New Testament Manuscripts

Numbers 1-500

Note: In the catalog which follows, bold type indicates a fullentry. Plain type indicates a short entry, which may occur under anothermanuscript.

Contents: 1eap and Family 1 *1r *2ap *4e *5 *6 *7e *13 and Family 13 *16 *18 *21 *22 *27 *28 *33 *35 *38 *42 *43 *59 *60 *61 *66 *69 *71 *81 *82 *83 *91 *93 *94 *104 *110 *115 *118 *124: see under 13 and Family 13 *131: see under 1 and Family 1 *138 *141 *157 *160 *162 *174 *175 *177 *179 *180 *181 *185 *189 *201 *203 *205 *206 *209: see under 1 and Family 1 *213 *223 *225 *229 *230: see under 13 and Family 13 *235 *245 *249 *251 *256: see under 365 and Family 2127 *262 *263 *265 *267 *270 *273 *280 *291 *304 *307 *314 *317 *319 *322 *323 *330 and Family 330 *346: see under 13 and Family 13 *348 *349 *356 *365 and Family 2127 *372 *383 *423 *424 *429 *430 *431 *436 *442 *443 *451 *453 and Family 453 *472 *473 *476 *477 *482 *485 *489 *495

Manuscript 1eapand Family 1

Location/Catalog Number

Basel. Catalog number: University Library A. N. IV. 2.


1 contains the entire New Testament except the Apocalypse. It is writtenon parchment, one column per page.


Usually dated paleographically to the twelfth century. (Scrivener, however,gives the date as the tenth century while noting that Burgon dated it to thetwelfth or thirteenth.) Originally contained a set of illuminations, but mostof these were extracted by 1862. One of the illustrations is thought to be ofthe Byzantine Emperor Leo VI (reigned 886-911) and his sonConstantine Porphyrogenitus (reigned 911-959), which seems to be the basis forthe tenth century date. Scrivener notes that Hebrews is the lastbook in Paul, and that as bound the gospels appear at the end of the volume.The writing style is described as "elegant and minute," and"fully furnished with breathings, accents, andι adscript. Theinitial letters are gilt, and on the first page of each gospel the fullpoint is a large gilt ball." Hatch reports, "Words writtencontinuously and without separation; accents and breathings; ruling with asharp point; letters pendents; high, middle, and low points, comma, andinterrogation point...." It has the Ammonian sections and lectionarynotes but not the Eusebian canons.

Description and Text-type

That 1 has a not-entirely-Byzantine text has been known at least since 1516,when Erasmus consulted it to compile the Textus Receptus.For the Gospels, Erasmus worked primarily from 1, 2e, and the vulgate,but he preferred the latter two as 1's text appeared to be aberrant.

In recent centuries, this "aberrant" text came to be recognizedas valuable; 1 was, for instance, one of the very few minuscules cited byTregelles, and Hort mentions it as having a relatively high number ofpre-Syrian readings. (All of this, it should be noted, applies only in thegospels; elsewhere 1 appears to be an entirely ordinary Byzantine text.)

A crucial discovery came in 1902, when Kirsopp Lake published Codex 1of the Gospels and its Allies. This work established the existence of thetextual family known as "Family 1" or "the Lake Group"(symbolized in NA26 as f1 and in earlier editions asλ; von Soden calls thegroup Iη).In addition to these basic four (1, 118, 131, 209),we now consider 205, 205abs,872 (Mark only), 884 (in part), 1582, 2193, and 2542 (in part) to be membersof the family.Within the type, 1 and 1582 form a close pair (they also seem to be the bestrepresentatives of the family). 205 goes with 209; in fact, Lake thought205 a descendent of 209; although Wisse disagrees, the only differencesbetween the two seem to be Byzantine corruptions, usually if not always in205.

The most obvious characteristic of the Lake Group is that these manuscriptsplace John 7:53-8:11 after John 21:25. In addition, 1 and 1582 contain ascholion questioning the authenticity of Mark 16:9-20.

Aland and Aland list 1 as Category III in the Gospels. Wisse lists it, not surprisingly, as a core member of Family 1, and "close to 1582." Von Soden classifies 1 as Iηa(i.e. Family 1) in the Gospels

This does not settle the question of what sort of text is found in Family 1.Here the name of B. H. Streeter is most important. Streeter, working largelyon the basis of data supplied by Lake, proposed that Family 1, along with theKoridethi Codex (Θ),Family 13, the minuscules 28, 565, 700, and the Armenian and Georgian versions,were the remnants of what he labelled the"Cæsarean Text."Streeter's theory, however, has become controversial in recent years,and cannot be discussed here. See the article onText-Types and Textual Kinship; also thevery brief mention in the entry on 13 and Family 13. It might benoted that even Streeter concedes Family 1 to be somewhat more Alexandrianthan the other "Cæsarean Text" witnesses.

In fact the relationship between Family 1 and the other "Cæsarean"witnesses is somewhat uncertain. While the other members of the type often do showsome sort of special relationship to each other, that of Family 1 to the others isslightly weaker. Streeter would define the "Cæsarean" witnesses in termsof non-Byzantine agreements. The following table shows the percentages of non-Byzantineagreements for certain leading "Cæsarean" witnesses (with B, D, andE thrown in for controls). The table is based on a set of 990 sample readings:

ΘFamily 1Family 13
Family 1125/156=80%-92/121=76%
Family 13115/145=79%92/121=76%-
28 (Mark)37/50=74%34/45=76%37/39=95%

The interpretation of these results is left as an exercise for the reader.

The following offers a brief summary of information about the various membersof Family 1:

MSDateLocationCatalog NumberSoden
Cited inComment
1XIIBaselUniversity Library A. N. IV. 2Iηa 1 core; close to 1582SQE13, Soden, Merk, Bover, Huck-Greeven Gospels, Acts, Epistles complete.
118XIIIOxfordBodl. Libr. Auct. D. infr. 2. 17 Iηb1 coreSoden, Merk, Bover, Huck-Greeven Gospels with lacunae; Matt. 1:1-6:2, Luke 13:35-14:20, 18:8-19:9, John 16:25-end from later hands. Many of the leaves are palimpsest, with 118 being the upper writing.
131XIV?RomeVatican Library Gr. 360Iη 1Soden, Merk, Bover, Huck-Greeven Gospels, Acts, Epistles complete. Dated to the eleventh century by Birch. "This copy contains many itacisms, and corrections primâ manu" (Scrivener).
XVVeniceSan Marco Library 420 (Fondo Ant. 5) Iη1; pair with 209SQE13 Old and New Testaments complete. Thought by Lake, and earlier Rinck, to be copied from 209. This is probably not true (Burgon considers 205 and 209 to be descended from the same uncial ancestor), but the two are very close. 205 was copied for Cardinal Bessarion, probably by his librarian John Rhosen.
209XIVVeniceSan Marco Library 394 (Fondo Ant. 10) Iηb1; pair with 205 SQE13, Soden, Merk, Bover, Huck-Greeven New Testament complete (gospels, acts, epistles are XIV century; r is XV century). Like 205, once belonged to Cardinal Bessarion, who used it at the Council of Florence in 1429. Many marginal notes in vermillion from the first hand. Writing style resembles 1 (Scrivener).
1582948AthosVatopediu 949Iηa 1; close to 1SQE13, Soden, Merk, Bover, Huck-Greeven Gospels complete. Evidently written by the same scribe as 1739.
2193XIηaSoden, Merk, BoverLost.

Note: Von Soden also classified 22 as a member of the Lake Group;however, Wisse considers 22 to be the head of a different group. 872 is consideredby von Soden to be part of Iηb, but Wissefinds it to be Kx. Two additional Family 1 witnesses found by Wisse,884 and 2542, are only weak and partial membersof the family. These four witnesses are therefore omitted.

Von Soden says that 1 is Ia3 in the Acts and Epistles. Alandand Aland list it as Category V outside the gospels. InActs, the Text und Textwert volume shows it with only one reading (out of 104 examined)that agrees with UBS against the Byzantine text (Acts 21:10, where it has a readingthat, although not the majority, is still common and could arise accidentally),and one other reading that agrees with neither (Acts 21:25, where it agrees exactlyonly with 226* although several other minuscules have similar texts). This raises theslight possibility that it is a little less mixed somewhere around Acts 21, but most ofits text is clearly very Byzantine.

Other Symbols Used for this Manuscript

von Soden: δ254.


Kirsopp Lake, Codex 1 of the Gospels and its Allies, Texts and Studies,volume vii, Cambridge, 1902, collates 1 with 118, 131, and 209.

Sample Plates:
Aland & Aland (1 plate)

Editions which cite:
Cited in NA26 and NA27 for the Gospels (usuallyas part of f1)
Cited, along with 205, 209, 1582, and 2542, in SQE13.
Family 1 is cited in all the UBS editions.
Cited by von Soden, Merk, and Bover.

Other Works:
Harvard Theological Review, July 1923, offers an article byR. P. Blake and K. Lake on the Koridethi Codex and related manuscripts.
B. H. Streeter, The Four Gospels: A Study of Origins (MacMillan, 1924)devotes considerable space to the relations between the various"Cæsarean" witnesses.

Manuscript 1r

Augsburg, University Library Codex I. 1.4.1. Labelled 1 in all previouscatalogs, but now renumbered 2814 in the new Aland list. Soden'sΑν20.Contains the Apocalypse only. Twelfth century. Has the Andreas commentary.Noteworthy primarily as the single Greek manuscript used by Erasmusto prepare the Apocalypse of his 1516 New Testament. It now ends(as it did in 1516) with 22:16, δαδ,forcing Erasmus to compile the remaining verses by retranslating the Vulgate.Erasmus borrowed the manuscript from Reuchlin, but it was lost for many yearsuntil rediscovered in 1861 by Delitzsch. Hort said of it, "it is byno means... of the common sort. On the one hand it has many individualismsand readings with small and evidently unimportant attestation: on the other ithas a large and good ancient element." Hort associates it with 38 [=2020].Other scholars have not placed it so high, however; the text (which often cannotbe distinguished from the commentary) seems to be fairly typical of the Andreasmanuscripts. Hodges and Farstad, following Schmid, place it in their"Me" group, a subset of the Andreas text containingsuch manuscripts as 181, 598, 2026, 2028, 2029, 2031, 2033, 2038, 2044,2052, 2054, 2056, 2057, 2059, 2060, 2065, 2068, 2069, 2081, 2083, 2186,2286, and 2302.

Manuscript 2ap

Basel, University Library A. N. IV.4. Labelled 2 in all previouscatalogs, but now renumbered 2815 in the new Aland list. Soden'sα253.Contains the Acts and Epistles complete. Generally dated tothe twelfth century, although Scrivener and Burgon list XIII/XIV.Classified as Ib1 by von Soden, but in Paul (the onlysection in which Von Soden cites it), this group (which includessuch manuscripts as 206, 429, 522, and 1891) is mostly Byzantine.That 2 is mostly Byzantine is confirmed by the Alands, who place themanuscript in Category V. Scrivenernotes that it has "short introductions to the books," butthese have no more critical value than those found in any othermanuscript. Thus the only real interest in 2 is historical; it isthe manuscript Erasmus used as the primary basis for his 1516 editionof the Acts and Epistles. (This, at least, is reported by mostexperts; Gary S. Dykes, however, claims that the TextusReceptus does not contain any of 2's distinctive readings. The Alandsin Text und Textwert note only one singular reading in Acts, at 10:12,and one reading which is entirely non-Byzantine, at 23:20; Dykes is right thatthe TR doesn't agree with 2 in either place -- although, curiously,it doesn't agree with 1 in the latter place either.)Scrivener quotes Hoskier to the effect that his (Erasmus's) bindercut off significant portions of the margin.

Manuscript 4e

Paris, National Library Greek 84. Soden'sε371.Contains the Gospels with minor mutilations (Matt. 2:9-20, John 1:49-3:11).Generally dated to the thirteenth century, although Scrivener andBurgon list the twelfth. Classified as I' by von Soden,but this group (containing among others P Q RΓ 047 064074 079 090 0106 0116 0130 0131 and a number of undistinguishedminuscules) is amorphous; most of its members are heavily if notpurely Byzantine. That 4 is mostly Byzantine seems to be confirmedby Wisse; who classifies it as Kmix/Kx/Kx.(The Alands do not assign 4 to a Category;this often means that themanuscript is heavily but not quite purely Byzantine.)In the past, Mill considered 4 to have some relationship to theLatin versions and the Complutensian Polyglot; this may, however,be simply an indication that it agreed with the Byzantine textwhere the latter differs from the Textus Receptus.The manuscript was included in the editions of Stephanus asγ'. It isdescribed as "clumsily written" and has extensive lectionaryapparatus.

Manuscript 5

Paris, National Library Greek 106. Soden'sδ453.Contains the Gospels, Acts, and Epistles complete. Usually datedto the thirteenth century; Scrivener gives twelfth century or later.In the Gospels, Soden lists it as Ak; other members of thisgroup include 15, 32, 53, 169, 225, 269, 292, 297, 416, 431, 448,470, 490, 496, 499, 534, 546, 558, 573, 715, 752, 760, 860, 902,946, 968, 976, 987, 1011, 1015, 1058, 1091, 1163, 1167, 1171,1211, 1227, 1291, 1299, 1321, 1439, 1481, 1484, 1498, 1566, 1800,2142, and 2176 -- an undistinguished group of manuscripts whichWisse generally classifies with Kx or its related groups(Wisse classifies 5 itself as Mix/Kmix/1519; seven otherAk manuscripts also go with 1519, butmany of the other manuscripts go with 1167 or have unique texts). That5 is largely Byzantine is confirmed by the Alands, who in theGospels place it in Category V.

Outside the gospels, 5 is much more interesting. The Alands promote itto Category III, and Von Soden placesit in Ia2 (along with such manuscripts as 467 489 623927 1827 1838 1873 2143). Some support for this is offered by Richards,as 623 is 5's closestrelative in his tests of the Johannine Epistles(so close that they might almost be sisters). The kinship of 5 with489 927 1827 2143, however, is not notable in Richards's lists; 5 agrees with all ofthese in the 60% range, which is fairly typical of its agreementwith Byzantine manuscripts. Richards classifies 5 and 623 as membersof his Group A3 (family 1739); even by his numbers, however,they are weak members, and should be discarded. Wachtel classified 5as a distinctly non-Byzantine (40+) manuscript, but withoutdistinguishing its kinship. The Alands in Acts find it to be closest to1101 among manuscripts with at least twenty readings; among those whichcontain the whole book or nearly, it is closest to 623 (81%), 2201 (81%),1893 (80%), 1595 (80%), and 1598 (79%). They list it with two singularreadings and 13 readings that agree neither with UBS nor the Byzantine text. In(X) of these it has nine or fewer allies. Its allies in these readings are:
Acts 21:25 -- 5 623
Acts 23:20 -- singular (but the variants here are forms of μελλοντα; I'd regard this as merely a copying error of no genetic significance)
Acts 24:15 -- 5 436 437 623 2464
Acts 18:16 -- (a special case; 5 is the only MS. with its reading, but ten others have very similar readings)
Thus, in Acts, the Aland data seems to indicate a fairly typical late Alexandrian manuscript with very few special readings; it probably is close to 623, but it is not really a distinctive text.

Scrivenernotes that it is "carefully written and full of flourishes."Colossians precedes Philippians.The manuscript was included in the editions of Stephanus asδ'. It is cited in NA28 for the Catholic Epistles,but was not cited in earlier Nestle editions.

Manuscript 6

Location/Catalog Number

Paris. Catalog number: Bibl. Nat. Gr. 112.


6 contains the entire New Testament except the Apocalypse. It is writtenon parchment, one column per page.


Dated paleographically to the thirteenth century. Scrivener writes of it,"This exquisite manuscript is written in characters so small that somepages require a glass to read them."

Description and Text-type

The quality of 6 varies in the various parts of the New Testament. Alandand Aland list 6 as Category V in the Gospelsand Acts and Category IIIin Paul and the Catholics.

Von Soden lists 6 as Ik(family Π) in the Gospels and asH in the Acts and Epistles. Wisse's classification as Π6 in Luke makes it a partof a subgroup of Πb also containing 515 and 1310. Still, this make it Byzantine overall.

In Actsit is also primarily Byzantine, although perhaps not so bad as the Alands imply --they show 14 readings which agree with neither the UBS nor the Byzantine text:
Acts 2:23 -- singular (probably a mis-copying of the Byzantine reading)
Acts 2:38 -- with 460 629 1409* 1722 (an omission of τον from the reading of ℵ A B 81 181 467 1409c 2774)
Acts 2:47-3:1 -- with 103 522 606 629c 630 641 945 1509* 1642 1739 1891 2200 2400
Acts 3:11 -- with 608 1744 2576 (but this is actually a complex variant with many sub-variants and perhaps should not be used for classification)
Acts 9:25 -- with 69 81* 468 915 935 1501 1548 1752 1874 1877 2143 2774
Acts 10:12 -- with 69 1857 2816 (but here again we have a complex variant)
Acts 10:29 -- with P74 ℵ A (C) 69 81 94 180 181 307 431 453 610 1175 1642 1678 1875 2818
Acts 10:47 -- with Ec and 32 other minuscules including 181 307 322 323 429 431 453 522 630 945 1739 1891 2200 2298
Acts 17:25 -- with P74 A 33 38 256 459 547 1319 1573 2127 2344 2544 2675 2737 2772 (and an easy mental slip from a reading supported by D Ψ and more than fifty minuscules including 181 307 436 441 453 614 623 1611 2138 2412)
Acts 15:34 -- with 87 other minuscules including 5 33 88 (181) 307 (322) 323 (326) 383 441 453 614 623 1175 (1611) 1739 1891 2298 2344 2412
Acts 23:20 -- with 85 other witnesses including ℵ2 Ψ 307 424c 436 441 483 614 629c 876 945 1175 1505 1611 1739 2138 2298 2412 2495
Acts 23:30 -- with 1409 2147 2541 2652
Acts 24:6-8 -- with 69 (and very close to 181 1875) (but this is an even more complex variant than 9:25 or 10:12)
Acts 25:5 -- with 68 other witnesses including Ψ 69 206 307 429 441 453 522 614 623c 1505 1611 1852 2138 2412 2495

It will presumably be evident that 6 isn't conspicuously close to anything (and it perhaps isn't as unusual as its high number of non-Byzantine/UBS readings would imply). Among manuscripts with which it shares at least thirty readings, the Alands list as its closest relatives 1526 (86%), 172 (85%), 1867 (82%), 483 (81%), 2080 (80%), 69 (80%), 1022 (80%), 1835 (80%), 1277 (80%). Those are pretty low agreement rates -- and, strangely, most of the higher percentages are from manuscripts which, although substantial, also have substantial gaps, which (given the Alands' too-small, biased sample) might serious bias the results. I think 6 in Acts needs more examination, but I don't see much evidence that it will really show us much.

The situation changes in Paul and the Catholic Epistles. 6 still possessesmany readings characteristic of the late phases of the Byzantine text,but it also has many distinct readings, many of which it shares with1739. Noteworthy among these are:

It will be observed that 6 shares all of these readings with 1739 -- and does notconsistently agree with anything else. Thispattern continues elsewhere; where 6 is non-Byzantine, it agrees with 1739over 90% of the time. (The connection of 1739 and 6 has been known almost sincethe discovery of the former, and recently was reaffirmed by Birdsall.)

6 also has a peculiar affinity with 424**; although these manuscriptsactually have fewer special agreements with each other than with 1739,this is because they are more Byzantine than 1739. 6 and 424** seem toform their own subgroup within family 1739 (note, e.g., their unique readingευωχιαις in Jude 12).

Other Symbols Used for this Manuscript

von Soden: δ356.Tischendorf: 6e; 6a; 6p.Cited in Stephanus as ε'.

J.N. Birdsall, A Study of MS. 1739 and its Relationship to MSS.6, 424, 1908, and M (unpublished Ph.D. dissertation, 1959)


Sample Plates:

Editions which cite:
Cited frequently in NA26 and NA27.
Cited in UBS4 for Paul.
Cited by von Soden, Merk, and Bover.

Other Works:

Manuscript 7e

Paris, National Library Greek 71. Soden'sε287.Contains the Gospels compete. Generally dated to the twelfth century;Scrivener quotes the eleventh. Classified asIφbby von Soden; other members of this group include115 179 267 659 827 and parts of 185 1082 1391 1402 1606. It is associatedwith Family 1424 (IΦa).Wisse classified 7 as "Cluster 7." This group contains 7,267 (Soden: Iφb),1651 (Soden: Kx), and1654 (Soden: Iα).Wisse describes the group as "close to Kx in Luke 1 and 10, but...quite distinct in Luke 20." The Alands do not assign 7 to aCategory; this is not inconsistent with Wisse'sclassification of the manuscript as often but not universally close to Kx.Physically, Scrivener describes 7 as having a "very full [lectionary apparatus]"and a metrical paraphrase. It is said to be "[i]n style not unlike Cod. 4,but neater." It is Stephanus's ς'.

Manuscript 13 and Family 13

Location/Catalog Number

Paris. Catalog number: Bibl. Natl. Gr. 50.


13 contains the Gospels with lacunae: lacking Matthew 1:1-2:20(two leaves lost), 26:33-52 (one leaf), 27:26-28:9 (two leaves),Mark 1:20-45 (one leaf), John 16:19-17:11 (one leaf), 21:2-end.It is written on parchment, two columns per page.


Dated paleographically to the thirteenth century. Scrivener says of itsappearance simply "it is not correctly written." Griesbach calledit "inelegant," and Ferrer/Abbott admit that there are numerousorthographic errors (though not exceptionally numerous). They say that the number of itacismsis especially noteworthy (the introduction, p. xii, counts 1523!). On page xiii,they add that the scribe "was somewhat prone to the omission of words,from carelessnes, sometimes even a part of a word being dropped...."

Description and Text-type

It was W. H. Ferrar who first brought widespread attention to 13. In aposthumous work published by T. K. Abbott in 1877, he pointed out therelationship between 13, 69, 124, and 346. For this reason, the groupFamily 13 (f13) is often called the Ferrar Group (symbolizedφ; von Soden calls thegroup Iι).

The most obvious characteristic of the Ferrar Group is that these manuscriptsplace John 7:53-8:11 after Luke 21:38.

Since the time of Ferrar, many more manuscripts have been added to theFerrar Group. The list as given in Nestle-Aland consists of 13,69, 124, 174, 230, 346,543,788, 826, 828, 983, 1689, and 1709. Von Soden broke the groupup into three subgroups, the a subgroup containing 983 and 1689; the b subgroupconsisting of 69, (124), 174, and 788; and the c subgroup containing 13, 230,346, 543, 826, and 828.

The Lakes offered a similar scheme (with slightly different nomenclature,essentially reversing the names of the a and c groups). In Colwell's opinion,this means that Family 13is not a true "family"; it is a "tribe" within which theLakes' Group a is a family. The Lakes' groups are as follows:

Wisse makes various adjustments to von Soden'slist, associating 174 and 230 with the uncialΛ rather than withFamily 13, describing 983 as "weak" in Luke 1, and listing 124 as"weak" in all chapters profiled. Wisse denies the existence ofsubgroups (p. 106 -- although we should note that conclusions based on justthree chapters don't prove much), and claims that either543 or 828 can represent the group as a whole. The studies of Geerlings, and theunpublished work of Geoffrey Farthing, also indicate that 826 stands near thecenter of the group.

It is widely believed that the Ferrar group is derived from a lost uncialancestor once located in southern Italy or Sicily (possibly Calabria; see, e.g., thenotes on 124, 174, and 346).

In the decades after the Ferrar Group was discovered, it was found to havecertain textual affinities with the Lake Group, the Koridethi Codex, anda handful of other minuscules. In 1924, B. H. Streeter suggested that thetwo groups, plus the Koridethi Codex, the minuscules 28, 565, and 700,and the Armenian and Georgian versions, were the remnants of a"Cæsarean" text-type.

In the following decades, the "Cæsarean" type was furthersubdivided. Ayuso, for instance, split it into a "pre-Cæsarean"group, containing P45 W (Mark) f1 f13 28,and the "Cæsarean" text proper, consisting of Θ 565 700 OrigenEusebius and the early forms of the Armenian, Georgian, and Syriac versions.

This was, in fact, the first step toward what appears to be anunraveling of the "Cæsarean" text. Hurtado has shown,for instance, that P45 and W are not as close to the other"Cæsarean" witnesses as Streeter and Kenyon claimed.(It should be noted, however, that Hurtado at no point addressesStreeter's definition of the "Cæsarean" text;only his own. For a comparison of the non-Byzantine readings of Family13 with those of other "Cæsarean" witnesses, see the item on1eap and Family 1.)

For whatever value the information may have, Aland and Aland (who are notenthusiastic about the "Cæsarean" text) rate 13 (and mostof the other members of its type) as Category III.The classifications of von Soden and Wisse have, of course, already been covered.

The following offers a brief summary of information about the various membersof Family 13:

MSDateLocationCatalog NumberSoden
Cited inComment
13XIIIParisNat. Libr. 50Iιc13aSQE13, Soden, Merk, Bover, Huck-GreevenGospels with several lacunae. Said to be "not correctly written."
69XVLeicesterRecords Office 6 DIιb13bSQE13, Soden, Merk, Bover, Huck-GreevenNew Testament with lacunae. Lacks Matt. 1:1-18:15. Rapidly and poorly written on bad materials. See separate entry
124XIViennaAustrian Nat. Libr. Theol. Gr. 188Iιbweak 13 bSoden, Merk, Bover, Huck-GreevenGospels. Missing Like 23:31-24:28. Written on parchment, one column per page. Ferrer/Abbott considers it to be more carefully written than the other manuscripts of the family they studied (p. xvii). Scrivener reports, "The manuscript was written in Calabria, where it belonged to a certain Leo, and was brought to Vienna probably in 1564."
1741052RomeVatican Libr. Gr. 2002Iιb Λ Soden, Merk, Bover, Huck-GreevenGospels with several lacunae, including John 8:47-end. Written by a monk named Constantine, and associated with "Georgilas dux Calabriae."
2301013?EscorialGr. 328 (Y. III. 5)Iιc Λ Soden, Merk, Bover, Huck-GreevenGospels complete, written by a monk/priest named Luke (who miscalculated or miswrote the indiction)
346XIIMilanAmbrosian Libr. S. 23 supIιc13 corea SQE13, Soden, Merk, Bover, Huck-GreevenGospels. Missing John 3:26-7:52. Bought in 1606 in Gallipoli. That is not, note, Gallipoli in Turkey, the site of the World War I battle, but the one in Apulia in Italy. Dr. Ceriani of the Ambrosian Library, who of course had much occasion to examine the volume, thought it was written in Calabria. It is written in a single column of 13 lines per page; some of the writing is badly faded. Ferrer and Abbott, p. xv, report that it shares very many of the scribal peculiarities of 13. It has a peculiar tendency to split compound verbs into a preposition and a simple verb.
543 XIIAnn ArborUniv. of Mich. MS. 13Iιc13 corea SQE13, Soden, Merk, Bover, Huck-GreevenGospels with several lacunae. Scrivener's 556
788XIAthensNat. Libr. 74Iιb13 coreb SQE13, Soden, Merk, Bover, Huck-GreevenGospels. Missing John 21:20-end
826XIIGrottaferratadella badia Libr. A a 3 Iιc13 coreaSoden, Merk, Bover, Huck-GreevenGospels complete. "A beautiful codex: written probably at Rhegium" (Scrivener)
828XIIGrottaferratadella badia Libr. A a 5 Iιc13aSoden, Merk, Bover, Huck-GreevenGospels complete
983XIIAthosEsphigmenu 31Iιa13c SQE13, Soden, Merk, Bover, Huck-GreevenGospels. Missing John 11:34-19:9
16891200?IιaSoden, Merk, Bover, Huck-Greeven Gospels (complete?). Lost.
Sample plate in Edward Maunde Thompson,An Introduction to Greek and Latin Paleography (plate 73).
1709XIITiranaStaatsarchiv Koder-Trapp 15 fol. 141-194Kx (John only)

Other Symbols Used for this Manuscript

von Soden: ε368.


W. H. Ferrar and T. K Abbott, Collation of Four Important Manuscriptsof the Gospels by the late William Hugh Ferrar, 1877, collated 13, 69, 124,and 346, establishing the Ferrar Group.

Sample Plates:
Aland & Aland (1 plate)

Editions which cite:
Family 13 is cited in NA26 and NA27 for the Gospels
Cited, along with 69, 346, 543, 788, and 983, in SQE13.
Family 13 is cited in all the UBS editions.
Cited by von Soden, Merk, and Bover.

Other Works:
B. H. Streeter, The Four Gospels: A Study of Origins (MacMillan, 1924)devotes considerable space to the relations between the various"Cæsarean" witnesses.
Kirsopp Lake & Silva Lake, Family 13 (The Ferrar Group): The Text Accordingto Mark, Studies & Documents 11, 1941
Jacob Geerlings, Family 13 -- The Ferrar Group: The Text According to Matthew,Studies & Documents 19, 1961
Jacob Geerlings, Family 13 -- The Ferrar Group: The Text According to Luke,Studies & Documents 20, 1961
Jacob Geerlings, Family 13 -- The Ferrar Group: The Text According to John,Studies & Documents 21, 1962(It should be noted that the Geerlings volumes suffer from significant methodologicalproblems.)
E. C. Colwell, "Genealogical Method: Its Achievements and its Limitations,"1947, reprinted in Studies in Methodology in Textual Criticism of the NewTestament, New Testament Tools and Studies IX, 1969, summarizes an attempt toapply Quentin's "Rule of Iron" to Family 13.
E. C. Colwell, "Method in Grouping New Testament Manuscripts,"1958, reprinted in Studies in Methodology in Textual Criticism of the NewTestament, New Testament Tools and Studies IX, 1969, illustrates thevarious sorts of textual groupings based on Family 13 among others.
Larry W. Hurtado, Text-Critical Methodology and the Pre-CaesareanText: Codex W in the Gospel of Mark, Studies and Documents 43, 1981

Manuscript 16

Paris, National Library Greek 54. Soden's ε449.The manuscript was once owned by the Medicis; it was acquired by the French NationalLibrary in 1599 along with the rest of Catherine de Medici's collection. Before that,it had been in the collection of Cardinal Niccolò Ridolfi, who may have brought it tothe west. (His number 34 may still be seen in the MS.)Its 364 folios contain the Gospels with minor mutilations (Mark 16:6-20 arelost and the manuscript was "never quite finished" -- hardlysurprising given the complexity of the copying process, as we will seebelow. The Ammonian Sections, for instance, are supplied only inMatthew and Mark, though the lectionary apparatus extends farther.). It hasa Latin parallel, but this is much less complete than the Greek, and they are notprinted in sense lines or synchronized; the Latin, which seems to be basedon the Vulgate but with many differences, lacks much of the second half of Markand almost all of Luke and John. (The beginning of John was eventually added,but by a presumably-later scribe who didn't try to match the colors of theGreek text.) Even in the parts of the Latin that are otherwise complete, thedecorated initials are often lacking. It is ironic, then, that the only Eusebian tablesare in Latin (with space left for a Greek parallel which was never supplied); thetables were a later addition to the text. In the Greek, the writing wascontinuous, with the scribe changing inks when it came time to writein another color. But it is clear, from the way the Latin was written, that thecolors in that part of the text were written at different times --that is, the text in each color was written in order (in the parts wheremultiple colors are used, obviously), and when the scribe finished with onecolor, he set that ink aside, moved back to the top of the section,and started writing in the nextcolor. (It appears the paint colors were also done this way in at leastsome of the illustrations.) What is more, there are many errorsin the Latin ink colors.
Artistic quality seem to have been given more weight than textual; when theLatin is too short to match the Greek, the text is filled out with minims (mmmmm orthe like) rather than simply left blank or using fewer words on the line.
Interestingly, there are relatively few captions on the illustrations -- perhapsa side effect of the fact that few at this time read both Greek and Latin, sohow would one choose which one to use for a caption?
The format is fairly large, although not large enough to qualify as anAtlantic Bible; it is 335 by 250 mm.The parchment is very fine, and the margins large; the text is unusuallyeasy to read. The quires are generally of ten folios rather than the usual eight.It was set up so that each synoptic gospel begins at the start of a new quire, and --most unusually -- both Mark and Luke are preceded by bifolios with the evangelistportraits on the second recto, meaning that it does not face the gospels. This mayalso have been the format of Matthew, since that portrait was a separate insertion.This implies (to me, at least) that this is another instance of the manuscript beingincomplete -- in addition to the missing Latin text, themissing portraits, and the missing canon tables, there was some other featureintended for those bifolia which was never even begun.
The text was classified asIβb by von Soden;other members of this group include 1216 1579 1588. Von Sodenconsidered this group to be weaker thanIαb(348 477 1279), but in fact both groups are largely Byzantine.Wisse, in evaluating 16, assigns it to its own group. Of this"Group 16" he remarks, "This group consists mainlyof MSS. classified by von Soden as the weak group ofIβ.However, the group is not simply a weakened form of Gr. 1216[=152 184 348 477 513(part) 555 752 829 977 1216 1243 12791579 2174 2726], though it stands closer to Kx. Ifthere is a relationship between Grs 16 and 1216 in Luke, it is a ratherdistant one." Other members of Group 16 include 119 217 330491 578(part) 693 1528 (which Wisse pairs with 16) 1588. In Matthew,the Alands link it with61 152 184 348 555 829 1148 1216 1279 1528 1579 2394 2726 -- althoughthey have only seven group readings, meaning that they don'thave enough data to group manuscripts (and the only one to agree with16 in all of them is 1528. 61 184 555 1579 agree in five of seven; othersagree less than that).In Mark, where the Alands have 26 group readings, the Alands list61 152 184 348 829 1216 1243 1279 1528 1579 2174 2726(i.e. adding 1243 and 2174 but dropping 555 1148 2394) -- although none of theseagree with 16 in more than 21 of 26 readings (that being the figure for 1279 and 1528),and 184 and 348 agree in only 16 of 26.In Luke, the Alands do not find 16's text un-Byzantine enough to seek itsrelatives, although given their small and biased sample size, this doesn't mean much.In John, they grouped it with 182 and 1528, but again, the sample of specialreadings is very small. Still, it is noteworthy that both Wisse and the Alandslink 16 firmly with 1528.
Interestingly, none of the manuscripts grouped with 16 is bilingual.
We can now go well beyond pairing 1528 and 16. James Dowden pointed out to meKathleen Maxwell's 2014 book on this manuscript, Between Constantinople and Rome, which offers proof that 16 is a copy of 1528(Princeton University Library, Garrett 3). Their texts, as we have alreadyseen, are very close (including,e.g., a number of long omissions not found in most Byzantine texts -- although thereis a curious and rare addition in 16's text of Matthew 10:29, του πατρος η μωνη του εντοις ανθρωποις, not found in 1528), but thestrongest single argument for kinship is that there are marginal markingsin 1528 corresponding to where spaces were left for illustrations in 16. Thesemarks are red crosses -- some now much faded or rubbed, but it appears that almostevery drawing in 16 corresponds to a mark in 1528,(Maxwell, p. 5, 80-81, suggests that the lead scribe of 16 added these markingsin 1528 to tell the copyists of 16 where they would be insertingdrawings.)Formally, this probably argues that 16 should be renumbered as 1528abs,but given the elaborate work that went into it, and the bilingual text,I doubt this will happen.
Despite Wisse's comments, Group 16 is much more Byzantine thananything else, though the Alands do not place 16 in anyCategory.
Much more interesting than 16's actual text is the appearance of thetext. Scrivener calls it "gorgeous and 'right royal,'" andthe reason is not hard to see, for the manuscript is written in fourcolours (as well as being illustrated). With the exception of a short sectionthat uses gold ink, narrative is copied in vermillion;the words of Jesus and of angels, along with the genealogy of Jesus, arein crimson; blue is used for Old Testament quotations (except that words ofGod in those quotations are in crimson) and for the speechesof those who might be regarded as sympathetic to Christianity: the disciples,Mary, Zechariah, Elizabeth, Simeon, John the Baptist; the words of evildoers(Pharisees, Judas, the Devil; also the mob) are in black (actually darkbrown), as are the words ofthe centurion and the shepherds (it is possible that these are by mistake, orperhaps a change of plans; Maxwell thinks the lead scribe changed his approachas the work proceeded. Certainly this happened at the beginning, where the firstfew sections of text written in blue were written over erasures of red; clearlythe idea of writing them in blue was not in the scribe's mind at first).The colors were very carefully executed; Maxwell counts only two significanterrors in their assignment in the first 150 folia.[For the the description of the colors, see the note at the end.]
Gregory believes that an Armenian had a hand in 16's preparation, as it hasArmenian as well as Greek quire numbers, and the Armenian numbering is more complete than the Greek.
The appearance of the manuscript is striking (there are many photographs,mostly black and white but some in color, in Maxwell). The Greek is a flowingcursive, obviously late, in a single wide column. The Latin column is much narrower,using an equally late script, but much darker, clearly written with a differentsort of pen. Even more curious is the fact that it is written by different scribeswith different writing styles. Maxwell shows a page (folio 121r) whereboth exist on the same page, but the differences are so dramatic that I saw themsimply glancing at pages in the two hands. My first thought is that the earlierwas thirteenth century and the later at least a century more recent.
Also noteworthy is that the colors of the two texts are not alwayscoordinated; much of the Latin text in Matthew matches the color of the Greek,but the opening of John is in red in the Greek and black in the Latin.
Space was left for 52 drawings in the text, plus portraits of theevangelists, but only 22 were completed and five more started but notfinished. In several cases the initial sketch was complete but the paintingonly partially done. It is widely believed that the illustrations are basedon those in Athos, Iviron 5 (Gregory/Aland 990) -- all of 16's paintingswhich also appear in 990 are found at exactly the same place in bothdocuments; what's more, in at least one case (the portrait of John), itappears that the portrait in 16 was intended to be a precise scaling-up of thatin 990 (since 16 used a larger page) -- but that the various parts were scaledunevenly, resulting in some peculiarities in the final image (Maxwell, p. 107).Several other manuscripts have illustrations associated with this school, butsince they are not textual relatives, I doubt this matters much.Curiously, although space was set aside in 16 for new illustrations (in addition tothose found in 990) in all four gospels,the plan seems to have been to add far more to Luke than the other gospels.Maxwell, p. 103, suggests that the added portraits were associated with thegreat festivals in the lectionary, which strikes me as not improbablealthough Maxwell's knowledge of the lectionary is based on rather inadequatesources. The added pictures have a strong tendency to involve Peter (Maxwell,p. 143, etc.) -- which makes sense for a Byzantine manuscript being made toappeal to the Roman church which associated itself with Peter.
According to Maxwell,there were three artists involved, and that A and B, who did the Evangelistportraits and the illustrations for Matthew and Mark, based their work on990, but that C, who did Luke and John, was more independent. (I will admitthat I'm not entirely convinced that A and B are different artists; Maxwell'sargument is based in part on thepalette of colors used -- but the main difference is in the blues, and bluepaints and pigments were rare andexpensive, so there may have been a problem with supply of some of the colors.Artist C's style is so distinct that I don't think there is much doubt thatthese paintings are by a different artist.) Maxwell believes (p. 37)that the extant drawings were made before the Latin text (but obviously afterthe Greek). Curiously, they do not seem to have been executed in order, even in Lukewhere only one artist was involved; in that book, excluding the portrait of theevangelist, #1 is finished, #2 and #3 started, #4, #6, and #7 complete,#8 and #9 started, #10 complete, #11 started, and the other ten not even begun.Maxwell notesthat the illustrations in 16 were not allotted the same format as those in 990(those in 16 are often short, very wide rectangles; those in 990, althoughalso wider than they are tall, are more square -- they approach the "goldenratio" that is widely regarded as the most desirable shape for a drawing.This change of dimensions often caused difficulties for the artists who workedon 16, because they had to somehow make their drawings fit in a strange shape).What's more, many of 16's illustrations have no parallel in 990. Maxwell, p. 119,calls artists A and B inept in their copying of the originals in 990 (while pointingout that it makes it easier to show that the manuscripts are related).
The manuscript is dated by the Kurzgefasste Liste to the fourteenth century,but as I noted above, the Latin looks more like thirteenth century writingto me, and several other art historians apparently also preferred a thirteenthcentury date. Maxwell, p. 25, speaking of the Greek "pearl" script style,says that it is a revival script based on an eleventh century style, with onlya few later elements. Manuscripts of this era are hard to date anyway, andan archaizing script would obviously make it even harder! Two scribes wereinvolved in the Greek text, but the primary scribe seems to have writtenalmost the entire text except for quire 23. Several writers have associatedthe scribe with a copy of the Octateuch, Athos, Vatopedi 602 -- another ornatemanuscript attractive enough to have been published in facsimile, but one withno textual significance that I've ever heard of.
Maxwell, p. 145, believes on the basis of elements other than the formof the writing that the manuscript must date from c. 1275-1295. Onpp. 8-9 she suggests that the manuscript was prepared duringthe Byzantine Emperor Michael VIII Paleologos's attempt to promote a reunionof the Greek and Latin churches, which ended with his death in 1282. (Michael,having forced out the Latins who had sacked Constantinople in the Fourth Crusade,wanted desperately to makesure they didn't come back, and thought that reuniting the churches would denythe Latins any excuse for attacking him. But his people hated the idea so muchthat when he died, the Orthodox church wouldn't even give him Christian burial.)Thus a strong case can be made that the manuscript was written in the thirteenthcentury -- and the reason that it is incomplete is that it was not finishedwhen Michael died. No one would want to finish the Latin side once the wholeidea of the union was dead! This would also explain the seeming attempt, later on,to restart the project (when the Latin text of the Gospel of John was started butnot carried on very far): Michael's successor Andronikos II briefly flirted with theLatin church in 1311 and the 1320s. This would explain why a different scribe didthe work and did not follow the color scheme of the original writing.
(Maxwell does allow other possibilities; one that she mentions on pp. 202-204 isthat it was made for a Catholic princess who married a Byzantine ruler or otherhigh lord for political reasons: Such a princess would likely have her own priest,or at least confessor, and a bilingual manuscript might make conversion, or at leastcommunication, easier. I must say that this idea strikes me as quite likely, giventhe number of such marriages.)
Maxwell, p. 29, suggests that the Latin scribes had little familiaritywith Greek, which helps explain the poor alignment between the two texts.
A full reproduction of the illustrations in the manuscript, although inblack and white, can be found in Henri Omont's Miniatures des plus anciensmanuscrits grecs de la Bibliothèque Nationale do VIe au XIVe siècle (1929);Maxwell has reproductions of many but not all, many of them in color.
[Note: Scrivener reported that 16 was written in "vermillion,lake, blue and black." This is very bad terminology -- lake is neithera color nor an ink. Indeed, it isn't even actually a pigment; lakeis a substrate into which pigments are added; it stabilizes them and changesthe way they reflect. But lakes themselves are essentially neutral in hue,although they are often used with red shades. But they are used inpainting, not writing, and had their greatest popularity after the manuscriptera anyway. So Scrivener's description really should not be followed.]
[Additional footnote: I have cited Maxwell's book repeatedly, but unlessyou are interested in the history of Byzantine art, it isn't much use. In readingthe early pages, I kept wondering why she never cited anyone I'd ever heard of --and why she never cited manuscripts by Gregory numbers. The section on the textof 16 does cite a few familiar sources, and does use Gregory numbers, but it isclear that her knowledge of textual criticism is limited and secondary, andnot particularly helpful; on p. 71, she admits that she had not studied textualcriticism before, and she got into it as part of her other research. Her field isart and art history, and it shows. The bookincludes neither a diplomatic edition of the text of 16 nor a proper collation; shecompares the text with 990 and 1528 for Matthew, and cites the "Nestle-Aland"text -- without even saying which edition! (if you dig through the footnotesenough, it appears it was the 27th edition, so she means the standard UBS3 text) --but unless you already know the text of either 990 or 1528, this non-collationis no help for determining the text of 16. Ifrankly don't see why she bothered.]

Manuscript 18

Paris, National Library Greek 47. Soden'sδ411;Tischendorf/Scrivener 18e, 113a,132p, 51r.Contains the New Testament complete. Dated by a colophonto 1364. Textually it is not noteworthy; the Alands list itas Category V (Byzantine)throughout. This agrees with Von Soden, who lists it asKr, and Wisse, who also describes it as Krin Luke. In Acts, it is one of the most thoroughly Byzantine of allmanuscripts of Acts, according to the Alands' data, with 98 Byzantine readings,four readings that agree with UBS against the Byzantine text, two singular readings(5:24 and 12:25), and no other readings at all which go against the Byzantinetext. Wachtel lists it as Kr in the Catholics.In the Apocalypse, Schmid places it in the dominant or "a"group of the Byzantine text headed by 046. In Merk's apparatus,it is part of the K1 group, most closely associatedwith 1835 2039 2138 2200. According to Scrivener, the manuscripthas two synaxaria between the Pauline Epistleas andApocalypse, and otherwise full lectionary equipment, but (typicallyof Kr manuscripts) does not have the Eusebianapparatus. It was written at Constantinople.

Manuscript 21

Paris, National Library Greek 68. Soden'sε286.Contains the Gospels with slight mutilations.Dated paleographically to the twelfth century (so Aland;Scrivener says tenth). Classified asIαby Von Soden -- that is, he regarded it as a mainstream"Western" or "Cæsarean" witness.More recent studies have not supported this classification. Wissefinds the manuscript to be Kx, and the Alandsaffirm this by placing 21 in CategoryV. The manuscript has pictures and most of the usualmarginalia; the synaxarion was added by a later hand.

Manuscript 22

Paris, National Library Greek 72. Soden'sε288.Contains the Gospels with some mutilations (lacking Matt. 1:1-2:24:20-5:25, John 14:22-16:27) and dislocated leaves.Dated paleographically to the twelfth century (so Aland, Gregory,Von Soden; Scrivener and Scholz preferred the eleventh). Classified asIηb,by Von Soden. Iηis what we now refer to as family 1;the b group contains the poorer witnesses to the type (118 131 209 872).This view has received partial -- but only partial -- supportfrom later scholars; Sanders (who published a "New Collationof Codex 22" in Journal of Biblical Studies xxxiii, p. 91)noted that Von Soden's collation is inaccurate, but in generalsupported the classification, and Streeter, whilehe believed 22 to be "Cæsarean," was not certainit was part of Family 1. The manuscript has a comment about theauthenticity of Mark 16:9-20; it is somewhat similar to, butdistinctly shorter than, that in 1. The Alands do not place 22 in anyCategory, implying that theydo not regard it as purely Byzantine but also do not regard itas a member of Family 1 or any other noteworthy type. Wisse'sconclusion is more interesting; he makes it a core member ofthe b subgroup of Group 22. Wisse does not analyse the natureof Group 22, but lists 660, 697, 791, 924, 1005, 1278,1365, 2372, and 2670 (part) as members of 22a while listing22, 134, 149, 351 (part), 1192, and 1210 as members of 22b. He alsolists some seemingly related groupings. Describing 22 itself,Scrivener reports that it is a "beautiful copy, singularly freefrom itacisms and errors from homoeoteleuton, and verycarefully accentuated, with slight illuminated headings to thegospels." The Eusebian apparatus is incomplete, andit lacks lectionary equipment.

Manuscript 27

Paris, National Library Greek 115. Soden'sε1023.Contains the Gospels with slight mutilations; in addition,the text has been lost from John 18:3, being replacedby a supplement (on paper) which Scrivener dates to thefourteenth century. The main run of the text is dated paleographicallyto the tenth century (so Gregory and Aland;Scrivener says the eleventh). Classified by von Soden asIφr;this is part of the amorphous group containing also Family1424(Iφa)as well as the groups headed by 7 and1010.This classification is largely affirmed by Wisse, who lists 27 as a memberof M27 (Wisse lists two basic M groups, M27 and M1386, along with a numberof subgroups). Wisse lists M, 27, 71, 248 (part), 447 (part), 518, 569,692, 750, 830 (part), 1914 (part), 1032 (part), 1170, 1222, 1228 (part),1413, 1415, 1458, 1626, 1663 (part), and 2705 as members of M27. (Note thatfew of the members of Soden's otherIφ groupsgo here; Von Soden's Iφr,corresponding to Wisse's M groups, stands distinct).It should be noted that the M groups are still Byzantine; the Alands place27 in Category V.Physically, 27 has pictures and most of the usualmarginalia including the Eusebian apparatus; the lectionary tables wereadded later, and Scrivener reports that it has been heavily corrected.

Manuscript 28

Location/Catalog Number

Paris. Catalog number: Bibl. Nat. Gr. 379.


28 contains the gospels with lacunae (missing Matt. 7:19-9:22, 14:33-16:10, 26:70-27:48,Luke 20:19-22:46, John 12:40-13:1; 15:24-16:12, 18:16-28, 20:19-21:4, 21:19-end).John 19:11-20:20, 21:5-18 are from a later hand.


Dated paleographically to the eleventh century (the added leaves arefrom the fifteenth century). 28 is written on parchment,one column per page. Scrivener says it was "most carelessly written byan ignorant scribe;" Streeter too calls the writer "ill-educated."Hatch comments, "Words written continuously without separation; accentsand breathings; ruling with a sharp point, letters pendent; high and middle points,comma, colon (:), and interogation point; initials red; initials at the beginning ofbooks ornamented with red, blue, green, and brown...." It has a synaxarion,but the lectionary indications are from a later hand. The Eusebian apparatusappears original.

Description and Text-type

Von Soden classifies 28 as Iα -- i.e. among theprimary "Western/Cæsarean" witnesses. However, Alandand Aland remark that it is "CategoryIII in Mark only; elsewhereV." Wisse generally agrees; although he labels 28 "mixed"in Luke 1, he puts it with Kx in Luke 10 and 20.

There is little doubt that most of 28's non-Byzantine readings are inMark (there are a few in John); in my sample of 889 test readings for which 28 exists,only 150 are non-Byzantine, and 92 of these are in Mark.

But what is this relatively non-Byzantine text of Mark? Streeter proposedthat it was "Cæsarean;" Ayuso further classified it as"pre-Cæsarean" (along with P45 W (Mark) f1f13). The "Cæsarean;" text has, however, come undersevere attack in recent decades (though the crucial study, that of Hurtado,does not cite 28). Therefore it is perhaps useful to cite the agreementrates of 28 -- in both overall and non-Byzantine agreements -- for Mark(the data set is the same as that cited above. In Mark, 28 exists for 211of the readings).


I would draw attention particularly to all three rates of agreement with f13,and also to the rate of near-singular agreements with 565. Whatever the type iscalled, there does appear to be kinship here. On the face of it, a common ancestral typebetween 23 and f13 seems nearly certain. Whether this was related toΘ, etc. is less clear,though the data does seem to lean that way.

Other Symbols Used for this Manuscript

von Soden: ε168.


Kirsopp Lake & Silva Lake, Family 13 (The Ferrar Group): The Text Accordingto Mark, Studies & Documents 11, 1941 (Mark only)

Sample Plates:
Hatch (1 plate)

Editions which cite:
Cited in NA26 for all four gospels, but in NA27only for Mark.
Cited by Von Soden, Merk, and Bover for the gospels.

Other Works:
B. H. Streeter, The Four Gospels: A Study of Origins (MacMillan, 1924)devotes considerable space to the relations between the various"Cæsarean" witnesses (though 28 receives relatively littleattention).

Manuscript 33

Location/Catalog Number

Paris. Catalog number: Bibl. Nat. Gr. 14.


33 originally contained the entire New Testament except the Apocalypse(as well as the LXX prophets, not including Daniel).Mark 9:31-11:11, 13:11-14:60, Luke 21:38-23:26 have been lost. In addition,the manuscript has suffered severely from damp; Tregelles said that, of all themanuscripts he collated (presumably excluding palimpsests), it was the hardestto read. The damage is worst in Acts, where some readings must be determinedby reading the offprint on the facing page. (In the Text und Textwert volumeon Acts, the Alands list eighteen out of 104 sample readings -- a sixth of the total! --as illegible.) In addition, Luke 13:7-19:44are on damaged leaves and contain significant lacunae. 33 is written on parchment,one column per page.


Dated paleographically to the ninth century (so Omont, Von Soden, Aland; Scrivenersuggests the eleventh, while Gregory thought the prophets and gospels to comefrom the ninth century and the rest from the tenth). Several scribes seemto have been involved; Von Soden suggests that one wrote the Prophets andGospels, another the Acts, Catholic Epistles, and Romans, and a third theremainder of Paul. Hatch supports this conclusion. The text supports thisopinion in part; the manuscript changes type dramatically between Romans and1 Corinthians. Hatch notes, "Words written continuously withoutseparation; accents and breathings; ruling with a sharp point; letters pendent;high, middle, and low points and comma; initials brown... O.T. quotationssometimes indicated; numbers and titles of chapters; no Ammonian sections orEusebian canons...." The Gospels have superscriptions and subscriptions;the Acts and Epistles have superscriptions but only occasional subscriptions and no στιχοι.

Description and Text-type

33 was christened "the queen of the cursives" in the nineteenthcentury. At that time, it was without doubt the most Alexandrian minusculetext of the New Testament. Today its title as "best minuscule"may perhaps have been usurped for individual sections (892 is perhaps slightlymore Alexandrian in the Gospels; 81 and 1175 rival it in Acts; in the Epistles,1739 is at least as good and more interesting). But overall there is nominuscule with such a good text over so many books.

In the Gospels, 33 is mostly Alexandrian, of a late type, with a heavyByzantine mixture (the extent of which varies from section to section).Wieland Willker, following a detailed analysis, is of the opinion thatit has most of the major Byzantine variants but few of the minor, whichhe believes means that it an ancestor started with an Alexandrian text butwas corrected very casually toward the Byzantine text (the corrector changingonly those readings he noticed on casual inspection to be incorrect). Thismatches my own unstatistical impression.

In Acts, it is Alexandrian, though with a significant mixture of Byzantinereadings. It appears closer to A than to ℵor B. It is very close to 2344; the two almost certainly have a common ancestor.(Based on Text und Textwert, among manuscripts which exist for thirty or morereadings in common with 33, 2344 agrees with 33 82% of the time, with 2718 agreeing62% of the time and nothing else agreeing as much as 60% of the time!)One might almost suspect 33 of being the ancestor of 2344 if it weren't fortheir differences elsewhere. It is noteworthy that the Alands list 33 as agreeingwith the UBS text 49 times, and with the Byzantine text 38 times (including 17 placeswhere it agrees with both) -- but as having twenty readings not found in either.This is an extremely high rate (B, for comparison, has 11 readings not found inUBS or Byz, ℵ 17, 1739 17), indicating a somewhat unusual text.

In Paul the manuscript falls into two parts. Romans, which is not in thesame hand as the other books, is mostly Byzantine; here again, Davies believes it to be akin to2344. Elsewhere in Paul, 33 is purely Alexandrian, with almost no Byzantineinfluence. It is, in fact, the closest relative ofℵ, agreeing withthat manuscript even more than A does.

In the Catholics, 33 is again purely Alexandrian; here it aligns mostclosely with A. These two are the main representatives of the main phaseof the Alexandrian text, which also includes (in more dilute form) 81,436, Ψ, bo, etc.

Von Soden lists 33 as H. Wisse lists it as Group B ("weak in [chapter]1"). Aland and Aland list 33 as Category IIin the Gospels and Category I elsewhere.

Other Symbols Used for this Manuscript

von Soden: δ48.Tischendorf: 33e; 13a;17p


Frequently collated in the nineteenth century (e.g. by Grisbach, Scholz,Tregelles); given the state of the manuscript, there is a real need for amodern collation using present-day resources.

Sample Plates:
Aland & Aland (1 page -- but this is of the ending of Romans, which is from the less valuable hand)
Hatch (1 page)
Facsmile in Scrivener

Editions which cite:
Cited in all critical editions since Von Soden, and frequently in Tischendorf.

Other Works:
M. Davies, The Text of the Pauline Epistles in MS. 2344 (Studies& Documents 38, 1968) briefly discusses the relationship of 33 with2344.

Manuscript 35

Paris, National Library Coislin Greek 199. Soden'sδ309;Tischendorf/Scrivener35e, 14a, 18p, 17r.Dated paleographically to the eleventh century.Contains the entire New Testament, without lacunae but withfairly heavy corrections. Von Soden classifies it as Kr in theGospels (based probably on the marginalia), and Wisse confirms thatit belongs to this group. Wisse places it (or, more specifically,the first hand) in subgroup 35 along with 141, 170, 204, 394, 402, 516c,521, 553, 660c, 758*, 769, 797, 928, 1250, 1482, 1487, 1493, 1559,1572, 1600, 1694*, 2204, 2261, 2554. (It is slightly peculiar to notethat Wisse attributes the Kr recension to the twelfthcentury while accepting the eleventh century date for 35.) In theActs and Epistles, Von Soden lists 35 as part to Ib2,though he cites it only in Paul (where the members of Ib2include 43 216 323 336 440 491 823 1149 1872 2298). This more orless corresponds to the judgement of the Alands, who do not placethe manuscript in a Category (whichusually implies a manuscript very strongly but not quite purelyByzantine).
In Acts, the Alands' detailed statistics show 35 having only 6 readingswhich agree with UBS against the Majority Text, and 69 which agree withthe Majority Text against UBS, plus five readings which agree with neither:
Acts 3:22 -- with Ψ 429 522 630 876 945 1611 1739 1891 2138 2200 2298 and twelve others
Acts 5:24 -- with 18 928 1107 1754 (but this is actually the Byzantine except for the omission of τε, which could easily happen by accident)
Acts 12:25 -- 35* with 41 other manuscripts including E 322 323 424c 630 1175
Acts 13:33a -- with 601 629 1175 2147c 2705 (but here again we have the Byzantine text minus a single short word, in this case ημιν)
Acts 24:6-8 -- (but this is a case where, although the Alands claim there is a majority text, there really isn't; it appears that slightly more than 40% of the manuscripts have a long text here, although the precise text varies greatly; somewhat less than 60% have the short text found in UBS which the Alands treat as the the majority text. 35 has the most common version of the long form. There is every reason to think that its text comes from the Byzantine tradition; it's just that that tradition is divided).
Thus I think the Alands are right to regard 35, in Acts at least, as not quite purely Byzantine but of very little independent value.
In the Apocalypse Von Soden places 35 in Ia3;Schmid places it in the "c" orComplutensian branch of the Byzantine text with manuscripts such as432 757 824 986 1075 1740 1957 2061 2352 (compare Merk's Kcgroup). Physically, like most Kr manuscripts, it hasextensive marginalia, including extensive lectionary equipment.

Manuscript 38

Paris, National Library Coislin Greek 200. Soden'sδ355;Tischendorf38e, 19a, 377p;Scrivener38e, 19a, 341p.Dated paleographically to the thirteenth century.Contains the Gospels, Acts, and Epistles with lacunae (lacking Matt.14:15-15:30, 20:14-21:27, Mark 12:3-13:4). Von Soden classifies it asIk in the Gospels, but Wisse lists it as Kx(Cluster 1053 in chapters 1 and 20; other members of this groupinclude 31, 113(part), 298, 407(part), 435, 552(part), 1053, 1186(part),1288(part), 1578(part), 2141(part), and 2724(part)).The Alands have little to add to this; they do not place 38in a Category (which generallymeans that it is heavily but not purely Byzantine), but weare not told whether it is non-Byzantine in some areas orin all (Wachtel classifies it as 10-20% non-Byzantine in theCatholics, but tells us no more). In the Acts and Epistles,von Soden lists the manuscript as a member of Ia3(the largest and most amorphous of the I groups, consistinglargely of late Alexandrian witnesses with moderate to heavyByzantine overlay). In Paul, it is cited after 1319 2127256 263, implying that it may be a weak member of Family 2127(Family 1319; see the entry on 365), and my own casualexamination of the data in Acts and Paul seems to imply that this is correctalthough I do not have the collations to be certain. Inthe Acts and Catholic Epistles, it still is listed with 1319and 256; these manuscripts, however, have much less valueoutside Paul. And, in Acts at least, the Alands list it as havingno readings at all that agree with UBS against the Byzantinetext, although it does have seven readings which agree with neither:
Acts 2:30 (where it has a singular minor variant on the Byzantine text)
Acts 2:43-44 (where it agrees with P74 ℵ A C 88 365 460 915 1175 1319 1573 1642 2127 2242, so it appears that it goes along with Family 2127 in agreeing with the actual Alexandrian text; it's just that B goes with the Byzantine text, and UBS followed B).
Acts 2:47, 3:1 (agrees with 1319 1573 2127, so again, it matches Family 2127)
Acts 10:10 (agrees with 38 other minuscules including 365 383 462 1243 1319 2127 2492, so this again appears to be a Family 2127 reading)
Acts 12:25 (with P74 A 6 33 256 459 547 1319 1573 2127 2344 2544 2675 2737 2772, so here again we have a Family 2127 reading that agrees with part of the Alexandrian text against B)
Acts 15:17-18 (with 256 365 1319 1573 2772 -- another clear Family 2127 reading, although 2127 itself defects here)
Acts 23:20 (with 24 other minuscules including 88 1319 1573 2127)
Thus it appears von Soden was right in putting it close to Family 2127; in Acts at least it appears to go with that family, although the family isn't worth much.
The manuscript has an interesting history; itwas written for the Byzantine Emperor Michael Paleologus(reigned 1259-1282), and was given to the French King Louis IX(St. Louis, reigned 1226-1270, who died of the plague whileon his way to lead what would be the Eighth Crusade). Scrivenercalls it "beautiful"; it is illustrated, but has onlylimited marginal equipment (Ammonian sections but no Eusebianapparatus or lectionary data).

Manuscript 42

Lost. Formerly Frankfurt on the Oder, Gymnasium MS. 17.α107;Tischendorf/Scrivener42a, 48p, 13r. A single leafof a lectionary is also bound in this manuscript; this isGregory l923;Tischendorf/Scrivener 287evl, 56apl.Dated paleographically to the eleventh century.Contained the Acts, Epistles, and Revelation with lacunae; Acts 2:3-34, 2 Pet. 1:1-2,1 Jo. 5:11-21, Rev. 18:3-13 were lost before the rest of the manuscript was catalogued.Acts 27:19-34 were a supplement fromanother hand. Von Soden classified 42 as Kc in theActs and Paul; K in the Catholic Epistles, and Io2 inthe Apocalypse. Schmid placed it in the in the main or "a"group of Apocalypse manuscripts -- the chief Byzantine group, headedby 046. It was missing for a time, which caused me to say that the Alandsdid not assign it to any Category.However, it was tested for the Text und Textwert volumes, andit appears the reason the Alands don't categorize it is that it is mostlybut not entirely Byzantine -- it has no readings which agree with UBS againstthe Byzantine text, but in nine of its 99 test readings it agrees with neitherUBS nor Byz.Scrivener describes it as"carelessly written, with some rare readings." Its textis said to resemble that of 51 and the Complutensian Polyglot;this appears to confirm Von Soden's classification in part, as51 is also a Kc manuscript. If we look at the nine non UBS/Byzreadings identified by the Alands, they are:
Acts 4:33 -- with 73 other manuscripts including D E 51 223 322 323 429 436 522 945 1739 1891 2200 2298 2495
Acts 12:25 -- with 60 other manuscripts including 51 206 223 429 522 876 945 1739 1891 2200 2298
Acts 13:42 -- with 101 other manuscripts including L 049 5 51 88 223 623c 2298 2492
Acts 15:17-18 -- with 16 other manuscripts including 51 223
Acts 15:34 -- singular reading (but it could arise from a single-word omission from the reading of 51 223 429 522 630 2200 etc.)
Acts 20:24b -- with 431 901 945 (but this reading is the Byzantine reading with one unnecessary word omitted)
Acts 24:6-8 -- with 51 223 234 912c 1003 1250 1405 1594 1753 1861 1863 2279 2511 2675 (there are many variations here)
Acts 25:5 -- with 68 other manuscripts including Ψ 6 51 69 206 223 307 429 441 453 522 614 623c 1505 1611 1852 2138 2412 2495
Acts 25:17 -- with 33 other manuscripts including B 056 51 181
The Aland statistics show that, among manuscripts which exist for at least thirty of the readings found in 42, its closest relatives (95% agreement) are 51 1594 2279; it agrees 94% with 234 1753 912, 93% with 390 1861 1863 2511, 92% with 223 1250. Thus there is every reason to believe that von Soden was right and that it belongs to Kc.

Manuscript 43

Paris, Arsenal 8409, 840. Soden'sε107,α270;Tischendorf/Scrivener43e, 54a, 130p.Variously dated; Scrivener lists the whole as eleventh century,Soden lists the gospels as eleventh and the rest as twelfth;Aland lists both parts as twelfth century.Contains the Gospels, Acts, and Epistles (in two volumes with slightlydifferent formats).Von Soden classifies it as Kx in the Gospels.Wisse concurs, specifying that it is part of Cluster 43(15, 43, 680, 1163, 1350, 1364, 1592, 2195(part), 2420, 2539) and pairs with 2420.The Alands do not explicitly concur, as they do not place the manuscript inany Category -- but this is probably based onthe text of the epistles, not the gospels. In the Acts and Epistles, von Sodenclassifies 43 as Ib (and cites it with Ib2 in Paul; themembers of this group, however, are not particularly distinguished). In Acts, the Alands' statistics show only two readings which agree with UBS against Byz, and only three others that agree with neither. This I would regard as a basically Byzantine text.Wachtel lists it as having between 10% and 20% non-Byzantine readings in theCatholics. Scrivener reports that, in the Gospels, the Eusebian apparatusis from the first hand but the lectionary notes are later; he speculatesthat it was written at Ephesus.

Manuscript 59

Cambridge (England), Gonville and Caius College MS. 403.Von Soden's ε272.Gregory gave the date as XIII; Scrivener says XII;James lists it as "Cent. xii, xiii, finely written." Scrivener,while noting that Tischendorf considered it to contain a "textunotabili," says that it is carelessly written; he counts 81 instancesof homoioteleuton.Contains the Gospels complete, although the pages (according to James) aremisnumbered. The manuscript seems not to have been finished; although thefirst pages (with the prologue to Matthew) are missing, the first page appearsto have left a space for an illustration, and this was never supplied. (Illustrationsof Mark and Luke are in the text; that of John seems to have been removed.) Thereare 237 surviving pages (plus an added page); there are 23 lines per page. Thefirst surviving page was written in red ink. The manuscript appears to havehad an interesting history. James says that it was "Doubtless one of thebooks left by Robert Grosseteste to the Franciscans of Oxford," (Grossetestebeing the famous and controversial Bishop of Lincoln who died in 1253). Jamesbelieves that at least one of the comments in the margin was written by Grossetestehimself. James notes that "Throughout the book are headlines in Latin,and chapter numbers, of cent. xiii. There are also interlineations in a similar handin Latin." Thus the manuscript was probably copied in the West, and certainlycame to England at an early date. It was given to Cambridge by Thomas Hatcher in1567.
In addition to Grosseteste's marginal scribbling, there are several othermarginal addenda, some quite interesting -- e.g. on page 2 we find the Latin names"Ioram and Ioatham" added to Matthew's genealogy of Jesus. Rendel Harristhought that the writer of some of the interlinear comments also wrote commented ona Greek psalter in the Corpus Christi library. Scrivenerpublished a collation in 1893 in Adversaria Critica Sacra.
Von Soden classifies 59 as Kx, which Wisse amends slightly toKmix/Kx/Kx. The Alands do not classify it, implyingthat it is not purely Byzantine but clearly heavily influenced by the Byzantine text.It is perhaps not of great value to New Testamentscholars, but might well be worthy of more examination by church historians.

Manuscript 60

Cambridge, University Library Dd. IX. 69. Soden'sε1321,α1594; Tischendorf/Scrivener60e, 10r. Contains the Gospels and Apocalypse complete,though probably written separately (Scrivener reports that "[t]he Gospelsappear to have been written in the East, the Apocalypse in the West of Europe".)A colophon dates it to 1297, but this probably applies only to the Gospels; theApocalypse appears more recent. Von Soden classifies it as Kx in theGospels, but Wisse elaborates this to Cluster 1685, "consisting of MSS 60,1454, and 1685, [and] closely related to Cl 7 and Kx Cl 1084." Thus,although the manuscript is perhaps not purely Kx, it is stronglyByzantine, which the Alands support by classifying it as Category V.In the Apocalypse it is also Byzantine; Von Soden places it in Ia7,with manuscripts such as 432 2067; Schmid places it in the "c" orComplutensian branch of the Byzantine text with manuscripts such as35 432 757 824 986 1075 1740 1957 2061 2352 (compare Merk's Kcgroup). Physically, Scrivener reports that it is an elegant copy, thatit has lectionary apparatus (added later), and that it has the Ammonianbut not the Eusebian apparatus. In the Apocalypse, "[it] has a fewscholia from Arethas about it."

Manuscript 61

Dublin, Trinity College A 4.21. Soden'sδ603; Tischendorf/Scrivener61e, 34a, 40p, 92r.Contains the New Testament complete. Generally dated to the sixteenth century (thoughScrivener admits that a fifteenth century date is possible on paleographic grounds, andan early scholar by the name of A. Clarke suggested the thirteenth century. Someonenamed Martin who worked in the seventeenth century in fact suggested an eleventhcentury date. But these early dates are surely ruled out by the fact that it uses theLatin chapter numbering as well as the Greek).Its text has not attracted much interest; Von Soden classifies it as Kx,though Wisse did not profile itdue to its late date. The Alands place it in Category Vin the Gospels and Acts (reinforcing that it is at least Byzantine if not a memberof Kx); in the Epistles and the Apocalypse they raise it to Category III.That it is non-Byzantine in the Apocalypse is confirmed by Schmid (though Von Sodenlisted it as a Koine witness); it is close to 69 (though not,as Dobbin thought, a copy of that manuscript).
James Dowden, however, tells me that Roger Omanson's Ph.D. thesisplaces it in Group 1216 in Mark -- still Byzantine, but not Kx.
It is possible that it is under-rated by the Alands in Acts; it has seven readingswhich agree with UBS against the Byzantine text, and a dozen others which agree with neither:
Acts 2:30 -- with 33 88 181 637 915 917 1874 1877 2737
Acts 2:31 -- with 876 1611 1718 1765 2138
Acts 2:43-44 -- singular, but very close to 181c and fairly close to several other manuscripts including 326 1837; this is a fractured variant that arguably should be considered more than one
Acts 3:22 -- with D 5 88 326 915 1827 1844 2344 2377
Acts 4:8 -- with 326
Acts 10:12 -- with 326 1837
Acts 10:25 -- 61* is formally singular here, but it is close to the Byzantine text and was corrected to the text of another Byzantine subgroup
Acts 12:3 -- technically singular, but the reading appears to be an error for the Byzantine text
Acts 15:17-18 -- 61* is formally singular here, but it is close to the Byzantine text and was corrected to the text of another Byzantine subgroup
Acts 18:21-22 -- singular reading which looks to me like a scribal error
Acts 18:27 -- with 326 1837
Acts 26:14 -- with 34 other manuscripts including H 049 322 323 326 629 1837
In all seven places where it agrees with UBS against Byz, 61 is in company with 326 and 1837; it will be seen that it agrees with the latter pair in most of its readings which agree with neither UBS nor Byz. Indeed, of manuscripts which are extant for at least half the readings of 61 in Acts, the Alands show 326 as 61's closest ally (88% agreement), followed by 1837 (84%). It seems pretty clear that these three form some sort of group, primarily Byzantine but with an overlay of something else, perhaps late Alexandrian; the group probably shouldn't just be written off as Category V.
What is noteworthy about this manuscript, however, isnot its text (which is at best mildly interesting) but the historical use to whichit was put. 61 is the manuscript which was presented to Erasmus to force him toinclude the "three heavenly witnesses" passage (1 John 5:7-8) inhis third edition of the Textus Receptus. It is believed thatthe codex was written for this express purpose, and in some haste; at least threeand possibly four scribes were involved in the project (the gospels having quitelikely been written before Erasmus's edition was published, then theActs and Epistles added to confute him; the Apocalypse may be later still; a dateof around 1580 has been conjectured for it). Dobbin thought the Acts and Epistlesmight have been copied from 326, although the latter manuscript seems somewhatmore interesting than 61. It has also been supposed that the gospels were takenfrom 56, but as 56 is a Kr manuscript, it is possible that anothercopy of that text was used. The haste with which 61was written is perhaps evidenced by its lack of lectionary apparatus (though ithas the κεφαλαια andAmmonian/Eusebian apparatus) and by the number of later corrections it required.It has been said that the only page of the manuscript to be glazed is that containing1 John 5:7-8, but in fact the paper is glazed throughout; it is simply thatso many readers have turned directly to that passage that the wear and tearhas caused the glazing to be visible on that page as on no other.

Manuscript 66

Cambridge, Trinity College O.viii.3. Soden's ε519.Contains the Gospels complete. Estimates of its date vary widely; Scrivener offersthe twelfth century, the Alands the fourteenth, von Soden the fifteenth. M. R. Jamesin his catalog of the Trinity manuscripts of class O says "Cent. xv. late, with neatbut not good pictures and ornaments. The hand is very good."Textually, Von Soden classifies it as Kr, and Wisse concursthough he notes that it has a "large surplus."The Alands, unsurprisingly, place it in Category V.It is unusual for a Kr manuscript in that it has the Ammonian andEusebian apparatus. It contains ten blank pages, on five consecutive leaves(for some additional material which was not supplied?). They follow the Eusebiantables. Scrivener believesthat two later hands have worked on it, the earlier making some correctionsin the text while the later added some scholia in the margin. Frankly it seemsa thoroughly unexceptional manuscript.

Manuscript 69

Location/Catalog Number

Leicester. Catalog number: Town Museum Cod. 6 D 32/1


69 contains the entire New Testament with many lacunae. Missing Matt. 1:1-18:15,Acts 10:45-14:17 (the manuscript skips from Acts 10:45 to 14:17 without break;it would appear the scribe did not realize there was a defect in his exemplarhere!), Jude 7-25, Rev. 19:10-22:21; Rev. 18:7-19:10 are fragmentary. The manuscriptalso contains five pages of assorted information about church history anddoctrine.


Dated paleographically to the fifteenth century, probably to the period1465-1472, since it was presented to George Neville, Archbishop of York,England during those years. The scribe is known from his other writingsto have been Emmanuel, a former resident of Constantinople who spent thesecond half of the fifteenth century in England copying Biblical andclassical texts. His writing style is absolutely peculiar; epsilons closelyresemble alphas, and accents are often placed over consonants rather thanvowels. Acute and grave accents are confused and hard to distinguish.Errors are also common; Scrivener counted 74 omissions of various sorts,and many words interrupted in the middle. The scribe also used theNomina Sacra in peculiar ways;Ιησους is consistentlyspelled out until John 21:15, when contractions begin to be used sporadically.The manuscript appears to have been written with a reed. Scrivener alsoremarks, "Though none of the ordinary divisions into sections, and scarcelyany liturgical marks, occur throughout, there is evidently a close connectionbetween Cod. 69 and the church service books, as well in the interpolationsof proper names, particles of time, or whole passages (e.g. Luke xxii. 43, 44placed after Matt. xxvi.39) which are common to both...." Ferrer/Abbottamplifies: "besides [a strange list of sections], there is no liturgicalmatter whatever, no division into sections, or Eusebian canons, or notes aboutlessons, except a marginal mark or two, and a few words, which are often illegible,scribbled at the foot of the first page of each leaf."

Ferrer/Abbott adds, "Many of the changes met with in this MS. arise frominversion of order, the substitution of simple for compound words, and viceversâ. 'A corrector's hand,' adds Dr. Scrivener, 'has been busy throughoutthis copy, whom Dr. Dobbin considers to have been the original scribe. I havedeemed the changes to be secundâ manu, but nearly as old as the first."

A number of marginal notes ("too many," Scrivener acidly remarks)are written in the hand of William Chark, who owned the manuscript probably inthe late sixteenth century.

69 is written on a mix of paper and parchment. Abbott/Ferrer, probablybased on Scrivener, speculate that the scribe had a limited supply of parchmentand estimated how much paper would have to be added to allow enough material fora whole codex. As a result, the quires are usually of five sheetsrather than four, with two parchment and three paper sheets per quire, theparchment leaves being on the outside of the quire. The paper is very poor --so bad that one side of four of the paper leaves had to be left blank. The manuscripthas one column per page. The books seem to have originally been in the orderPaul (with Hebrews last), non-Biblical materials, Acts, Catholic Epistles,Apocalypse, Gospels.

Description and Text-type

The text of 69 varies significantly. In the Gospels it was identified byFerrar with Family 13, and this has been affirmed by everyone since (Wisseclassifies it as 13, and von Soden put it in Iib). However, somehave thought it one of the best Family 13 manuscripts, and others countit one of the poorer. Probably the peculiar readings generated by scribalerrors had something to do with this. Within the Ferrar group, it has beenplaced in the "b" group (along with 174 and 788) by scholarsfrom von Soden and Lake to Colwell. The Alands, interestingly, classify69 as Category V (Byzantine) -- despite the fact that its profile(1341 631/2 222 50s) seemsto be fairly typical for the Ferrar Group (e.g. 13 is1501 711/2 312 54s; 346is 1721 821/2 242 53s).

In the Acts even Scrivener concedes the text to be "less valuable."Von Soden classes it as Ia3, but places it among the lower membersof the group. The Alands classify it as Category V.Based on their numbers, that classification is not justified. The Alands find only four readingsin Acts where 69 agrees with UBS against the Byzantine text -- but fourteen readings whichagree with neither:
Acts 1:14 -- (singular, but close to the Byzantine reading; probably an error by the scribe)
Acts 2:23 -- (singular, but again close to the Byzantine reading)
Acts 4:33 -- with 321 1448* (but this could be the result of a misunderstood correction to the word order)
Acts 9:25 -- with 6 81* 468 915 935 1501 1548 1752 1874 1877 2143 2774
Acts 10:11 -- (singular, but a minor variant on the Byzatine reading)
Acts 10:12 -- with 6 1857 2816
Acts 10:19 -- with P74 ℵ A 6 81 94 180 181 307 431 453 610 1175 1642 1678 1875 2818
Acts 15:7 -- with 102 189 1102 1360 1736* 2492
Acts 16:33 -- with 451 (but this reading differs by just one letter and a stroke from the Byzantine reading)
Acts 19:14 -- (singular, but an easy error for several other readings)
Acts 20:15 -- with 45 other manuscripts including D* 2412 but no other manuscripts thought to be of note; this looks like the reading of a Byzantine subgroup.
Acts 24:6-8 -- with 6 (but there are many, many fairly similar readings; this is a very fractured reading)
Acts 25:5 -- with 68 other manuscripts including Ψ 6 206 441 453 522 614 623c 1505 1611 1852 2138--98 2412 2495 (although this is the reading of Family 2138, it is also the Kc reading, so this probably counts as Byzantine despite being listed by the Alands as a special reading)
Acts 28:16 -- singular but very close to the Byzantine reading
This data requires some examination. To begin with, the four readings that agree with UBS against Byz are perhaps misclassified; in those four readings, 69 agrees with 104, 71, 99, and 22 other manuscripts (and, in the last case, is a minor variant on 42 more); 69 agrees with UBS against the majority, but it appears it is Byzantine even so. A for its 14 readings that agree with neither, we observe that 69 doesn't seem to agree with anything very often. (In the Alands' list, its closest relative among manuscripts with substantial texts are 6 and 172, and they agree only 80% of the time -- extremely low rates for a manuscript which agrees with the majority text itself about 80% of the time! I do note a certain tendency for the non-Byzantine readings to clump. My first thought, before I examined the readings, was that 69 might be block mixed. But the more likely conclusion is that 69 was very poorly copied -- and that the copyist had better and worse patches, probably due to sleep and/or lighting issues. Most of 69's non-Byzantine readings do not appear genetically significant; it just has readings from lesser branches of the Byzantine text, copied by a not-very-good scribe. So the Aland Category V classification is probably correct even though their numbers don't really support it. (Which shows why the classification based on just four numbers isn't much good.)

It is generally agreed that 69 and 462 are closely akin in the PaulineEpistles. Their combined text is, however, only slightly removed from theByzantine. The Alands classify 69 as Category III in Paul (they do notcategorize 462). Von Soden places 69 and 462 next to each other in Ia3.Davies links 462 (and so by implication 69) with 330,436, and 2344; her technique, however, makes theseresults questionable. There is as yet no clear evidence that 69 and 462should go with any of the stronger members of the Ia3 group,such as Family 330 or 365 and Family2127.

In the Catholics the Alands again classify 69 as Category V, and vonSoden again classifies it as Ia3. Wachtel lists it as having10-20% non-Byzantine readings. Richards classifies it as Mw,which makes it a mixed manuscript that does not seem to have any closerelatives. This seems to conform with the results of Wachtel.

In the Apocalypse, the Alands classify 69 as Category V.Von Soden lists it as I', grouping it with 61 and 046.

Other Symbols Used for this Manuscript

von Soden: δ505. Tischendorf: 31a, 37p, 14r


W. H. Ferrar and T. K Abbott, Collation of Four Important Manuscriptsof the Gospels by the late William Hugh Ferrar, 1877, collates 13, 69, 124,and 346 in the Gospels. Much of the material in the introduction, and the collationare based on Scrivener's collation.
F. H. A. Scrivener, An Exact Transcription of Codex Augienses, 1859,collates and discusses the manuscript.

Sample Plates:
Metzger, Manuscripts of the Greek Bible (1 page)

Editions which cite:
Cited in SQE13 where it differs from Family 13 and the Majority Text.
Cited by von Soden, Merk, and Bover.

Other Works:"Origin of the Leicester Codex of theNew Testament, 1887.
M. R. James, "The Scribe of the Leicester Codex," Journalof Theological Studies, v (1903/4).

Manuscript 71

London, Lambeth 528. Soden'sε253.Contains the Gospels complete. Generally dated to the twelfth century;Scrivener offers the exact date 1160 C.E. based on the colophonwhich dates it ετει απο Χριστου ᾳρξ. Scrivener's Introduction reports that "This elegant copy, which once belonged to an Archbishop of Ephesus,was brought to England in 1675 by PhilipTraheron, English Chaplain at Smyrna." It is almost complete,but Scrivener reports the loss of Matthew 14:13-15:16; the loss is recent, sinceTraheron collated it.Classified by Von Soden as Iφr,along with M 27(part) 692(part) 1194; Iφ as a whole iswhat Streeter calls Family 1424. Wisse partly corroborates Von Soden, making 71a core member of the M27 group (while pointing out that M is not really a goodexample of the M type). Other members of M27 include M 27 71 248(part) 447(part)518(part) 569 692 750 830(part) 1014(part) 1032(part) 1170 1222 1228(part)1413 1415 1458 1626 1663(part) 2705. The Alands give this their usual half-heartedendorsement by refusing to place 71 in a Category;this generally means that the manuscript belongs to the Byzantine text but not one of themainstream Byzantine groups. It has a lectionary apparatus, and issaid to have "many" later corrections. Scrivener also notes that "thiscopy presents a text full of interest, and much superior to that of the mass ofmanuscripts of its age." Mill thought its text similar to that of 29, thoughWisse's analysis does not confirm this in Luke. Scrivener collated it under theletter g in A Full and Exact Collation of about Twenty Greek Manuscripts of theHoly Gospels.

Manuscript 81

Location/Catalog Number

57 folios are in the British Library in London (Catalog number: Add.20003); 225 folios are in Alexandria (Patriarchal Library MS. 59). TheBritish Museum portions were taken from Egypt, where Tischendorf"discovered" the manuscript.


81 contains the Acts and Epistles. Acts 4:8-7:17, 17:28-23:9 havebeen lost. It is written on parchment, one column per page.


Dated by its colophon to April 20, 1044, and written by a scribe named John.

Description and Text-type

81 has been called "the best minuscule witness to Acts." Itis consistently Alexandrian (although with some Byzantine corruptions).Von Soden lists 81 as H. Aland and Aland describe it as "at leastCategory II."

Curiously, although 81 has ben highly praised, at least in Acts, few seem tohave attempted to classify it beyond calling it Alexandrian. Based on the Alands'percentage agreements, it doesn't have any particularly close relatives -- amongsubstantial manuscripts, it is listed as closest to P74 and ℵ,but it agrees with those two only 73% of the time. It agrees with both A and B 71%.It agrees 66% with 1175, 65% with C, 60% with 1739, and with no other substantialmanuscript as much as 60%. The sample sizes here are small, and the Aland samples areimperfect, but it appears 81 has a fairly independent text.

In Paul, 81 is clearly closer to ℵ A C 33 than to P46 B or 1739.Its text seems to fall somewhere between the early and late formsof the Alexandrian text, and may represent a transitional phase in theevolution of that text (most late Alexandrian witnesses -- e.g. 436, 1175,family 2127, 2464 -- seem to be closer to 81 than they are to each other).In the Catholics it is again Alexandrian with some Byzantine mixture; itseems to be a slightly less pure form of the A/33 text.

Other Symbols Used for this Manuscript

von Soden: α162. Tischendorf: 61a; also loti andpscr (the latter after Scrivener's symbol p)



Sample Plates:

Editions which cite:
Cited for the Acts and Epistles by all editions since Von Soden.

Other Works:

Manuscript 82

Paris, National Library Gr. 237. Soden's O1;Tischendorf/Scrivener 10a, 12p, 2r.Contains the Acts, Epistles, and Apocalypse complete. Universallydated to the tenth century.Includes a commentary (listed by Von Soden as that of Oecumenius, i.e. thepseudo-Oecumenius;Scrivener describes it simply as "scholia and other matter.")Von Soden did not classify it beyond listing it among the Oecumeniusmanuscripts, but Scrivener believed that "its value in theApocalypse is considerable." This has not been confirmed byfurther research; Schmid places it in the main or "a"group of Apocalypse manuscripts -- the chief Byzantine group, headedby 046. This is confirmed by the Alands, who place 82 inCategory V in all sections. (Indeed, their datashows that, in Acts, it has only two readings not supported by the plurality ofmanuscripts, and even those two have dozens of supporters; it appears to be aboutas Byzantine as a manuscript can be.) Scrivener describes82 as "neatly written," and notes that it contains non-Biblicalmatter (including the treatise of Dorotheus of Tyre mentioned in theentry on 177). The manuscript was included in the editions of Stephanus as ιε'.

Manuscript 83

Munich, Bavarian State Library Gr. 518.Soden's ε1218;Contains the Gospels complete. Dated by both Scrivener and Aland tothe eleventh century; Von Soden prefers the twelfth. Von Sodenclassifies it as Kr, and Wisse concurs, listing it asa perfect member of the type. The Alands list it asCategory V (Byzantine). Scrivener describesit as "beautifully written." It has all the marginaliaexpected of a Kr manuscript, even though (or perhapsbecause) it is one of the earliest examples of this type.

Manuscript 91

Paris, National Library Gr. 219. Von Soden's O14;Tischendorf/Scrivener 12a, 16p, 4r.Contains the Epistles, and Apocalypse, with commentary. (Somesources say it contains Acts also, but that text is no longer extant.)Dated paleographically to the eleventh century.The commentary on the Acts and Epistles is that of the (pseudo-)Oecumenius;that on the Apocalypse is that of Arethas. As an Oecumenius manuscript,Von Soden does not really classify the text (beyond listing it asKo in the Apocalyse), but the Alands do not list it in anyCategory. This implies that it islargely but not quite purely Byzantine -- but it may simply be that theirdata is imperfect; . In the Apocalypse,Schmid places it in the dominant or "a"group of the Byzantine text headed by 046. Scrivener describesit as "neat," with lectionary tables but no apparatus.It once belonged to the Medicis.

Manuscript 93

Paris, National Library Coislin Gr. 205.Von Soden's α51;Tischendorf/Scrivener 17a, 21p, 19r.Contains the Acts, Epistles, and Apocalypse with lacunae (lacking1 Cor. 16:17-2 Cor. 1:7; Heb. 13:15-25; Rev. 1:1-2:5 isan addition by a later hand).The colophon, written by a monk named Anthony, dates it to the year1079 (though for some reason the Kurzgefasste Liste simplygives the manuscript's date as XI). The text is described byVon Soden as a mix of I and K types in the Acts, and as purelyK (Byzantine) elsewhere. The Alands do not place 93 in anyCategory, but this implicitlysupports Von Soden, as uncategorized manuscripts are usually veryheavily but not quite purely Byzantine. Indeed, in Acts, they find only threereadings which agree with UBS against Byz, plus one singular reading(Acts 28:16) and one reading which agrees with neither (Acts 19:3, where it may belongto a Byzantine subgroup; 23 other minuscules, none of them well known, also have itsreading). I do findit interesting that its three readings which agree with UBS against Byzare close together (readings 96, 98, 99=Acts 25:16, 26:14, 26:28;it agrees with both UBS and Byz in readings 97 100 = Acts 25:17, 27:5).Thus there would appear to be a possibility that it is block-mixed, with a purelyByzantine text in the first 24 or so chapters, but with a different sort of textfor some part of chapters 25-28. Possibly this extends to the Catholics;Wachtel lists it as beingbetween 20% and 30% non-Byzantine in the Catholic Epistles. In the Apocalypse,Schmid places 93 in the dominant or "a"group of the Byzantine text headed by 046. The manuscript has theusual lectionary equipment, prologues, etc.

Manuscript 94

Paris, National Library Coislin Gr. 202 (folios 27-328; this numberalso includes a portion of Hp).Von Soden's O31 andΑν24;Tischendorf/Scrivener 18a, 22p, 18r.Contains the Acts, Epistles, and Apocalypse complete and with commentary.The Apocalypse is dated paleographically to the twelfth century; the Actsand Epistles to the thirteenth (so the Kurzgefasste Liste; Scrivenerlists eleventh and twelfth, respectively. The change in script correspondsto a change in writing material; the first portion is on parchment, the rest onpaper). The commentary on the Apocalypse is that of Andreas; Von Soden liststhe rest as having the commentary of the (pseudo-)Oecumenius, thoughScrivener describes it simply as "scholia to the Acts and CatholicEpistles... [prologues] to St. Paul's Epistles." Von Soden, as usual,classifies the text by its commentary; the Alands list it asCategory III in the Acts and CatholicEpistles "but clearly lower for Paul and Revelation."In the Catholic Epistles, Wachtel lists it as having from 30% to 40%non-Byzantine readings. Its numbers in Acts are also quite high -- perhapshigh enough to imply a status above Category III.The Alands list it as having 40 readings that go with the Byzantine text againstUBS, and only 22 which go with UBS against Byz -- but what is reallynoteworthy is its 23 readings that agree with neither. That's more than a fifthof all its readings! (And only three of them are singular.) They include:
Acts 2:7b -- singular (but close to the Byzantine reading)
Acts 2:30 -- with 307 453 610 2818
Acts 2:31 -- with 94 other manuscripts including C3 E 5 51 104 223 307 322 323326 429 453 522 629c 630 945 1739 1891 2200 2344 2412
Acts 2:43-44 -- with 180 307 453 610 1501 1678 2818
Acts 3:21 -- with ℵ2 B2 E 363 610 945 1106 1409 1678 1884 2818
Acts 4:25 -- singular
Acts 4:33 -- with 94 180 307 453 610 629 1678 2818
Acts 8:37 -- with 29 other manuscripts including 429* 630 876 945 1678 1735 1739 1891 2200 2298 and very close to 323 429c 180 307 453 610 2818
Acts 8:39 -- with 30 other manuscripts including Ac 180 307 322 323 453 610 876 945 1678 1739 1891 2298 2818
Acts 9:25 -- with 049 209* 307 429 453 522 610 628* 630 636* 1066 1490 1642 1721 1749 1751 1758 1831 2200 2511 2818
Acts 10:12 -- singular (but a minor variant on the readings of C* 5 33 322 323 424c 623 2298 etc. and that of 180 307 453 1678 2818)
Acts 10:19 -- with P74 ℵ A 6 69 81 180 181 307 431 453 610 1175 1642 1678 1875 2818
Acts 10:47 -- with 33 other manuscripts including Ec 6 180 181 307 322 323 429 453 522 610 630 945 1678 17391891 2200 2298 2818
Acts 12:25 -- with 53 other manuscripts including D Ψ 180 181 307 436 441 453 610 614 623 1611 1678 2138 2412 2828
Acts 15:34 -- with 87 other manuscripts including 5 6 33 88 180 307 323 383 441 453 610 614 623 1175 1678 1739 1868 2298 2344 2412 2818
Acts 16:28 -- with 180 307 431 468 489 610 629 927 1319supp 1678 1729 1780 1868 2242 2502 2818
Acts 18:17 -- with 180 307 431 453 610 1319supp 1678 2818
Acts 19:3 -- with D Ψ 180 206 429 522 630 945 1490 1509 1704 1739 1751 1758 1831 1891 2080 2200 2298 2718
Acts 19:14 -- with P74 ℵ A B 33 206 429 522 945 1175 1490 1704 1735 1758 2298 2344
Acts 20:15 -- with 94 206 630 1175 1891 2200
Acts 23:20 -- with 85 other manuscripts including ℵ2 Ψ 6 180 307 424c 436 441 483 610 945 1175 1505 1611 1678 1739 2138 2298 2412 2495 2818
Acts 24:6-8 -- with 46 other manuscripts including Ψ 5 33 88 307 436 610 623 630 945 1319c 1739 1891 2200 2298 2818
Acts 25:5 -- with 68 other manuscripts including Ψ 6 69 180 206 307 429 441 453 522 610 614 623c 1505 1611 1678 1852 2138 2412 2495 2818
A tendency to agree with the members of Family 453 (=180 307 453 610 1678 2818=36a) will surely be obvious, and the Alands' overall data affirms this; among substantial manuscripts, 94 agrees most closely with 610 (76%), then 36 1678 (74%), 307 (73%), 180 (72%), 453 (71%). This might explain also why 94 is less valuable in Paul -- it might retain the Family 453 text, but in Paul, Family 453 isn't worth much.
I do note that in the last few chapters of Acts, 94 seems to shift away from the 453 group to be slightly closer to Families 1739 and 2138 -- two families which, in Acts, seem to have an odd kinship. This might well reward further investigation.

Manuscript 104

Location/Catalog Number

British Library, London. Catalog number: Harley 5537.


104 contains the Acts, Epistles, and Revelation complete. It is writtenon parchment, one column per page.


Dated by its colophon to 1087.

Description and Text-type

Generally listed as an Alexandrian witness, and it does have Alexandrianreadings in the Epistles, although it is more Byzantine than anything else.There are also hints of other text-types -- e.g. 104 shares a certain numberof readings with family 1611. On the whole, the best description of themanuscript is probably "mixed."

Von Soden lists 104 as H in the Acts and Epistles; he lists is at Ib2in the Apocalypse. Merk places it in the Anr group (a sub-groupof the Andreas text). Aland and Aland describe it as Category III in Pauland the Catholics, Category V in Acts and the Apocalypse.As regards the Apocalypse, this is probably accurate, but seems unfair in Acts. The Alands' own data for that book shows 104 as having eight readings which agree with UBS against Byz, and ten which agree with neither (out of 102 readings tested); none of the ten special readings is singular. That is surely a high enough rate to justify a Category III rating. Of course, that doesn't tell us what sort of text 104 has. If we look at the ten non-UBS-or-Byz readings, we find:
Acts 2:30 -- 104 with Ψ 296 429 457 460 467 617 876 1251 1490 1501 1611 1718 1758 1765 1780 2138
Acts 2:31 -- 104 with 94 other manuscripts including C3 E 5 94 180 307 322 323 326 429 453 459 460 467 522 610 617 629c 630 945 1251 1490 1501 1722 1739 1891 2200 2344 2412 (and close to several more including 1838)
Acts 2:43-44 -- 104 with E 33 181* 459 1409 1722c 1838 1842c 1844 2344 2774
Acts 12:25 -- 104 with 680c 1352 (but probably an error for a very common reading; this variant is probably not genetically significant)
Acts 18:20 -- 104 with 459 1838 (1839 also similar)
Acts 20:4 -- 104 with D 1838
Acts 21:8 -- 104 with 459
Acts 24:1 -- 104 with 256 459 1704 1838 2143
Acts 24:6-8 -- 104 with 254 308 322 323 398 400 424c 459 464c 489 592 621 910c 927 1521 1524 1609 1702 1727 1729 1842 1843 1852 1868 1873 2086 2143 2201 (and close to about a dozen others include 1838)
Acts 28:29 -- 104 with 256 365 996 1405 1719 2243 2625 (and close to several others including 459)
It will be evident that 104's readings in these variants are not very distinguished -- they are not, e.g., Alexandrian readings that the UBS editors ignored. But there are interesting hints of kinship, especially with 459 and 1838. And, indeed, among substantial manuscripts it is closest to 459 (90% agreement), followed by 172 (84%, but 172, although substantial, is far from complete), 1526 (83%, but again, 1526 is incomplete), and 1838 (82%). These manuscripts might reward further investigation.

Other Symbols Used for this Manuscript

von Soden: α103. Tischendorf: 25a; 31p;7r



Sample Plates:

Editions which cite:
Cited by NA26 for Paul.
Cited by NA27 for Paul.
Cited by UBS3 for Acts, Paul, and the Catholics.
Cited by UBS4 for Paul.
Cited by von Soden, Merk, and Bover.

Other Works:

Manuscript 110

London, British Museum Harley 5778. Soden'sα204;Tischendorf/Scrivener 28a, 34p, 8r.Contains the Acts, Epistles, and Apocalypse with some mutilations: Acts 1:1-20,Rev. 6:14-8:1, 22:19-21 "and perhaps elsewhere" (so Scrivener, whocollated the Apocalypse). Dated paleographicallyto the twelfth century. Classified as K by Von Soden, and the Alandsconcur by placing it in Category V.In the Apocalypse, Schmid places it in the dominant or "a"group of the Byzantine text headed by 046. Scrivener describes it asbeing in "wretched condition, and often illegible." All thesedescriptions seem likely to be true; the Aland sample shows 110 as havinga large lacuna in Acts 6-7 and as having seven other readings where it wasunreadable -- and, of the 94 legible readings, only three disagreed with themajority text.

Manuscript 115

London, British Museum Harley 5559. Soden'sε1096.Contains the Gospels with extensive mutilations: Matt. 1:1-8:10,Mark 5:23-36, Luke 1:78-2:9, 6:4-15, John 11:2-end are all lost, though afew additional words of John 11 can be read. Generally dated to the tenth century;though Scrivener gives a twelfth century date. Classified asIφbby von Soden; other members of this group include7 179 267 659 827 and parts of 185 1082 1391 1402 1606. Wisse, however,does not concur; he finds the manuscript to be Kmix/Kx/Kmix.The Alands do not assign 115 to a Category; thisis not surprising for a manuscript with a text close to but not identical to Kx.The manuscript has only a limited set of reader aids; according toScrivener, it offersκεφαλαια,"some"τιτλοι,the Ammonian sections, and "frequently" the Eusebian apparatus; Scrivenerspeculates that the manuscript was "never quite finished."

Manuscript 118

Oxford, Bodleian Library Auct. D. infr. 2.17 (was BoldeianMisc. Gr. 13). Soden'sε346.Contains the Gospels with some defects; later hands supplied Matt. 1:1-6:2; Luke13:15-14:20, 18:8-19:9, John 16:25-end. The binding also contains portionsof the Psalms on paper. Dated paleographically to the thirteenth century.118 is one of the manuscripts found by Lake to belong toFamily 1; every examination since has confirmedthis. Von Soden listed it asIhb,i.e. part of the b subgroup of Family 1; other manuscripts he placesin this group include 22, 131 (in Mark and Luke), 209, and 872 (in Mark).Wisse concurs with the assignment to the Lake Group, listing 118 as a core member of Family 1.The Alands, interestingly, do not place 118 in anyCategory, but do list it with Family 1.Most seem to agree with Von Soden in placing 118 closer to 209 than to1 and 1582.Scrivener reports the manuscript to be a palimpsest, but with thegospel text uppermost. It has the full set of scribal aids, thoughthe lectionary tables were added later. For more details on the text,see the entry on Family 1.

Manuscript 138

Rome, Vatican Library Greek 757. Soden's A201 andCι24.Contains the Gospels with a commentary and minor lacunae. Universally dated to thetwelfth century. The commentary on Mark is that of Victor; elsewhereScrivener lists it as being primarily from Origen, though Von Soden considersit to be the "Antiochene commentary" (Chrysostom on Matthew,Victor on Mark, Titus of Bostra in Luke) in the Synoptic Gospels whileJohn is listed as having the "Anonymous Catena." The text itself VonSoden places in the Ac group -- a generally undistinguished groupcontaining such manuscripts as 127, 129, 137, 139, 143, 151, 374, 377,391, 747, 989, 1312, 1313, 1392.In any case Wisse's classifications do not accord with von Soden's; themanuscripts von Soden lists as Ac appear to belong to almostevery Byzantine subgroup. 138 itself was profiled only in Luke 1, butthere Wisse lists it as Kx This is supported by the Alands,who classify 138 as Category V. Scrivenersummarizes Burgon's report on the manuscript by saying that thecommentary is "mixed up with the text, both in a slovenly hand."
James Dowden, however, tells me that, in John at least, 138 may bemore interesting than this. Bruce Morrill in his Ph.D. thesis found that138 357 994 2575 2684 seem to form a group which may be a subgroup ofthe Lake Group (Family 1). This probably deservesdeeper investigation.

Manuscript 141

Rome, Vatican Library Greek 1160. Soden'sδ408; Tischendorf141e, 75a, 86p, 40r.Contains the New Testament complete. Dated paleographically to thethirteenth cetury by Gregory, Aland, Scrivener; von Soden prefersthe fourteenth. The text of the manuscript is not noteworthy;both Von Soden and Wisse declare it to belong to Krin the Gospels, and the Alands classify 141 asCategory V throughout (and their datasupports this; in Acts, e.g., it has only five readings which disagreewith the Byzantine text). In the ApocalypseSchmid places it in the dominant or "a"group of the Byzantine text headed by 046. It is in two volumes, withthe two volumes numbered separately. In the Acts and Epistles it hasthe Euthalian apparatus, though it does not appear to have the text.The full lectionary equipment is supplied, and it has pictures, butlike most Kr manuscripts it lacks the Eusebian apparatus.

Manuscript 157

Rome, Vatican Library Urbin Gr. 2. Soden'sε207.Contains the Gospels complete. Universally dated to the twelfth century, based bothon the writing and on a pair of pictures, of the Emperor Alexius Comnenus(Byzantine Emperor 1081-1118) and his son John (II) Comnenus (1118-1143). Itwas apparently written for John Comnenus, and waswas brought to Rome by Pope Clement VII (1523-1534).Classified as Iσby von Soden, the other members of this group being 235(part) 245 291713 1012. Wisse's data, however, paints a completely different picture;he finds 157 to be a member of Kx in Luke 1, mixed withsome relationship to the Alexandrian text ("Group B") in Luke 10,and Alexandrian in Luke 20. The other manuscripts ofIσ do notshare this profile, and in fact do not seem to be related to each other at all.That 157 is mixed is confirmed by the Alands, who list it asCategory III, and by Hort,who considered it mixed but still the most important minusculeof the gospels other than 33. Streeter thought it Alexandrianwith "Cæsarean" influence -- but it should be notedthat Streeter thought everything had "Cæsarean" influence.Zahn thought it might have had Marcionite influence. Hoskier, who collated it(J.T.S. xiv, 1913), thought there were points of contact with thePalestinian Syriac. 157 is noteworthy for having theJerusalem Colophon after each gospel.Scrivener observes that 157 is "very beautifully written... [with]certain chronicles and rich ornaments in vermillion and gold." It hasother pictures in addition to the portraits of the Emperors, as well as lectionaryapparatus.
Sample plate in EdwardMaunde Thompson, An Introduction to Greek and Latin Paleography(plate 68).

Manuscript 160

Rome, Vatican Library Barb. Gr. 445. Soden'sε213.Contains the Gospels complete. Dated by its colophon to the year 1123.Classified as Iφcby von Soden, the other members of this group being 945 990 1010 1207(part) 1223 1293.Iφ is Streeter'sFamily 1424, but the c branch, if it is part of the family at all, is veryweak. Wisse lists 160 as Mixed in Luke 1 and Kx Cluster 160 in Luke10 and 20. It is interesting to note, however, that all three manuscripts whichWisse lists in Cluster 160 (160, 1010,and 1293) were called Iφc by von Soden; this would seem to imply atleast some unique traits to that group. Given theconnection of this group with Kx, it is surprising to note that theAlands do not list a Category for 160, implying thatthey find it to be somewhat less than purely Byzantine. Themanuscript itself has the full lectionary equipment and the Ammonian Sections,but no Eusebian apparatus.

Manuscript 162

Rome, Vatican Library Barb. Gr. 449. Soden'sε214.Contains the Gospels complete. Dated by its colophon to May 13, 1153.Classified as I by von Soden, but with no subgroup specified; itis not one of his regularly cited manuscripts. It would appear thatthis was a casual classification -- based, perhaps, on the manuscript'sreading in Luke 11:2, where it hasελθετω σου τοπνευμα το αγιονκαι καθαρισατωημαςforελθετω ηβασιλεια σου-- a reading shared, in its essentials, by 700, Marcion (orTertullian in talking about Marcion), Maximus, andGregory of Nyssa but no other known witnesses. In any case, Wisse does notconcur with von Soden's classification of 162 as an I witness (which would make itWestern or "Cæsarean"); he lists162 as Kx/Kmix/Kx, and the Alandsconfirm its Byzantine nature by placing it in Category V. The manuscript, written by oneManuel, has the Eusebian apparatus but no lectionary equipment at all.

Manuscript 174

Rome, Vatican Library Gr. 2002. Soden'sε109.Contains the Gospels complete with major lacunae; Matt. 1:1-2:1, John 1:1-27,8:47-end are gone. Dated by its colophon to September 7, 1052.Classified as Iιb --that is, as part of Family 13 -- by von Soden, butonly in Matthew is it cited. Wisse confirms that its text shifts, for he placesit in Group Λ in Luke.The Alands seem to confirm this as well; although they list 174 as a member of Family 13in NA27, they do not assign it to a Category(most members of Family 13 are Category III; the fact that 174 is not implies thatit is weaker than other members of the family). For more details onFamily 13, see the entry on that manuscript. 174 itself waswritten by a monk named Constantine under the authority of "Georgilas duxCalabriae" [Scholz]. It has the full Ammonian and Eusebian apparatus, pluslectionary indications, but the lists of readings, if it had any, have not survived.

Manuscript 175

Rome, Vatican Library Gr. 2080. Soden'sδ95;Tischendorf/Scrivener 175e, 41a, 194p,20r.Contains the entire New Testament except for Matt. 1:1-4:17.Dated paleographically to the tenth century (so Gregory, Aland, von Soden;Scrivener would allow any date between the tenth and twelfth).Von Soden classifies the Gospels as Kx, but Wisse lists themas weak Πa.The Alands seem to agree with the latter judgement, as they do not place175 in any Category (which usually meansthat the manuscript is strongly Byzantine but not a member of Kxor Kr). In the Acts and Epistles, Von Soden lists the textas K (Byzantine), and there is no reason to doubt this. In the Apocalypse Schmid places it in the dominant or "a"group of the Byzantine text headed by 046. The Alands again don't list acategory, but their data for Acts shows it as having only four readings whichdo not agree with the Majority Text; in that book, at least, we can surelyregard it as Byzantine.The arrangement of the sections isunusual; Scrivener notes that the book places them in the order Gospels,Acts (with scholia), Apocalypse, Catholic Epistles, Paul. The book has"some" marginal corrections from the first hand. Paul hasthe Euthalian subscriptions, but otherwise the marginal equipment islimited.

Manuscript 177

Munich, Bavarian State Library Gr. 211. Soden'sα106;Tischendorf/Scrivener179a, 128p, 82r.Contains the Acts, Epistles, and Apocalypse complete.Dated paleographically to the eleventh century (so Soden, Scrivener, and theListe; Delitzsch suggested the thirteenth century).Von Soden classifies it as Ia3 in the Acts and Paul;in the Catholic Epistles he lists it as K. If it is a member ofIa3 (a group consisting mostly of late Alexandrian witnesseswith greater or lesser degrees of Byzantine mixture), it must be aweak one, as the Alands list 177 asCategory V (Byzantine) throughout, and thisseems likely, given that it has only five readings which disagree with theByzantine text -- and only one of those agrees with UBS; two are sub-singular(Acts 2:43-44, 5:24) and the other two, although they are not the majorityreading, are clearly the reading of a Byzantine subgroup (Acts 10:10, 13:42).In the Apocalypse Schmid places 177 in the dominant or "a"group of the Byzantine text headed by 046. In addition to the NewTestament material, it contains the treatise by Dorotheus of Tyre (fl. c.360) on the Twelve and the Seventy (found also in 82,459, etc.). Scrivener reports that the text is "very near thatcommonly received." It also contains fragments of Eusebius'scanon tables (perhaps implying that it was once a complete New Testament);there are marginal scholia on Paul from a later hand.

Manuscript 179

Rome, Angelicus Library 11. Soden'sε211.Contains the gospels with lacunae. Dated paleographicallyto the twelfth century. Classified asIφb;other manuscripts of this group include 7 115 179 185(part) 267 659 8271082(part) 1391(part) 1402(part) 1606(part). This classification isnot confirmed by Wisse, who lists 179 as Mix/Kx/Kxand seems to dissolve theIφ groups(except for Iφr).The Alands do not place 179 in any Category,implying that they agree with Wisse's classification as mostly but notpurely Byzantine. The lectionary lists in 179are in a later hand (fifteenth or sixteenth century) on supplied leaves.Seven other leaves (five at the end) are also from later hands.

Manuscript 180

Rome, Vatican Library Borgiae Gr. 18. Soden'sε1498,α300; Tischendorf/Scrivener180e, 82a, 92p, 44r.Contains the New Testament complete. Includeslectionary apparatus. The gospels, which were written by oneAndreas, are dated paleographically to the twelfth century (so Aland;Scrivener says XI, and Gregory proposed XIV). The remainder of the NewTestament (with some additional material) were written by John, evidentlyin November 1273. The gospels are classified as Kxby von Soden (this seems to have been the only section he examined), and thisis confirmed by Wisse, who places it in Kx Cluster 180 inthe two chapters profiled. Other members of Cluster 180 are 998 and 1580.The Alands also confirm that 180 is Byzantine in the Gospels, where theyplace it in Category V. They also classifyit as Category V in Paul, the Catholic Epistles, and the Apocalypse(in the latter it goes with the largest "a" Koine group headedby 046); in the Acts, however, they raise it to Category III.
If anything, that figure is low. 180 has 24 readings which agree withUBS against the Majority Text, as opposed to 38 readings which agree withthe Majority against UBS -- by itself, a ratio high enough to probablyindicate Category II. But 180 has fully 21readings (four of them singular) which agree with neither. Thus 180 has45 non-Majority readings, only 38 Majority readings. Those 21 readings aresurely worth examining.
Acts 2:30 -- singular (but very close to 94 307 453 610 2818)
Acts 2:31 -- with 94 other manuscripts including C3 E 5 94 104 307 322 323 326 429 453 522 610 629c 630 945 1739 1891 2200 2344 2412 2818
Acts 2:43-44 -- with 94 180 307 453 610 1501 1678 2818
Acts 3:21 -- singular (but close to ℵc Bc E 94 363 610 945 1106 1409 1678 1884 2818 and not too far from the Alexandrian reading, which is also that of 453 1739 etc.)
Acts 4:33 -- with 94 307 453 610 629 1678 2818
Acts 8:37 -- have the longer reading, using the specific form found also in 307 453 610 2818
Acts 8:39 -- with 30 other manuscripts including Ac 94 307 322 323 453 610 876 945 1678 1704 1739 1891 2298 2818
Acts 10:12 -- with 307 453 1678 2818 and close to about 30 others including C* 5 33 94 322 323 424c 623 2298
Acts 10:19 -- with P74 ℵ A 6 69 81 94 181 307 431 453 610 1175 1642 1678 1875 2818
Acts 10:47 -- with 33 other manuscripts including Ec 6 94 181 307 322 323 429 453 522 610 630 945 1678 1704 1739 1891 2200 2298 2818
Acts 12:25 -- with 53 other manuscripts including D Ψ 94 181 307 436 441 453 610 614 623 1611 1678 2138 2412 2818
Acts 15:34 -- with 87 other manuscripts including 5 6 33 88 94 307 323 383 44 453 610 614 623 1175 1678 1739 1891 2298 2344 2412 2818
Acts 16:28 -- with 94 180 307 431 468 489 610 629 927 1319supp 1678 1729 1780 1868 2242 2502 2818
Acts 18:17 -- with 94 180 307 431 453 610 1319supp 1678 2818
Acts 19:3 -- with D Ψ 94 180 206 429 522 630 945 1490 1509 1704 1739 1751 1758 1831 1891 2080 2200 2298 2718
Acts 20:24a -- singular (adding one or two short words to various common readings)
Acts 23:20 -- with 85 other manuscripts including ℵ2 Ψ 6 94 307 424c 436 441 610 614629c 876 945 1175 1505 1611 1678 1704 1739 2138 2298 2412 2495 2818
Acts 23:30 -- singular
Acts 24:6-8 -- with E 628 1161 1409 1884 2541 (but similar to a great many other readings; the agreements here probably aren't genetic)
Acts 25:5 -- with 68 other manuscripts including Ψ 6 69 94 206 307 429 441 453 522 610 614 623c 1505 1611 1678 1852 2138 2412 2495 2818
Acts 25:17 -- with C 307 431 453 610 614 1292 1505 1611 1678 1890 2138 2147 2412 2495 2652 2818
Looking at this, and at the Alands' overall agreement rates, it seems pretty clear that 180 is a member of family 453 along with such manuscripts as 36 (the Alands' 2818) 94 307 453 1678.

Manuscript 181

Rome, Vatican Library Reg. Gr. 179. Soden'sα101,α1578; Tischendorf/Scrivener40a, 46p, 12r.Contains the Acts, Epistles, and Apocalypse. The basic run of the text,containing the Acts and Catholic Epistles, plus Paul through Titus 3:3, isdated to the eleventh century. The remainder of the text (Titus 3:3-end,Philemon, and the Apocalypse) was supplied in the fifteenth century.The text is arranged according to the Euthalian edition, and so isclassified by Von Soden as Ia1 -- most of the other members of this group(which contains 88 917 1898 throughout the Acts and Epistles, plus in the Acts andCatholics 36/2818 307 431 610 453 915 1829 1874, in Paul and the Catholics 1838,and 1912 in Paul alone) are also Euthalian (see Von Soden i.674). In Paul,however, 181 does not seem to be a good representative of the type; samplesindicate that its text is about 80% Byzantine, and there are hints ofblock mixture with the Byzantine text. In the Acts the text is noticeably better, and has a numberof Alexandrian readings. The Alands place 181 in Category IIIin the Acts and Epistles, V in the Apocalypse (though their numbers in the Catholicsbarely qualify it for that category, and it does not appear in Wachtel's lists. Clearly181 is better in the Acts than elsewhere). The later additions to the manuscriptare classified as Ia2 by Von Soden; in the Apocalypse it has anAndreas type of text, forming part of the groupwhich also contains 1 598 2026 2028 2029 2031 2033 2038 2044 2052 20542056 2057 2059 2060 2065 2068 2069 2081 2083 2186 2286 2302. 181 itself,however, does not have the text of the commentary. It does have lectionaryapparatus but no synaxarion. We first hear of the manuscript during thepapacy of Alexander VIII (1689-1691), when Christina presented it to thatpope.
I question von Soden's association of the manuscript with Ia1, even inActs. Glancing at the Alands' agreement numbers, the only substantial manuscript it agreeswith more than 55% of the time is 1875 (with which it agrees a still-low 81% of the time). 181's statistics are fascinating: it has 34 agreements with UBS against Byz, 18readings which agree with both, just 28 readings that go with Byz against UBS --and 24 readings which agree with neither (four of them singular); this is usually a signof an interesting manuscript. They can also be a fast way to look for relatives. The 24 readings are as follows:
Acts 2:23 -- singular (probably a copyist's error)
Acts 2:30 -- with 33 61 88 637 915 917 1874 1877 2737
Acts 2:31 -- with 88 296 390 431 432 614 915 1162 1292 1751 1827 1831 1838 2718 2805 2815
Acts 2:43-44 -- 181* with E 33 104 459 1409 1722c 1838 1842c 1884 2344 2774 (181c singular)
Acts 2:46 -- with 1319
Acts 2:47-3:1 -- with 914 1398
Acts 5:24 -- singular but close to Byz
Acts 10:19 -- with P74 ℵ A 6 69 81 94 180 307 431 453 610 1175 1642 1678 1875 2818
Acts 10:25 -- with 1875 (but the reading is essentially that of the Byzantine text except for a small portion of a long expansion found also in D and only D)
Acts 10:39 -- singular but close to 1875
Acts 10:47 -- with 43 other manuscripts including Ec 6 94 180 307 322 323 429 453 522 610 630 945 1678 1704 1739 1891 2200 2298 2818
Acts 12:25 -- with 53 other manuscripts including D Ψ 94 180 307 436 441 453 610 623 1611 1678 2138 2412 2818
Acts 15:17-18 -- with 1877
Acts 15:34 -- with 1875 (but close to the readings of at least 89 other manuscripts including most notably 33 453 614 1175 1739 2344 2412)
Acts 16:28 -- singular (but most of the variants here are just changes in word order and may not be genetic)
Acts 16:33 -- with 1875
Acts 19:3 -- with 1875
Acts 19:14 -- with E Ψ 629 1319supp 1409 1875
Acts 20:24a -- with ℵ2 A 1735c 1875 2344
Acts 21:25 -- with C E Ψ 103 307 453 606 641 913 919c 1162 1505 1611 1830 1853 1884 2138 2473 2495 2544
Acts 23:30 -- with ℵ A E 81 206 429 436 441 522 621 1175 1505 1509* 1739 1751 1842* 1852 1875 1884 1890* 1891 1894 2138 2200 2495
Acts 24:6-8 -- with 1875 and close to 6 69; this is a very fractured reading where exact agreements probably have limited genetic significance
Acts 25:17 -- with 33 other manuscripts including B 056 and a lot of minuscules with no known significance; it appears to be an accidental omission picked up by some Byzantine subgroups
Acts 27:14 -- with Bc 256 365 676 1405 1829* 1875
It should be said that the Alands' readings are not a good classification tool in Acts -- too many of the readings fracture in non-genetic ways, and there is no good way to test the strands of the Byzantine text. But the impression I get of 181 is of a badly-copied manuscript (although the errors might be in 181's exemplar, not 181 itself) with block mixture -- it seems to be close to 1875 in the middle chapters of Acts, but not in the early (although 1875 is defective for some of those). Certainly 181 does not belong to any established group, whether Alexandrian, Western, 1739-type, 614-type, 453-type, or other. Its lack of agreements with ℵ and (especially) B in its non-UBS/Byz is particularly noteworthy, but I suspect that's an artifact of the Alands' biased sample.

Manuscript 185

Florence, Bibl. Laurenz. VI.16. Soden'sε410.Contains the gospels complete. Dated paleographicallyto the fourteenth century (Scrivener says twelfth). Classified by von Soden asIφb(but in John only); other manuscripts of this group include 7 115 179185(part) 267 659 8271082(part) 1391(part) 1402(part) 1606(part). This classification isnot confirmed by Wisse, who lists 185 as Cluster 1531 along with suchmanuscripts as 1531, 2291, 2387, and 2771.The Alands list 185 as Category V(Byzantine). It should be noted, however, that neither Wisse northe Alands examined readings in John; thus its text has notbeen fully examined. It is at least possible that it is block mixed,with a better text in John than elsewhere.Physically 185 is not noteworthy; it haslectionary indications and the Ammonian Sections but not theEusebian apparatus.

Manuscript 189

Florence, Bibl. Laurenz. VI.27. Soden'sε1401,α269;Tischendorf/Scrivener 189e, 141a, 239p.Contains the Acts and Epistles complete and the gospels with lacunae(lacking John 19:38-end). The Acts and Epistles are dated paleographicallyto the twelfth century, and the Gospels to the fourteenth (exceptthat Scrivener dates the whole to the twelfth century). The gospelsare classified as Kr by Von Soden, and this is confirmedby Wisse (who further classifies 189 as Cluster 189 along with 1236,1625, and perhaps 825). This is consistent with the marginal apparatusof 189, which lacks the Ammonian/Eusebian material. The Alands alsoconcur in regarding it as unremarkable,describing 189 as Category V(Byzantine). Outside the gospels, the Alands still list 189 asCategory V, agreeing with Von Soden's"K" classification. Even among Category V texts, it seemsvery un-notable; it has only one reading which agrees with UBS against themajority, and nine readings which agree with neither -- but in six of the ninecases it agrees a major fraction of the Byzantine text against the majority,and the others are mostly readings where there are many very similar readings.So those allegedly "non-Majority/non-UBS" readings are mostly Byzantine.The manuscript has the Euthalianapparatus (though not the arrangement or text). Scrivener describesthe manuscript itself as "minute [certainly true; it measures12 cm. x 7 cm.] and beautifully written."

Manuscript 201

London, British Museum Add. 11837. Soden'sδ403;Tischendorf/Scrivener 201e, 91a, 104p,94r; also mscr (Gospels); pscr (Acts/Paul);bscr (Apocalypse).Contains the compete New Testament. Dated by a colophon to 1357. The gospelsare classified as Kr by Von Soden, and this is confirmedby Wisse (who notes that it is a "perfect member" of the group).The Alands also concur, listing 201 as Category Vin all sections, and their data for Acts seems to confirm this -- it has only fournon-majority readings -- and in those four, it has 110, 162, 87, and 181 allies, meaningthat although it sometimes lacks the majority reading, it always has the reading of atleast one large Byzantine subgroup; clearly it is entirely Byzantine.Wachtel lists it as a member of Kr in theCatholics. In the Apocalypse, Schmid places it in the dominant or "a"group of the Byzantine text headed by 046.Scrivener says of it that it has "many changes by a later hand;"it also has a very full marginal apparatus, including prologues, subscriptions,and stichoi lists, plus "some foreign matter." Rather curiouslyfor a Kr manuscript, it has the Ammonian Sections and "some"of the Eusebian numbers.

Manuscript 203

London, British Museum Add. 28816. Soden'sα203;Tischendorf/Original Gregory 203a, 477p, 181r;Scrivener 232a, 271p (Acts/Paul), 107r.Contains the Acts, Epistles, and Apocalypse with lacunae (lacking 1 Cor. 16:15-endplus the prologue to 2 Corinthians; Eph. 5:3-6:16 is supplied in a fifteenthcentury hand). At the end of the volume are ten pages of non-Biblical material(in the original hand).These include a list of the errors condemned by the seven ecumenical councils;Scrivener says that this resemble the exposition in 69.Dated by a colophon to 1111.Von Soden classifies the manuscript as Ic2 in the Acts and Epistles(though he cites it only in Paul, where the other members of the group include221 257 378 383 385 506 639 876 913 1610 1867 2147). This group is of someinterest in the Catholic Epistles (where many of its members are part ofFamily 2138), but in Paul theyseem generally to be of limited value. In addition, 203 at least has very limitedvalue in Acts; according to the Aland statistics, it has only three non-Byzantinefive non-Byzantine readings -- three that agree with UBS and two that do not.In all but one of these five readings (Acts 25:5), 203 has atleast 80 manuscript allies (meaning that it belongs to with a Byzantine subgroup evenif its reading is not that of the majority of manuscripts) -- and even the one exception,in 25:5, is a mere change in word order. Its Byzantine nature is confirmed by the Alands, who place 203 in Category V.In the Apocalypse, Schmid places it in the dominant or "a"group of the Byzantine text headed by 046.Scrivener says of it that it is "a splendid copy," with"many marginal glosses in a very minute hand." It hasthe κεφαλαιαnumbers in red in the margins and the entries themselves beforeeach epistle. It has the Euthalian apparatus, and Arethas'sprologue and tables on the Apocalypse. It has lectionary indicationsbut no τιτλοι. The scribe was namedAndreas.
Sample plate in Edward Maunde Thompson,An Introduction to Greek and Latin Paleography (plate 67).

Manuscript 205

Venice, Bibl. San Marco 420 (Fondo ant. 5). Soden'sδ500;Tischendorf/Scrivener 205e, 93a, 106p88r.Contains the complete New Testament and the Greek Old Testament. Datedpaleographically to the fifteenth century.The text of 205 has long been recognized as being very close kin to theearlier 209 (at least in the Gospels). The two are such close kin thatseveral scholars, starting with Rinck, have believed that 205 is acopy of 209. Burgon offered the theory that both were copied from thesame uncial ancestor. While the manner has not been definitively settled,the modern opinion seems to be that 205 is not copied from 209, butthat they have a close common ancestor.209, of course, is known to be a member of Family 1;it therefore follows that 205 must also be part of this group. VonSoden acknowledges this by placing 205 in theIηgroup (Family 1; 209 is a member of the "b" subgroup), andWisse concurs, going so far as to say "Pair with 209." (Curiously,the Alands do not list 205 as a member of Family 1, and even insist onciting 205 separately in SQE13. They do list both manuscriptsin the same Categories: Category III inthe Gospels and Apocalypse; Category V in the Acts and Epistles.)In the Acts and Epistles, 205 is listed by Von Soden as Ia(again agreeing with 209, which is Ia3). The data of theAlands, however, clearly implies that 205 is Byzantine (rather thanlate/mixed Alexandrian, as Von Soden's classification would imply).This also means that we cannot determine the manuscript's relationshipwith 209 without detailed examination.
Looking at the Aland data in Acts, we find that 205 and 209 each hasonly two readings which agree with UBS against the majority text, and 205 and 209share only one of these:
* 209* agrees with UBS against the majority in Acts 10:32, but 209c 205 agree with the majority
* in Acts 15:7, 205 and 209 both go against the majority reading, but 161 other manuscripts share this reading, so while it is not the Majority reading; it is clearlythe reading of a substantial subgroup of the Byzantine text
* 205 agrees with UBS against the 209* and majority in Acts 21:8 -- but here 209c with 205, and in any case 180 other manuscripts against the majority, so again, it is a Byzantine reading even if not the majority reading.
If we look at the readings of 205 and 209 which agree neither with UBS nor the majority, they are as follows:
Acts 2:46 -- 205 209 with 3 582 796* 914 1240 1767 2494 (but there are many minor variants in this text; it probably has limited genetic significance)
Acts 3:21 -- 205 209 with 3 450 582 592 914 1240 1886
Acts 9:25 -- 209* with 049 94 307 429 453 522 610 628* 630 636 1066 1490 1642 1721 1749 1751 1758 1831 2200 2511 2818; 205 209c with Byz etc.
Acts 10:47 -- 209 agrees with only 2501; this is a word order inversion of the reading of 205 and 33 other manuscripts including Ec 6 94 180 181 307 322 323 429 453 522 610 945 1490 1739 1758 1891 2200 2298 2818
Acts 15:34 -- 205 209 with 86 other manuscripts including 5 6 33 88 94 180 307 323 383 441 453 610 614 623 1739 1891 2298 2344 2412 2818; this is probably another case where the Byzantine text splits and 205 209 derive their reading from a Byzantine source that just didn't end up being the majority text
Acts 20:15 -- 209 with the Majority Text; 205 with 45 other manuscripts including D* 69 582 796c 2412; it looks as if 205 agrees with a Byzantine subtype against the majority.
Acts 20:24a -- 209 with the Majority Text; 205 is singular
Acts 21:20 -- 205 209c with the Majority Text; 209* with ℵ 3 97* 457* 1409 1642 1717 2243 2816*
Acts 23:20 -- the Byzantine text here splits into two strands, with 209 agreeing with one of the two and 205 not, but the two differ by only one letter μελλοντες versus μελλοντας, and 205 has 24 allies which read the latter; I doubt this reading has genetic significance.
Acts 25:17 -- 205 209c with the majority text; 209* with 33 other manuscripts including B 056 3 42 181 1642 1886
This isn't much to go on, but it gives me the feeling that 205 and 209 were both Byzantine texts, but of slightly different types, with one of 205's ancestors corrected toward the same type that was used to correct 209, and that this involved a slightly unusual Byzantine group that might have involved such manuscripts as 3 582 1642 1886. But this is very tentative, and in any case, both 205 and 209 are too Byzantine in Acts to be of much importance for criticism. 97
In the Apocalypse, Von Sodenlists 205 as an Andreas manuscript, even though it lacks the commentary.Physically, 205 is a rather large volume but with limited marginalia;it lacks the entire Eusebian apparatus (209, by contrast, has theAmmonian sections but not the Eusebian canons) as well as all lectionarydata. It has the κεφαλαιαin both Greek and Latin, subscriptions, and prologues to the Pauline andCatholic Epistles. It was written for Cardinal Bessarion, probablyby his librarian John Rhosen. A copy of 205 exists; now designated205abs, it is Tischendorf/Scrivener 206e,94a, 107p, 101r. (Note: It is theopinion of most examiners that 205 is the original and 205absthe copy; Maurice Robinson, however, based on the text in the storyof the Adulteress, believes that 205abs is the originaland 205 the copy.) For more details on the text of 205, see the entry on1 and Family 1.

Manuscript 206

London, Lambeth Palace 1182. Soden'sα365;original Gregory 214a, 270p;Scrivener 182a, 252p, ascr;Hort 110.Contains the Acts and Epistles with minor lacunae and many latersupplements; Acts 1:1-12:3, 13:5-15, 2 John, 3 John, and Jude arefrom a later (fourteenth century) hand.Dated paleographically to the thirteenth century (except that Scrivener,who probably examined it most fully, says twelfth). Scrivener reports that the readings in Acts"strongly resemble those of [429], and [81] hardly less,especially in [chapters 13-17]."Von Soden lists the text of 206 as Ib1, placing it with242 429 491 522 536 1758 1831 1891 in Acts (1739 2298 323, itshould be noted, are key members of Ib2, so von Soden is placing206 and its relatives in a group similar the the 1739 text); in Paul thegroup members include 2 242 429 522 635 941 1099 1758 1831 1891;in the Catholics 206 is listed along with 216 242 429 440 522 1758 1831 1891.
This classification (rather typically of Von Soden's groups) is indicativebut probably not entirely accurate.
More recently, Thomas C. Geer, Jr., in Family 1739 in Acts, studies 206 (amongothers), and finds that 206 is a member ofFamily 1739 (along with323 429 522 1739 1891; Geer does not examine the other members ofvon Soden's Ib group). Within Family 1739, the closestrelatives of 206 are 429 and 522. (This is confirmed by the Alands, who list429 and 522 as the closest relatives of 206, agreeing fully 95% of the time;nothing else agrees as much as 90%). Geer does not compare the firsthand of 206 with 206supp, but he does compile separatestatistics for the first and second halves of Acts. It is worth notingthat, in chapters 1-14, 206 agrees only 81% of the time with 429, and75% of the time with 1739 (Geer, p. 69), while in Acts 15-28, itagrees with 429 fully 93% of the time (though still only 77% of thetime with 1739). Thus it appears quite likely that the supplement in206, while having perhaps some kinship with the family text, has beenheavily influenced by the Byzantine text. The original hand, by contrast,clearly goes with 429 522 630 2200.
To this we can add that 206 429 522 630 2200 are also akin in the Catholic Epistles. But in the Catholic Epistles,instead of being members of Family 1739 (which, it should be noted,is even more distinctive in the Catholics than in Acts), the 206-groupare members of Family 2138.This kinship has been confirmed by all who have investigated the matter;Wachtel places 206 in his group Hkgr along with 429 522630 2200 (plus such important manuscripts as 614 1505 1611 1799 2138 2412 2495,which are not related to the 206-429-522-630-2200 group in Acts).Similarly, Richards places 206 in his A1 group along with614 1611 1799 2138 2412 (in 1 John; the supplements in 2 and 3 John Richardsfinds to be Byzantine). And Amphoux places 206 in Family 2138 (along withnearly all the above manuscripts, plus such others as 1108 and 1518).
Knowing that, we need to step back from Geer's conclusion, because in Acts, however, there is a kinship between Family 1739 andFamily 2138. They are notthe same family, but there is some overlap. This relationship definitelyneed more examination than it has gotten. The result mostly shows a methodologicaldefect in Geer -- because he compared 206 (and 429 522 etc) to 1739 etc., butnot to 614 1505 1611 2138 2412 etc., he came to a false result.
In Paul, 206 has not been as heavily studied; our best information comesfrom the Alands, who list 206 as Category V(they list it as Category III in the Catholics -- along with allthe other members of Family 2138; in Acts, they list 206 as Category V,but here we see their prejudices -- or the effects of the supplement, which intheir sample is almost purely Byzantine --affecting them; what survives of the original text of206 has only five readings which agree with UBS against theMajority Text, but of its 61 readings, fully 13 agree with neither UBS nor themajority; clearly it deserved better than to be branded Byzantine). 429 and 522 are alsoCategory V in Paul; it thus appears likely that these three manuscripts arerelated throughout. (630 and 2200 are not wholely Byzantine inPaul; in the latter books, they are Byzantine, but in Romans throughGalatians they are weak members of Family 1739. In addition, they appearto be closer to 1739 in Acts. Thus 630 and 2200 might possibly representa forerunner of the 206-429-522 text, but are not actually part of it.)Physically, Scrivener reports of 206 that it has Paul before theCatholic Epistles, that it is illustrated, that it has full lectionaryapparatus, and that it includes antiphons for Easter and "otherforeign matter." It is said to have come from a Greek island.See also the discussion on 429 or on522.

Manuscript 213

Venice, Bibl. San Marco 542 (Fondo ant. 544). Soden'sε129.Contains the Gospels with mutilations (John 18:40-end have beenlost). Universally dated to the eleventh century. Classified by Von Soden asI0 -- a group which contains a very mixed bag ofmanuscripts: U X 443 1071 1321(part) 1574 2145.Wisse classifies 213 as mixed throughout. The Alands do notassign it to any Category.Some of the confusion may be due to a poor scribe; 213 hasmany strange properties. Scrivener notes "heroic versesas colophons to the Gospels," "[l]arge full stops inimpossible places," the Ammonian/Eusebian apparatus"most irregularly inserted," and only scatteredlectionary indications.

Manuscript 223

Location/Catalog Number

Ann Arbor. Catalog number: University of Michigan MS. 34. Formerly Burdette-Coutts III 1; the University acquired it from the Burdette-Coutts estate in 1922. It was originallyacquired at Janina in Epirus.


223 contains the Acts and Epistles, with some minor defects (in Paul,2 Corinthians 1:1-3, Eph. 1:1-4, Hebrews 1:1-6 are missing; Scrivener believesthey were cut out for the sake of the illuminations). Philemon precedes Hebrews.It has κεφαλια for Acts only, but υποθεσις (often more than one) for all books,and concludes with a lectionary table.It is writtenon parchment, 1 column per page. The parchment is of excellent quality,and the manuscript has many colorful illuminations (apparently before everybook except 2 Corinthians, Ephesians, and Hebrews, all of which, we note, havehad their first leaves removed), implying that unusualeffort and expense was devoted to its preparation. According to Clark,the book titles are in gold. Scrivener says of it,"This is one of the most superb copies extant of the latter part of theN.T., on which so much cost was seldom bestowed as on the gospels."


Dated paleographically to the fourteenth century. A colophon at the endof Jude states that it was written by Antonios of Malaka, who is also creditedwith writing 1305 (dated by its colophon to 1244) andℓ279 (datedpaleographically to the twelfth century). The dating of the manuscript isthus problematic. Gregory, perhaps on the basis of the colophon, gave a dateof the eleventh or twelfth century. It is noteworthy, however, that the colophon of223, although in gold, is not in the hand of the original scribe. Clark,after considering multiple explanations, tentatively suggestson p. 9 of Eight American Praxaposoloi that the explanation is that thecolophon was copied in its entirety from some other manuscript (possibly but notcertainly the exemplar) but suggests that there is a need to compare 223, itscolophon, 1305, and ℓ279 in detail.

Description and Text-type

Von Soden lists 223 as Kc. Clark and his collaboratorsquestioned this, since von Soden's collation was highly inaccurate.However, spot checks indicate that 223 possesses about 70% of the characteristicreadings of Kc. Thus it is likely that it is at least a weakKc witness. Indeed, looking at the Acts data, I think it is a goodwitness of the type and the problem is von Soden, not 223.

Aland and Aland list 223 as Category V,i.e. Byzantine. This is clearly correct (at least if one accepts Kcas Byzantine); in Acts, for instance, it has no readings at all that agree withUBS against the Majority text -- although it has nine readings which agree withneither. However, in six of these, it belongs has many allies although they do notrepresent the majority (Acts 2:31, 94 allies; Accts 4:33, 73 allies; Acts 12:25, 60allies; Acts 13:42, 101 allies; Acts 15:34, 33 allies; Acts 25:5, 68 allies); the otherthree readings (Acts 10:11, 15:17-18, 24:6-8) represent places where it either differstrivially from the majority or where the tradition is very fractured. It is the pictureof a manuscript belonging to a Byzantine subgroup, in this case surely Kc.

Richards lists 223 as belonging to his B3 group in the JohannineEpistles, having all nine of the characteristic readings in 1 John. Othermembers of this group, with von Soden's classification of them, are97 (K), 177 (rather weakly, K), 1597 (Kx), 1872 (Ib2,but Kc in r), and 2423.

Other Symbols Used for this Manuscript
von Soden: α186. Scrivener: 220a; 264p.Tischendorf: 223a; 278p


K.W. Clark, Eight American Praxapostoloi (1941).
F. H. A. Scrivener, Adverseria Cristica Sacra (1893); cited as "a"

Sample Plates:
Metzger, Manuscripts of the Greek Bible (1 page)

Editions which cite:

Other Works:
Kenneth W. Clark, A Descriptive Catalogue of Greek New Testament Manuscripts in America (1937), pp. 312-313 (description of form and contents)

Manuscript 225

Naples, Bibl. Naz., Cod. Vein. 9. Soden'sε1210.Contains the Gospels complete. Dated by its colophon to 1192.The manuscript is among the smallest known, measuring less than 14cm. by 10 cm. Perhaps to accommodate such a pocket edition,the Eusebian and Ammonian apparatus are omitted, as are most otherreader helps except the lectionary markings (the manuscript issupplied with pictures, however). Classified by Von Soden asAk -- a group which also contains5, 15, 32, 53, 169, 269, 292, 297, 416, 431, 448,470, 490, 496, 499, 534, 546, 558, 573, 715, 752, 760, 860, 902,946, 968, 976, 987, 1011, 1015, 1058, 1091, 1163, 1167, 1171,1211, 1227, 1291, 1299, 1321, 1439, 1481, 1484, 1498, 1566, 1800,2142, and 2176. These manuscripts are, however, mostly Byzantine,and Wisse largely disregards this group. 225 itself he classifiesas Kmix/1167/1167; other members of Group 1167 include 75 116(part)245(part) 431 496 546 578(part) 843 896 951 1015 1167 1242(part) 14381479(part) 1511(part) 1570 2095(part) 2229 2604. The Alands more orless confirm that 225 is Byzantine but not a mainstream witness to thetype by refusing to assign it to a Category.The most noteworthy thing about 225's text, however, is where it placesthe story of the Adulteress (John 7:53-8:11). Alone among all knownwitnesses, it places the story after John 7:36.

Manuscript 229

Escorial X.IV.21. Soden'sε1206.Contains the Gospels with lacunae (lacking Mark 16:15-20, John 1:1-11).Dated by its colophon to 1140.Classified by Von Soden asIkc -- i.e. as a offshoot of FamilyΠ; othermembers of this group include 280 473 482 1354.Wisse, however, reports that 229 is block mixed; it isΠain Luke 1, Kx in Luke 10 and 20. The Alands do notassign it to a Category; thisperhaps implies that the Family Πelement predominates, as they usually classify Kx witnessesas Category V but leave Family Πwitnesses unclassified. Scrivener notes that it was written by "BasilArgyropolus, a notary." It includes pictures. A later handhas added lectionary indications and retraced parts of the text, aswell as correcting various readings (apparently correcting theFamily Π texttoward the Byzantine mainstream, as Scrivener reports that theoriginal readings resemble those of A and K, both of which areassociated with that family.)

Manuscript 235

Copenhagen, Kgl. Bibl. GkS 1323, 40. Soden'sε456.Described by Scrivener as "written by theιερομοναχοςPhilotheus, though very incorrectly; the text agrees much withCodd. DK. i. 33 and the Harkleian Syriac.... [T]he words areoften ill-divided and the stops misplaced." The kinshipwith these manuscripts is, however, at best very weak; Von Sodenlists it as Iσ(along with 157 245 291 713 1012), but cites it only for John. Wisselists it as Kmix/Kx/Kx, and the Alands alsoregard it as Byzantine, listing it asCategory V.

Manuscript 245

Moscow, Historical Museum V.16, S.278. Soden'sε1226.Dated by its colophon to the the 1199. Written by"John, a priest" and formerly kept at the monastery ofBatopedion. Von Soden categorizes its text asIσ;other manuscripts of this type include 157 235(John) 291 713 1012.Wisse lists the text as Kmix/1167/1167. The members of Group 1167do not correspond to those of Von Soden's group.Whatever its exact type, it seems certain that the manuscriptis primarily Byzantine, and this is reflected by the Alands,who list it as Category V.

Manuscript 249

Moscow, Historical Museum V. 90, S.93. Soden'sNι10.Contains the Gospel of John (only), with a catena. Its datingvaries wildly; Aland says XIV, Scrivener XI. Von Soden's numberimplies that he agrees with Scrivener. Von Soden lists it as havingNicetas's commentary on John, assigning its symbol on this basis(other manuscripts with this commentary include 317 333 423 430 743).Merk lists the text-type as K (Byzantine). Little else can be said ofit; the Alands do not assign it to a Category(presumably because it contains only John, and they tested only Matthewthrough Luke), and Wisse of course does not profile it. Originallyfrom Mount Athos.

Manuscript 251

Moscow, Russian Gosud. Library Greek 9. Von Soden'sε192.Contains the gospels complete. Dated paleographically to the eleventh(Scrivener, von Soden) or twelfth (Aland) century. Von Soden lists itas a member of I' (the vaguest of all the I groups, containing a handfulof Byzantine uncials, assorted uncial fragments -- not all of which areByzantine -- and many mostly-Byzantine minuscules). Wisse lists 251 asa member of Cluster 1229, the other members of this group being 1229(which, like 251, von Soden lists as I') and 2487. The Alands do notassign 251 to a Category, implying thatit contains at least some readings (though not many) which are notpurely Byzantine. Physically, 251 has the Eusebian tables and Ammoniansections, but not the Eusebian marginalia; these perhaps were neverfinished. 251 has illustrations, but no lectionary equipment.

Manuscript 262

Paris, National Library Greek 53. Soden'sε1020.Contains the Gospels complete, though the marginalia seem notto have been completed; Scrivener reports that it has"some" τιτλοι.The Ammonian and Eusebian apparatus (including harmonizations) arecomplete in Matthew and Mark, but only partial, and in a laterhand, in Luke and John. 262 is universally dated to the tenth century.Scrivener observed a similarity to Λ,and this is confirmed both by Von Soden (who places it in the Irgroup with Λ 545 11871555 1573) and Wisse (who makes it a core member of Group Λ).The Alands assign it to Category Vas Byzantine.

Manuscript 263

Paris, National Library Greek 61. Soden'sδ372.Contains the Gospels, Acts, and Epistles complete. Dated paleographicallyto the thirteenth century. The text is generally uninteresting; in theGospels, von Soden listed it as K1, which Wisse corrects minimallyto Kx, and the Alands list it asCategory V (Byzantine). The Alands alsoplace it in Category V in the Acts and Catholic Epistles, though VonSoden listed it as Ia3, based probably on the text of Paul.The Alands appear to be right, though; in Acts, 263 has only two readingswhich agree with UBS against the Majority Text, and three others which agreewith neither (one of them singular).The one exception to this trend of ordinariness is in Paul. Here theAlands promote it to Category III, and Von Soden's Ia3classification makes somewhat more sense. Bover, in particular,specifies it as a member of "Family 1319" (for which see theentry on 365 and Family 2127) -- and while 263 doesnot seem as good as the leading members of the family (256, 365, 1319, 2127),there does seem to be kinship. Scrivener believed the manuscriptcame from Asia Minor, and this is perhaps reasonable for a text somewhatrelated to the Armenian version. In the Gospels, it has Ammonian Sectionsbut not the Eusebian equipment, and lectionary indications but no tables.

Manuscript 265

Paris, National Library Greek 66. Soden'sε285.Contains the Gospels complete. Generally dated to thetwelfth century, though Scrivener lists the tenth. Classified byVon Soden as Ika,i.e. as a member of the main Family Πgroup, along with such manuscripts as A K Y Π This is confirmedby Wisse, who lists it as a core member of the mainΠagroup. The Alands do not place it in any Category;this is fairly typical for Family Πmanuscripts. Physically, the manuscript has the Eusebian apparatusbut not much else; lectionary equipment is lacking.

Manuscript 267

Paris, National Library Greek 69. Soden's ε1289.Contains the Gospels with minor lacunae (missing Matt. 1:1-8,Mark 1:1-7, Luke 1:1-8, Luke 24:50-John 1:12 -- perhaps cut outfor the sake of illustrations or the like?). Generally dated to thetwelfth century, though Scrivener lists the tenth. Classified byVon Soden as Iφbalong with such manuscripts as 7 115 179 185(part) 659 827 1082(part) 1391(part)1402(part) 1606(part). That it is close to 7, at least, is confirmed by Wisse,who places 267 in Cluster 7 along with 7, 1651, and 1654.The Alands place 267 in Category V(Byzantine). The manuscript is slightly unusual in having the Ammonian andEusebian numbers in the same line.

Manuscript 270

Paris, National Library Greek 75. Soden'sε291.Contains the Gospels complete. Generally dated to thetwelfth century, though Scrivener lists the eleventh. Classified byVon Soden as Ikb(i.e. as a member of one of the weaker subgroups of Family Π)along with such manuscripts as 726 1200 1375.Wisse confirms its kinship with the Πgroups, listing it as part of the b subgroup in Luke 1 and the a subgroupin Luke 10 and 20.The Alands place 270 in Category V(Byzantine). Curiously, Scrivener reports that the manuscript has bothsynaxarion and menologion (along with illustrations and the Eusebianapparatus), but no lectionary indications in the text.

Manuscript 273

Paris, National Library Greek 79. Soden'sε370.Contains the Gospels with some slight damage, most of it made good bya supplement. Dated paleographically to the thirteenth centuryby Aland and von Soden; Scrivener lists the twelfth century and datesthe supplements (which are on paper; the rest of the manuscript is vellum)to the fourteenth century. Classified byVon Soden as I', i.e. as one of the miscellaneous weak "Western"witnesses. Wisse, however, finds it to be mostly Byzantine; he listsit as Kmix/Kx/Kmix.The Alands do not place 273 in any Category,which usually means it is strongly but not quite purely Byzantine; thisperhaps supports Wisse's analysis. Scrivener lists it as having avery full marginalia (though some of the lectionary material is fromthe later hand), and says of it that is "contains also somescholia, extracts from Sererianus's commentary, annals of theGospels, a list of gospel parallels, with a mixed text."

Manuscript 280

Paris, National Library Greek 87. Soden'sε294.Contains the Gospels with some damage (Mark 8:3-15:36 aremissing). Dated paleographically to the twelfth century. Classified byVon Soden as Ikc(i.e. as a member of one of the weaker subgroups of Family Π)along with such manuscripts as 229 473 482 1354.Wisse confirms its kinship with the Πgroups, but lists it as a core member of the primary groupΠa.The Alands place 280 in Category V(Byzantine); this may indicate that it it less pure in the othergospels than it is in Luke (since the Alands usually do not assignΠamanuscripts to any category). However, it could also be an indicationof the Alands' lack of control of their Categories.

Manuscript 291

Paris, National Library Greek 113. Soden'sε377.Contains the Gospels complete. Dated paleographically to the twelfth (Scrivener)or thirteenth (Aland, von Soden) century. Written with silver ink, butwith relatively few reader aids (lectionary markings but no tables;no Ammonian or Eusebian apparatus). Classified byVon Soden as Iσ-- a strange mixed group containing also 157 235(part) 245 713 1012.Wisse however places 291 in its own Group 291, which he associates looselywith the Πgroups; other members of this group are 139 371 449 597 1235 13402346 2603 2728.The Alands place 280 in Category V(Byzantine).

Manuscript 304

Paris, National Library Greek 194. Soden'sCμ23,A215. Contains the gospels of Matthew and Mark (only),with commentary interspersed with the text. Dated paleographicallyto the twelfth (von Soden, Aland) or thirteenth (Scrivener)century.Classified by von Soden based on the commentary: He lists it as having the"Anonymous Catena" on Matthew (one of only three manuscriptsto have this commentary, the others being 366 and 2482) and the"Antiochene Commentary" of Victor on Mark. (Scrivener quotesBurgon to the effect that the commentary on Mark is a "modificationof Victor's," however.) The Alands list 304 asCategory V (Byzantine). Since the manuscriptdoes not include Luke, it has not been studied by Wisse, but there is noparticular reason to doubt the Alands' judgement. Thus there is no reasonto consider 304 particularly unusual -- except for the fact that it iscommonly cited in critical apparati (NA27, UBS4,etc.) as omitting the longer ending of Mark (16:9-20). Maurice Robinsonhas examined a microfilm of the end of the manuscript, however, and offersthese observations: "[T]he primary matter [in 304] is the commentary.The gospel text is merely interspersed between the blocks of commentary material,and should not be considered the same as a 'normal' continuous-text MS. Also, it is often verydifficult to discern the text in contrast to the comments....
"Following γαρ2at the close of [16:8], the MS has a mark like a filled-in 'o,'followed by many pages of commentary, all of which summarize theendings of the other gospels and even quote portions of them.
"Following this, the commentary then begins to summarize theετερον δε ταπαρα του Μαρκου,presumably to cover the non-duplicated portions germaneto that gospel in contrast to the others. There remain quotes andreferences to the other gospels in regard to Mary Magdalene, Peter,Galilee, the fear of the women, etc. But at this point the commentaryabruptly ends, without completing the remainder of the narrative or theparallels. I suspect that the commentary (which contains only Mt and Mk)originally continued the discussion and that a final page or pages at theend of this volume likely were lost.... I would suggest that MS 304 shouldnot be claimed as a witness to the shortest ending...."

Manuscript 307

Paris, National Library Coislin Greek 25. Soden'sAπρ11;Tischendorf/Scrivener 15a.Contains the Acts and Catholic Epistles complete. Dated paleographicallyto the tenth (Aland) or eleventh (Scrivener) century. Commentary manuscript,described by both Von Soden and Scrivener as that of Andreas the Presbyter.Von Soden classified it as Ia1 (along with 36ac 88 181 307 431 453610 915 917 1829 1836(caths only) 1874 1898). Some of these manuscriptsprobably are not allies of 307, but at least some are; an examination of thedata in the UBS4 apparatus to Acts shows that 36, 307, 453, 610, and1678 (all Andreas manuscripts) agree over 90% of the time (and 100% or nearly in non-Byzantinereadings; for details, see the entry on 453). Geer, based on thedata compiled by the Alands (who classify 307 as Category III), notes a very high agreementof 307 with 453 and 2818 (the new number for 36). The Alands' numbers show that 307'sclosest ally among substantial manuscripts is another Family 453 manuscript, 610 (94%agreement), followed by 36=2818 (93%), 453 (92%), and 1678 (88%); no other substantialmanuscript agrees more than 80% of the time in their sample. Thus there appears nodoubt that 307 is a member of the 453 group. The situation is slightly morecomplicated in the Catholic Epistles; here Wachtel identifies a groupcontaining 36 94 307 453 720 918 1678 2197, but does not place 307 in thesame subgroup as 453. The text of 307 itself is said to have been "compared withPamphilius'[s] revision" [Scrivener]. It is cited in NA28 for the Catholic Epistles, but was not cited in earlier Nestle editions.

Manuscript 314

Oxford, Bodleian Library Barroc. 3. Soden's O11;Tischendorf/Scrivener 23a, 28p, 6r.Contains the Acts, Epistles, and Apocalypse with severe mutilations.Losses include Acts 1:1-11:12 (with 1:1-3:10 replaced by a later hand),14:6-17:19, 20:28-24:12, 1 Pet. 2:2-16, 3:7-21, 2 Cor. 9:15-11:9,Gal. 1:1-18, Eph. 6:1-19, Phil 4:18-23, Rev. 1:10-17, 9:12-18,17:10-18:11. Dated paleographically to the eleventh century. Commentary manuscript;Scrivener describes it as having "scholia on the Epistles"(identified by Von Soden as the commentary of (the pseudo-)Oecumenius) and "afull and unique commentary on the Apocalypse." As usual, Von Sodensimply describes it as an Oecumenius manuscript; in the Apocalypse helists it as being of type K0, but Merk modifies this toplace it among the Arethas manuscripts. Schmid grouped it with the"a" or primary Byzantine group (headed by 046) in theApocalypse. The Alands simply list it as Category V (i.e. Byzantine), thoughone wonders if they really had enough text of Acts for the determinationto be reliable there (they have only nineteen test readings, although it istrue that every one of these agrees with the Byzantine text).Scrivener calls it "a beautiful little book,"and it certainly is small (13 cm. x 10 cm.), and in a small hand. Apartfrom the commentary, the only marginal equipment are theκεφαλαια; it alsohas prologues and τιτλοιbut no lectionary or other apparatus.

Manuscript 317

Paris, National Library Greek 212. Soden's Nι31.Contains somewhat more than half of John (10:9-end), with a commentary reported by von Soden to be that of Nicetas. Dated paleographically to the twelfth century. Textually, relatively little is known about the manuscript. Wisse did not examine it, as it does not contain Luke, and von Soden simply listed it among the Nicetas manuscripts (the other manuscripts with the Johannine portion of this commentary include 249 333 423 430 743). The Alandsdo not assign 317 to any Category, becausethey examined test readings only from the Synoptic Gospels. Thus 317 hasnever been subjected to any systematic textual evaluation.

Manuscript 319

Cambridge, England, Christ's College MS. Dd.1.9 (so the first edition of theKurzgefasste Liste) or Gg.1.9 [MS. 9] (so the Acts volume of Text undTextwert) or F.1.13 (James, Scrivener). Checking the online Cambridge catalog(if I understand it correctly), it appears that neither Aland citation is right;Dd.1.9 is "A Collection of Medical Recipes" and Gg.1.9 is "The Register Book of theCity of Westminster."Soden's α256; Tischendorf 24a, 29p.Collated by Scrivener and cited as l in his Codex Augiensis.313 pages with 22 lines per page. James describes it as "Cent. xii, in a fineclear hand, hanging from lines ruled with the dry point." The Listeand Scrivener agree with the twelfth century date although Scrivener notes that there aremany later corrections. James observes that the quires are "wrongly numbered byold hand," adding "There are a good many errors in the numbering of the quires."There are several cancels. Several leaves have been lost; Acts 18.20-20.14,James 5.14-1 Peter 1.4 are lacking. The order of the books is Acts, Catholics,Paul. Hebrews follows Philemon. Two leaves in 2 Timothy are out of order.The Alands list it as Category V (Byzantine).Based on an examination of Scrivener's collation of 319 in Galatians, thisappears correct. There are only 35 differences from the Textus Receptus.One of these is corrected, and in one the corrector has messed up a Byzantinereading. Of the remaining 33, 17 are places where 319 is in fact Byzantineand the TR is not, or places where the Byzantine text divides and 319 followsone of the strands. Of the remaining 16 readings, about half are differencesof spelling or accent. Almost all the remainder appear to be simple scribalslips. The situation appears to be similar in Acts; of the 95 readings testedby the Alands, 94 agree with the Majority text -- and even in the one exception,Acts 10:10, 319 has 48 allies, all minuscules, most if not all of them Byzantine;it appears this is the reading of a Byzantine subgroup that is not the majority.

Manuscript 322

London, England, British Library MS. Harley 5620.Soden'sα550 ; Tischendorf 27a, 33p.Dated paleographically to the fifteenth century.It doesn't seem to have been properly examined by von Soden; he listed it as Ib?, based presumably on its extremely close kinship with the older manuscript323 -- a kinship acknowledged by almost all authorities,who seem to follow von Soden in largely ignoring 322.Aland and Aland list it as CategoryII in the Catholics and Category III elsewhereand explicitly note that it is the same category as 323, which they call asister manuscript. The kinship is easily confirmed for Acts based on the Aland data;the two have only four disagreements out of 102 tested, which is by far the highest rateof agreements for either manuscript; they disagree only forActs 2:30 (where both have singular readings which differ by just one letter; their ancestor surely had the same reading, although it's not clear which it was), 3:11 (where 323 goes with ℵ B and 322 has a minor variant), 8:37 (both add the verse, but inslightly different forms, both singular), 15:34 (where both add the verse, but 322 hasa minor singular variant). Thus even the Alands' figure of 96% agreement probablyunderstates their similarity; if 323 is not the ancestor of 322, they must come froma recent common ancestor, with 322 perhaps having varied from it slightly more (notsurprising, given that it must have had several centuries of time to add errors).For a British manuscript, Scrivener's description is surprisinglyshort: 134 folios; fifteenth century; paper; "of some weight: there are nochapter divisions primâ manu; the writing is small and abbreviated."See also 1739 and Family 1739 and 323.

Manuscript 323

Location/Catalog Number

Geneva. Catalog number: Bibliothèque Publique et Universitaire,Gr. 20.


323 contains the Acts and Epistles. Acts 1:1-8, 2:36-45 are from a laterhand; there are a few other minor defects. It is written on parchment,one column per page.


Dated paleographically to the eleventh century. Scrivener calls it"beautifully but carelessly written, without subscriptions."

Description and Text-type

323 is very closely related to the fifteenth century minuscule 322;the two are evidently sisters. Beyond that, 323's closest affinity is withthe members of Family 1739 and with the Byzantine text.

323 stands closest to 1739 in the Catholic Epistles, particularly in2 Peter-Jude. In those books it might almost be a copy of 1739 with somecorruptions. In James and 1 Peter it still has affinities with family 1739,but the ties are weaker and the Byzantine text more prominent.

The situation is similar in Acts. 323 appears to belong with family1739, but the Byzantine element is very strong. (So strong that Geertried to classify it as a Byzantine member of family 1739!) For detailson Geer's analysis, see the entry on 1739.The Aland sample numbers are similar: 1739 has 53 non-Byzantine readings out of 104total readings in the sample, but 323 has only 32 non-Byzantine readings. 323, in theAland sample, agrees with 322 in 96% of the sample readings, but doesn't agree with anyother substantial manuscript more than 80% of the time -- its next closest relative among manuscripts extant for at least fifty readings is a 76% agreement with 2289. Among manuscripts generally regarded as valuable, it is closest to 1852 (68%); its rate of agreements with 1739 is so low that 1739 does not even show up in the Alands' list of its relatives. In other words, Geer is right that the text is more Byzantine than anything else. But it is not purely Byzantine; it is Byzantine mixed with something else -- and the something else is quite clearly Family 1739. If we look at those 54 readings where 1739 is non-Byzantine, we find that 323 is Byzantine in 23 of them (and approximates the Byzantine text in a 24th), but of the other 29 readings, it agrees exactly with 1739 in 21, and approximates it in four more; three readings are so fractured and complex that differences are probably due to imperfect corrections; only once (Acts 3:11) does 323 have a non-Byzantine reading clearly distinct from 1739's -- and even there, the difference is slight enough that a careless scribe might have missed it. Since 1739 exists, 323 has little independent value, but it does at least serve to confirm that there is a Family 1739 text.

In Paul, 323 is almost entirely Byzantine. The few non-Byzantine readingshint at a family 1739 text (perhaps related to 945), but they are so fewthat no definite conclusions can be reached.

Von Soden lists 323 as Ib2. Aland and Aland list it asCategoryII in the Catholics and Category III elsewhere.Richards lists 323 as amember of Group A3 (Family 1739). Amphoux also associated itwith 1739.

Other Symbols Used for this Manuscript
von Soden: α157. Tischendorf: 29a; 35p



Sample Plates:
Aland & Aland (1 page)

Editions which cite:
Cited in NA26 for the Acts and Catholic Epistles.
Cited in NA27 for the Acts and Catholic Epistles.
Cited in NA28 for the Acts and Catholic Epistles.
Cited in UBS4 for the Catholic Epistles.
Cited by von Soden, Merk, and Bover, but very rather sketchily (especiallyin Paul).

Other Works:
Thomas C. Geer, Jr., Family 1739 in Acts (Society of BiblicalLiterature Monograph Series, 1994). Consists mostly of tables comparingmanuscripts 206, 322, 323, 429, 453, 522, 630, 945, 1704, 1739, 1891, 2200.The analysis is flawed, but the results are generally valid.

Manuscript 330 and Family 330

Location/Catalog Number

Saint Petersburg. Catalog number: Public Library Gr. 101.


330 originally contained the entire New Testament except the Apocalypse.It is now slightly damaged. 330 is written on parchment, one columnper page.


Dated paleographically to the twelfth century.

Description and Text-type

For the most part, 330 is a quite ordinary Byzantine manuscript. Inthe Gospels, for instance, Von Soden listed it as Kx and Wissespecifies it as Group 16 (a group close to Kx). Colwell describes 330as part of Family 574 (=330 574 [Mix/KxCluster 585 according toWisse] and 1815+2127[Π473 according to Wisse]) in the Gospels. TheAlands classify it as Category V (Byzantine). Although there is obviously somedoubt about the exact Byzantine group to which 330 belongs, there is noquestion but that it is Byzantine.

The same is true in the Acts and Catholic Epistles, where the Alandsagain list 330 as Category V; their datashows only four readings which go against the Majority text: it agrees with UBSin Acts 10:10 (where, however, it has 67 allies; this is pretty definitely a Byzantinereading although not the majority reading), 16:14 (where it has 64 allies or near-allies,so again, the reading of a Byzantine subgroup although not the majority group) and has singular readings in Acts 16:33 (where it differs from the majorityonly in having αι for οι, so it's probably just an error), 17:41 (a simplemis-spelling, καματων for κυματων). Thus, there is no reason to think any of itsancestors was anything other than Byzantine in Acts.

In the Johnannine Epistles, Richards lists330 as Byzantine, assigning it specifically to Group B1 (whichalso contains 319, 479, 483, 635, 1829, and 1891).

The situation is entirely different in Paul. Here the Alands upgradethe manuscript to Category III. But 330's contentis, perhaps, even more interesting than the typical manuscript of that category.

330 has a unique type of text shared by only three other known manuscripts:451, which outside of Hebrews is almost close enough to 330 to be a sister; 2400(according to Gary S. Dykes);and 2492, which seems to have a slightly more Alexandrian-influenced versionof the same text. The text of family 330, as we have it, is largely Byzantine,but the remaining readings do not belong purely to either the Alexandrianor "Western" texts. The following list shows some of the uniqueor nearly unique readings of 330:

Von Soden lists 330 as Ia3 in the Acts and Epistles. Thisis interesting, since Ia3 also contains 462 and 436, which Davieslinks to 330. Even Davies, however, admits that the strength of the link"varies," and 436 and 462 do not belong to Family 330.Von Soden appears to be correct, however, in believing the family to belinked, very loosely, with Family 2127 (often called Family 1319). The linkprobably comes via the Euthalian recension;330 has the Euthalian apparatus.

There are also hints, although only very slight ones (due to 1506'sfragmentary nature), that Family 330 shouldbe linked to the text of 1506.Given 1506's extraordinary text, the matter deserves examination.

330 is not the best of the Family 330 texts. It is almost purelyByzantine in Hebrews. However, it is the only member of family 330 to havebeen published, and deserves fuller study.

The other members of Family 330 are as follows:

Other Symbols Used for this Manuscript
von Soden: δ259. Tischendorf: 330e; 132a; 131p.Also cited as 8pe


M. Davies, The Text of the Pauline Epistles in MS. 2344 (Studies& Documents 38, 1968) collates 330 for Paul, and discusses its relationshipwith 436, 462, and especially 2344.

Sample Plates:

Editions which cite:
Cited in UBS3 for the Acts, Paul, and the Catholic Epistles,but omitted from UBS4.
Cited by von Soden, Merk, and Bover for Paul, but this collation is verybad.

Other Works:
E. C. Colwell, The Four Gospels of Karahissar I, History and Text,Chicago, 1936, examines assorted manuscripts in the gospels, placing 330 inFamily 547

Manuscript 348

Milan, Ambrosian Library Barb. B. 56 Sup. Soden'sε121.Contains the Gospels complete. Dated by its colophon to December 29, 1022.Classified as Iβaby von Soden, the other members of this group being 477 1279. Wisse listsit as a core member of Group 1216 (which corresponds to Von Soden'sIβ), andthough Wisse expels many of Soden's family members from the group (and listsno subgroups), he shows all three of theIβamanuscripts as part of Group 1216. Colwell also affirmed the existenceof Iβ.The Alands do not place 348 in any Category; thisis fairly typical for manuscripts with a largely but not purely Byzantine text.Scrivener notes that it is in two columns, with Old Testament citations markedwith an asterisk (a somewhat unusual notation). It has full lectionary andEusebian equipment.
Sample plate in Edward Maunde Thompson,An Introduction to Greek and Latin Paleography (plate 62).

Manuscript 349

Milan, Ambrosian Library F. 61 Sup. Soden'sε413.Contains the Gospels complete. Dated by its colophon to 1322.Classified as Iφaby von Soden, i.e. as a member of Family 1424 (the other membersof this group are 517 954 1188(part) 1424 1675). Wisse does not quite agree;rather than placing 349 in Cluster 1675 (the approximate equivalent ofFamily 1424), he places 349 in M349, pairing it with 2388. (The M groupsare roughly equivalent to von Soden'sIφr). TheAlands do not place 349 in anyCategory; this is fairly typical for manuscriptsof this type. Physically, 349 has relatively little equipment: Ammoniansections but no Eusebian apparatus; lectionary tables but no indicationsin the text. It was taken from Corfu.

Manuscript 356

Cambridge (England), Emmanuelle College I.4.35. Soden's α255.Tischendorf's 53a, 30p (but note that it does notcontain the Acts). It is n in Scrivener's Codex Augiensis.M. R. James says of it "Cent. x, xi (Gregory says xii), in a most beautifulminute hand, hanging from lines ruled with a dry point." It needed to bein a minute hand; it is a tiny codex, 3-5/8" by 3-1/8" or 9.2 cm. x 7.8 cm.Despite this tiny size, there are 24 lines to the page. Not too surprisingly,it uses a very large number of abbreviations. There are 144 regularpages and one added page.According to an inscription, it was given to the college in 1598 by SamuelWright. It contains the Catholic and Pauline Epistles, with several lacunaeand disarrangements. The Catholic Epistles open the manuscript, but pageshave been lost at the beginning. The first extant page probably begins with2 Peter 2:1, but is so illegible that the first really readable verse is2 Peter 2:4. Many other portions of the manuscript are also very hard toread. The text continues to 1 John 3:20, then breaks off, to resumein the middle of the prologue to Romans (which is from Oecumenius).There is another lacuna from 1 Cor. 11:7-15:56.A second hand, "not quite so good," takes up the writing at 1 Tim.6:5, and is responsible for the rest of the surviving text, which breaks off atHebrews 11:27. There are pencilled notations in Latin from about century xiii. This obviouslyimplies that it was brought into the Roman Catholic regions by that century.Scrivener, despite collating it, says nothing about its text, and von Soden didnot classify it. The Alands put it inCategory V (Byzantine).

Manuscript 365 and Family 2127

Location/Catalog Number

Florence. Catalog number: Laurentiana library. VI.36.


365 originally contained the entire New Testament except the Apocalypse, plus the Psalms. Rom. 1:18, 7:18-21, 8:3-31 have been lost. It is written on parchment, one column per page.


Dated paleographically to the thirteenth century.

Description and Text-type

365 first examined by Scholz, then declared "lost" by Burgon when a librarian assured him there was no such manuscript. It was "rediscovered" by Gregory.

365 is primarily Byzantine in the Gospels, Acts, and Catholics. Aland and Aland list 365 as Category V everywhere except Paul, where they upgrade it to Category III.

Von Soden lists 365 as Ik in the Gospels and K in the Acts and Epistles. Wisse lists it as Πb in Luke. 1319 and 2127, which are associated with 365 in Paul, also belong to Family Π.

In Paul 365's text is significantly different from the text in other areas. Although it still has more Byzantine readings than anything else, there are a number of Alexandrian readings as well. The vast majority of these readings are shared with 2127 and other texts of what Bover, following the lead of von Soden, calls "family 1319" (a subgroup of the Ia3 text, containing 1319, 2127, 256, 263, etc.; also evidently 1573. A better name would probably be Family 2127, as 2127 is probably the best manuscript of the type. There are hints of a connection with the Armenian; 256 is a Greek/Armenian diglot). 365 agrees with 2127 about 85% of the time (90% of the time in non-Byzantine readings), including such noteworthy readings as

Other important agreements with family 2127 (although not with 2127itself) include:

Von Soden, as noted, considered family 1319 to belong to the I type.However, it has many more Alexandrian than "Western" readings.365 seems to be a slightly mixed member of the group (it is more Byzantinethan, e.g., 2127), perhaps closest to 1573.

The situation for Acts is a bit complicated. The Alands' data for 365 in Acts shows no readings at all which agree with UBS against the Majority Text -- but eight readings which agree with neither:
Acts 2:43-44 -- with P74 ℵ A C 38 88 460 916 1175 1319 1573 1642 2127 2242 (obviously this is the Alexandrian reading; it's just that B doesn't have it and UBS doesn't follow it).
Acts 2:47-3:1 -- with 218 1359 1409 1505 2718 (but this is a very fractured reading and 365 agrees with 38 1319 1573 2127 for part of it)
Acts 10:10 -- with 48 other witnesses, all minuscules, including 38 383 462 1243 1319 1735 2127 2492 (presumably this is the reading of a Byzantine subgroup)
Acts 15:17-18 -- with 38 256 1319 1573 2772 (and cf. the reading of 2127 below)
Acts 16:28 -- with 621 (but this differs from the Majority text only in a matter of word order)
Acts 18:27 -- with 29 other manuscripts, all minuscules, of which only 876 seems to be of any note at all; this again looks like a Byzantine subgroup reading
Acts 27:14 -- 365* with Bc 181 256 676 1405 1829* 1875; 365c with Byz, and the readings are similar enough that this could be a simple mis-reading
Acts 28:29 -- with 104 256 996 1405 1719 2243 2625

Similarly 2127 has 95 Byzantine readings (25 of which are also found in UBS), two readings which agree with UBS against the majority, and seven readings (two of them singular) which agree with neither:
Acts 2:43-44 -- with P74 ℵ A C 38 88 365 460 916 1175 1319 1573 1642 2242
Acts 2:47-3:1 -- 38 1319 1573 2127 (but this is a very fractured reading and 365 agrees with 38 1319 1573 2127 for part of it)
Acts 3:22 -- with P74 ℵ A B C 81 94 180 307 453 610 629 886 1175 1678 1895 2818 (256 is defective and 38 365 1319 1573 Byzantine)
Acts 10:10 -- with 48 other witnesses, all minuscules, including 38 365 383 462 1243 1319 1735 2492 (presumably this is the reading of a Byzantine subgroup)
Acts 12:25 -- with P74 A 6 33 38 256 459 547 1319* 1573 2344 2544 2765 2737 2772 (365 Byzantine)
Acts 15:17-18 -- singular, but the reading is an h.t. error for the reading of 38 256 1319 1573 2772
Acts 22:20 -- with P74 ℵ A B C E 181 629* 1409 1875 1884 (38 256 1319(*) 1573 Byzantine and 365 defective)
Acts 23:20 -- with 38 88 205 254 312 385 467 469 567 641 914 996 1240 1319 1524 1573 1721 1742 1757 1842 1886 2085 2712 2716 (365 defective; 256 has a minor variant similar to the Byzantine text)
Acts 28:16 -- singular (but this is another fractured reading)

Looking at this, it appears clear to me that the 365/1319/2127 type still exists in Acts (and possibly in the Gospels also, since the three manuscripts are grouped together by Wisse); it's just that the type has so few non-Byzantine readings that no one has noticed it. But it might be worth a detailed study of 38 256 365 1319 1573 2127 in Acts.

The following offers a brief summary of information about the various membersof Family 2127 (note: Citations are for Paul, although von Soden, Merk, andBover generally cite the same manuscripts in the Acts and Catholics):

MSDateLocationCatalog NumberSoden
Cited inComment
256XI/XIIParisNational Libr. Armen. 9Ia3II Soden, Merk, Bover, UBS4Contains the Acts, Epistles, and Apocalypse with lacunae. Greek/Armenian diglot. The Alands list it as Category II in Paul only; V elsewhere. This classification appears to be correct in Acts at least; the Alands list only one reading which agrees with UBS against the Majority text and five which agree with neither.
263XIIIParisNational Libr. Gr. 61Ia3III Soden, Merk, Bover, UBS4Contains the Gospels, Acts, and Epistles. The Alands list it as Category III in Paul only; V elsewhere. Von Soden lists as K1 in the Gospels; Wisse lists it as Kx. "Probably from Asia Minor" (Scrivener).
365XIIIFlorenceLaurentiana Libr. VI.36.KIII NA26, NA27, UBS4Contains the Gospels, Acts, and Epistles with lacunae. Valuable only in Paul.
1319XIIJerusalemTaphu 47.Ia3III Soden, Merk, Bover, UBS4Contains the Gospels, Acts, and Epistles. The Alands list it as Category III in Paul only; V elsewhere. Von Soden lists as Ik in the Gospels; Wisse describes it as Πb.
A large section of Acts (from roughly the middle of chapter 16 to the middle of chapter 21) is a supplement, and this seems to my casual inspection to be a more interesting text than the rest of Acts; it looks as if it may be associated with family 453.
1573XII/XIIIAthosVatopediu 939IIIUBS4 Contains the Gospels, Acts, and Epistles. The Alands list it as Category III in Paul only; V elsewhere. Von Soden lists as Ir in the Gospels; Wisse describes it as Mix in Luke 1 and Group Λ in Luke 10 and 20.
2127XIIPalermoNational Libr. Sep. Mus. 4; also Philadelphia, Free Library, Lewis CollectionIa3IISoden, Merk, Bover, UBS3, UBS4Contains the Gospels, Acts, and Epistles. The Alands list it as Category II in Paul only; V elsewhere. Von Soden lists as IB in the Gospels and K in the Catholics; Wisse describes it as Π473. The number 1815 was also assigned to this manuscript. Probably the best manuscript of the family, although it seems to be prone to occasional short omissions.

Other Symbols Used for this Manuscript
von Soden: δ367. Tischendorf: 145a; 181p


M. Davies, The Text of the Pauline Epistles in MS. 2344 (Studies& Documents 38, 1968) collates 365 for Galatians (only).

Sample Plates:

Editions which cite:
Cited in NA26 and NA27 for Paul.
Cited in UBS4 for Paul.

Other Works:

Manuscript 372

Rome, Vatican Library Greek 1161. Soden'sε600.Contains the Gospels, breaking off at John 3:1. Dated to the fifteenthcentury by Scrivener, the sixteenth century by von Soden and Aland.Classified as Ia by von Soden, which would make it "Western"or "Cæsarean." Wisse does not find a relationship to the majormanuscripts of either group, but concedes that it has a mixed text, whichhe describes as "very strange." The Alands do not assign 372 to anyCategory; this at least seems to confirmthat it is not purely Byzantine. Scrivener describes it as "beautifullywritten," but lists it as having almost no marginal equipment ( lectionary information or Eusebian apparatus), and what it has is inLatin. One wonders if the Latin did not somehow influence the Greek.

Manuscript 383

Oxford, Bodleian Library E. D. Clarke 9. Soden'sα353; Tischendorf'sand Scrivener's 58a, 224p.Contains the Acts and Epistles (Heb. 13:7-end have been lost). Universallydated to the thirteenth century.Classified as Ic2 by von Soden. In Acts, this places 383with manuscripts such as 6142147, with 1108 1245 1518 1611 2138 (Ic1) at a greater distance.This corresponds with conventional wisdom that makes 383 a secondarywitness to the "Western" text of Acts. (Though it should benoted that it has not clearly been demonstrated that Family2138, to which 383 evidentlybelongs, is actually "Western.")However, the Alands declined to place 383 in a Category,which often indicates a manuscript largely but not purely Byzantine.
The Aland data makes it clear that 383, if a member of Family 2138, is a weak one; they showonly nine non-Byzantine readings (three of which agree with UBS, six which agree withneither UBS nor the majority). It's so Byzantine that the Alands did not even list itsclosest relatives. The readings where it does not agree with the majority are an interesting list:
Acts 2:7B (with P74 ℵ A B C* 81 321 1175 1409 2400 2508 and UBS; not with Family 2138)
Acts 10:10 (with 48 other minuscules; not with Family 2138)
Acts 15:23 (with 614 1292 1501 2147 2412)
Acts 15:34 (with 87 other minuscules including 614 2412 although not 1611 1505 2138 2495; also 33 323 623 1175 1739 1891)
Acts 17:13 (with 104 other manuscripts including 206 429 453 522 614 630 1505 1611 2138 2412 2495; also P74 ℵ A B D1 Ψ 33 81 436 945 1739 1891 2298 and UBS)
Acts 18:21-22 (with 2147 2401c and close to several others but not to any Family 2138 texts)
Acts 18:27 (with 2147 and close to several others, but not to any Family 2138 texts)
Acts 20:32 (with 2147 2652)
Acts 21:8 (with 181 other manuscripts including almost all Alexandrian texts, Family 2138, and 2147; clearly 383 has a Byzantine reading here although not the majority reading)
This is not the picture of a Family 2138 witness; the manuscript may need further examination, but based on this, it looks very close to 2147. Beyond that, I would label it an almost purely Byzantine manuscript with a few independent readings of no clear type.
I would add that, looking at the Valentine-Richards collation of Acts 15, I count 44differences between 614 and the Textus Receptus. 383 agrees with 614 in 19 of these, with the TR in 25; it has 11 readings which agree with neither. But I also note that most of the differences with the TR are early in the chapter, and most of the disagreements with 614 are toward the end. This led me to check chapter 13, where 58 has 18 readings with 614, 27 with the TR, and 21 that stand alone. In chapter 14, it's 17 readings with 614, 12 with the TR, and 12 that stand alone. The general conclusion: 383 is probably block-mixed, with some parts that are not as Byzantine as the Aland numbers wouldimply, but probably not a primary Family 2138 witness -- indeed, I would regard it as anopen question whether it goes with Family 2138 at all. It might, but the collation dataisn't very indicative.
In Paul, 383 and its Ic2 allies appear to be much more Byzantine.In the Catholics, 383 is again grouped with 614 2147 etc. by Von Soden, butneither Wachtel nor Amphoux lists it as a member of Family 2138. It seemslikely that it is again Byzantine in these books. Collated byAugust Pott in Der abendlädische Text der Apostelgeschichte unddie Wir-Wuelle, and has been used by many others such as Clark and Ropesin determining the "Western" text of Acts.
The text of Acts 13-22 is collated in A. V. Valentine-Richards, The Text of Acts in Codex 614 (Tisch. 137) and its allies which gives a complete text of 614 along with collations of 383 431 1518. However, although this book was published in 1934 (and reissued in paperback in 2014), the work was done long enough ago that the collations use Tischendorf numbers rather than New Gregory numbers.
According to the notes in the Valentine-Richards book, 383 belongs withFamily 2138only in Acts 13-22,with the rest being Byzantine, which is why it is collated only for those books.

Manuscript 423

Munich, Bavarian State Library 36, 37. Soden'sNμ60,Nι60;Tischendorf/Scrivener 423e+425e.Two volumes, the first containing Matthew (complete) with the catenaof Nicetas (this is Tischendorf 423e) and the secondJohn (also complete and with what Scrivener calls a "very full"catena of Nicetas). The first volume contains a colophon dating it to1566. The scribe is unnamed, but wrote two manuscripts which were inthe Tischendorf list (424e, a commentary on Luke, and 432e,a commentary on Mark) which Gregory deleted from the catalog. It is notcertain that the manuscript was ever intended to include Mark or Luke;the Matthew volume is marked Tomos A and the John volume is Tomos B.Little is known of the text; Von Soden simply listed it as a Nicetasmanuscript, and of course it did not contain Luke, so Wisse could notclassify it. The Alands do not place it in any Category,but it is not clear whether this is because of its text or because ofthe limited sample size.

Manuscript 424

Location/Catalog Number

Vienna. Catalog number: Nat. Bibl. Theol. Gr. 302, folios 1-353.


424 contains the Acts, Epistles, and Revelation (the latter missing 15:6-17:3,18:10-19:9, 20:8-22:21). It is written on parchment, one column per page.


Dated paleographically to the eleventh century. The original run ofthe text does not contain an unusual number of obvious errors, but the manuscript has beenheavily corrected (see below).

Description and Text-type

The original text of 424 is of the ordinary Byzantine type of the period,and is in no way worthy of note. However, the manuscript has been subjectedto a complete revision in the Pauline and Catholic Epistles, constitutingmany hundreds of alterations (with three hands reportedly involved; see alsothe entry on correctors).Some noteworthy examples include:

It will be observed that 424** shares all of these readings with1739.This pattern continues in the uncited readings; apart from trivial corrections,the corrections agree with 1739 over 90% of the time -- and even wherethey do not agree with 1739, other members of family 1739 (e.g. 6, 1881)can be found which agree with 424**. (The connection of 1739 and 424**has been known almost since the former was discovered, and more recently wasreaffirmed by Birdsall.)

Within family 1739, 424** is perhaps closest to 6 (see, e.g., theirunique readings χαριτος for πιστεως in Rom. 12:3and ευωχιαις in Jude 12). The two are by no means identical(as the list above shows), but 6 424** appear to me to form a subfamily withinfamily 1739.

This does not mean that the corrected text of 424 is as important atext as 1739. It remains more Byzantine than anything else. But where 424**presents us with a non-Byzantine reading, it should be treated as veryimportant, especially when supported by some other member of family 1739such as 6, 1739, 1881, or 0243.

Von Soden lists 424** as H in the Acts and Epistles (with the(pseudo-)Oecumenius commentary on the Praxapostolos); in the Apocalypsehe describes it as Io1. Aland and Aland list 424* as CategoryV and 424** as Category III (in Paul and the Catholics). Richards lists424* as belonging to group B6 and 424 as corrected as belongingto group M2 in 1 John and MW in 2 and 3 John. (This,of course, ignores the obvious facts that 2 John and 3 John are too shortto allow textual classification, the fact that "mixed" is nota text-type, and the fact that we should treat the corrections in 424 asdistinct from 424 as corrected; Richards simply doesn't understand manuscriptclassification). In the Apocalypse, Schmid placed it inthe "b" group of the K type.

Although the Alands value 424c only in Paul and the Catholics(and don't cite it very often even in those books), the situation in Acts is,at minimum, more interesting than they would imply. It is true that 424* agrees withUBS against the Majority text only twice (Acts 10:10 and 23:20, and in the former it still has 51 allies and in the latter it has 44 immediate allies and the reading is very fractured) Further, it has 74 readings that agree with theMajority against UBS, and the number of corrections in Acts is smaller than inPaul. Still, there are seven corrections in Acts:
10:12 -- with 29 other manuscripts including C* 5 33 322 323 623 2298
12:25 -- with 41 other manuscripts including E 322 323 630 1175
13:42 -- with 76 other manuscripts including P74 ℵ A B C D E Ψ 097 81 94 104 181 206 307 322 323 326 429 436 441 453 522 610 614 623* 629 630 876 945 1175 1611 1739 1891 2138 2344 2412 2818; also UBS
21:25 -- with 206 322 323 429 547 630 1040 1490 1509 1751 1758 1831 1891 2200 2298 2464
23:20 -- with 86 other manuscripts including ℵc Ψ 6 94 180 307 436 441 610 614 629c 876 945 1175 1409 1505 1610 1611 1739 2138 2298 2412 2495 2818
24:6-8 -- with 28 other manuscripts including 104 322 323 (but this is a very fractured meeting)
24:14 -- with 99 other manuscripts including P74 ℵ* B E Ψ 6 69 81 94 104 181 307 322 323 436 453 610 614 623 630 876 945 1505 1610 1611 1739 1852 1891 2138 2200 2298 2412 2464 2818

Note that all but one of these agrees with 322 323, and that one exception (23:20) is a case where 424 probably agreed with one strand of the Byzantine text. Interestingly, 424c does not agree with 1739 particularly often. Still, 322 323 are weak members of Family 1739. It looks as if 424 in Acts was again corrected toward a text of the same sort as in Paul and the Catholics; it's just that that text had been more thoroughly conformed to the Byzantine text.

Other Symbols Used for this Manuscript
von Soden: O12. Tischendorf: 66a; 67p;34r

J.N. Birdsall, A Study of MS. 1739 and its Relationship to MSS.6, 424, 1908, and M (unpublished Ph.D. dissertation, 1959)


Sample Plates:
Aland & Aland (1 page)

Editions which cite:
Cited in UBS4 for Paul.
Cited by von Soden, Merk, and Bover, but very imperfectly.
Also cited frequently by Souter.

Other Works:

Manuscript 429

Wolfenbüttel, Herzog August Bibliothek 16.7 A0. Soden'sα398 (Acts and Epistles),α1471 (Apocalypse);Tischendorf/Scrivener 69a, 74p, 30r.Contains the Acts, Epistles, and Apocalypse complete. The Acts and Epistleswere written by a monk named George in the thirteenth (Scrivener) orfourteenth (Aland) century. The Apocalypse was added later in afourteenth (Scrivener) or fifteenth (Aland) century hand. The manuscript hasrelatively little in the way of reader aids, but has "manymarginal readings." The text is an interesting mix; Von Sodenclassifies it as Ib1 in the Acts and Epistles (grouping itwith 2065221758 1831 1891 etc.) and as K in the Apocalypse, but infact the matter is much more complicated. The Alands correctly assessit as Category III in the Acts andCatholic Epistles and as Category V in Paul and the Apocalypse. In Acts, the Alandslist its closest substantial relatives as 206 (95% agreement, although 206 has onlypart of Acts), 522 (92%), 1758 (875), 1509 (84%), and 1490 (83%). Also in Acts,Geer has argued that 429 belongs withwith Family 1739(206 322 323 429 522 630 945 1704 1739 1891 2200), being closest to 206 522.This is somewhat dubious, and shows the dangers of not having control overone's methodology; the Alands show 429 agreeing with 1739 just 65% of the time --less than its agreement with some Byzantine manuscripts.Like 206 and 522 -- and also 630and 2200, with which 429 seems to form a group -- 429 definitely belongs to Family2138 in the CatholicEpistles (where its classification has been confirmed by both Amphouxand Wachtel). There is an odd kinship between Family 1739 and Family2138 in Acts; it is likely that429 goes with the latter, and it is the kinship between the 1739 and 2138 typesthat Geer detected.

Contrary to Von Soden, the manuscript (again like 206 522, but unlike 630 2200)loses almost all value in Paul; the Alands correctly assess itas Byzantine. In the Apocalypse, 429 falls within the main or "a"Byzantine group headed by 046.See also under 2138 and Family 2138and 1739 and Family 1739as well at the extensive discussion under 206.

Manuscript 430

Munich, Bavarian State Library 437. Soden'sNι11.Contains only a fragment of the Gospel of John (1:1-8:14), with thecommentary of Nicetas. Dated to the eleventh century by all authorities.Its text, unfortunately, has never been properly assessed; Von Sodensimply lists it as a Nicetas manuscript, and Wisse and the Alandsdid not profile the text of John.

Manuscript 431

Stasbourg, Seminary 1. Soden'sδ268;Tischendorf/Scrivener 4312, 180a, 238p.Contains the Gospels, Acts, and Epistles. Dated to the eleventhcentury by Scrivener, to the twelfth by von Soden and Aland. Kenyon datedit around 1200. It has a curious non-history: Gregory thought it hadperished in a fire in Strasbourg in 1870, but later discovered that ithad been moved before the fire.Valentine-Richards declares it to be "a beautiful little book, measuringonly 13½ x 10 centimetres. The writing, which is small and regular, occupies a spaceof about 9½ x 6½ centimetres. The number of lines on a page varies in differentparts, but in Acts is usually twenty-eight."
In the Gospels, von Soden lists 431 as Ak and Wisse as1167 (indicating rough agreement, as six of Von Soden's Akwitnesses are listed by Wisse as part of 1167). The Alands list itas Category V, i.e. Byzantine.In the Acts and Catholic Epistles, the text is more interesting;here the Alands raise it to CategoryIII, and von Soden lists it as Ia1 (which in Acts includesthe "Western" text, but clearly von Soden is actuallyplacing it with the rather amorphous but interesting group ofminuscules 36 88 181 307 453 610 915 917 1829 1874 1898 -- many but not all of whichare members of Family 453. Apparently,though, he based this on the text of Acts 13 only). The Alands' data bearsthis out only in part -- the closest substantial relative of 431, in their list,is 610, but the agreement is a mere 76%. 2818 (i.e. 36) is next, but that agreementis only 74%. On the face of it, 431 looks like a mixed member of the 453 type, butwhat it is mixed with is not clear.
Amphoux mentions it as a member ofFamily 2138 (though thisis perhaps on the basis of its affinities in the Catholic Epistles).This is not supported by Wachtel, who lists it simply asa manuscript with 20-30% non-Byzantine readings -- and indeed, hisevidence makes it highly unlikely that it is a member of Family 2138.
In Paul, von Soden still reports the manuscript to be Ia1, butthe Alands return it to Category V.Scrivener simply says that the manuscript has "many unusual readings,"but it is not clear which part of the manuscript he is referring to.
A. V. Valentine-Richards, The Text of Acts in Codex 614 (Tisch. 137) and its allies, gives a complete text of 614 in Acts along with collations of parts of 383 431 1518.(Note, however, that although this book was published in 1934 (and reissued in paperback in 2014), the work was done long enough ago that the collations use Tischendorf numbers rather than New Gregory numbers.) 431 (cited as 180) is collated only for Acts 1-9. Inspot-checking this collation, I found 431's relationship with 614 rather dubious. InActs 2, there are twelve places where 614 disagrees with the Textus Receptus (not aperfect representative of the Majority Text, but close enough to have some value); 431agrees with 614 in six of these, with the TR in six. It also has two readings notfound in either. In Acts 5, 431 agrees with 614 in 17 instances, with the TC in 13, and had two independent readings. In Acts 8, it has 20 readings with 614, nine with the TR,seven that agree with neither. Clearly 431 is not closely akin to 614, although there might be a distant relationship. Since Family 453 bears some kinship to both the 1739and 2138 types, this would account for 431's relationship to 614.

Manuscript 436

Location/Catalog Number

Rome. Catalog number: Vatican Library Gr. 367.


436 contains the Acts and Epistles. It is written on parchment, one columnper page.


Usually dated paleographically to the eleventh century; NA27moves it up to the tenth century.

Description and Text-type

436 is generally regarded as a mixed Alexandrian manuscript (so, e.g.the Alands place it in Category III). Wachtel lists it in the leastByzantine (40%) category in the Catholic Epistles, pairing it with 1067.

Von Soden classifies 436 as Ia3, but this group in factconsists mostly of mixed Alexandrian witnesses. Thus von Soden's classificationimplicitly agrees with that of the Alands.

Detailed investigation seems generally to support Wachtel's conclusionsin the Catholics. It is one of the better minuscules, and agrees most stronglywith A, 33, and the Bohairic Coptic, making it a primary witness to thedominant form of the Alexandrian text. It has very few unique readings.

In Paul the manuscript is somewhat less good; it agrees with the Byzantinetext more than anything else. Apart perhaps from 1067, it seems to fall closest to104. Even this kinship is rather distant. Overall, the ancestry of the textseems to belong with 1962, family 2127, and the other late Alexandrian manuscripts(this agrees generally with von Soden's results).

As far back as the nineteenth century, 436 was linked with 69, and Daviesextends this group to include 462 (known to be very closely related to 69),330, and 2344. The link to 330 appears false; their similarities lie simplyin late Alexandrian readings. The tie to 69 and 462 appears stronger; 436 and462 have high rates of agreement where both are non-Byzantine. However, theyare not immediate kin; an examination of Davies's collations shows thatthey do not share many special readings, and that they have each suffereddistinct patterns of Byzantine corruptions (with 462 being much the moreByzantine of the two; it is closer to the Byzantine text than to 436).

In Acts, 436 is clearly more Byzantine than anything else; in the Aland sample set, it agrees with the Majority text in 80 of 104 readings and with UBS in just 36 (22 of thesebeing readings that agree with both; it has ten readings which agree with neither (two ofwhich are singular). It is interesting to find that it doesn't agree with any other manuscripts very closely; among manuscripts which have at least fifty test readings, its closest ally is 1856, with which it agrees 84% of the time. Next on the list are 142 and 1103, with which it agrees 78% of the time.

There is, however, an interesting pattern to its readings. Of its ten special readings, eight are in the second half of Acts, and six are in the last quarter. All of its 14 readings that agree with UBS against the Majority are in the last 60% of the book. This strongly suggests either block mixture or imperfect correction (that is, the tendency thatwe see in manuscripts such as L 579 of the Gospels and 33 in Paul for the first part of a corpus to be carefully corrected but the rest much more poorly handled). The question then becomes what we find if we look at 436's non-Byzantine readings.
Acts 4:33 -- with 73 manuscripts including D E 322 323 429 522 945 1739 1891 2298 2495 (but this has support from so many manuscripts that 436's reading might be from a Byzantine subgroup)
Acts 12:25 -- with 53 other manuscripts including D Ψ 94 180 181 307 441 453 610 623 1611 2138 2412 2818 (here again, 436 might have gotten its reading from a Byzantine source)
Acts 15:34 -- singular reading although somewhat similar to C
Acts 19:3 -- with P41 P74 ℵ A E 5 33 623 1409 1642 1735 1873 1884 2344 2374
Acts 23:1 -- with 42 other manuscripts including B 307 453 610 614 623* 1505 1611 2138 2412 2464 24952818
Acts 23:20 -- with 85 other manuscripts including ℵc Ψ 94 180 307 424c 441 614 629c 876 945 1175 1505 1611 1739 2138 2298 2412 2495 2818 (this variant again probably is not meaningful)
Acts 23:30 -- with ℵ A E 81 181 206 429 441 522 621 1175 1505 1509* 1739 1751 1842* 1852 1875 1884 1890* 1891 1894 2138 2200 2495
Acts 24:6-8 -- with 46 other manuscripts including Ψ 5 33 88 94 307 610 623 630 945 1739 1891 2200 2298 2818
Acts 24:15 -- with 5 437 623 2464
Acts 24:22 -- singular reading, but this is a very fractured reading
This perhaps hints at a kinship with 5 623, but the sample is too small and too vague for any certainty. The only real conclusion seems to be that we should look at 436's relationships only for Acts 15-28.

I can't do this for every manuscript, but here are the percentages of agreement for key manuscripts, as well as the weighted agreements (that is, giving extra value to agreements where few other manuscripts agree. The result is not a percentage; it's just a number, the higher, the more significantly the manuscripts agree). The three closest relatives of 436 in each list are shown in red; the closest relative of each is in bold red.

Manuscript% agreementWeighted agreement

Thus there does seem to be a special relationship with 623, but it's rather weak; there is also perhaps an even fainter relationship to Family 2138, but that is't much closer than to the Byzantine text, which is the next-closest relative.

Other Symbols Used for this Manuscript
von Soden: α172. Tischendorf: 73a; 80p.


M. Davies, The Text of the Pauline Epistles in MS. 2344 (Studies& Documents 38, 1968) collates 436 for Paul, and discusses its relationshipwith 330, 462, and especially 2344.

Sample Plates:

Editions which cite:
Cited in NA28 for the Catholic Epistles.
Cited in UBS3 for the Acts and Epistles, and inUBS4 for Paul and the Catholics.
Cited in von Soden, Merk, and Bover for the Acts and Epistles.

Other Works:

Manuscript 442

Uppsala, University Library Greek 1. Soden'sO18; Tischendorf's 73p, 68c.A very curious manuscript, because it is really twomanuscripts in one volume, designated 441 and 442. The most likelyexplanation is that the two manuscripts were both damaged and sowere bound together -- even though this resulted in some overlap.441 includes leaves 3-182 (with Acts 8:14-end plus Romans and1 Corinthians 1:1-15:38); 442, which consists of leaves 183-440,contains 1 Corinthians 13:6-end (thus overlapping 441 for two anda half chapters), all the rest of Paul from 2 Corinthians on,and the Catholic Epistles. Both are commentary manuscripts,listed by von Soden as Oecumenius's commentary. Thus they area logical pair of manuscripts to bundle together -- but theirtexts are not the same. What is more, their texts are not at alltypical of most of the other manuscripts von Sodden lists asOecumenius commentary manuscripts, most of which (e.g.056 and 0142)are essentially Byzantine. According to Scrivener, it waspurchased in 1673 at Venice by Sparvenfeldt; this perhapsimplies that it was taken to Venice from the east, but thereis no clear indication of its source. Scrivener says that itresembled 33 in Paul, and it is one ofthe minuscules most often cited by Tischendorf in thatcorpus. But the similarity to 33 seems exaggerated to me;based on the readings cited by Tischendorf, it seemscloser to the Euthalian group as represented, e.g., by365 and Family 2127. The Alandsrate 442 as Category II(while rating 441 as CategoryIII). Thus the decision to cite it for the CatholicEpistles in NA28 was appropriate.

Manuscript 443

Cambridge, University Library Nn.ii.36. Soden'sε270.Contains the Gospels complete. Dated paleographically to the twelfthcentury by all authorities. Classified by von Soden asIo; this amorphous group also contains U X 213 10711321(part) 1574 2145. This is not confirmed by Wisse (who dissolvesIo, and evidently with good reason); he reports 443 asa memberof M159 (along with 159 and part of 1557).The Alands list 443 as Category V.Scrivener reports that the ordinaryκεφαλαια havebeen subdivided in this manuscript. It has the Eusebian apparatus,but the lectionary data is partial, coming from another, apparentlylater hand.

Manuscript 451

Rome, Vatican Library Urbin. Gr. 3. Soden'sα178;Tischendorf/Scrivener 79a, 90p.Contains the Acts and Epistles complete. Universally dated to theeleventh century. Von Soden lists it as a K witness, and this appearsto be true in the Acts and Catholic Epistles. Certainly theAlands concur, placing 451 in Category Vin those books, with only three non-Byzantine readings (out of 105) in Acts(one of them a subsingular reading that is an easy error for the Byzantinereading); the others are Acts 21:10 (where it agrees with 67 other manuscripts,so even this is the reading of a Byzantine subgroup) and 26:14 (where it has 22immediate allies, including E 5 69 330 441 623 1175 1243 2464, and many othersthat are not far different).It has 8 non-Byzantine readings (out of 98) in the Catholics.Matters change entirely in Paul,and the Alands reflect this by upgrading the manuscript toCategory III. Here 451 is a clear andobvious member of family 330; the two agree infully 436 of 464 test readings, including 75 of 77 readings whereboth are non-Byzantine. Over a third of their 28 differences arein Hebrews, where 330 is largely Byzantine. (The third member ofthis family, 2492, is by no means this close to the other two.) It ispossible that 451 and 330 are sisters, with the common exemplarhaving some corrections between the time 451 and 330 were copied.Certainly the two have a common ancestor not far back in theirancestry. It is conceivable that 451 is the ancestor of 330, butthis seems somewhat unlikely, as the following readings from the apparatusof GNT3 demonstrate:

For more about family 330, see the notes on 330.

Manuscript 453 and Family 453

Location/Catalog Number

Rome. Catalog number: Vatican Library Barb. Greek 582.


453 contains the Acts and Catholic Epistles complete, with a commentary (reportedby Von Soden to be that of Andreas).


Dated by the Kurzgefasste Liste, following Gregory, to the fourteenthcentury. Scrivener, however, listed an eleventh century date. (We should note thatScrivener's information was incomplete. Scholz was unable to see the manuscript,and Scrivener's list says that the manuscript "contains but one chapter of theActs and the Catholic Epistles.")

Description and Text-type

Von Soden lists 453 as a member of Ia1 in Acts, a diverse groupcontaining, e.g., D 88 181 431 915 917 1829 1874 1898. But this group alsocontains 36 (now renumbered 2818) 307 453 610. All of thesemanuscripts, according to Von Soden, have the Andreas commentary, and they arecertainly closely related. The following shows the percentage agreements ofthese manuscripts, and certain control manuscripts, in the variants noted inUBS4. Agreements over 90% are highlighted:


(We should note that Von Soden lists several other Andreas manuscripts:K/018, 437, 832, 886, 1895, 2186. K, however, does not contain the Acts --and is Byzantine in any case. 832 2186 also lack Acts. 437 887 1895 containActs, but based on the information compiled by the Alands, they cannotbe true members of Family 453; either they are severely mixed or they belongto another text-type.)

These numbers demonstrate clearly that there is a group consisting ofthese five manuscripts. The question then becomes, what is the nature of theFamily 453 text? The Alands esteem it highly; in Acts, they list 36 asCategory II and 307 453 610 1678 asCategory III (we should note, however, thatthere is no reason, based on their numbers, to separate 36 from the other four;all have almost exactly the same ratio of Byzantine readings to UBS readings).Also, they fail to note that 94 appears to be a member of the group.But the Alands' classification does not characterize text-types; it simply tellsus how non-Byzantine a manuscript is. If we look at the above list, it wouldappear that the members of Family 36 fall closer to1739 than to any of the otherprimary manuscripts (e.g. ℵA B D L P 614). And indeed, we find Thomas C. Geer, Jr., who studiedFamily 1739 in Acts, labelling 453 as a weak member: it is "somewhatsignificantly related to [the leading manuscripts of Family 1739]" --but he adds that it "does not have a strong enough relationship to beconsidered a leading member of the family... it is already clear that itis a 'cousin' at best" (Family 1739 in Acts, p. 100). Geerdid not study the other members of Family 453, but there is every reasonto believe that he would have regarded the other members similarly.The evidence listed in the table above is also inconclusive; while 453 andits relatives agree with 1739 on the order of 75% of the time in the sample(which those who follow the ColwellDefinition would regard as close enough to belong to a text-type),it should be noted that the above sample is biased; it contains manyreadings where D opposes the entire Greek tradition -- readings whichshould not be counted under the Colwell definition. If these are omitted,the agreement between 1739 and Family 453 falls well below the 70% threshhold(on the order of 65%). It's also noteworthy that 453 agrees more with1739's more Byzantine relatives (945 1891) than with 1739 itself. Finally,if we examine the number of non-Byzantine agreements in the above sample,453 does not stand all that close to 1739; it has 37 such agreementswith 1739, but 37 also with P74 and B (even though P74is not complete), 36 with ℵ --and, by comparison, 53 non-Byzantine agreements with 36, 57 with 307, 50with 610, and 53 with 1678. Thus it would seem likely that 453 andFamily 453, while they may share common influences with Family 1739, arenot truly members of the same text-type (though a fuller study wouldbe needed to make this certain; Geer's work, even if one ignoresseveral methodological problems, did not examine Family 453 as a whole,and the data for Acts given above is based on too small a sample).

In the Catholic Epistles, the situation changes somewhat. The Alands'data implies that 453 and its relatives are much more Byzantine in theCatholic Epistles than in Acts. Wachtel elaborates this analysis of thedata considerably. 453 and its relatives are listed among the manuscripts witha text 30-40% non-Byzantine. Within this class (not really a text-type),we find 453 heading a group of eight manuscripts: 36, 94, 307, 453, 918,920, 1678, 2197. 36, 307, and 1678 we of course recognize as members ofFamily 453 in Acts. 94 is reported by Von Soden to have Oecumenius's commentaryon the Acts and Epistles, but has Andreas on the Apocalypse. 918 is listed asanother Oecumenius manuscript by Von Soden (though the Kurzgefasste Listedoes not show it as having a commentary); it does not contain Acts. 920 is nota commentary manuscript, but Von Soden lists it as another Iamanuscript (although von Soden assigns it to the Ia3 group ratherthan Ia1). 2197 contains only Paul and Catholic Epistles,and Von Soden does not seem to have classified it outside Paul (sincehe lists it simply as a Theophylact/Paul manuscript).

Other Symbols Used for this Manuscript

von Soden: Aπρ40
Tischendorf/Scrivener: 453a



Sample Plates:

Editions which cite:
Cited in UBS4 for Acts.

Other Works:
Thomas C. Geer, Jr., Family 1739 in Acts, Scholars Press, 1994,discusses 453 in the context of Family 1739.

Manuscript 472

London, Lambeth Palace 1177. Soden'sε1386; Scrivener's 511e.Contains the Gospels with extensive lacunae (lacking Matt. 4:1-7:6,20:21-21:12, Luke 4:29-5:1, 16:24-17:13, 20:19-41, John 6:51-8:2,12:20-40, 14:27-15:13, 17:6-18:2, 18:37-19:14).Dated paleographically to the thirteenth century by Alandand von Soden; Scrivener says eleventh or twelfth. Classified by von Soden asI', that is, among the miscellaneous "Western"/Byzantinemixed manuscripts. Wisse's data would seem to at least allow thepossibility that it is mixed with something not quite Byzantine; helists it as "Mix/Kmix/Mix; pair with 1009." This isgiven some additional support by the Alands, who do not assign472 to any Category. It hasrather incomplete equipment: Ammonian sections but no Eusebiandata; lectionary markings and Synaxarion but no Menologion;partial κεφαλια. Scrivener'sIntroduction declares that it is "for valuable readings by far the mostimportant at Lambeth [presumably of the gospel minuscules],shamefully ill written, torn and much mutilated." He collated it ashis c in A Full and Exact Collation of about Twenty Greek Manuscriptsof the Holy Gospels, where he adds some history: It was first examinedby Burney, who gave up on it because it was so strange and ill-written.Scrivener himself agrees that he "certainly never met with a copyof the gospels written with such irreverent and scandalous negligence"but adds that it abounds with "novel and remarkable readings, which(in spite of its unpromising appearance) would have amply repaid all thediligence [Burney] could have bestowed upon it." In particular, it obelizesLuke 23:39-41 and John 6:4. The writing is a "miserablescrawl on the coarsest parchment"; Scrivener suspects but is not certain thattwo hands were involved. Given the number of errors in endings, as well asthe writing and the parchment, I rather suspect it was a private copy madeby someone who didn't know Greek very well.

Manuscript 473

London, Lambeth Palace 1178. Soden'sε1390;Scrivener's 512e.Contains the Gospels, now complete (the first few leaves, containingintroductory matter and Matt. 1:1-8, were lost for a time).Dated paleographically to the thirteenth century by the Listeand von Soden; Scrivener's Introductionoffers the curious dating "xi or xiv" although A Full andExact Collation of about Twenty Manuscripts of the Holy Gospelsclaims x or xi.Classified by von Soden as Ikc,that is, as part of the third group of Family Π witnesses,along with such manuscripts as 229 280 482 1354. Wisse's resultsgenerally confirm this; 473 is listed as a member ofΠ473 --although it should be noted that none of von Soden's otherIkc witnesses are part of Π473. TheAlands classify 473 as Category V.Scrivener reports that it obelizes Luke 22:43-44, John 5:4.Physically, Scrivener's Introductiondescribes the manuscript as "A noble-lookingcopy" and written "in a fine hand, splendidly illuminated,and with much curious matter in the subscriptions;" the ExactCollation calls it "one of the most splendid manuscripts extant."The one exception is Luke 5:30-6:4, where the original page has been lost andreplaced by a paper page that is far less imposing. Ithas the usual Eusebian apparatus and lectionary equipment. Scrivener'sExact Collation... cites it as d (with the paper supplement beingd).

Manuscript 476

London, British Museum, Arundel 524. Soden'sε1126; Scrivener's 566e.Contains the Gospels complete. Gospels, lacking John 11:18-41.Dated paleographically to the eleventh century by all authorities sinceForney, who suggested the tenth. It is said to be on coarse vellum,now discolored, with pale ink; the handwriting is (with someexceptions) neat and clear, but it is probably not easy to read.Classified by von Soden as K1. Wisse almost agrees,listing the manuscript as Kx (to Wisse, K1is part of Kx). As one would expect, theAlands classify 476 as Category V.Scrivener claims unique readings at Mark 2:8 (γνους), 3:10 (τουτου for αυτου),Luke 2:51 (ηλθον), 13:11 (omit ασθενειας), 17:27 (και εγαμ...), 21:29 (παλιν beforeπαραβολην), John 19:12 (πας ουν), and it obelizes Matt. 6:13. Most of thesereadings are indeed very rare, but many might well be just errors in 476 orone of its ancestors; I doubt they give it much value.Scrivener calls it "somewhat partial to glosses."Physically, 476 is rather small (just more than 17x13 cm),but otherwise un-noteworthy; it has the usual Eusebian andlectionary apparatus.Collated by Scrivener as h in A Full andExact Collation of about Twenty Manuscripts of the Holy Gospels.

Manuscript 477

Cambridge, Trinity College B.X.17. Soden'sε350;Scrivener's 508e, which he included in his collations as i.Contains the Gospels, on vellum with some leaves of paper. The KurzgefassteListe says it is complete, but M. R. James notes that quire IH* has lost itsfourth leaf, and quires ΛH and M have lost their eighth leaves. In addition,there are twelve leaves which have been replaced by cancels. Despite the cancels.James calls it "well written."Dated paleographically to the thirteenth century by all recent authorities(Bentley, who gave it to Trinity College -- it was originally from the monasteryof Pantocrator on Athos --dated it xi). Classified by von Soden asIβa;other members of this group include 348 and 1279; and the "b"group of this type contains 16 1216 1579 1588(part). Wisse givesa similar classification, placing 477 in Group 1216 (one of two groupsWisse associated withIβ,Group 16 being the other). Wisse calls Group 1216 clearly distinctfrom Kx, but the Alands classify 477 asCategory V.
It appears, however, that the manuscript is block mixed; James Dowdentells me that Roger Omanson finds it to be Kx in Mark.
477 has only limited marginalia: Ammonian Sections but no Eusebianapparatus, and while the lectionary information is present, thereis no menologion. Even the synaxarion may be an afterthought, as it(and the hypotheses to Matthew) are on six paper leaves at thebeginning while the remainder ofthe manuscript (including the list of readings at the end)is parchment.

Manuscript 482

British Library Burney 20. Soden'sε329; Scrivener's 570e.Contains the Gospels complete.Dated by its colophon to 1285 -- although, in an interesting but clumsy forgery,this has been altered to read 985 (the two have the sameindiction; the change was from ψ to υ).Classified by von Soden as Ikc,that is, as part of the third group of Family Π witnesses,along with such manuscripts as 229 280 473 1354. Wisse's results partlyconfirm this; he lists 482 asKxaa.Scrivener's Introduction comments that it is "quite equalin value to Cod. cscr [472, which shows in Wisse's list asprimarily mixed]... and often agrees closely with wscr [489,which is listed by Wisse as pureΠa]." Scrivener's list of readings which agree with theuncials is long, and include texts of many Byzantine types, but A and K arevery often cited, which would also support the affiliation with Family Π.The Alands, however, assign 482 toCategory V. As members of Family Πmore often thannot are uncategorized in their lists, they would seem to supply some faintsupport for the Wisse's contention that 482 has some Kx.The manuscript was written by a monk named Theophilus, who was not verygood; there are many transpositions and substitutions of forms, and not a fewsingular readings; Scrivenerreports that it has "many corrections" from a later hand,which also added the lectionary lists (though the lectionary markingsin the text, like the Eusebian apparatus, are from the first hand).Scrivener's A Full andExact Collation of about Twenty Manuscripts of the Holy Gospelscalls the writing clear but inelegant, and often faded.Collated by Scrivener as p in A Full andExact Collation....

Manuscript 485

London, British Library Burney 23. Soden'sε1386; Scrivener's 572e.Contains the Gospels with major lacunae (lacking Luke 5:22-9:32,11:31-12:25, 27:24-28:4, John 8:14-end) and a few disordered leaves.Dated paleographically to the thirteenth century by von Soden andAland; Scrivener suggests the twelfth. Classified by von Soden asI', i.e. in the miscellaneous vaguely "Western" witnesses.Wisse classifies it as Kx, and this is supported by theAlands, who list it as Category V.Von Soden may have been confused by the way it was written; Scrivener'sIntroduction describes the manuscript as"boldly but carelessly written" --though he also comments "with many later changes and weighty readings."In A Full and Exact Collation of about Twenty Greek Manuscriptsof the Holy Gospels Scrivener says it is "carelessly written in abold unequal hand" and that the vellum is now soiled. Accents and breathingsare inaccurate. It has been fairly heavily corrected, no doubt because the firsthand is so inaccurate; Scrivener considers both hands to be interesting andworthwhile. It has illustrationsbefore each gospel, which Scrivener considers "poor."It has full lectionary equipment and the Ammonian Sections, but not theEusebian apparatus. Scrivener collated it as s inA Full and Exact Collation....

Manuscript 489

Cambridge, Trinity College B.X.16. Soden'sδ459; Tischendorf's 489e, 195a,252p. Scrivener's 507e, 224a,260p; he collates it in Codex Augiensis,,using the symbol w in the Gospels and k in the Epistles. Hort used the number102 for it.Contains the entire New Testament except the Apocalypse, plus much additionalmaterial. The order is Gospels, Acts, Catholic Epistles, Paul, with Hebrewsat the end. It has just about every reader help ever invented -- lectionarylists, prologues (although some of these are slightly disarranged), Eusebianapparatus, you name it. At the end, it also has the Pseudo-Dorotheus's treatiseon the 72 disciples, a brief vita Jacobi, and some other odds and ends(all this, including the synaxarion and calendar of readings, occupies onlya couple of dozen pages, although there are lacunae and the end seems to be lost).The colophon credits it to Jacobus, on Mount Sinai, and dates itto 1316. Scrivener describes it as "inelegantly written."It is notable that there are fully a dozen cancels in the last several quires (althoughno fewer than nine of these are in quire κα, so it may just be that this quire wasmessed up beyond fixing). Classified by von Soden asIκa in the gospels and Ia2 in the Acts and epistles.The former assessment is confirmed by Wisse, who describes 489 as Πa inall chapters of Luke. He pairs it with 1219 -- an interesting assessment, since 1216(which is dated XI century) remains at Sinai. The relationship between the two mightbe worth examining. Scrivener also mentions a similarity to his 570e, i.e.482 (XII century). This relationship may not be genetic; 482 is listed by Wisse asΠa in Luke 10 and 20, but Kx in Luke 1; probably what Scrivenernoticed was simply the Family Π readings. The Alands do not place 489 in anyCategory; this is fairly typical of their treatmentof Family Π manuscripts. The text of the Acts and Epistles might warrant furtherexamination -- in Acts, it has eight readings which agree with UBS against the majoritytext, and eleven which agree with neither. In Acts at least, it appears to pair with 927.

Manuscript 495

London, British Library Add. 16183. Soden'sε243;Scrivener's 581e.Contains the Gospels complete, though some of the introductorymaterial has been lost. Dated paleographically to the twelfth century.Classified by von Soden asI', i.e. in the miscellaneous vaguely "Western" witnesses.Wisse classifies it as Kmix, while the Alands do not list it asbelonging to any Category.All of these descriptions, diverse as they sound, imply much thesame thing: A manuscript clearly Byzantine, but with some readingsnot associated with Kx. Whether these readings have anyreal value must await a more detailed study.It has a full apparatus (Eusebian materials, lectionary equipment,etc.), though the Eusebian tables were not finished. The hand isdescribed by Scrivener as "minute." The manuscript isthought to have been taken from Sinai.