watermarks are listed and dated as follows: Siena 1297, Bologna 1342, 
Perpignan 1330, varia between 1314 and 1336, Aldenbiesan 1345, Siena 1325 
with varia between 1322 and 1361, Bologna 1324-32, Decizes 1386, Siena 
1331-33, Sion 1349-75: Hunger 218. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

204 MANUSCRIPTS OF XENOPHON'S SYMPOSIUM 

There are readings that show that ZQRH2q derive from a lost 
codex ~, a gemellus of S. The following is only a selection of ~ 
readings.60 

1.1 'tu tv] tv ZQRH2 2.3 ne; 11J.1tV tVE-yKat] 11J.1tV ne; tVEYKat 
ZQRq 2.11 ~t<p&v op8&v] op8&v ~t<prov ZQRH2 2.13 1tuv't(XC; 
om. ZQRH2 4.30 ixavoe; dl1v] £1l1V bmvoe; ZQRH2q 4.44 Kat 
J.1ltV Kat] Kat J.1"V ZQRH2 4.45 on om. ZQRH2 4.61 £<P11 J.10l 
oOKel 'Avno8tvl1e;] J.10l oOKel 'Avno8tvl1e; £<P11 ZQRH2 8.36 
1tO'tEP~ natol. qnA118Evn] no'tEp~ <ptA118EV'tt 1tatOl ZQH2 (R 
om. 1tatOl ... ad Convivii finem). 

Laurentianus 55.19 (Z) is a parchment codex 230 folios in 
length, in quinternion gatherings. 61 Codex Z contains only three 
works; all three of Xenophon: Symp. (f. 1), Oee. (f. 21), and 
Cyrop. (f. 55). Part of the subscription at the end of the codex (f. 
230v) appears in verse form and reveals that Cyrop. was written 
in Constantinople by Chrysococces for Francisco Filelfo; the 
date given is 23 November during the fifth indiction, 1426: 

"'noe 1tEpae; Aupev 11 3evo<proV'toc; PtPAOe; aplo'tll 
1tatodl1v ye KUpOLO KaA&c; J.1uAa ole~ouoa 
Xnp1. feropyiou ypa<petoa 'tOU XPUOOKOKKll 
<l>tAEA<pOU 0' avaAWJ.1aOt 'tou <l>paYKtOKOLO KA:110LV. 
heAnffi8l1 J.1l1V1. NoeJ.1ppt<[l KY' ivo e' thoue; 
~ Ae' tv Krovo'tavnvou1toAn. 

At the bottom of the first folio can be seen a circular emblem 
with Filelfo's initials FR and PH written in golden letters on 
either side of the circle. An elaborate rubric begins each work. 
After Oee. (f. 54; 54v is blank) Cyrop. begins on folio 55 with 
another elaborate rubric very similar to the one on folio 1; 
Filelfo's emblem and initials reappear at the bottom of this same 
folio. The similarity between the first folio and folio 55 may 

60 There are four readings that might, at first glance, seem to deny this 
grouping; they can, however, be easily explained: 4.27 Kpno!)ouA'P R H2: 
KPl'tOj3oVAou ZQ rell. 4.52 au Of QR: au E of) ZH2 rell. 5.10 Ota<pEPElV R ola<p[ .... ] 
H2: ota<p8dpElv ZQ rell. 8.32 na'\)aavia~ QR: na'\)aavia~ 't£ ZH 2 rell. It is 
easy to see how at 4.27 a genitive is attracted by surrounding datives into the 
dative case; it is also easily corrected. At 4.52, a rather superfluous of) is omitted 
independently. The entire word at 5.10 may not have been legible in the 
parent of ZRH2; Z guessed the correct reading, R made a poorer choice and 
H2, this time, did not hazard a guess. At 8.32, while the tradition reads 't£ 

Marchant uses Athenaeus' ye; Q and R again independently omit an 
extraneous particle. 

61 Bandini (supra n.27) I 283f. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

JOHN CIRIGNANO 205 

indicate a lapse of time between the copying of the Symp. and 
Oec. section, and Cyrop., which begins after a blank folio within 
the last gathering of Oec. 

Ambrosianus A 157 sup. (Q) is a paper codex, 1+128 folios in 
length, in quaternions, with a parchment folio at the beginning 
and at the end of the manuscript (Martini and Bassi I 65). Codex 
Q contains only works of Xenophon: Anab., Symp. (f. 91), and 
Oec. (104v). Brief notes appear in the margins of Anab. in a hand 
resembling that of the marginalia in Laurentianus 85.9 (D). The 
letter corresponding to the appropriate book of Anab. is written 
at the top of each recto. A subscription at the end of Anab. (f. 
90v) states that George Chrysococces wrote the Anab. of codex 
Q in October, 1425: E'tEAElWSll 'to 1tapov ~t~Aiov EV JlllVt 
6K'tW~piq> iv(5o~ 'tE'tap'tll~ E'tOU~ ,~ot>~ ot> ACll (5' XElpt 'Ypa<jl£v 
btaKOvou 'YEWP'YtOU 'tou XpuO"oKOKKll. Since the subscription 
appears at the end of Anab. rather than at the end of the codex, it 
is not certain whether Symp. and Oec. were written in October 
of 1425. The rest of folio 90v is blank; Symp. begins at the top of 
91, ends on 1 04v, and is followed on the same folio by 0 ec. 
Symp. and Oec. are without marginalia except for an occasional 
0"11 and words marking the contents of the adjacent passage. The 
hand is unmistakably that of Chrysococces throughout the 
codex. 

Ambrosianus E 119 sup. (R) is a fifteenth-century composite 
paper codex I+87+XII folios in length (Martini and Bassi I 365). 
The quire signature a' appears on folio 59v and probably marks 
the end of what was originally the first quinternion of a separate 
codex. Codex R then falls into two halves: folios 1-49 and 50-87. 

The manuscript opens with two works of Xenophon: Gee. is 
followed by an incomplete Symp. (f. 25), which ends at the 
bottom of folio 37r with 'tip 1to'ttpw~ <jltAllStv'tt (8.36) and is 
followed immediately by a brief excerpt from a letter attributed 
to Euripides. Codex R continues with Theophrastus' Charac­
teres I-XV and Demosthenes' Epitaphius (f. 42v-47); two more 
orations of Demosthenes follow, De corona (f. 50) and Contra 
Cononem (f. 84v), each preceded by a brief hypothesis. Most of 
Oec. and Symp. were copied by one scribe who was relieved 
for folios 21r and 22v. A third hand is evident on folio 48; folios 
48v and 49r-v are blank. De corona is written by a fourth hand, 
with marginalia in the third hand, which returns on folio 84v to 

complete the codex. 
Vindobonensis phil. 37 (H 1, H2), as already noted, is a 

composite manuscript and contains two copies of Symp.: HI 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

206 MANUSCRIPTS OF XENOPHON'S SYMPOSIUM 

derives from Laurentianus 80.13 (E); H2 (ff. 148v-163r) also 
dates to the sixteenth century, was quickly and shoddily written, 
and contains many errors. 

Ambrosianus A 157 sup. (Q) has all the errors of Z plus plenty 
of its own; a selection of Q errors follows: 

2.3 oucrCH rocr1tEp] oucrat rocr1tE~ Kai Q 3.4 KaAoKa:yaSia] 
KaAoKuyaSta £<Pll Q 4.18 rov] rov av Q 4.28 1tAEtOV] 1tAdro 
Q 4.45 uvqEAacrav] EYEMcrav Q 4.54 OMS YE] oAaS Q 4.63 
A icrXUAov] Kat A icrXuAov Q ~ll'tOUV'tES] E~ll'tOUIlEV Q 5.8 xpTj] 
XPl, i1 Q 6.2 av 'ttS om. Q 6.8 \jIuAAa] \jIuAAaS Q 7.5 0 ouv] 
Kat 0 Q 8.35 0pEX8n 'ttS] 'ttS 0pEX8n Q 9.5 crKW1t'tov'tas] 
crKT]1t'tov'tas Q. 

There are no significant errors in Z that are not also found in 
codex Q; only the following differences were found: 4.27 
'YpaJ.1J.1ancr-rn] 'YpaJ.1J.1a'toKOJ.1tcr-rn Z t corr. in margo 4.38 
acppo3tcruicrat] acppo3tacrat Z 4.57 EP'YOV dyad dvat EP'YOV dvat 
Z. Q then is derived from Z, despite the conflicting dates that 
appear in the two manuscripts. If the dating of Q's Anab.-Oc­
tober, 1425-is correct, it seems that folios 91-96 of Q were left 
blank for approximately thirteen months during which time Z's 
Symp. and Gee. were written; then Cyrop. was written and 
completed 23 November 1426. Soon after, Chrysococces wrote 
Symp. and Gec. into Q. It is also possible that Chrysococ­
ces-since it was so early in the new year (which began in 
September)-recorded the previous year in Q. If so, Anab. in Q 
was really copied in October, 1426; then Z's Symp., Gee., and 
Cyrop. were finished by 23 November 1426, after which time 
Chrysococces returned to Q and copied Symp. (and Gee.?) 
from Z 

Ambrosianus E 119 sup. (R) has an abundance of separative 
readings, which demonstrate that R is not the source for any of 
the existing Symp. manuscripts. Only a selection of R errors 
follows: 

1.14 E1tt'tEAOtll] E1ttll£AOill R 2.4 EV YUllvacriotS] EV 'tOtS 
YUllvacrtOtS R 2.25 ayav u8poroS] u8poroS ayav R 3.7 
IlEytcr'tOV] IlEya R EAot06pouv IlE] 4.32 EAOt8opOullllV R 4.53 
1tav'ta.1tacrt ... au'tos up' £<Pll om. R 8.10 'tllS \jIuXllS 't£ Ka1. 'tllS 
<ptAias] ~S <ptAias Kat 'tllS \jIUXllS R 8.22 xaA£1tov] xaAat1tov 
R. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

JOHN CIRIGNANO 207 

Nor is H2, a sixteenth-century text, the source for any extant 
Symp. ~anuscripts. The following is only a small selection of H 2 

separatIve errors: 

1.1 a~toJlvllJl6v£u'tu] a~toJlvT}JlOV£u'tou<; H 2 1.9 U 1t ' 
£Kdvou ... £crXllJlu'ti~ov'to om. H 2 2.4 il8£iat] il8Elvat H 2 

2.23 qnclAllv om. H 2 4.7 £VEyKn] £V EYJlll H 2 4.47 
vOJli~oJl£v YE] VOJli~OJlEVOV yE H 2 5.2 1tPOcrEVQKo."CW] 
1tPOcrqKcl'tO) H2 7.3 JltJlOUJlEVOU<; ... roPUlOU<; om. H2 
8.311t£1toill'tat] mo1toill'ta H2. 

The source for the excerpt manuscript, Laurentianus Acquisiti 
e Doni 37 (q), appears to be codex Q. Codex q shares the 
following errors with Ambrosianus A 157 sup. (Q): 4.30 
cru KO<pclV "Cw;] cruKo<pav'touv'ta<; Q q 4.34 a.o] o.ov Q q vOJl i~w] 
vOJli~O) Eq>ll Qq 4.47 JlEV Kai] JlEV Qq'tOu'tO] 'tolno 811 Qq 4.49 
acral ()1tocra Qq. Laurentianus Acquisti e Doni 37 (q ) is a paper 
manuscript 138 folios in length, which dates to the sixteenth 
cen tury. 62 Codex q contains excerpts from various Greek 
authors including Plato, Polybius, Xenophon, Aeschylus, 
Euripides, and Isocrates; there are also several folios (118v-121) 
written in Latin. Excerpted passages from Symp. occur on folios 
22-28. These excerpts have not previously been consulted by 
editors of Symp. 

The source for the Symp. excerpt in Bernensis 690 (r), which 
begins at 8.12 and ends at 8.39, is not clear; there is only one 
reading that possibly connects it to Laurentianus 85.9 (D): 8.37 
d58T}AOV] &81lAOV DFa.c·rt: OUK &8T}AOV rm.63 Codex r also has the 
following separative readings: 

8.16 iiya'tat] iiya8at r 8.18 'ta om. r 8.22 avoO"ta] aVEO"ta r 
8.31 nEnOtll'tat om. r 8.33 alO"xuvoV'tat] alO"xuvouV"Cat r 8.37 
EVEpaAov] avtpaAov r 8.38 "Cponata] 'tponaLOv r 8.39 Kat 
nota] nota r. 

62 For a brief description of Laurentianus Acquisti e Doni 37 see E. 
Rostagno and N. Festa, "Codici greci Laurenziani non compressi nel catalogo 
del Bandini," StIt 1 (1893) 197. 

63 Codex r sides with the tradition against the M group at 8.15; also there are 
no A. or E group readings in the section of Symp. in r. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

208 MANUSCRIPTS OF XENOPHON'S SYMPOSIUM 

Bernensis 690 (r) is a paper manuscript 62 folios in length, 
which dates to the sixteenth century.64 Codex r contains some 
orations of Demosthenes (ff. 1-51 v), followed by excerpts from 
Plato's Phaedo (ff. 54-58v) and excerpts from Xenophon's 
Symp. (ff. 58v-60v); folios 52-53v are blank; the manuscript 
ends with notes in different hands that include the Greek alpha­
bet with Arabic numerals subscripted, and notes in Greek and 
Latin concerning the Greek months. B ernensis 690, like 
Laurentianus Acquisti e Doni 37, has not been consulted 
previously by editors of Symp. 

Monacensis 495 (g) is a fifteenth-century paper codex 235 
folios in length. Codex g does not contain a complete Symp., but 
does contain the same fragments of Symp. as does codex V; 
there are no significant differences between the fragments of the 
two manuscripts. Folios 125 through 155 in codex g are identical 
in content and order to folios 49r through 83r in V. Codex V is 
the source for these works in g, which include excerpts from 
Mem., Oec., and Symp., and also a complete Hiero. 65 In 
addition, codex g contains a diverse collection of letters, 
speeches, monographs, prayers, hymns, etc. 66 Particularly 
interesting in g are works of Pletho (ff. 61-123v; 170-72) and 
other references to him. One note (f. SOy) gives the exact time 
of death of the master, but misspells his name: Jlllv. iouv. xC;' N. 
lE' E'tEAE1Yt:ll<JEV 6 <h8u<JXUAOC; 6 YOJlo<J"COC; TWEP't 8EU'tEP't, wP't 
U' 'tlic; TlJlEpUC;. Codex g also includes two odes to Pletho: one by 
Charitonymus Hermonymus (f. 213), the other by Gregory the 
monk (f. 218)-possibly both are autographs. 67 There is also a 
short ode to Sparta by Cyriaco d' Ancona (155v), an Italian 

64 For a brief description of Bernensis 690 see H. Hagen, Catalogus codicum 
Bernensium (New York 1974) 502. Omont, Zentralblatt fur Bibliothekswesen 
3 (1886) 427, also gives a brief description, but dates Bernensis 690 to the 
seventeenth century. 

65 Hiero in codex g is derived through V from Laurentianus 80.13. See 
Leverenz 17-23. 

66 For a complete description of g and its contents, see 1. Hardt, Catalogus 
codicum mss. graecorum Bibliothecae Regiae Bavaricae (=J. C. L. DeAretin, 
Cat. cod. mss. Bib. Bav. [Munich 1806-12]) IV 151-68. 

67 Folios 213-217v are enough like Marcianus gr. 206 (Harlfinger [supra n.57] 
56) to be an autograph of Charitonymus Hermonymus in a more calligraphic 
form. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

JOHN CIRIGNANO 209 

traveler who visited Pletho in Mistra in the summer of 1447 and 
again in the winter of 1448; the ode is not in Cyriaco's hand.b8 

Of at least ten hands in codex g, Charitonymus Hermonymus 
seems to me to have done the most work. Noteworthy also are 
the Attic and 'Hellenic' alphabets that a scribe was practicing (f. 
156); an apparently illiterate scribe is also evident in different 
places throughout the manuscript. 69 These oddities may indicate 
that at least a part of M onacensis 495 was a practice manuscript 
for scribes in training. 

Stemma codicum 

Having evaluated all twenty-three Symp. manuscripts, we can 
now arrange the four families in a complete stemma (see the 
illustration on the next page). A common origin for M and 'Y has 
been demonstrated for Xenophon's Hipparchicus Qackson 
182-85). The portion of Symp. in 'Y is too short to warrant the 
same observation, but given their similarity in contents it is 
likely that Symp. in M and 'Y also originated from the same 
source (9). The A. group is free of errors passed on to M and 'Y, 
and is therefore not a derivative of 9. Because 9 lacks the errors 
that A. passed on to G and E, neither can 9 derive from A. And 
because A and 9 agree against the unique readings of the a 
group, the two lost manuscripts must have shared a common 
source (£). The lost a is reconstructed where J3 agrees-against 
the other groups-with codex V, the closest of all extant 
manuscripts to the archetype. Agreement between a and £ is 
the basis upon which the text of Symp. should be built. Only in 
the six places where a and £ disagree must cp's readings be 
divined. 

68 The works of Pletho, the note giving the exact time of his death, the two 
funeral odes, and Cyriaco's ode to Sparta, all support codex g's connection to 
Mistra. Urbinas 95 (V) probably came to the Peloponnesus with John 
Eugenicus, who made several visits there. While in Mistra, codex V could then 
be partially copied by scribes of Monacensis 495 (g). See Cirignano (supra 
n.22) 50-53. 

69 See folios 5Ov, 59v, 61r, 109r, 156r, 225r+v. 



CIRIGNANO, JOHN, The Manuscripts of Xenophon's "Symposium" , Greek, Roman and 
Byzantine Studies, 34:2 (1993:Summer) p.187 

210 MANUSCRIPTS OF XENOPHON'S SYMPOSIUM 

Stemma Codicum Convivii Xenophontis 

SIGLA 

v = Vaticanus Urbinas 95, saec. xiv, xv 
D =Laurentianus 85.9, saec. xv 
F= Vindobonensis phil. 109, saec. xiv 
Z=Laurentianus 55.19 saec. xv 
Q=Ambrosianus A 157 sup., saee. xv 
R =Ambrosianus E 119 sup., saec. xv 
H2= Vindobonensis phil. 37, ff. 148v-

163, saec. xvi 
q = Lauren.tianus Acquisti e Doni 37, 

saec. XVI 

r= Bernensis 690, saee. xvi 
g=Monacensis 495, saec. xv 
G = Vindobonensis hist. 95, saec. xv 
E= Laurentianus 80.13, saec. xiv 

UNIVERSITY OF VERMONT 

January, 1994 

A =Parisinus 1643, saee. xv 
X= Leidensis Vulc. 2, sacco xv 
B=Parisinus 1645, sacco xv 
b= Britannicus 5110, sacco xv 
HI= Vindobonensis phil. 37, ff. 112v-

124v, sacco xvi 
C =Laurentianus com). sopp. 110, 

saec. xv 
P=Parisinus 2955, saec. xv 
K=Perusinus B34, sacco xv 
M =M arcianus 511, sacco xv 
0= Marcianu5 369, sacco xv 
L= Laurentianus 55.22, saec. xv 
N = Marcianus 368, sacco xv