The Text of
The New Testament

Divinity Library
EEL- US 

Li 5

"/give theft Boots
hfei-ikei.feji^'tngtif^

'¥^LE«¥IMlMEI^SIIinfe DIVINITY SCHOOL
TROWBRIDGE LIBRARY

flDrfot'H Church %txt Boofcss
The Text of
The New Testament

BY

THE REV. K. LAKE, D.D.
PROFESSOR OF EARLY CHRISTIAN HISTORY IN THE
UNIVERSITY OF HARVARD

FIFTH EDITION- REVISED
{Sixth Impression)

NEW YORK
EDWIN S. GORHAM
II WEST 45TH STREET
1916

CONTENTS
CHAP. PAGE
i. The Object and Method of Textual Criticism, 1
n. The Apparatus Criticus of the New Testament
— Greek mss., ...... 11
kn. The Versions,  22
i\. Patristic Quotations — Liturgical Evidence —
Chapter Divisions and Stichometry, . . 47
v. History of Modern Criticism, ... 59
vi. The Western Text,  73
Appendix — A. Tischendorf s System of Notation, 92
B. Summary of Greek and Latin mss., 94
C. Books valuable for Textual Criti
cism,  98
D. The new researches and theory
of Professor von Soden . . 99
Index  104

PREFACE TO THE FOURTH EDITION
The appearance of Professor von Soden's important book
on the Text of the New Testament rendered it necessary
to consider in what way it was possible, in a new edition,
to give some account of his researches. The most
feasible method seemed to be the addition of a short
appendix devoted to a statement of his new notation and
theory. This will be found on pp. 100 ff. A somewhat
important addition has also been made, in the light of
his work, to the statements on p. 20 f, concerning the
minuscule mss. of the Gospels. After some hesitation
I have decided not to alter Chapters V. or VI. Obviously
I should now express many points somewhat differently
in the light of the great German book, if I were writing
afresh ; but the balance of advantage seems to be in favour
of leaving them as they are, until there has been some
more complete discussion of the new facts and hypo
theses. It will, I think, not be until after the publica
tion of Professor von Soden's critical text, and the elapse
of some considerable time for its study, that a secure
judgment will be possible on his views, and I do not feel
that I should be justified in using a 'textbook' for the
promulgation of criticisms which have not yet been
tested by those who are competent to decide. Kirsopp Lake.
Leiden, 1908.

CHAPTER I
THE OBJECT AND METHOD OF TEXTUAL CRITICISM
One of the most necessary parts of the investigations
of historians is to criticise the documents on which
their researches are based, in order to be certain that
the text which they are using really represents the
original writing of the author. This criticism is usually
known as Textual criticism, for the obvious reason that
it deals with the text as opposed to the subject-matter.
It is less commonly termed the Lower as opposed to the
Higher criticism, which deals not with the text as written
by the author or editor of the document in question, but
with the sources and methods used by him in. making
the text. Thus Higher criticism approaches the subject
at a point higher up the stream of its existence.
The object of all textual criticism is to recover so far
as possible the actual words written by the writer. But
in order to do this properly the critic has to explain how
each successive deviation from the original came to be
currently adopted, and frequently he finds the clue
enabling him to do this in the history of some later
period, which gives some reason for a textual variation.
In these researches it sometimes happens that the dis
coveries of the textualist are of great value to the his
torian ; for the corrupt reading of some important
document often explains otherwise inexplicable pheno
mena in the history of ideas or the conduct of a con
troversy. The problem, then, which faces the textual critic is to
remove from a number of manuscripts of varying date
the corruptions which have crept into the text and to

2 THE TEXT OP THE NEW TESTAMENT
assign to each variation its appropriate cause, thus
obtaining in the end the original pure text.
Let us assume, then, what as a matter of fact is never
more than approximately the case, that the critic has
at his disposal all the known mss. of a given work. He
begins his work knowing nothing about the character
of the mss., and from them he has to find out and re
construct the original text. His work falls into four
stages, which in practice necessarily pass imperceptibly
into each other, but which in theory are distinguishable,
and ought not to be confused : —
I. The study of each manuscript by itself, correcting
obvious mistakes which are due to slips of the
pen and cognate reasons, and such readings as
seem clearly to be corrupt forms of other re
corded readings.
II. A comparison of the manuscripts to which this pro
cess has been applied, and their arrangement into
groups, according to similarities of reading, the
rule being followed that, speaking generally,
community of error implies community of origin.
This process is carried on until all the known
mss. have been put into groups, each with a
presumably distjnct ancestor or archetype.
III. These i archetypes are then compared, and a pro
visional text is constructed out of them, the arche
type of the archetypes being arrived at as closely
as possible.
IV. This provisional text is finally subjected to the
process known as conjectural emendation. That
is, an attempt is made to explain and emend all
the passages which still seem corrupt.
These four stages in the work of textual criticism call
for a little fuller explanation.
_ 1. The investigation of individual mss. and the detec
tion of scribes' mistakes or alterations demand the know
ledge and application of the laws which obtain in these
matters.

METHOD OF CRITICISM 8
The chief point to be remembered is that mistakes and
corruptions are of two classes : —
1. Unintentional, due to natural error.
2. Intentional, due to a desire for improvement.
1. Unintentional alterations.— Many instances of this
source of error are quite easy to detect and remedy;
such, for example, are cases where a word or phrase is
senselessly repeated twice, e.g. in the Latin of the
Laudian ms. of Acts ii. 4 the scribe has written cet re-
pleti sunt et repleti sunt omnes spiritu sancto,' where
the omission of the second ' et repleti sunt' is an obvious
and certain correction.
This is technically called dittography ; similar causes
of error are homoioteleuton — the confusion of words ending
in similar syllables ; this cause often leads to the omis
sion of a complete line of the archetype ; and if many
examples of it occur in the same ms., it is sometimes
possible to deduce from them the length of the lines in
the archetype ; haplography — writing a word once when
it ought to be repeated, e.g. icvpie for Kvpie, tcipie ;
itacism — strictly a tendency to replace other vowels by
iota, but loosely used of other vowel changes. In later
Greek mss. almost any vowel seems changeable for any
other, nor does the same ms. always observe the same
spelling, e.g. \eyerai is often spelt Xeyere ; &rj\v is written
Brjki ; ai yvva'iKfs becomes e yivaiKais, and so on.
There are many other technical phrases for similar
kinds of mistakes, most of which explain themselves.
The important thing is that they classify to some extent
the slips of the pen and misspellings of scribes. A
slightly different form of error is where the scribe seems
to have preserved the right order of letters, but produced
the wrong word from them, e.g. in Col. ii. 18 we read
a iupaicev ijx^aTeiav, where a possible explanation of an
otherwise hopeless passage is that an early scribe thus
divided up aia>paKevep.f3aTeva>v (altering t to e) instead
of thus, alapa Kevcp-fiareiav, being deceived by the rarity
of the word KcvepftaTeveov. It must, of course, be re
membered that the earliest mss. have no accents or
breathings. A similar form of mistake is due to misunderstanding

4 THE TEXT OF THE NEW TESTAMENT
of contracted words. A possible example of this is the
curious reading in Matt, xxvii. 16, rlva afKere cmo rtav Sio
airokio-a Vfiip 'lnaovp ftapafifiav r\ 'Introvv k.t.X. The usual
way of writing 'lyo-ow is w, i.e. the first and last letters.
It is suggested that the origin of the reading is that an
early scribe was guilty of dittography, and wrote itp.iviv
for bfi.iv, but saw his mistake and deleted the second iv
by dots— iv. This was taken for a contracted word by
some later reader — the more easily because Bapafifiav
seems to be a patronymic. (Such an explanation is, of
course, double-edged ; the omission of at can be explained
equally well as an example of haplography. )
It is very important to collect the examples of this
kind of mistake, not simply because their detection is
a first step towards the purifying of the text, but because
they are an important clue to the history of the manu
script in which they occur. The more senseless the
mistake, the more important it sometimes is, e.g. in
Matt. xiii. 54, Cod. Sinaiticus reads els rr/v avrmarpiha
for els tj)v narpiSa, where Dr. Rendel Harris has pointed
out that this is a clue to the birthplace of the ms., just
as we might imagine an Oxford scribe of Shakespeare
writing —
" I come to ' Banbury ' Caesar " for ' bury ' Caesar,
and mistakes in spelling, especially if repeated, often
give a hint as to the pronunciation, and so nationality,
of the scribe. For example, if a scribe of early date
is found to write consistently 'michi' for cmihi,' it is
probable that he is a Spaniard.
All these forms of mistake and similar ones are fairly
easy to detect, and their classification is the first thino-
that a critic has to do. Some of them, such as ditto-
graphs, are obvious at once, others are only recognised
when several other mss. have been seen, and a roughly
provisional text exists at least in the mind of the critic.
It must, however, be remembered that great caution is
required in deciding whether a reading is certainly
corrupt or only possibly so. And the critic has always
to be ready to revise his judgment. He ought always

METHOD OF CRITICISM 6
to be suspicious of readings, but far more suspicious of
his own conclusions.
2. Intentional alterations. — As was said, these are due
to a desire to improve the text, either because the arche
type had an obvious mistake, which the scribe wished to
emend, or because he wished to simplify a difficulty. It
is clear that often a scribe made an easier text than the
original, and therefore one rule of criticism is that when
two variant readings are obviously connected with one
another, the more difficult is to be preferred.
It would be a profitless task to attempt to classify the
possible causes of intentional alteration. But some of
those which especially affect the New Testament text
are : —
(a) The influence of translations known to the scribe.
This is especially the case with bilingual mss.,1
which are Greek in one column and Latin in
the next. The texts of the Greek and Latin in
these cases are almost always accommodated to
each other, partly in order to have as little
divergence of reading as possible, partly for the
mechanical reason of wishing to keep one line
of Greek equal to one line of Latin. As it must
have often happened that unilingual mss. were
made by copying the appropriate column of a
bilingual ms., it will be seen that the influence of
translations has always to be remembered.
(6) A cognate cause is the influence of harmonies,2 or
even the study of the comparison of the four
Gospels. This inevitably led to a tendency to
assimilate the Gospels to one another, and to
remove discrepancies and contradictions.
(c) It is probable, though not certain, that dogmatic
reasons may have caused alterations. It is known
1 As will be seen, there are some erities who believe that the
oldest form of the Latin version was bilingual, and possibly even
interlinear. Bilingual mss. are not only Graeco-Latin, there are
also examples of Graeeo-Thebaie and Graeco- Arabic.
a i.e. texts in which one continuous narrative is made by
compilation from the four Gospels.

6 THE TEXT OF THE NEW TESTAMENT
that Marcion about a.d. 140 'edited' a New
Testament to suit his views, and on a priori
grounds we may say that both orthodox and
heretics would often have wished to make altera
tions. The fact that so few variants can certainly
be traced to this cause is probably due to the
vigilance with which the orthodox and heretics
regarded each other's efforts in this direction.
The elimination from the text of all the errors which
are obviously due to the above-mentioned causes forms
the first stage of textual criticism ; and since it implies
at least in practice some degree of knowledge of other l
recorded readings, it passes easily into the second stage
with no perceptible break in actual practice.
II. At this point the critic has as it were before him
a number of mss., the text of which has been roughly
cleansed from its more obvious impurities. A list of
the various mistakes in each codex lies in front of him.
Inspection will probably show that it is now possible to
group the mss. according to their agreement and dis
agreement in possessing these mistakes. Now, it is
obvious that whereas agreement in a correct reading is
no criterion of similarity of origin, agreement in
erroneous readings is a very good criterion. So that
by arranging mss. according to their common mistakes,
it is possible to form a kind of genealogy. That is to
say, it is possible to argue that Codex A and Codex B
are both copies of the same ms., because both have the
same mistakes (although each has also mistakes peculiar
to it), and it is possible to reconstruct this original ms.,
if it be not extant, by putting into it all the common
mistakes, and leaving out those which are peculiar to
one alone. By going through this process one probably
finds the mass of mss. beginning to assume some order,
and one is able to say as follows : —
1. Codd. A, B, C, D, represent an archetype X.
2. Codd. E, F, „ „ Y.
1 That is to say, we may find a sentence in some mss. which ia
hopeless nonsense. Its correction is necessarily made simple and
certain if we know that all other mss. have a reading which is
good sense, and of which the nonsense is clearly a corruption.

METHOD OF CRITICISM 7
3. Codd. G, H, I, J, K, L, M, N, represent an arche
type Z.
Or we may represent the same by genealogical trees
thus: —

ABCD E F GHIJKLMN
It must, however, be observed that it is often impossible
to group mss. quite so decisively as this. For we often
find Mixed variants, and consequently cross grouping.
That is to say, mss. were not always copied directly
from one source ; but, on the contrary, scribes often
produced an eclectic text, so that the same ms. may
sometimes represent more than one archetype, and be
found sometimes in one group, sometimes in another.
Such a result may genealogically be illustrated thus : —
X Y
I  |__
I I I
BAD
in which A represents a- mixture of archetypes X and Y,
while B and D are unmixed representatives of each
respectively. In reconstructing, therefore, the archetype
X the critic would argue that where A arid B agree
their text represents X ; where they differ, if A agrees
with D, probably B represents X, and the reading of A
is due to the mixture with archetype Y. If A does not
agree with D, there are three alternatives : (1) B
represents X, D represents Y, A represents some peculiar
variant due to its own scribe ; (2) A represents X, and B
has the peculiar error, D still representing Y ; (3) B
represents X, A represents Y, and D has the peculiar
error. Between these three possibilities choice is guided
by the consideration of the general characteristics of the
mss. and their archetypes which have been acquired by
the study of other variants in them.
III. At this point the third stage is reached. This is
the consideration of those numerous cases where in the

8 THE TEXT OF THE NEW TESTAMENT
reconstructed archetypes there are two or three variants
in a reading, which are all possible, and none obviously
wrong. To some extent the reading in the majority of
mss. has a claim to be adopted; to some extent it is
equally clear that the reading in the oldest mss. has a
claim. But the genealogical process mentioned shows the
limitations of these presumptions. One ms. now lost
may have been copied more often than another, and so
its readings may have obtained in the next generation
a fallacious majority in their favour ; or a very old ms.
may be represented best by a recent copy, which thus
represents a generation closer to the archetype than
other copies made long before.
Therefore instead of considering merely numbers and
age in the mss. attesting a reading, we have to consider
the numbers and ages of the archetypes of groups.
First we judge what was written in the archetype of a
group, and then comparing it with others, decide which
is the right reading. This is done by the consideration
of two points with regard to each reading : —
(1) Intrinsic probability.
(2) Transcriptional probability.
That is to say, it is necessary first to consider which of
the two or more variants makes the best sense ; which is
most in accordance with the general style of the author,
and so on ; which, in short, the author is most likely to
have written. And secondly, to consider how the scribe
is likely to have arrived, by the kind of mistakes
mentioned above, at any of the various readings.
This process is not always easy, for sometimes intrinsic '
and transcriptional probability seem to point different
ways. To some extent the harder reading has always
got transcriptional probability. But there is a point at
which a reading is so hard that intrinsic probability is
decisively against it. In such a case it sometimes is very
difficult to judge.
The result, therefore, of the earlier work at this stage
is to remove variants which, though not obviously wrong,
like those removed in the first stage, are nevertheless
shown to be wrong by the test of the two kinds of

METHOD OF CRITICISM 9
probability. And a further criterion can now he applied.
It will be possible to say what are the characteristic
mistakes of each group and its archetype. It will be
recognised that while Group A frequently interpolates
it rarely omits, while the contrary is true of Group B.
The application of this criterion will often decide between
readings, the evidence for which is otherwise equal.
And in this way the process is carried on, successive
layers as it were of corruptions being scraped off and
explained, and the number of archetypes gradually
reduced in number and thrown back in age, until it is
possible to construct a provisional text which represents
the archetype of all known mss. The construction of
this archetype would be the completion of the third
stage. New Testament criticism has not yet reached it,
and therefore the fourth stage is not yet of any great
practical importance.
IV. This consists in conjectural emendation of those
passages which seem to be corrupt even in the archetype
of ali known mss. It calls for a knowledge of palaeo
graphy, in order to discover how the scribes' errors may
have arisen, and for a certain ingenuity in making up a
theory as to how a mistake may have crept in. For
instance, Dr. Hort suggests for the impossible 6£kav iv
TaTreivo<ppoo~vvrj of Col. ii. 18, cdeXorafffivo^poirvvT], which
makes sense, is distinctly Pauline, and is a rare (if not
unknown) word which a scribe might easily misread.
Such, roughly speaking, is the method of modern
textual criticism. Three points are worth noting in it : —
(1) The difference between Biblical and classical textual
criticism. In classical textual criticism, the archetype
of all the extant mss. is often obtainable with com
paratively little work, but often is very corrupt. There
is therefore scope for much conjectural emendation. In
Biblical textual criticism, on the other hand, it is still
doubtful what is the archetype of the existing manu
scripts. But at least we may be sure that it is an
exceedingly early one, with very few corruptions, and
therefore the work of the conjectural emendation is very
light, and scarcely ever necessary.
(2) It is impossible to separate the history of the text

10 THE TEXT OF THE NEW TESTAMENT
from the general history of the Church. The local
history of a district, the monasteries of the country,
local heresies, and certainly local pronunciations and
dialects with their variations at different times, all act
on the text, and are influenced by it in turn. The
perfect textual critic will have to be an expert palaeo
grapher and the possessor of a complete knowledge of
all the bypaths of Church history.
(3) It also follows from this that judgments on a
difficult question are really valueless unless they are
made not merely with regard to the probabilities of the
individual case, but with a distinct grasp of the family
relations as it were of the mss. concerned, and their
characteristics. The object to be aimed at is to find the
right reading by way of tracing the history of the wrong
reading through the various ramifications of the MS.
genealogical tree, until a point is reached at which it
first appears, and before which it is not found.
After thus roughly establishing the method which is,
and ought to be, usually adopted, the next step naturally
is to ask what is the material to which the student of
New Testament Textual criticism has to apply his
method. An attempt will be made in the three next
chapters to answer this question, and in Chapter V. to
sketch the outlines of the attempts which have been
made to apply this and other methods to the material,
commonly called the 'apparatus criticus.' But it has
been found neither possible nor desirable to avoid re
ferring in Chapters II. III. and IV. to ideas and termino
logies which cannot be fully explained until later.

CHAPTER II
THE APPARATUS CRITICUS OP TnE NEW TESTAMENT —
GREEK MSS.
The apparatus criticus is the whole collection of docu
ments which is at the disposal of the critic.
It consists ultimately entirely of mss., though practi
cally a large proportion of them can be used conveniently
and adequately through printed editions.
These mss. fall into four groups ; —
(1) Greek mss.
(2) mss. of versions.
(3) Lectionaries.
(4) mss. of other works, practically all of them
'Fathers,' which quote the New Testament.
Of course, for practical purposes the student of any
one of these groups often takes some standard edition of
the others as a working text, and verifies it in detail only
when necessary.
A complete knowledge of the whole apparatus is more
than any one possesses, but a fair working knowledge of
the chief documents is a necessity for any critic.
Greek MSS. — These are divided into two classes —
Uncials and Cursives.
(1) Uncials. — The exact derivation of this word is un
known. But the meaning is that style of writing in
which each letter is separate, and, 'roughly speaking, of a
' capital ' shape, with some degree of curving.
(2) Cursives. — Those written in a ' running hand ' in
which letters are ligatured together.
The old theory about these two kinds of writing seems
to have been that uncials were used in the earliest times,
and that the cursive hand was adopted in the eighth or

12 THE TEXT OF THE NEW TESTAMENT
ninth century. Certain discoveries, however, show that
this is not the case. The earliest division of writing, which
is found long before the era of Christ, is into literary and
private 'hands.' The former is at first always an uncial,
the latter a cursive type.^ These two 'hands' must have
existed side by side throughout the first eight centuries.
But about that time a literary cursive hand was adopted
by professional scribes which gradually supplanted the old
uncial writing. The word cursive is therefore a little
ambiguous. It means a running private hand in the
earliest times, and it is also used for the literary hand in
which small connected letters are used. It would perhaps
be better if the word cursive were kept for the private
hand, and minuscule adopted for the literary cursive.
There are more than seventy-two uncials of the New
Testament, denoted by the capital letters of the English,
Greek, and Hebrew alphabets.
The number of mss. which we possess at present is,
however, so large that even with the help of three
alphabets it is scarcely possible to find letters for all the
mss. The plan has therefore been adopted of using the
same letter for different1 mss. of different books of the
Bible. This method is based on the fact that complete
Bibles (' Pandects,' they are called) are very rare, and even
complete New Testaments are not common, the usual
plan having been to have one ms. volume of Gospels,
another of Acts and Catholic Epistles, and a third of the
Pauline Epistles. It is usually obvious which ms. is
meant, but in doubtful cases the recognised practice is to
write a numeral in the right-hand bottom corner of the
letter — e.g. E = Cod. Basileensis of the Gospels; E2 = Cod.
Laudianus of the Acts; E3 = Cod. Sangermanensis of the
Epistles. The most important uncials are the following : —
N (Aleph) Codex Sinaiticus, an uncial of the fourth
century, now in the Library of St. Petersburg. This
ms. was found by Tischendorf in 1844, z at the monastery
1 In a few cases also different letters for the same us. in
different books.
' The discovery was made in this year, but it was not com
pleted until 1859.

APPARATUS CRITICUS 13
of St. Catherine, on Mount Sinai. The story of his
adventures and difficulties is most interesting, but is
scarcely an essential part of textual knowledge. (It can
be seen in Scrivener, ed. iv. vol. i. p. 90 ff.) The ms. is
13£ inches by 14$, and at present has 346J leaves of thin
beautiful vellum. It is written in four columns, with
forty-eight lines in each. There are practically no
accents or breathings, and very few contractions or
abbreviations. The margin contains the Ammonian sections and
Eusebian canons (v. p. 54), but not by the first scribe.
The text itself, according to Tischendorf, was written by
four scribes, of whom one, who wrote the last part of
S. Mark, is identified with the scribe of Cod. Vaticanus
(B). It has been corrected several times : —
(1) In a few places by the first hand, or by the hiopBitrrji,
i.e. the corrector of the same date, who, according to
custom, revised the ms.
(2) In the sixth century by an elegant writer whose
notes are often important. Known as Xa.
(3) In the sixth century by another scribe a little
later. Known as Xb.
(4) In the seventh century in many places by a scribe
known as N° or as Nca.
(5) In the same century by Ncb Ncc and NCQ which are
less important.
(6) In the twelfth century in a few places by N°.
The ms. contains the Gospels, the Pauline Epistles, the
Acts, the Catholic Epistles, the Apocalypse, Barnabas,
Hermas, and has the Epistle to the Hebrews after 2
Thessalonians. The origin of this great Codex is one of the nicest
Problems of criticism. It can, however, be discussed best
after Liie description of Codex Vaticanus (B), with which
it is inextricably involved. The text contained in N is
of a mixed character, as is explained in Chapter V. The
basis of it is ' Neutral,' but there is a large admixture
of 'Western' and 'Alexandrian,' without, however, any
clear traces of 'distinctively Syrian' readings. The
text of N° n the Pauline Epistles is peculiar; and,
as is shown in Chapter IV., forms a group with HPaul,

14 THE TEXT OF THE NEW TESTAMENT
Euthalius, and, possibly, the Armenian version. The
same hand has a note at the end of the Book of Esther,
to the effect that the text has been compared with a ms.
belonging to Pamphilus. This, it will be seen, is of
great importance in attempting to fix the birthplace
of N.
B = Codex Vaticanus. At Rome. An uncial of the
fourth century. It appears in the Vatican catalogue of
1475, but no one knows how it came there. It is written
in three columns, and forty-two lines to a column. In
the tenth century some one inked over the writing,
sometimes altering a word, and sometimes emending by
omitting to ink a word or phrase which he wished to
leave out. It has also been corrected by the usual
diopdarvs, and by one other scribe of no importance.
The text itself was written by three scribes, of whom one
alone wrote the New Testament. Tischendorf identified
him with the scribe 'D' of X, who wrote the end of
Mark and part of the Old Testament in the latter
ms. (v. p. 13). The text in the Gospels and Acts is the
purest known specimen of ' Neutral,' but in the Pauline
Epistles there is a distinctly ' Western ' element.
It is now necessary to ask what is the birthplace
of KB? This is a question which has to be answered
for both together, not because they have an extra
ordinary similarity of text, although that is a marked
phenomenon, but because of certain facts which show
that they were originally both together at the same spot.
This spot is Caesarea. Almost all critics now accept this
conclusion, though Drs. Westcott and Hort in their
Greek text were inclined to think that some peculiarities
of spelling in proper nouns point rather to the West.
The case for Caesarea is this : —
(1) The colophon of Esther in N°, which seems to show
that in the seventh century at least N was at Caesarea,
and was compared with a ms. written in that place by
Pamphilus. (2) The curious reading in N in Matt. xiii. 64,
'AvmrarptSa, which Dr. llendel Harris describes as the
mistake of a local writer. Cf. p. 4.
(3) The identity of hands in part of N and B (if

APPARATUS CRITICUS 15
Tischendorf be right, and his view is generally allowed
to be extremely probable).
(4) A curious chapter-division in Acts, which can be
traced through Euthalius to Pamphilus and Caesarea.1
More must be said on this point in Chapter IV. in
connection with systems of chapter- divisions.
Thus it will be seen that there is evidence to connect
N with Pamphilus and Caesarea, X with B, and X B and
Euthalius with Pamphilus and Caesarea.
Can we say any more? Some critics think that we
can, and connect both N and B with a definite edition,
which is mentioned in Eusebius' life of Coustantine.
Eusebius says that he sent to Constantine's new city fifty
aatpaTia iv bi(p8ipais, . . . iv irokvTeXas rjaKr/fievois revxeai
rpio-a-a Kai Terpacro-d. No one knows quite what is meant
by this last phrase. Rival views are : (1) Bound up in
quires of three and four sheets ; (2) written in three and
four columns ; (3) in cases of three or four. Those who
accept the second explanation point to the fact that N
and B are in four and three columns, and that X has the
Eusebian canons, by the first or a contemporary hand.
They therefore regard the two great uncials, so closely
connected with each other and with Eusebius' home, as
part of Eusebius' present to Constantine.
On the other hand Rahlfs l has pointed out that the
order of the books in B corresponds to the Canon of
Scripture given by Athanasius in the 'Festal Letter' of
367 a.d., and thinks that this points to Alexandria
rather than Caesarea (see also p. 53).
A = Codex Alexandrinus. Now in the British Museum.
An uncial of the fifth century, which was given to Sir
Thomas Roe, the British Ambassador of Charles i. to
Turkey, by Cyril Lucar the Patriarch, who tried to re
form the Eastern Church on the Geneva mqdel. There
is an Arabic note at the beginning which says that it
was written by Thecla the Martyr. It has a mixed text,
worse in the Gospels than elsewhere, with a considerable
Alexandrian element.
C. An uncial of the fifth century at Paris. Known as
1 A. Rahlfs Alter unci Heimat der Vatikanischen Bibelhand-
schrifl, in the Nachrichten der Gesell. der Wits, zv, Gottingcn,
1S89, i. pjj. 72-79.

16 THE TEXT OF THE NEW TESTAMENT
the Palimpsest of Ephraem Syrus, that is to say, it is a ms.
which has been used twice — once for a copy of the New
Testament, and later, when the original writing had been
rubbed or washed off as far as possible, for the writings
of Ephraem the Syrian. The lower writing has been re
vived to some extent by the use of chemicals, but it is
always a serious task to read even a good palimpsest. It
seems to have been brought from the East by Andrew
John Lascar, who gave it to the Medici family, and so
through Queen Catherine de Medici it came at last to
the Bibliotheque Royale (now Nationale) in Paris. Its
text, like that of A, is mixed ; but it has, in the Gospels
especially, a considerable number of Neutral and Alexan
drian readings.
D = Codex Bezae. At Cambridge. A Graeco-Latin
ms. of the sixth century, containing the Gospels and Acts.
It was at Lyons1 in the sixteenth century, whence by
some not entirely understood means Theodore Beza the
Reformer obtained it and presented it to the University
of Cambridge. It has been published in extenso by the
late Dr. Scrivener, and a complete photographic repro
duction has now been published. It is a ms. of the
greatest importance and interest, not because it possesses
a necessarily good or sound text, but because it is tne
earliest Greek form that we possess of what is known as
the Western Text (see Chapter VI.).
The questions which are raised about D independently
of the general question of the Western text are these : —
(1) Where was it written?
(2) Has the Greek been assimilated to the Latin, or
the Latin to the Greek ?
(3) Are there any traces of Syriac or Aramaic in
fluences ?
The books which deal with the questions best are
Scrivener's edition of the ms., Dr. Rendel Harris's Study
of Codex Bezae, and Dr. Chase's two books on the Syro-
Latin Text.
1 So Beza said, but there is also reason for doubting whether
Beza was in a position to know, and some evidence that it was
really in Italy.

APPARATUS CRITICUS 17
(1) The first question is hard to answer, but on the
whole the Rhone valley is the most probable place.
This is shown as clearly as it can be by Dr. Rendel
Harris. He bases his arguments on various marginal
notes in the ms., and on certain philological indica
tions in the text itself. The first argument is doubtful,
for the marginal notes show signs of Byzantine litur
gical usages of the ninth century,1 and suggest that at
that time Codex Bezae was in South Italy, where the
Creek rite was used. But the philological indications
suggest that at the time when the Codex was written the
Latin language was in a state of transition into the
Romance languages, and more especially into French, so
that the Rhone valley, and possibly Lyons itself, is a very
probable place for the origin of D. It must, of course, be
remembered that this does not at all imply that the text
represented by D also came from that locality. Dr. Rendel
Harris himself points out various traces of other localities,
including especially Carthage, which seem possibly to
have influenced some of the archetypes of D.
(2) Dr. Rendel Harris is the great advocate of assimi
lation of the Greek to the Latin. Before his work
appeared the usual view was that accepted by Drs.
Westcott and Hort, that when the two texts agreed the
Latin followed the Greek, and not vice versd. It is im
possible to summarise Dr. Rendel Harris's case against
this theory.
Many of his examples probably fall to the ground ;
in some cases the converse of his theory appears pro
bable ; but a residuum remains, and the most probable
view is that neither text of Codex Bezae has entirely
escaped from the influence of the other.
(3) It has been suggested that there are Syriac forms and
idioms in D. The case in favour of this theory is elabor
ately worked out by Dr. Chase in the Syriac Element in
Codex Bezae. A little more must be said on the point in
Chapter VI. Scholars are not agreed on the amount of
Eastern or Syriac influence which they trace in Codex
Bezae, but most of them agree that there is some. For
instance, the spelling of ijX«, rj\ei \apa (a<p8avti, in Matt
xxvii. 46, seems to imply a knowledge of Hebrew and
1 See Dr. Rendel Harris's Annotators of Codex Bezae.

18 THE TEXT OF THE NEW TESTAMENT
Aramaic. A brief account of Dr. Chase's arguments will
be found on p. 87 in connection with the general question
of the origin of the Western text.
There are no other mss. of equal importance to these.
A list ot all the uncials will be found in Appendix B.
But it is perhaps as well to say something about a few.
L Codex Regius, at Paris, a ms. of the eighth century.
Said to be probably written by an Egyptian scribe. Its
text is allied to that of B, but it has many readings in
common with Origen. It is remarkable for possessing
the ' shorter conclusion ' of S. Mark.
A. Codex Sangallensis, in the monastery of S. Gall,
probably of the eighth century, a Greek uncial with an
interlinear Latin version. It has an inferior text as a
rule, but in Mark there are many readings of an
Alexandrian, and sometimes Neutral type. It is the same
ms. as G3 of the Epistles.
T. Cod. Borgianus (v.) is a collection of Graeco-Sahidic
fragments, which are remarkable for a Neutral text, and
according to Westcott and Hort approach more nearly to
B than any other ms.
S. Cod. Zacynthius (viii.) is a palimpsest fragment of
Luke, belonging to the British and Foreign Bible Society,
with a Neutral and Alexandrian type of text.
E3. Codex Laudianus of the Acts (vii.), a Latino-Greek
ms. probably written in Sardinia, and thence taken to
Britain, when it was almost certainly used by Bede in
his Betractationes. It afterwards passed into the possession
of Archbishop Laud, who gave it to the Bodleian, where
it is at present. It has a remarkable Western text, allied
to D, but more closely connected with the Latin ms.
'gigas,' which is at Stockholm, and so with Lucifer of
Cagliari in Sardinia, whose text is almost identical with
gigas. It has the same chapter-numeration system as
KB am. Euthalius (v. p. 65).
D?. Cod. Claromontanus(vi.); E,. Cod. Sangermanen-
sis (ix.) ; F2. Cod. Augiensis (ix.) ; G3. Cod. Boeraerianus
Six.) of the Pauline Epistles, form a group of Graeco-
.atin mss. which are certainly closely connected. E« is
almost certainly copied from D2 ; and Fj, if not copied
from Gj, is probably an inferior copy of the same arche-

APPARATUS CRITICUS 19
type ; while D2 and G3 themselves probably have a
common and not very distant ancestor, D2 being pro
bably the better representation as a rule. The text of
their group is called Western, because its characteristics
are similar to that of the Western text (cf. Chapter VI.)
of the Gospels and Acts. They are, however, so much less
striking that there is room for legitimate doubt as to
whether the Western text of the. Gospels and Acts is
not due to some special cause which did not affect the
Epistles. II3. Codex Coislianus of the Pauline Epistles (of which
12 leaves are at Paris, 9 in the Laura S. Athanasii on
Mount Athos, 2 at Moscow, 6 at various libraries at
S. Petersburg, and 2 at Turin), an uncial of the sixth
century, which seems from the subscriptions and notes
to have been originally in the great monastery at
Caesarea. Its text is that of the group represented by
N°H3, the Armenian version1 and Euthalius, and often
termed the Euthalian text. There is an elaborate and
most interesting account of it in Dr. Armitage
Robinson's Euthaliana in Texts and Studies.
Turning to the cursives, or more properly minuscules,
we find that their numbers are greater than their interest.
The list is constantly being increased by the cataloguing
of new mss., and there are now about three thousand.
They are denoted by figures. There is no fixed system
of cataloguing them which is universally received ;
Gregory's catalogue probably is used most often, but
some use Scrivener's Introduction, which has not always
the same numeration. As things stand at present, it is
a superhuman task to remember in the case even of
important mss. the differing notation of Gregory and
Scrivener, and also in some cases the antiquated and
provisional notations which were used by Tischendorf
and by Westcott and Hort in their editions previous
to the catalogues of Gregory and Scrivener. This can
be seen at once by looking at Appendix A on Tischen-
dorf's Notation.
There are, however, a few minuscules which are inter-
1 The Armenian version may have to be omitted from this
group.

20 THE TEXT OF THE NEW TESTAMENT
esting and important, though it is scarcely possible in the
present book to do more than indicate which they are.
33. Paris Nat. Gr. 14, probably of the tenth century
or earlier, in bad condition, and most difficult to read.
Its text is more like that of K B than is that of any other
minuscule. It may probably be described as Alexandrian
and Neutral, with some Syrian elements.
13-69-124-346, at Paris, Leicester, Vienna, and Milan
respectively, are all twelfth century mss., except 69,
which belongs to the fifteenth at the earliest. They are
known as the Ferrar group, because Dr. Ferrar of Dublin
proved their origin from a common archetype. The Abbe
Martin has also shown, partly on palaeographical grounds,
partly on the hints conveyed by the names of the saints
in menologies or calendars which are found in each, that
they all (except 69, which is later) originated in Calabria
or Sicily.
Dr. Rendel Harris has also shown that the curious
text which they preserve has many affinities with the old
Syriac, and especially with Tatian's Diatessaron. They
are therefore valuable evidence for Western readings,
and in some cases preserve readings which are not found
in any other Greek mss., though attested by the Syriac
and Latin versions. There are several other mss. which
belong to this group, notably 543 a Burdett Courts ms.
of the twelfth century ; 826 and 828, also of the twelfth
century, at Grotta Ferrata ; and 788 at Athens,1 and
von Soden (see Appendix D) adds several others which
he thinks represent more or less attenuated forms of the
same text.
Besides the Ferrar group, there is a family of four mss.,
codd. 1, 118, 131, and 209, the text of which is edited
in Texts and Studies, vii. 3. These have unquestionably
a common archetype, and it is suggested in Texts and
Studies that this archetype, the Ferrar group, and codd.
22, 28, 565 and 700 represent, at least in Mark, an
ancient text which has been corrupted in different ways
1 There are also certain other mss. {e.g. 211, 709) which have
some points of affinity with the group, though they are not
primary members of it. The Abbe' Martin thought that 348
belonged to the group, but this is not the case.

APPARATUS CRITICUS 21
by a varying admixture with the later types. Codd. 22
and 28, are mss. of the eleventh or twelfth century at
Paris ; 505 (Scrivener, 473) is a St. Petersburg ms.
which Tischendorf called 2"° because it is second in a
catalogue made by Muralt ; and 700 (Scrivener, 604) is
an eleventh century ms. in the British Museum, which
has been very carefully edited by Mr. Hoskier, in his
Collation of Cod. Ev. 604, 1890.
Recently von Soden has presented a somewhat different
view. The Ferrar group, codd. 28, 565 and 700 are, he
thinks, independent authorities for the earliest form of
his J text (see Appendix D) which roughly corresponds
to W. H's. Western text, and he has found another ms.
(his e-050) which largely agrees with cod. 700. The group
headed by cod. 1 is one of the four sub-families which
represent an early branch of I. The other sub-families
are represented by (1) cod. 872, a twelfth century ms. in the
Vatican. (2) A group headed by cod. 22, and (3) two
new mss. which he calls t-183 and e-1131. To this early
branch of I he gives the name of H1.
Minuscules of some importance for the criticism of the
Acts are 58 and 137, as they frequently preserve Western
readings (vide Pott's Der Abendlilndische Text der Apostel-
geschichte), 61, on the other hand, has a Neutral and
Alexandrian text.
In the Pauline Epistles, 17 ( = 33 of the Gospels) and 67
are conspicuous for often preserving early and good read
ings, and there are many minuscules of the Apocalypse
which are important not so much for themselves as for
preserving the commentaries of Andreas and Arethas,
which had a great influence on the text.
Von Soden's results on the Acts and Pauline Epistles
are not yet published ; no doubt they will throw new
light on the text, especially with regard to the evidence
of the minuscules.

CHAPTER III

THE VERSIONS

Next in importance to the Greek mss. come the Versions,
that is to say, the translations into various languages
which have been made at different times from the Greek
text. But before describing any of them separately, it
may be well to consider what is the value of a version
and what are the necessary limitations to its use for the
purposes of textual criticism. It is obvious that the
exact text of any given version has to be recovered in
the manner described in Chapter I. Assuming that this
can be done, the value of the version is that of the Greek
text underlying it. If the version is an early one, and
has been made from a good Greek text, its value is great;
otherwise it is small. If, therefore, we possess a version
which seems to have been made at a very early time
from a good text, we have nexAto ask the question, How
far can we reconstruct the Greek text which was used
by the makers of the version ? In considering this point,
the first problem is, How far is the translation a literal
one? For, of course, if we can assure ourselves that we
have a word-for-word translation, and that the same
word in the version always implies the same word in the
underlying Greek, we can restore the latter with ease
and certainty ; while, on the other hand, if we cannot so
assure ourselves, any restored text will be hypothetical
and tentative. In other words, before we can properly
use the evidence of versions, we must try to wrest from
them some information as to the method which the trans
lator employed in making them. Few things, however
are more remarkable in the study of textual criticism
22

THE VERSIONS 23
than the really solid advance which has been made in
this sphere of work, especially perhaps with regard to
the Latin version ; and probably all who have studied
the subject would agree that the general trend of recent
criticism has been to show that the early versions are
singularly faithful to their underlying Greek, in spite of
an occasional tendency to paraphrase.
The student of versions, therefore, has four distinct
tasks before he can use them for the purposes of textual
criticism — (1) By the application of the methods de
scribed in Chapter I. he must reconstruct the archetype
of the version with which he is dealing ; and here he
needs to exercise caution to avoid the mistake of suppos
ing that all mss. of the same book in the same language
represent the same version. There may be more than
one version in the same language.
(2) By careful comparison of renderings in different
places he must form an induction as to the methods of
the original translator, and decide whether he adopted
a paraphrastic or literal style.
(3) He must then proceed to reconstruct the under
lying Greek text ; and he will do this with confidence if
he found in the previous stage of his work that the charac
teristic of the translator was a literal style, so that
divergence of rendering may be safely taken to imply
divergence of reading, while he will do it with diffidence
and with alternative possibilities kept in view if he found
the reverse to be the case.
(4) He must finally form a judgment as to the value
of this reconstructed ms. on exactly the same grounds
as he would judge a Greek ms. of the same date and
character. It must be remembered that although the text of a
version has as great a chance of being corrupt as a Greek
ms. has, yet it is not often likely to have been corrupted
in the same way ; e.g. in 1 Tim. iii. 16, the Greek mss.
differ between Beos and St — i.e. between 62 and Q2.
Here a scribe's confusion is so probable that decision is
difficult ; but in Latin the difference is Deus and qui, so
that we can take the Latin as free, in this case, from a
";ause which may have contaminated the Greek,

24 THE TEXT OF THE NEW TESTAMENT
There are three versions of the New Testament of
first-class importance— Latin, Syriac, and Egyptian; and
possibly similarity of language in each case conceals the
fact that we have really more than one version in each.
And there is an immense number of translations of smaller
and varying importance, some of which have been made
from one of the three great versions, some independently
from later Greek mss. Such are the Armenian, Aethiopic,
Gothic, Georgian, Arabic, Persian, Sclavonic, and many
others. The Latin Version. — The history of the Latin version,
or it may be versions, bristles with difficult and disputed
points. The best way of describing it would, of course, be to
begin with the oldest form and trace its gradual growth
and development. But this is impossible. Its origin
is shrouded in mystery, and therefore it is necessary to
begin with the earliest historical statements about the
Latin version and work back from them to an earlier
time. The development after that point demands separate
treatment. The first statements on which we can rely are those of
Jerome and Augustine.
Jerome. — Jerome's information is given in the open
letter which he wrote to Pope Damasus in 384 a.d. as
a preface to his edition of a revised Latin text. This
revised text is that which is known as the Vulgate,
which, as will be seen, has been the text of Latin-speak
ing Churches of the West ever since. In this letter to
Damasus, Jerome explains :—
(1} Why he found it necessary to issue a revised text ;
(2) On what principles he conducted his revision.
His answer to the first point is that there was a great
difference between different renderings. ' Si enim Latinis
exemplaribus fides est adhibenda, respondeant quibus.
Tot enim exemplaria pene quot codices.' (So in Cod. Am.)
To the second, that he had revised the Latin by means
of the oldest Greek mss. which he could find, removing
harmonistic and clerical corruptions, but so far as pos
sible preserving the renderings which were familiar to
Latin ears.

THE VERSIONS 25
' Igitur haac praesens praefatiuncula pollicetur quattuor
tantum evangelia quorum ordo iste est Mattheus Marcus
Lucas Iohannes codicum graecorum emendata conlatione
sed et veterum. Quae ne multum e lectionis latinae con-
suetudine discreparent ita calamo temperavimus ut his
tantum quaa sensum videbantur mutare correctis reliqua
manere pateremur ut fuerant.'
So far, then, as Jerome's evidence goes, we are told
that at the end of the fourth century there was a great
variety of Latin renderings which differed both among
themselves and also from the Greek text.
Turning to Augustine, we find that he attests practically
the same facts. The locus classicus is his tract De Doctrina
Christiana. In this he speaks in the strongest terms
of 'Latiuorum interpretum infinita varietas,' which he
attributes to the effects of separate translations by early
Christians. ' Ut enim cuique primis fidei temporibus in
manus venit codex Graecus, et aliquantulum facultatis
sibi utriusque linguae habere videbatur, ausus est inter-
pretrari.' The result, therefore, of the evidence. of Augustine, as
of Jerome, is to niake us look for a pre- Vulgate version
or versions characterised by many variations of reading
and rendering. And the proper method will obviously
be to inquire whether we can group these variations
geographically and chronologically by identifying them
with the text used by any definite Father or group of
Fathers. First of all, then, we put on one side for the moment
all the mss. of the Vulgate or Jerome's edition, to be
dealt with later on. It is the residuum which is im
portant for the early history of the Latin version. It is
usual and convenient to cali this residuum the Old Latin
version. The Old Latin. — There are about twenty-seven mss. of
this of the Gospels, about seven mss. of the Acts, some
fragments of the Catholic Epistles, about six mss. of the
Pauline Epistles, and fragments of the Apocalypse.
lite first glance at these confirms the evidence of
Augustine and Jerome. The variations are immense.
But after a time two manuscripts of the Gospels single

26 THE TEXT OF THE NEW TESTAMENT
themselves out as particularly differing from the rest
and agreeing between themselves. These are —
k. Codex Bobiensis,1 now at Turin, but once in the
great monastery of Bobbio. It is a ms. of the sixth century,
and tradition connects it with S. Columban, the founder
of the monastery, who died there in 615.
e. Codex Palatinus, now at Vienna, either of the
fourth or fifth century. A beautiful fragment of purple
vellum, with gold and silver letters.
Their text is most peculiar both in reading and render
ing; and on examining the text of the earliest Fathers,
we find that we can identify a locality in which it origin
ally prevailed, although it does not follow that it was con
fined to this district. It is clearly the same as the text
used by Cyprian,2 the African father of the third century,
so that it has been called the African Latin. And using
the key given by Cyprian's use, we find the same type
of text in the Acts in a fragment known as h (cf. p. 97),
or the Fleury palimpsest, now at Paris, and in m or
Mai's edition of the so-called Speculum Augustini, a serie?
of quotations from the whole New Testament, except
Philemon, Hebrews, and 3 John. Whether this form
of text was originally used solely in Africa is a question
which we can hardly decide. It involves the whole ques
tion of the genesis of the Latin version. Possibly the
geographical limitations suggested by the name African
may prove to be baseless. But if we bear this possi
bility always in our minds, the name African is a useful
one, for it draws attention to the connection of the text
with the great African fathers, Tertullian and Cyprian.
Their date (second and third century) shows that the
Latin version is at least as old as they are. Zahn doubts
whether Tertullian had a written version ; but few agree
with him (cf. Salmon).
Having thus eliminated from the mass of Latin mss.
the Vulgate and African Old Latin, we have to examine
the few which remain, and see whether they all preserve
1 The edition of k in ' Old Latin Biblical Texts ' contains in its
introductions the most valuable collection of facts that the
student of the Old Latin possesses.
> Qne would add Tertullian but for the doubt raised belpw,

THE VERSIONS 27
the same type, or whether we can still further divide
them into groups. And it is generally agreed that we
can find two groups : —
(1) That known as the European.
(2) That known as the Italic.
The European is represented chiefly by two mss., known
as a, b, which are mss. of the fourth or fifth century, at
Vercelli and Verona respectively, and perhaps i, of the
seventh century. On the whole, too, the Latin text of
Codex Bezae is European, and the text used by the Latin
Irenaeus seems to belong to the same group.
The Italic is represented properly only by f Codex
Brixianus of the sixth century, and q Codex Monacensis
of the seventh. It probably was a modified form of the
European, perhaps owing to attempts, made before
Jerome's time, to revise the European text with the help
of the Greek. It is so called because it used to be thought
that it was to it that Augustine referred when he spoke
of the ' Itala.' For among the diversity of translations
which he mentions, he specifies one as in his opinion the
best : ' In ipsis autem interpretationibus Itala ceteris
prseferatur.' And it has been thought that this meant
that one of the pre- Vulgate texts was called Itala. But
Mr. Burkitt, in The Old Latin and the Itala, shows that
this is improbable. Augustine would probably use himself
the text which he considered preferable, and beyond all
doubt the text he was using in the Doctrina Christiana
was that of the recently published revision of Jerome.
So that all that the passage means is that Augustine,
who, as living in Africa, naturally termed Jerome's version
the 'Itala,' considered it to be better than any of its
predecessors. At the same time, it is probably true that f, q, repre
sent the type of Latin mss. most used by Jerome. So far,
then, we can roughly divide the Old Latin in three
groups, represented in the Gospels chiefly as follows : —
(1) The African in k, e.
(2) The European in a, b, 1.
(3) The Italic in f, q.
A fuller list, and grouping of the mss. is given on pages
32, 96 f. Many represent, as may be seen, the corrupt

28 THE TEXT OF THE NEW TESTAMENT
and mixed forms of the three families, the basis of most
of them being European.
Problems of the Old Latin version : —
(1) Are these three groups separate in origin, or do
they go back to one original version of which no other
trace remains ?
(2) What was the character of the Greek text which
underlies the Old Latin ?
No final answer can be given as yet to either problem.
As to (1). Probably we have scarcely digested the
material at our disposal sufficiently to form a decisive
opinion. Few, however, would refuse to admit that the Italic
family is a later form of the European, and cannot be
separated from it in origin.
And perhaps most critics would also agree that at least
in large portions of the Gospels we can reconstruct
with considerable certainty a text which was current in
Africa in the third and probably the second century.
We can with rather less certainty reconstruct an almost
contemporary text of a European kind ; and when the
critical edition of the Novum Testamentum S. Irenaei is
published, we shall do so with more confidence. But,
even so, much of these reconstructions is tentative. Nor
does a comparison of them lead to many definite results.
Sometimes it seems as though we had traces of original
differences of rendering which necessitate the theory of
at least two original versions.
But against these have to be set other cases of extra
ordinary constancy in maintaining the same rendering,
and sometimes even a wrong rendering. For instance,
in any reconstructions, such as those suggested above,
both African and European would read gaudentes in
Mc. ix. 15, a rendering which is due to a mistake made
by misreading rpixovres as xaiP0VTfs- It ls this kind of
phenomenon which drives one to believe that ultimately
the African and European Latin are traceable to one
original version. But, as was said above, the point is
still doubtful, and, of course, the presumption on a priori
grounds is in favour of many original versions rather
than one only. Perhaps it may ultimately turn out that

THE VERSIONS 29
there were originally many local versions, but that an
official version was adopted at a very early date. This
version may lie behind both African and European, while
the continued local use of the original attempts may
explain some of the striking differences which we find.
Many critics also believe that the earliest form of the
Latin version was a bilingual one. That is to say, it
consisted of a Greek text with a Latin translation either
in parallel columns or in alternate lines. This view is
supported by the existence of the early bilingual mss., such
as D A of the Gospels, E of the Acts, and D, E, F, G of
the Pauline Epistles ; and also by the exact verbal agree
ment of the Latin and Greek, in which the translator
seems to have had some reason for wishing to make the
sentences and the words in one language exactly corre
spond in number with those in the other.
As to (2). The question really is whether all the
oldest variants found in the Latin are to be traced back
to a Greek source. The point is this : If we assume that
the Latin represents an accurate translation of the Greek,
we have to suppose a lost Greek original of a definitely
' Western ' type, such as is now found in Greek only in
D (which is suspected of being influenced by the Latin),
and in a few places by the Ferrar group and some others.
Was there ever such a Greek text? This is really the
great problem of modern textual criticism — the origin of
the Western text. Here it is only necessary to say that
on the whole it is inconceivable that the interpolations of
the Latin versions do not go back to a Greek original for
two reasons : —
(a) Many of them, though not all, are also found in
the Old Syriac (v. p. 33).
Whatever view is taken of the close relationship which
exists between the Latin and Syriac, no one has suggested
that either has been made from the other, therefore a
common Greek source is demanded for all the interpola
tions, etc. , which are common to the Latin and Syriac.
(b) Many of the Western readings are traceable in the
text used by Justin Martyr and by Marcion. In these
cases it is obvious that a Greek and not a Latin text was
used.

30 THE TEXT OF THE NEW TESTAMENT
At the same time, it is quite likely that some of the
Western readings may be originally purely Latin, and
due to corruptions of the Latin text rather than the use
of a corrupt Greek original, assuming, what is not certain,
that no reading of purely ' Western ' attestation is primi
tive and correct.
One more point calls for mention. If we believe that
in any sense the Old Latin mss. are traceable to one
original version, the question may be legitimately raised,
Where was this original version first made ? This, again,
is a question which is not yet answerable. The primitive
character of the Cyprianic text might suggest Africa ; but,
on the other hand, the a priori probability in favour of
Rome is very strong ; while, paradoxical as it may seem,
the close textual connection subsisting between the Latin
and the Syriac versions has led Dr. Sanday and Dr. Chase
to consider seriously the possibility whether the Latin
version was not made originally in the East, perhaps at
Antioch (v. p. 87).
Having thus traced the history of the Latin backwards
from Jerome to the earliest times of which we can say
anything even with probability, it is necessary to return
to the starting-point and trace the history of the revised
version, which, as was seen (p. 25), he founded on the
Old Latin mss. of his time.
The Vulgate. — When Jerome was alive, the characteristics
of the Old Latin must have been very like those which
we have seen in the mss. which are still extant. And
when at the instance of Pope Damasus he undertook the
task of revising the version, he must have had as a pre
liminary to decide what he should take as the basis of his
work. To judge from the result, he adopted for his
Latin authorities the Italic type now represented by f
and q, and revised it by his knowledge of other types and
of the original Greek. It is an interesting question what
Greek mss. Jerome used, for no one ms. now known covers
all the readings which seem to have come from a non-
Latin source. But, on the whole, Jerome's Greek mss.
were* probably of a ' Neutral ' rather than a ' Western '
type. He published the Gospels in 384, and the rest of
the New Testament probably before 386, and his version

THE VERSIONS 31
gradually became accepted as the standard Latin Bible.
But, naturally, as years went on the text of the Vulgate
itself became corrupt; it was contaminated not only by
the ordinary causes of corruption in the course of repro
duction, but also by the influence of the Old Latin which
was still extant.
The result is that we possess mss.1 with a Vulgate base
which contain sporadic readings which have crept in from
Old Latin mss. of every kind. There is scarcely any Old
Latin reading which cannot be found in some Vulgate ms.
Therefore revisions of the Vulgate became necessary.
The earliest of these which were important were made in
Gaul, by Alcuin in 801, and a few years later by Theodulf.
Various other attempts at revision were made later on,
but perhaps the most important are the lists of variants
collected by various students and called ' correctoria.'
But when printing was invented the publication and
reproduction of better editions became easier. Famous
among these are the Complutensian Polyglot of Cardinal
Ximenes, and Erasmus' notes on the Latin translation
which he appended to his famous edition of the Greek
Testament ; but the first really critical edition was that
of Robert Stephanus in 1528. At about this period the
Roman Church began to recognise the importance of pro
ducing a pure Vulgate text. Pope Sixtus v. (1585-90)
accordingly undertook the publication of a pure and
authentic text. The result was the publication in 1690
of an edition accompanied by a Bull declaring that this
was the authentic and only trustworthy version. Sixtus,
however, died soon after this, and 'in 1592 Clement vm.
called in all the copies of the Sixtine edition, for a reason
which is somewhat obscure. Some say merely because it
was inaccurate, some because the Jesuits, whom Sixtus
had offended, desired it. In the same year the Clementine
Vulgate was published, but, apparently to avoid the ap
pearance of dissension, under the name of Sixtus ! The
Bull which accompanied it, and which has never been
repealed, makes it the standard Roman text; no word of
it may be altered, nor may variants be printed in the
1 The best ms. of the Vulgate is Cod. Amiatinus, now at Flor
ence. A list of a few of the more important is given on p. 98.

32 THE TEXT OF THE NEW TESTAMENT
margin, so that officially, at all events, textual criticism
in the Roman Church ended in 1592. Modern editions
of the Vulgate are being made in England by the
Bishop of Salisbury (Dr. Wordsworth) and the Rev. H.
J. White, and in Germany by Dr. P. Gorssen.
The relations of the Latin versions to each might be
graphically represented thus : —

Greek Text
(perhaps Neutral
Type).

Greek
(probably Western
Type).

African Latin
(k, e, Cyprian, etc.).

The Vulgate of S. Jerome
(am. fuld., etc.).
¦ It is also possible that the European Latin ought to be con
nected directly with the African.
2 The influence of this source ought perhaps to be confined to
the Italio Latin.

THE VERSIONS 33
Syriac Versions. — As is the case in the Latin version,
so with the Syriac : the main body of the evidence repre
sents a revised recension. This is admitted on all hands,
though some consider that the amount of revision implied
is considerable, others that it is quite insignificant.
This revised text is known as the Peshitto (or simple)
version. It is, like the Latin Vulgate in the case of the
Latin version, the starting-point for criticising the Syriac
version. But there is this difference: we know all about
the making of the Latin Vulgate from accurate and
contemporary information. We know who made it, and
to a large extent why, when, and how he made it.
None of this can be said of the Syriac. Here we are
obliged to fall back entirely on the mss. and the attempt
to- make them tell their own story.
The Old Syriac. — As far back as the beginning of the
last century Griesbach and Hug perceived that the
Peshitto was, in the form in which we possess it, a revised
version, and postulated an Old Syriac which lies behind
it, though they could not point to any ms. evidence for
it. But since their time three important discoveries and
much able work have thrown light on the matter. These
discoveries are : —
(1) The discovery in 1847 and the publication in 1858,
by the late Dr. Cureton, of some fragments of a fifth
century Syriac ms. of the Gospels, brought in 1842
from the monastery of S. Maria Deipara, in the Nitrian
desert by Archdeacon Tattam.
This ms. is now B. M. addit. 14,451, and is known as
the Curetonian Syriac. It contains large fragments of all
the Gospels, and at the beginning of S. Matthew has the
curious and important title, ' Evangelion da-Mepharreshe
Mattai.' The meaning of ' da-Mepharreshe ' is not quite
certain, but there seems to be an almost universal consent
among Syriac scholars that in all probability it means
'separate' in the sense of not harmonised into a con
tinuous narrative with the other Gospels on the plan of
T?tian s Diatessaron (cf. Payne Smith, Thesaurus Syriacus).
(2) A series of partial discoveries connected with
Tatian's Diatessaron. This work, as its name implies,
was a continuous harmony of the lour Gospels which
o

34 THE TEXT OF THE NEW TESTAMENT
Tatian put together in the second century, as Eusebius
tells us (iv. 29). It had been supposed that this work
was lost, but in 1836 the Armenians of the Mechitarist
monastery of San Lazaro at Venice published a copy of
i commentary on the Diatessaron by Ephraem, a Syriac
leather of the fourth century, which they possessed in an
\rmenian translation. Some years later Moesinger
translated this, and Zahn and other scholars have re
constructed the text on which the commentary is based.
The interest aroused by this discovery drew attention to
two other sources of information on Tatian's work :
(a) an Arabic translation of the Harmony; (3) a Latin
translation ; both of these seem to preserve the order of
the sections in Tatian though both have been corrected
in text by an ordinary copy in their respective languages.
The former is published with a Latin translation by
Ciasca, and the latter is known as Codex Fuldensis of
the Vulgate (cf. p. 98).
(3) In 1892 Mrs. Lewis and Mrs. Gibson found some
palimpsest leaves of a Syriac ms. of the Gospels of the
fifth century in the monastery of S. Catherine on Mount
Sinai, where Tischendorf had found his Codex Sinaiticus
(K). These contain in a legible form about three-fourths
of the Gospels. The ms. itself is still at Sinai, but it has
been photographed. It is usually known as the Sinaitic '
Syriac, or the Lewis 2 Syriac.
These three discoveries give us all the ms. evidence
which we possess for the Old Syriac.
Problems of the Old Syriac Version: —
(1) What are the mutual relations of the Curetonian,
the Sinaitic, and Tatian ?
(2) What are the relations between this group and
the Peshitto ?
(3) How do we explain the rareness of this type of Syriac
version, as compared with the numbers of Peshitto mss. ?
(4) What type of Greek text does the Old Syriac imply?
_ It seems to be certain that the Curetonian and the
Sinaitic ought to be grouped together against Tatian,
i
1 This is usually represented by Syr"In, SyrS.
* Sometimes represented by Syr1.

THE VERSIONS 35
and that they may be regarded as two mss. of an original
Old Syriac version, the Sinaitic being greatly the better
of the two. The question is does this version represent
an attempt to keep the text of Tatian while abandoning
his arrangement, or is Tatian a harmony based on a
previously existing Syriac text, or, as a third alternative,
do Tatian and the Old Syriac represent independent
attempts to render the Greek? Various scholars have
supported the two first suggestions, but Professor
Burkitt's recent Evangelion Da-Mepharreshe (or the
' separated ' gospels) makes it appear probable that the
third view is preferable, for he thinks that the Diatessaron
represents the type of Greek in use in Rome in the
middle of the second century, and that it was made
before the Old Syriac, which represents a translation
based by some one who knew the Diatessaron on the
Greek text in use at Antioch about the second half of
the second century.
Two other points bearing on the relation of the
Sinaitic to the Curetonian are worth notice.
(a) It has been suggested that the Sinaitic shows traces
of a text which has been altered to support the view of
the Ebionites, who held that our Lord was the human son
of a human father, and that the Divine Spirit entered him
at the Baptism. This is supported by the reading in
Matt. i. 16, where the Sinaitic reads — Joseph, to whom
was betrothed Mary the Virgin, begat Jesus, who is called
Christ. Whether this variant is really due to this cause
is very doubtful, and this point is not one which com
mands confidence.
(/3) Far more probable seems a suggestion that the
Curetonian has a later form of the Old Syriac than the
Sinaitic, and has been contaminated by a Latin source.
The evidence for this statement is that, as compared with
the text of the great Greek Uncials (known as the Neutral,
cf. p. 65), both the Latin and Old Syriac have many
curious, interpolations (cf. pp. 79, 80), and of these some
have both Latin and Syriac attestation. But there are
some which are peculiar to the Syriac, some to the Latin.
Now it would seem that in many places the Curetonian

36 THE TEXT OF THE NEW TESTAMENT
agrees with the Latin against the Sinaitic ; and as there
is no evidence to show that the Sinaitic was connected
with the SB type of text with which it, of course, agrees
in these places, the probable suggestion is that the Cure
tonian is indebted for these readings to a Latin source.
An important corollary follows from this point. The
text of the Old Syriac, as will be seen later on, is closely
connected with the Old Latin, but it would seem from
the above-mentioned point that this connection is not
explainable merely by a common origin in Greek for the
versions, but by some kind of contact at more than one
time. In other words, there are various strata in the text
of the versions which are common to both, not the bed
rock alone. So that in reconstructing the textual
history of the versions, we must adopt some theory which
will allow for some period of time in the earliest stagei
of their history in which the two, as it were, lived side by
side. The importance of this will be seen in Chap. VI.
(2) What are the relations of the Old Syriac group to the
Peshitto? That is to say, does the Sinaitic-Curetonian
Tatian group represent an old text which was revised
into the Peshitto, or is it essentially a different text?
A long controversy has been waged on this point, and
cannot be said to be finished. At one time a few
scholars were inclined to maintain that the Old Syriac,
as we have called it, is really a corruption of the Peshitto.
But probably no one would now maintain that position ;
certainly the great majority of Syriac scholars have de
cided against it. But it is a more delicate question
whether the original of the Peshitto is to be sought in the
Old Syriac. The choice of theories may be genealogically
shown thus : —
(1) Old Syriac (Syr8ta Cur. Tat.)
Peshitto,
which is the more generally accepted view.
or (2) Original Syriac

Peshitto
Old Syriac

THE VERSIONS 37
The latter theory claims that the Peshitto of to-day is
nearer the original Syriac than is the Old Syriac. But
the possibility of holding this view has been greatly
reduced by Mr. F. C. Burkitt's S. Ephraim's Quotations
from the Gospel; Texts and Studies, vii. 2. Mr. Burkitt
shows that there is no reason for thinking that Ephraem
used the Peshitto, and that it is therefore very probable
that the Peshitto was made by Rabbula, who was appointed
bishop of Edessa in 411. It has hitherto been believed
that Ephraem' s knowledge of the Peshitto proved that
Rabbula's translation was some other quite unknown
version. Mr. Burkitt has shown that this belief is
erroneous, and, as he points out, we are now free to
believe that the Peshitto was made at the beginning of
the fifth century by Rabbula, bishop of Edessa, — a date
and historical setting which suit it best.
(3) The question thus arises, How is it that the MSS. of
the Old Syriac are so few in number, compared with those
of the Peshitto, which, as will be seen, are very numerous ?
Those who deny that the Peshitto is the result of a revi
sion of the Old Syriac, explain that this is due to the fact
that the Old Syriac is a merely accidental survival of a
curiously corrupt text, which never had a large circula
tion. This is not supported by the quotations in Aphraates
and Ephraem, and therefore probably a better theory is
that which postulates a revision of the Syriac text in "the
fifth century, which corrected the Old Syriac into the
Peshitto (as S. Jerome corrected the Old Latin into the
Vulgate), and insisted on the abandonment of the Old
Syriac. It is said that there is no external evidence for
this theory. In view of Mr. Burkitt's book this is
scarcely true. But even if it were quite true it would
not be necessarily fatal ; early Syriac literature is
extremely scarce. And also it must not be forgotten
that there is another small piece of evidence which sup
ports the idea of a revision. Theodoret, who wrote in
the middle of the fifth century, mentions that he himself
found over two hundred copies of Tatian in use in his
district, and replaced them by the four Gospels. This
vigorous line of action will easily explain the rapid
extinction of Tatian's Diatessaron, and it is at least pro
bable that the copies supplied would be made as accurate
as could be, which implies something very like a revision

38 THE TEXT OF THE NEW TESTAMENT
(4) What type of Greek text does the Old Syriac repre
sent? This question is complicated both by the differ
ences between the Curetonian and the Sinaitic, and by
the difficulty of discovering their relations to the
Diatessaron, the influence of which is especially important
if, as Professor Burkitt suggests, it represents a Greek
text from another locality and with other characteristics
than that which lies behind the Old Syriac. On the
whole, however, taking the Sinaitic as the best guide
available the answers to the question raised would be :
(1) The text implied is a ' Western ' one (cf. p. 65). (2) It
is a short text. The first point is shown by the additions
which it makes to the ordinary Greek text, as represented
by W. H., some of which, e.^r. ' quia Deus est Spiritus ' in
Jo. iii. 6, are common to it and the Old Latin, and belong
to that stratum of the Western text which is common to
Latin and Syriac, and is probably the earliest corruption
of the true text. Some, not differing in internal char
acteristics, are peculiar to the Syriac, e.g. in the statement
that Barabbas' name was Jesus.
The second point is shown by a series of striking
omissions, some of them of considerable length, e.g. the
story of Herod and Pilate's reconciliation in Luke xxiii.
12; some of them — notably perhaps, a series in S. Mark
— omissions of redundant phrases, e.g. Mark i. 32,
where Syi-Bin omits 6-tyias yevop.evns. Omissions, of a
character which can scarcely be called dissimilar, are
found in the Old Latin, but not often in the same
passages as the Syriac omissions; and W.H., writing
before the discovery of SyrBin, accepted most of the
Latin omissions as representing the true text. The
question therefore arises, Are we to say the same of the
Syriac omissions, and greatly increase the number of
passages suspected of being Western non-interpolations,
or, in other words, non- Western interpolations? This
question is not yet settled ; indeed, it has hardly been
discussed. It will, however, be hard to resist the Syriac
omissions, unless it be possible to show adequate reason
for thinking that the Sinaitic text represents a deliber
ately shortened recension. In the face of the Syriac
additions, this will be hard to do. It is also worth while
to remember, in the case especially of the shortening of
readings in S. Mark such as the omission of o^iat ytvo-
Itivntf mentioned above, that the bearing on the Synoptic

THE VERSIONS 89
question must be considered. Higher critics, assuming
as they do that the second Gospel, or something very like
it, was used by the compilers of" the first and third, always
find a difficulty in those passages which, like that quoted,
seem to be conflations from S. Matthew and S. Luke.
The Peshitto. — We now return to the Peshitto. There
are many extant mss., ranging in date from the fifth
century onwards. A critical edition has recently been
published at the Clarendon Press by Mr. Gwilliam. It
is said that some of the later mss. show signs of readings
not found in the earlier copies, but existing in the Old
Syriac. This would be expected on the analogy of the
Latin version, where it is known that the later mss. of the
Vulga*te show a text contaminated by the survival of Old
Latin readings in the popular use. The date of the
Peshitto is less doubtful now than it was formerly. It
used to be alleged that the absence of the disputed
Catholic epistles (the avriheyopeva) points to a date not
later than the third century, when these writings were
received. This certainly shows that the original Syriac
version was in existence at that time, and the evidence
of Tatian suggests that it was in existence in the second
century, but it does not show that the Peshitto in its
present form was in existence at the time. Rabbula's
revision of the Syriac text may have affected only the
Gospels. The analogy of the Latin, where the Vulgate
text of the Catholic epistles seems to have undergone but
little revision, suggests this. Therefore all that can be
said is that a Syriac version existed, probably in the form
which we have called Old Syriac, in the second century ; and
that the Peshitto, as we have it, was based on a revision
which was probably made by Rabbula in the fifth century.
We do not know yet whether it affected all the New
Testament. That it affected the Gospels is certain. That
it affected the Acts is probable (cf. Dr. Chase, Syro- Latin
Text). That it affected the epistles is quite doubtful.
Whatever may prove to be the case with this supposed
revision, there certainly were two later revisions of the
Peshitto itself. These were : —
1. The Philoxenian. — This was made for Philoxenus, or
Xenaias, the bishop of Hierapolis, in 508, by Polycarp,
Rural-Bishop. There are apparently no mss, extant of
this revision except in the Apocalypse.

40 THE TEXT OF THE NEW TESTAMENT
2. The Harklean. — This is really not so much a revision
of the Peshitto, as of the Philoxenian. It was made in
616 by Thomas of Harkel, who was afterwards bishop of
Hierapolis. It is literal to the point of obscurity, and its
text as a whole is that of the later Greek mss., but it has
many striking readings, usually in the margin, and an
elaborate system of asterisks and obeli, imitated probably
from Origen's use of such signs in the Hexaplar Septua-
gint. It is an interesting problem whence Thomas of
Harkel obtained these readings. Practically- all that we
know is what he says himself in his subscription, which
is also the chief authority for the facts given above con
cerning the Philoxenian. He says : ' This book of the
four holy Gospels was translated ... in the cfty of
Malry (Hierapolis) in the year of Alexander of Macedon
819 (a.d. 508). . . . Afterwards it was collated with much
diligence by me, the poor Thomas, by the help of two or
three accurate Greek manuscripts in Anton ia of the great
city of Alexandria in the monastery of the Antonians
(probably the monastery of the Enaton, i.e. of the ninth
milestone), it was again written out and collated in the
aforesaid place in the year of the same Alexander 927
(a.d. 616),' etc. The results of these labours are the
marginal readings mentioned above, and it seems pro
bable that they represent a Greek ms. of a markedly
Western type and early character. So that the margin
of the Harklean is more valuable than its text. What
exactly is the meaning of the signs used is unknown.
It is suggested that an asterisk implies that the reading
is an addition to the text, and that the obelus means that
the words marked ought to be taken out. It must be
remembered that this version is often called ' Syra pos
terior,' and so Tischendorf always quotes it as Syri>,
naming the Peshitto Syraoh from Schaaff, one of the
earliest editors of the Peshitto.
So far, all the Syriac versions have in all probability
been connected genealogically. It remains to consider
two Syriac versions, which probably stand apart
(1) The Jerusalem1 or Palestinian.— This is a Lectionary
(cf. p. 51) made, according to a subscription of a monk
i It has been shown by Mr. Burkitt, in the Jowrnal of Theo.
logical.Studies, ii. 6. p. 174, to be connected with Antioeh rather
than Jerusalem

THE VERSIONS

41

named Elias of Abud, in tho monastery of the Amba
Musa at Antioch in the year a.d. 1029. It is written in
a peculiar hand and a peculiar dialect, and its text is a
curiously mixed type. It often agrees with B against
the mass of mss.
It is remarkable as the only Syriac authority for the
pencope adulterae, with th exception of one ms. of the
Harklean version. It is not mentioned by any Syriac
writer, and we do not know when it was used. There
are only three mss. of it known at present.
(2) The Karkaphensian. — This used to be called a version,
but it is more probably a kind of Massorah, an attempt
to preserve the best traditions of the Syriac text by a
catalogue of readings, etc. There are seven known uss.
— six Jacobite, and one Nestorian. It is called Karka
phensian, which is an adjective of the Syriac word for
skull, apparently because it tried to perpetuate the text
favoured in the ' Monastery of the Skull. ' It does not
seem to have any great value, and is hardly ever quoted.
The relations of the Syriac versions, so far as we can tell,
to each and to the Greek, may be graphically represented
thus : — Greek from Antioch. Greek from Rome.
(?) Special Latin
influence.

Greek (probably later Anti-
ochene type.

A separate group altogether
would be : — Greek.
The Jerusalem Syriac

Harklean, text and margin.

42 THE TEXT OF THE NEW TESTAMENT
The Egyptian Versions. — Very much less is known at
present about the Egyptian versions than either the Latin
or Syriac. They are found in various forms of that
debased type of the ancient Egyptian language, written
in characters borrowed from Greek, which was current in
the earliest centuries of the Christian era, and is usually
known as Coptic.1 The questions which are raised are
very complicated, and perhaps they may be best divided
into those which are concerned with the history of the
dialects, and those which are concerned with the history
and character of the versions in those dialects.
The Egyptian Dialects. — Apart from the evidence of mss.,
the only testimony which seems important is that of Atha
nasius of Kos in the Thebaid (eleventh century), who
published an Arabic-Coptic grammar. He says that there
were three dialects : (1) Sahidic ; (2) Bohairic ; (3) Bash-
muric, representing three districts in Egypt. Of these
we can identify the Sahidic as the southern, and Bohairic
as the northern dialect, and we have a fair amount of
mss. in each. Bashmuric is more difficult to identify.
It was at one time thought to belong to a district near
the Delta, but it is now generally believed that what
Athanasius was referring to is the form now known as
Middle Egyptian, which is a form intermediate to the
Bohairic and Sahidic, but nearer to the latter than the
former. It lacks, however, the definitiveness of the two
other versions, and it is still an unsettled point among
Egyptologists whether it is not possible to divide Middle
Egyptian into more than one dialect. Notably there are
fragments which have been found at Akhmim and the
Fayoum, to which some have been inclined to give the
names of the Akhmimic and Fayoumic dialects ; and
Mr. Headlam, in the fourth edition of Scrivener, seems
to think it possible that Akhmimic is the parent of the
Sahidic. The dates of these dialects are doubtful. The Arabic
historian Macrizi (fifteenth century) says Sahidic is the
oldest, and Bohairic and Middle Egyptian offshoots of it.
1 'Coptic' is a corruption of Alytiirrtos, just as in modern.
military slang 'Gyppy' is a corruption of 'Egyptian.'

THE VERSIONS 43
Bohairic is a more literary language, and is said to show
more signs of Greek influence in grammar and style.
Geographically it is more probable that it developed
out of Middle Egyptian than out of Sahidic. It may be
doubted whether it was not always far more a literary
than a popular language. It died out earlier than the
other dialects, except for ecclesiastical purposes.
We have fragments of all the dialects in mss. as early
as the fourth or fifth centuries ; but it is considered
probable that as far back as the second century attempts
were made to use Greek characters such as were used in
the later developments of Coptic.
The Versions.— We have remains of versions in each
of the three great dialects, but the Bohairic alone is
complete. The Sahidic. — The text of the version in this dialect1 has
not yet been worked out, or properly edited. A great
deal was published by M. Amelineau, in 1886 and the
following years, in the Zeitschrift fur Aegyptische Sprache.
It is remarkable for a number of 'Western' readings;
e.g. in Luke xxiii. 53 it preserves the curiously Homeric
addition that the stone laid at the mouth of the sepulchre
was such that twenty men could not move it (cf. p. 74 f.).
The Middle Egyptian. — Even less is known of this than
of the Sahidic. It is not yet accessible to any except the
Egyptian specialist. It would seem that there is con
siderable doubt, as was suggested above, whether there
is not more than one dialect preserved in the fragments
which are usually classed together as Middle Egyptian.
Nor is it certain whether it represents a separate version,
or is merely another form of the Sahidic.
The Bohairic. — Far more is known of this version,2 which
is now easily accessible in Lagarde's edition and in the
Clarendon Press edition of the Gospels, by Mr. Horner.
Probably it is a translation separate from those pre
served in the other dialects. The text is Neutral and
Alexandrian rather than Western.
The broad questions which remain as to the Egyptian
versions are naturally these : —
1 Also called the Thebaic. ,„,.,,,„
s It is sometimes called the Memphitic, and Tischendorf calls
it the Coptic vorsion.

44 THE TEXT OF THE NEW TESTAMENT
(1) What is the date of these versions ?
(2) What was the underlying Greek Text ?
As to (1), the earliest evidence which we possess is
contained in the life of S. Antony, attributed to S.
Athanasius. S. Antony lived in the latter half of the
third and first half of the fourth century, and appears to
have been in the habit, when a young man (c. 270), of
hearing the Gospels read in Coptic. This would be
conclusive evidence if it were not for the doubt whether
what he heard was a written or an oral translation.
But from the emphasis which Pachomius in the fourth
century lays on the study of Scripture, in his monastic
rules, it seems certain that there were written translations
in his days, so that on the whole the end of the thirds
century is the most probable date for the beginning of •
the native versions. Which version is the earliest we
cannot say, but Mr. Forbes Robinson in Hastings' Bible
Dictionary, vol. i. p. 668, thinks that the Sahidic _is the
oldest, and the Bohairic the most recent. His reasons
seem to be that (1) the Sahidic dialect, as a dialect, is.
probably the oldest ; (2) on a priori grounds the needs
of that part of the country which knew least Greek
might be expected first to produce a native translation,
and this would suggest a district further from rather than
nearer to Alexandria. Other scholars, e.g. Mr. Headlam,
regard the Bohairic as the earliest, but their chief reason
seems rather too largely to consist in the character of
the text, which is a very dangerous argument.
. As to (2), the Greek Text underlying the Sahidic
version is, beyond doubt, an early Western one, and the
Middle Egyptian, seems to be of the same character. . But
the Bohairic represents a Neutral and Alexandrian text.
It is the only version of firstrate importance in any
language of which this can be said. Therefore one of
the most interesting problems of the future will be to
explain this difference between the Sahidic and Bohairic.
Possibilities are : —
(a) The Western element in the Sahidic is secondary
and not primitive. This is most improbable.
(/3) The Bohairic Neutral element is to be traced to
the conviction which obtained in the neighbourhood of
Alexandria that the Neutral text was the better. It is

THE VERSIONS 45
possible that the working out of this theory may throwliglu
on the question raised in Chapter V. as to the possibility
that the Neutral text is a recension made by the school
of Origen. This could only be done by a careful com
parison of the Bohairic with the quotations of Clement of
Alexandria, Origen, Cyril of Alexandria, etc.
The Secondary Versions. — The versions of secondary
importance are those which have been made either from
one of the great versions or from a late Greek text. The
most important are : — -
(1) The Armenian version. — The early history of this
version is lost in obscurity. According to Mr. Cony-
beare's account in the fourth edition of Scrivener, there
were two translations made in the fourth century: (a) By
S. Mesrop, with the aid of a certain Hrofanos (? Rufinus),
based on a Greek text ; (/3) by S. Sahak, from Syriac.
After the Council of Ephesus in 430, Sahak and Mesrop
compared and revised their translation with Greek
manuscripts from Constantinople, and it is this recension
which is represented by the Armenian mss. which we
possess. Therefore we should naturally expect to find in the
Armenian text three distinct elements : (1) An Old
Syriac element ; (2) an earlier Greek text ; (3) a later
Greek text, and we should expect the last to be the most
prominent. Roughly, this is what we do find. The
bulk of the readings attested by the version are of the
' Syrian ' type, but there are also two distinct and earlier
strata :
(a) A series of readings, agreeing with the Old Syriac
text, e.g. in the omission or practical omission of the
spurious conclusion of S. Mark.
(b) A curious group of readings in the Pauline epistles
especially agreeing with NCH3 and Euthalius, and there
fore (cf. p. 66) probably representing a Caesarean Greek
text of an early type.
The special importance of these two groups of readings
is that the former sometimes gives us, in the Acts and
Epistles, an opportunity of reconstructing the Old
Syriac text which is extant in mss. in the Gospels only,
and that the latter is possibly a guide towards the re-

46 THE TEXT OF THE NEW TESTAMENT
construction of the ' Codex Pamphili ' which is mentioned
on p. 14.
There are interesting possibilities suggested by the
Armenian version of Mark xvi. 9-20, where one ms. con
nects the verses with Ariston Eritzu,1 or Presbyter, who
may perhaps be tne Aristion mentioned by Papias.
(2£ The Aethiopic Version. — There are two recensions of
this version made apparently in the fifth and twelfth
centuries. The latter naturally represents a late text,
but the earlier has many early readings. It frequently
agrees with the Egyptian version, though it is said to
have signs of being in the main an independent transla
tion. A collection of readings found in Aethiopic mss.
is given in Dr. Sanday's ' Appendices ' ; but very few
people know anything about it, and what they do know
does not seem to be very important for the purposes of
textual criticism.
(3) The Gothic Version. — The story of the Gothic migra
tion from Scandinavia to Italy is to be found in Gibbon
(D. and F. ch. x.-xxvi. etc.). While they were still in
Moesia (318-388), Ulfilas, or Wulfilas, a Cappadocian,
who was their second bishop — an Arian — translated the
whole Bible into Gothic. It is almost certain that he
used the lxx. in the Old Testament, and the Greek in the
New Testament; but there are considerable signs of
the influence of Latin readings on the text. The Arian
heresy has left few or no traces in it.
There are three mss. of the version, of which the most
important is the Codex Argenteus at Upsala, of the fifth
or sixth century, of purple vellum with gold and silver
letters. The Gospels are in the 'Western' order
(Matthew, John, Luke, Mark), which may be due to Latin
influence or the use of Greek mss. of a ' Western ' type.
"Other versions which are sometimes quoted are the
Arabic, the Persian, the Sclavonic, the Georgian, the
Anglo-Saxon, and a few others. None of them have any
claim to importance. The Georgian is probably the most
important of them, as it is possible that, like the Armenian,
it may contain traces of an Old Syriac base.
1 Cf. The Gospel according to S. Mark, ed. Swcte. p. oiil.

CHAPTER IV
PATRISTIC QUOTATIONS — LITURGICAL EVIDENCE —
CHAPTER DIVISIONS AND STICHOMETRY
So far the only evidence dealt with has been that con
tained in mss. of different languages, which give a con
tinuous text of the various books of the New Testament,
and claim to provide direct evidence as to the words
used by the writers. The evidence which remains is
indirect. It is that which is provided by : —
(1) Quotations of varying length and importance in
the early Fathers.
(2) Ecclesiastical or liturgical usage and service-books.
(3) Chapter divisions and stichometry.
I. Patristic Quotations. — This is a branch of textual
criticism which is full of difficulties. There are three
points which have always to be discounted in consider
ing the bearing which a quotation in a Father may
possess : —
(1) The text of the Father is itself a subject for criti
cism. We cannot always be sure that we have the text
in a sound condition. And especially is this the case
with quotations. Scribes could scarcely be expected to
refrain from correcting the text into agreement with
their own ideas of what the true text ought to be.
Therefore as a practical rule we can attach less weight
to the quotation of a Father if it agrees with the ordinary
text than if it differs.
(2) Quotations from memory are obviously less to be
trusted than those copied from a book. And the great
majority of quotations are undoubtedly made without

48 THE TEXT OF THE NEW TESTAMENT
consulting any authority. It is only in the case of long
quotations which cannot be made from memory, or
where some definite reading is emphasised, that we can
feel able to claim that the quotation really represents
the ms. of the Scriptures.
(3) It is always necessary to ask whether we can be
quite certain that the quotation is not made from a
harmony or lectionary.
But after making every allowance for these factors,
the value of Patristic quotations remains almost as great
as that of mss. or versions, though of a different kind,
and greater than that of lectionaries or divisions of the
text. Their value consists in the opportunity which they
afford us of localising and dating various kinds of texts
in mss. and versions. For instance, if we find a certain
well-defined type of text in the Old Latin mss., and also
in the quotations of certain African Fathers of the second
and third centuries, we are obviously justified in saying
that this form of Latin version was used in Africa in the
second and third centuries. Whereas if we had not the
quotations, we should have very little certain evidence
either as to date or place.
There are of course more Fathers than one can count,
but fortunately the number can be considerably reduced
for the purposes of criticism. With few exceptions, all
the Fathers later than the fifth century who write in
Greek seem to use a late or 'Syrian' text. They there
fore merely confirm the well-known fact that by the
fifth century this type of text was to be met with almost
everywhere. Similarly the Latin Fathers almost invari
ably use the Vulgate, and so add nothing but confirma
tion to our previous knowledge of the wide use of that
version. The Fathers therefore who are important are
those who are earlier than the fifth century.
Taking, then, this selection of Fathers, we find that for
practical purposes we can divide them into a few more or
less easily recognisable groups : —
1. A group of Latin Fathers, of whom the earliest and
most important are Tertullian and Cyprian representing
Africa, and Cyprian's contemporary Novatian of Rome.

PATRISTIC QUOTATIONS 49
These three give us in their numerous quotations a
fairly complete record of the third century Latin text,
and Tertullian takes us to the last days of the second
century. The character of their quotations is Western.
2. A small group of Western Greek writers, contain
ing Justin Martyr and Marcion of the second century,
Irenaeus of the second and early third, and Hippolytus
of the third.
Of these it is also true that they present a definitely
Western type of text, with many agreements though
with some differences from the Greek text implied by
the Latin group. It is noticeable that although their
final home was the West, yet Justin Martyr (Samaria
and Rome), Marcion (Pontus and Rome), and Irenaeus
(Ephesus, Rome, and Lyons) can all be traced back to
the East.
3. A group ot Eastern Greek Fathers, who are chiefly
represented by Methodius of Lycia and Tyre, and Euse
bius of Caesarea. This group probably represents the
tradition of Pamphilus, an earlier Father of the same
locality. These, again, have a Western text, though
of a somewhat later type. In this group it is to be
noticed that Methodius and Eusebius, though using
much the same text, did not belong to the same school
of thought. Methodius was an opponent of Origen,
Eusebius a disciple.
4. A group of Syrian Fathers, especially Tatian (so
far as we can judge from his Diatessaron), Aphraates, and
Ephraem. These also have many Western readings ; but it is
still a doubtful point how far they represent a text
independent of the Diatessaron.
The Diatessaron of Tatian has been discussed on p. 33.
It is therefore only necessary to repeat here that the pro
blem of its relations to the old Syriac text as represented
by the Sinaitic and Curetonian mss. and the quotations
of Ephraem and Aphraates is still unsolved.
5. A group of Alexandrian Fathers extending over
several generations, represented primarily by Clement of
Alexandria, Origen, and Cyril. The text represented
by these writers is a most difficult problem.
D

50 THE TEXT OF THE NEW TESTAMENT
The main points are these : — Clement gives a text
which, though perhaps more Neutral in character than
that of the Latin or Syrian Fathers, yet has a great
number of Western readings. Origen in many ot his
works gives us an almost purely Neutral text, with some
traces of the Alexandrian ; in others he has a Western
element, which may perhaps be due to his connection
with Caesarea, or may be due to the same source as that
which affected Clement.
Cyril has an almost purely Alexandrian text.
It must be remembered that Origen is known to have
been a critic and editor of the Old Testament, and may
have extended his labours to the New Testament, though
there is no proof of this.
These groups of Fathers, of which only a few repre
sentative names have been given, are the foundations of
patristic evidence. The main facts which are clear from
them are : —
(1) The Syrian text as such was unknown in the third
century.
(2) The Western text in some form or another was
predominant everywhere except in Alexandria.
(3) Even in Alexandria the earliest known text was
probably more Western than Neutral, though this
point is still somewhat disputed.
The bearing of this evidence will be seen in Chapter VI.
Besides these there are many other Fathers whose
readings are interesting and valuable, e.g. Primasius, who
quotes nearly the whole Apocalypse in an 'African'
text; Augustine, who uses the Vulgate in the Gospels,
but an African text in the Acts ; Gregory Nazianzen,
Theodore of Mopsuestia, Chrysostom, and many others.
It would, however, be neither easy nor profitable to say
anything of them separately, and to some extent the
value of their writings for purposes of textual criticism
still remains for future critics to determine. l
II. Liturgical Evidence. — This is chiefly to be found in
those service-books which contain texts of the New
Testament complete, but not continuous, being really
sections adapted for the use of the Church for public
i New critical editions are also needed.

LITURGICAL EVIDENCE 61
reading. These may be conveniently termed ' Lection-
aries,' though they are strictly known by various names,
according to the name of the New Testament from which
they have been compiled, and services at which they were
intended to be used. A collection of passages to be used
as the Gospels at the Eucharist was known as an Evangeli-
arium, selections from the Acts as an Apostolos, and so
on. But the use and method of compilation in each case
was the same. In the Evangeliarium, for example, the
selection was based on one of the four Gospels, and in
the main the text chosen was adhered to faithfully. But
there were two exceptions to this faithfulness : —
(1) At the beginning and end of a passage a few words
would often be altered in order to make a more intelligible
commencement or conclusion, e.g. avrbs might be altered
into the name of the person to whom it referred, pcra
Taiira might be replaced by some two or three words
summarising the events referred, or the whole might be
prefaced by the words iv i<eivio tw Kaipa, or threv 6 mipios,
or some such phrase.
(2) If the parallel passages in the other Gospels sup
plied some interesting detail which was not found in the
Gospel which was being used, the requisite verse would
often be inserted, e.g. the story of the angel in the
Garden of Gethsemane was inserted in this way into the
narrative of S. Matthew.
It is this general faithfulness to an originally continu
ous text, taken in connection with two exceptions on the
one hand, and with the well-known verbal conservatism
of Church services on the other, that gives to the evidence
of lectionaries both its value and its limitations.
Most lectionary systems go back to a great antiquity,
though very little is known at present as to their origin or
history ; and as they have not changed, in all probability,
with any rapidity, they are therefore sometimes valu
able evidence. But it will be seen at once that not only
is their testimony almost valueless on small points of
wording, but it also carries no weight when narratives
are in question which have parallels. Indeed, many
harmonistic readings in continuous texts are probably
due to the influence of lections, e.g. there is a group of

52 THE TEXT OF THE NEW TESTAMENT
minuscules (including the Ferrars^ which does actually
read the story of the angel at the Garden of Gethsemane
in Matthew instead of in Luke.
The evidence to be gained from this source has not
been fully worked out; it supports, on. the whole, the
view which is suggested by the other evidence, that be
hind the ecclesiastical text (Syrian) of the times sub
sequent to Chrysostom there is a period in which the
Western text was almost, if not quite, supreme.
Traces of the Western text are found in the Luxeuil
lections and in the Liber Comicus of the Acts, and it
is possible that a search among the oldest liturgies of
other languages would reveal the same facts.
III. Division into Chapters and Stichometry. — Most of
this kind of evidence is valuable only as giving a clue
to the genealogical connections of mss., for there is of
course a probability that mss. with the same chapter divi
sions come from the same part of the world. It is some
times possible to argue as to the original division or
addition of a passage of fair length from the evidence
which the numeration of lines or oti'xoi affords.
The handling of this branch of criticism is all compara
tively modern, and is at present confined to a very small
body of scholars.
It will perhaps be best to divide the subject somewhat
arbitrarily, and to start with the most ancient of the
known chapter or capitulation systems and work down
wards, afterwards returning to the stichometry proper.
Chapter Divisions. — At an early period it became
necessary to divide the books of the New Testament into
chapters, in order to facilitate the convenience of readers
and writers. There are four ancient systems of division
which have been preserved in mss.
(1) The Vatican Sections. — The oldest system which is
known to us is that preserved in Codex Vaticanus (B).
The origin of these is unknown, but it is obvious that
they are made with reference to breaks in the sense.
There are 170 in Matthew, 62 in Mark, 152 in Luke, 60
in John. An interesting point in connection with them
is that the Pauline Epistles are numbered continuously
as if they were regarded as one book. And this fact

CHAPTER DIVISIONS AND STICHOMETRY 53
enables us to say something about the order of the Epistles
in a ms., now lost, from which the capitulation was
copied, for as at present arranged there has obviously
been some dislocation. Sections 1-58 cover regularly
Romans, 1 and 2 Corinthians, and Galatians ; but
Ephesians, instead of beginning with 59, begins with
70, and then there is no further break in order until 93,
which brings us to the end of 2 Thessalonians, after
which follows the Epistle to the Hebrews, beginning with
59 and going on to 64 in ix. 11, after which the ms. is
defective. Obviously the archetype of the capitulation
system contained the Epistle to the Hebrews between
Galatians and Ephesians. It has also been pointed out
that this is the order given in the Sahidic version of the
'Festal letter' of Athanasius in 367 a.d., while the order
found at present in B agrees with that of the other
versions of this letter, placing Hebrews immediately
before the Pastoral Epistles. It has therefore been
argued that we ought to place the home of the Vatican
ms. in Alexandria. Against this must be set the argu
ment on pp. 14-15, which shows good reason for con
necting both it and tf with Caesarea. It is, however,
quite possible that the archetype of both mss. came
from Alexandria to Caesarea — a theory which perhaps
does justice to both arguments.
One other ms., S (Cod. Zacynthius of Luke), has the
same divisions in the Gospels, and it is significant that
both textual and palaeographical considerations suggest
that this is an Alexandrian ms.
Other divisions also found in B are dealt with on p. 57.
(2) K£(|>oXata majora. — Next in antiquity to the Vatican
sections, if not quite as old, come the K£<pa\aia majora,
or breves, or Wtaoi, as they are interchangeably called,
though KecpaXaiov is strictly the chapter itself, and ti'tAos
the technical name for the summary heading describing
its contents, e.g. iced), d of Mark begins Mark i. 23,
and its titKos is ircpt rov haip.oviiop.ivov.
It should be noted that in no case does K«fi. a begin
with the beginning of the book, almost certainly because
the custom of early scribes was to call the first section,
as we should call it, the wpooiptov or preface.
These divisions, of which there are 68 in Matthew

54 THE TEXT OF THE NEW TESTAMENT
48 in Mark, 83 in Luke, and 18 in John, are not
found in KB, but exist in A, so that their use in
the fifth century is quite certain.
(3) The Ammonian Sections and Eusebian Canons. — The
name Ammonian, usually given to these divisions, prob
ably perpetuates a mistaken theory. Ammonius was an
Alexandrian who lived at the beginning of the third
century, and appears to have published a harmony of the
Gospels, as Tatian had done, only instead of making a
new continuous text he adopted some arrangement of
parallel columns.
Eusebius of Caesarea describes this attempt of Am
monius in a letter to Carpianus ; and says that, taking the
hint from this attempt, he had himself worked out a new
plan of divisions for making parallels and references
easy to find.
The basis of this plan, which with various modifications
is found in the majority of mss., and is invaluable to the
collator or examiner of mss., was to divide the Gospels
into sections ; and it has been supposed, probably errone
ously, that Eusebius adopted the divisions used by
Ammonius. Having thus divided up the text, Eusebius next pre
pared the indices or tables (xdvoves) of sections. The
first table contained those sections which related events
common to all four Gospels, and the numbers of the four
sections were noted in the table or 'canon,' and the
number of the canon (i.e. in this case a) was written in
red alongside or underneath the section number in each
Gospel ; so that if any one were reading, for example,
the account of the superscription on the cross in Mark
which is in section o-tS (214), and wished to consult the
parallels, he would look at the margin and see that the
section number was accompanied by the figure a (1) in
red. He would therefore turn to the first table or
canon, and run his finger down the sections of Mark,
until he came to_o-i8, when he would find opposite it Mt.
rXe (335), Lc. tk8 (324), Jo. p0T(197), and on looking up
these sections he would find his parallels. In the same
way the second canon contained narratives found in Mt,

CHAPTER DIVISIONS AND STICHOMETRY 55
Mc, Lc, and so on, until all the combinations were
exhausted. (4) Euthalius. — Who Euthalius was, and exactly what
he did, are questions which are much disputed ; but con
nected with his name there is found in many mss. a
considerable amount of work on the Acts and Epistles
which is obviously intended for the help of students and
others of the index-using and reference-seeking tribe.
A full account of the problem can be found in Dr.
Armitage Robinson's Euthaliana, vol. iii. 3 of the
Cambridge Texts and Studies; and some interesting,
though rather difficult contributions, are made in E.
von Dobschiitz's Euthalius-Studien, published in the
Zeitschrift fur Kirchengeschichte (see footnote, page 58).
The outlines of the difficulty are these.
The account given, apparently by Euthalius himself, in
a short prologue to the Pauline Epistles, is that he first
of all published, as we should say, an edition of the
Pauline Epistles, giving (1) a list or index of the quota
tions from the Old Testament, (2) at the beginning of
each Epistle a list of the chapters into which it is
divided, (3) ttjv to>v dvayvi>o~ttt>v dKpi(3eoTdrrji> ropr/v, the
meaning of which is doubtful, avayvao-tav is, of course,
our 'lessons,' but what 'the most accurate division' means
is obscure.
We get further information in a similar prologue to
the Acts and Catholic Epistles, which appear to have
been published by Euthalius at a later date. Here
Euthalius adds that he had divided the text of the
Pauline Epistles into short sentences so as to assist
intelligent reading. He now promises to treat the Acts
and Catholic Epistles in the same way, and to add the
corresponding list of chapters. So that, putting aside
the question of ori'vot, we find that Euthalius
(1) Divided the text into short sentences.
(2) Gave tables of chapters.
(3) Gave tables of quotations.
Turning, then, to the mss. which profess to give
this Euthalian apparatus, we find that not only do
they give these three lists or tables, but add a great
rjeaj more.

66 THE TEXT OF THE NEW TESTAMENT
The list given in Euthaliana, p. 14, is as follows : —
(1) Mapripiov Tlaiikov rov otootoAov.
(2) avaKf<pa\aia>cris ra>v dvayvaxreav Kai S>v expvo-i Ke-
(pakaiav ko\ paprvplatv Kaff iKaarnv iiriaTokt\v
tov airoaroXov Kai claw e/euo-rn roirav orix^"
rvyxdvei.
ttpaypappa, explaining,
dvaKC (paXaiairis deiwv papTvpi&v — a shortsummary.
(5) irpoypappa, explaining,
(6) dvaKecpdkaiaa-is deiav paprvpimv.
(7) List of Pauline epistles.
(8) ujrocWir Ilpcorr/s 7Tpot 'Papaiovs eVto-roXfjs.
(9) e/ccWir Ke(pa\alav of the Epistle to the Romans.
As the reader of Euthaliana will see, much of this has
no claim to be considered as genuinely the work of
Euthalius. But the points of direct value for an
elementary knowledge of textual criticism are connected
with (1) the chapter divisions; (2) the stichometric
edition of the text. It will be perhaps most convenient
to deal with the second point when discussing sticho
metry generally. The chapter divisions remain. It is
only possible to summarise the points of interest and
difficulty. 1. It was apparently a double system, i.e. each division
was subdivided into smaller sections. It is doubtful how
these subdivisions were indicated, perhaps originally not
by numbers but by asterisks. But it is obviously in
evitable that confusion would soon spring up, and that
at last the double system would become a single one in
which the numeration would run straight on.
2. Without saying at present anything about the date
or locality of Euthalius himself, it is safe to say that the
Euthalian apparatus at an early period in its history was
well known in the great library at Caesarea. This is
proved by certain colophons : —
(a) In H3, saying that the Pauline Epistles were com
pared with a copy in that library.
(6) In some other mss., saying the same of the Acts
and Catholic Epistles,

CHAPTER DIVISIONS AND STICHOMETRY 57
These do not prove that Euthalius went to Caesarea,
but show that his apparatus did.
And it is possible to point to the results of this j ourney.
(a) The third hand of (Kc) which is already connected
with Caesarea, and the second (probably) hand of B,
has adopted a variation of the Euthalian chapters
in the Acts, treating the divisions and subdivisions as
a continuous series. This is the other system in B
referred to on p. 63.
(6) Some editions of the Euthalian apparatus . have
also adopted the earlier numeration of B, and connect
it with the second colophon mentioned above and so
with Caesarea definitely.
Stichometry proper.— It was the custom in ancient
times to measure books by the line, and this line was
usually sixteen syllables or the equivalent of an hexa
meter. This system is also found in the New Testament,
not in the oldest mss., but in many of the earliest of
the later ones. It used to be thought that this system
both in the Gospels and in Acts and Epistles was the
work of Euthalius. But there is no evidence to that
effect, as Euthalius only claims to have arranged the
Acts and Epistles in a-rlxoi ; and as he was by no means
the inventor of the system, there is no reason to assert
that he also applied it elsewhere. In the case of the
Acts and Epistles the measuring line as in classical books
seems to be the sixteen-syllable hexameter, but in the
Gospels it appears probable that the standard was a
fifteen-syllable line.
The value of these reckonings is the evidence that
they bring to bear on the length of the text, and there
fore on the question of long interpolations like the con
clusion of Mark, or the pericope adulterae.
The most ordinary system in the Gospels gives 260C
for Mt., 1600 for Mc, 2800 for La, and 2300 for Jo.;
but these are probably corruptions of 2560, 1616, 2750,
2024 respectively, which are found in several mss., and
imply the presence of xvi. 9-20 in Mark, and the omission
of vii. 63 — viii. 12 in John.
There was, however a constant tendency to substitute
sense-lines for syllable-lines, consulting the convenience

58 THE TEXT OF THE NEW TESTAMENT
of the reader rather than the scribe, and all modern
verse-divisions are based primarily on that plan. There
is also one survival of what seems to be at least
probably an ancient sense-line division in the prjpara
which form a stichometric reckoning found in the Ferrar
group and a few other mss. Their probable history has
been traced by Dr. Rendel Harris in his book on
Stichometry. The question of their origin is obscure,
but it seems certain that they represent a retranslation
of a Syriac stichometry, which is, as said above, probably
founded on a sense-line system in the Gospels, but in
the Epistles (where the prjpara are not found, but the
Syriac system is) may be connected with the Euthalian
stichometry.1 There are no other ancient systems of chapter
divisions and stichometry known to us at present, and it
must suffice to mention in the briefest manner later
developments. The oldest of these developments is probably the Greek
dvayvao-para, which seem to be ecclesiastical in origin,
and no doubt there are local divisions both in Greek and
other mss. which may be ancient.
The system of chapters now in use was invented by
Cardinal Hugo de S. Caro in 1238, and soon became
universally used in the West. He also divided each
chapter into paragraphs by means of letters, but this
part of his work has been superseded since 1551 by
the modern verses which were invented by Robertus
Stephanus. None of them has any critical value.
1 Since the paragraph on Euthalius was written, von Soden has
published in his Die Schriften das Neuen Testaments a long dis
cussion on Euthalius. On the ground of a dogmatic fragment
preserved in Mt. Athos, he thinks that Euthalius lived in the
seventh century. If so, the 'Euthalian' apparatus, which is
undoubtedly earlier, does not belong to this Euthalius, but to
some one else, perhaps to Evagrius, whose name occurs in some
'Euthalian' mss. The whole question is very difficult; besides
the books mentioned in the text, reference should be made to
articles by Th. Zahn, Neuc Kirchl. Zeitschr., xv. 4. 5, F. C.
Conybeare, Zeitschr. f. d. N. T. Wissenschaft, v, and by J. A.
Robinson, Journal of Theological Studies, October 1904, and to
the article on Patristic Commentaries by C. H. Turner in the
supplementary volume of Hastings' Dictionary of the Bible,

CHAPTER V
HISTORY OF MODERN CRITICISM
The history of the criticism of the text of the New
Testament is the history of the attempts, successful and
unsuccessful, to apply the principles and methods ex
plained in Chapter I. to the materials discussed in the
three succeeding chapters. Much indeed of what has
been said in these chapters has only been found out and
added to our store of material owing to these efforts, so
that the collection of material and the criticism of it
always react on each other.
Of course, critics have consciously and unconsciously
differed in their methods. A full consciousness of the
proper method was not possessed by the earliest critics ;
and it is not wonderful if many of their conclusions
prove untenable, when we remember that they had the
triple task of collecting the material, discovering the
method,' and applying it.
The result of the work of the successive generations of
critics is found in the printed editions of the text. In
considering these critics it is best for us not to go back
to a date earlier than the invention of printing ; for
although men like Origen were no doubt in some sense
critical editors, we now regard their work rather as
material for criticism than as anything else.
The first printed text of the Greek Testament dates
from the early part of the sixteenth century. In 1514
Cardinal Ximenes printed a Greek text of the New
Testament, which was followed by an edition of the Old
Testament in three columns, giving the Septuagint, Vul
gate, and Hebrew ; the Vulgate is put in the middle,
because, says the editor, Christ hung on the cross between
two thieves.
It is not known what mss. the Cardinal used for his
6»

60 THE TEXT OF THE NEW TESTAMENT
edition of the New Testament, but it is obvious from
his text that they were not valuable mss., and his
edition has no critical value. It is known as the Com
plutensian, from Complutum, the Latin name of Alcala,
where Cardinal Ximenes founded an university.
The next edition to be printed was that of Erasmus,
undertaken at the request of Froben, the printer of
Basel, and actually published earlier than the Complu
tensian. It is based on the cursives Evan. 2, Act Paul 2,
Act 7, none of which is critically valuable, with occa
sional use of the valuable cursive 1 (cf. p. 20). He also
seems to have occasionally translated back from Latin mss.
of the Vulgate into Greek. This edition was published
in 1516, and was re-edited four times by Erasmus.
These two editions — the Complutensian and Erasmus' —
were the basis of all the early editions. A fuller account
of them and their immediate successors can be found in
Scrivener, ed. 4, but here it is only necessary to notice
a few of the latter, and those cursorily.
1. Robert Stephanus' third edition. — Stephanus was a
Paris printer ; his real name was Estienne, which,
according to the custom of the time, he Latinised into
Stephanus. He published four editions based on the
Complutensian and Erasmus' edition, and the use of
fifteen mss. His third edition was published in 1550, and
has been taken as the standard text in England.
2. Beza's editions (1565-82-88-98 in folio, and 1565-67-
80-91-1604-11 in octavo), which differ but slightly from
each other or from Stephanus' fourth edition.
3. Elzevir's editions, 1624, 1633, which differ but
slightly from each other, were based on a comparison of
-Beza's and Stephanus' editions. The edition of 1633 is
the continental standard, and from the preface to it we
get the name 'Textus Receptus.' 'Textum ergo habes
nunc ab omnibus receptum,' says the editor, cheerfully
assuming the fulfilment of his hopes.
The publication of these editions gave a necessary start
ing-point to criticism, and from their time onwards there is
a steady sequence of attempts to improve the text by the
comparison of mss. and the other sources of information.
Much good work was done in this direction by Bishop

HISTORY OF MODERN CRITICISM 61
Walton of Chester (1667), Archbishop Ussher, who
collated sixteen mss., Courcelles, and Bishop Fell
(1675) ; but the first work of great importance was the
edition of Dr. John Mill, a work which employed him
for more than thirty years. It was published in 1707 in
folio, and its author, felix opportunitate mortis, died
suddenly a fortnight later, before he could be grieved
by the unfair and foolish criticisms of Whitby and
Collins. For the first time critics were given something
like an apparatus criticus, with a fairly full list of
variants, though the actual text was not a new one. Of
coursej the amount of information given is small com
pared with that in modern editions; but with the possible
exception of Tischendorf, probably no one person has
added so much material for the work of criticism.
Bentley, 1662-1742. — The next work of importance was
that of Richard Bentley, the famous Master of Trinity
(Cambridge), who projected (1716) an edition of the
New Testament, a complete collation of all the known
mss., and a text based on the consent of the Greek and
Latin Vulgate, by which he believed it would be possible
to reconstruct the fourth century text. This work was
never completed, partly because Bentley was involved
in college business and other dilatory work, partly
because he found that the task he had set himself was
more complex than he had imagined.
Bengel. — The Abbot of Alpirsbach, 1734, is the next
person of importance. In some ways he is the father of
modern criticism, for he was the first to suggest a
simplification of the mass of mss. evidence by that
classification into families which has been adopted in
some form or another by all modern critics. He adopted
a classification into Asiatic and African, corresponding
more or less to the modern Syrian and Alexandrian
(pp. 65, 66). This important factor in criticism was
developed by Griesbach, between whom and Bengel
there was a line of scholars who added nothing worthy
of memory to the general theory of criticism, but col
lated and examined a great number of mss. Prominent
among these are Wetstein, Alter, Birch, Matthaei.
Griesbach, 1745-1812. — He not only collated a great

62 THE TEXT OF THE NEW TESTAMENT
number of mss., but developed the 'family' theory of
Bengel, sketching out certain groups which, although
enlarged and modified by later researches, have always
been recognised by other critics since his time. His
theory was that there are three groups : — *
(1) The Alexandrian or Origenian, so called because
it is found chiefly in Origen's quotations,' AB C L, and
the Egyptian versions. (2) The Western, so called
because it is found in the Latin Fathers and versions,
and in D. (3) The Byzantine, so called because it is
found in the mass of Greek mss., and may be taken as
representing the text of the Byzantine Empire and the
Patriarchate of Constantinople. This last he considered
less valuable than the other two ; but his principle was
that, unless internal evidence forbade it (and he allowed
great weight to this exception), the reading found in two
groups ought to be preferred to that found in only one.
Contemporary with Griesbach was Hug, a scholar who
was more remarkable for ingenuity than sound judgment.
He paid great attention to the fact that there seem to
have been three recensions of the LXX. — one in Egypt
by Hesychius, another in Antioch by Lucian, and a third
in Caesarea by Eusebius and Pamphilus, based on the
work of Origen ; and he considered that there were
similar recensions of the text of the New Testament in the
same places and by the same people. This was an attractive
theory ; but it has, except perhaps in the case of Lucian,
too little evidence supporting it. A pupil of Hug's is the
next critic who stands out as eminent, J. M. A. Scholz,
whose great claim to fame is the number of mss. which
he examined and partially collated — nearly a thousand in
all. But he was somewhat careless, and his results are
not always verified by subsequent investigators.
Neither Hug nor Scholz is comparable to Griesbach,
but in 1842-50 the published works of Lacfunann gave
a fresh turn to the progress of criticism. Since Gries-
bach's day the amount of material for criticism had
increased very greatly, and Lachmann found it
necessary to simplify the bulk of the apparatus. This
he did by putting aside the mass of late mss. as palp
ably containing a late text, and in so doing he has

HISTORY OF MODERN CRITICISM ^63
been followed by nearly all critics since his time. On
the other hand, he probably went too far in this direction,
and overlooked the certainty that there are early read
ings imbedded in late texts. The discarding of late
evidence was a necessary step towards making a tentative
text, but that text cannot be regarded as final until all
the evidence has been considered. The weak point,
however, of Lachmann was that he did not pay attention
to the groupings of the mss. which he retained, but to
a large extent adopted the reading which a majority of
his mss. preserved.
Following on Lachmann came the editions of TregelleB
and Tischendorf. Little need be said of their critical
principles, which in the main are those of Lachmann ;
but their great value is that they supply us with the
most convenient and accessible form of all the evidence
which has been collected. Tischendorf, edition 8, is far
the fuller ; Tregelles is the easier to read and understand.
The possession of Tischendorf is almost « necessity to
any one who wishes seriously to study the text ; at the
same time, it is not always easy to understand his state
ments of the evidence, so full a use has he made of
abbreviation and symbols (vide p. 92).
No critical edition with apparatus has been published
since Tischendorf, edition 8 ; but there are many perma
nent contributions to criticism which are more recent.
There are the researches and discoveries of mss. made
by Dr. Scrivener, Dean Burgon, and others, not to
mention the various contributions of more recent scholars
which are alluded to in various places in this book ; but
more important than anything else is the publication of
the critical text and introduction of Drs. Westcott and
Hort, usually referred to as W.H. This work is the
foundation of nearly all modern criticism, and demands
a somewhat close attention.
The theory of W.H. is this. Judging from the evi
dence of patristic quotations, all important changes in
the text are earlier than the fourth century, by which
time the text had become stereotyped. The question,
therefore, is, Can we distinguish the history of the text
at an earlier period?

64 THE TEXT OF THE NEW TESTAMENT
W. H. propose three criteria : —
(1) Conflation, i.e. if one variation is shown to be due to
tho mixture of two others, clearly it is later than its sources.
(2) Patristic evidence : if we can show that the earliest
Fathers always use one variation in preference to another,
clearly the former has the superior claim.
(3) As a last resort, there is the somewhat subjective
evidence of critical probabilities.
W.H. urge that these three lines of evidence all lead
in the same direction, and they put aside one large body of
variations as later than the rest. This body of variations
they call Antiochene or Syrian. It agrees chiefly with
the quotations of Chrysostom. It will be well to look
at one of the examples which they give : —
(1) Mc. ix. 38. The text found in the mass of mss. is : —
bs oiiK aKoXovdel rjpiv Kai iKGikvaapcv avrov on ovk
aKoXovBei r/plv.
But two other variations are found : —
(a) Kai iKcoKioptv airov on ovk dKo\ov8cl f/plv, which
with a few small variants is found in NBC LA
Syr-pesh ar boh aeth.
(/3) 6s ovk aKoXovdei pf8' fjpiv Kai iKa>\vopev airov,
which again, with a few variants of no im
portance, is found in DX 1-209 ferr. gr. 28.
al pauc, k.a.b.c. ff2 i. vg. arm.
Now it is obvious that either (a) and (/S) are rival attempts
at simplifying the longer reading, or the latter is a ' con
flation' made up of (a) and ((3). W.H. maintain the
conflation theory, on the ground that either of the two
shorter readings has demonstrably earlier evidence than
the longer.
They quote seven other examples of the same kind,
and express the view that many more might be found.
Now, the important thing to notice is that in all the
cases which they give, the division of authorities is much
the same.
We get (a) a short reading found in X B boh.
(/3) a short reading found in D lat syr70'.
(8) a longer reading found in the mass of later
mss. and versions, which seems to unite
(«) and (/3).

HISTORY OF MODERN CRITICISM 65
W.H. call (8) the Syrian text, (/3) the Western (using
Griesbach's name), (a) the Neutral.
The keystone of their theory is in the passages where
we get this_triple variation, and the point of the argu
ment lies in the assumption that the longer reading is
made by uniting the two shorter ones — not the two
shorter by different dealings with the longer. This
point can be tested only by an appeal to Patristic evidence
and general probability.
The hitter argument is precarious because subjective,
so that the ultimate and decisive criterion is Patristic
evidence. On the whole, this is in favour of the shorter
readings. For taking Origen, Clement, Irenaeus, Cyp
rian, and Tertullian as representing the earliest Fathers
whom we can consult on textual points, we find prac
tically no instances in which they support the Syrian
form. On these grounds, therefore, W.H. convict the
Syrian readings of being later than those which are
Western or Neutral in places where there is a triple
variation. The next step is the application of this result to
merely double variation. Here it is found that we
really get the same three groups of evidence represented
by KB., D., and the late mss. The apparent absence of
one group (making double instead of triple variation)
is only due to the agreement of the Syrian text with one
of the others, sometimes with the Neutral, sometimes
with the Western.
The theory which W. H. base on these facts is that at
some point earlier than the fourth, and probably later
than the middle of the third century, the text of the New
Testament consisted in the main of two great branches,
Neutral and Western, and that then a certain group of
critics — probably in Syria, possibly in Antioch, conceiv
ably connected with Lucian's recension of the LXX, text
— produced an eclectic text, sometimes following one
branch, sometimes the other, sometimes combining both
by conflation, and very occasionally producing a new
variant probably by conjectural emendation of corrupt
or difficult passages. These various choices produced
readings which W.H. for obvious reasons call Neutral

66 THE TEXT OF THE NEW TESTAMENT
and Syrian, Western and Syrian, Conflate Syrian, and
distinctively Syrian.
Had the Syrian revisers any other sources which we
can trace? W.H. think that we can see signs of one
other group of readings, which they call Alexandrian.
These readings represent, in their opinion, a scholarly
revision on small points which seems to have been due
to the school of Alexandrian criticism working on the
Neutral text. It is never very important in itself, and it
is chiefly verbal and grammatical. No one ms. preserves
it in a pure form, but the largest remains are found in
CL boh. Origen, with occasional readings in AA.
W.H.'s analysis of the text may therefore be presented
thus : —
1(a) Neutral text. KB. boh. Orig.
(0) Western text D., Old Lat., Old Syr., Iren.,
Tert., Cyp.
(y) Alexandrian. CL Orig., and traces in other
MSS.
(8) Syrian based on selection and mixture of
(a) (fi) (y), EFGHK, etc., the later versions
and Fathers.
The grouping given here is that which is found in
the Gospels. It is also roughly true of the Acts, so far
as the mss. are the same, and with the important addition
of E2 (Cod. Laudianus) and Aug. to the AVestern
text. A fuller grouping of authorities, and including
the Pauline and Catholic epistles, is given on pp. 71-2.
Having thus roughly analysed the grouping of mss.
and established the relative lateness of the Syrian text,
W.H. went on to investigate the mutual relations of
the three Pre-Syrian groups.
Of these the Alexandrian is easily dealt with; the
readings of the Alexandrian authorities are never found
in any widely spread group, they never rise above the
level of scholarly emendation, and they seem almost
certainly to depend on the Neutral text, though there
is an important but small class which show signs of
Western influence.
The Western and Neutral texts, on the other hand,
are both traceable to the earliest periods of which we

HISTORY OF MODERN CRITICISM

67

have knowledge. For the Western text there is the
second century (probably) authority of Old Latin, Old
Syriac versions, Irenaeus and Tertullian, and possibly
Justin Martyr and Marcion.
For the Neutral there is the evidence, almost equally
ancient, of Origen, and according to the dating of some
critics, of the Bohairic version. On the mere enumera
tion of evidence the Western text claims the majority,
but W.H. decide that on internal evidence the Neutral
deserves the preference. They consider that the
AVestern text is a very early and very corrupt text
due to a process of free interpolation and paraphrase
in an age when the preservation of the true text had
not suggested itself to most people as a desirable object.
The Neutral text therefore (which is best preserved
in KB., esp. B.) is AV.H.'s idea of the true authentic
text. But they make one exception. If the Western
omits anything, they consider that such omission
deserves great consideration, because the genius of the
Western text is so inclined to addition that, if it omits
any reading found elsewhere, the probability is that it
does so because the omission is primitive ; in other
words, they regard Western omissions as not omissions
so much as non-interpolations, and consider that these
passages are to be regarded as corruptions which have
affected all texts except the AVestern.
Their theory then may be presented graphically thus : —
The Original Text. Non-Western additions
Western additions. (Western non-interpolations).

Neutral text KB., etc.

y
Western text D., etc.

Alexandrian text CL, etc.

Syrian Revisers.
Syrian text EGHKU, etc.

68 THE TEXT OF THE NEW TESTAMENT
No one would claim that this theory is final ; but
certainly, whenever the history of the century is written,
it will be found that in the field of textual criticism the
work of Drs. Westcott and Hort is a landmark which,
whether for agreement or disagreement, forms the
necessary point of departure for the1 next generation,
and in parts at least will be the foundation of all suc
cessful work.
It only remains to say something of two lines of
investigation which have been followed up since the
publication of \V. H.
I. That of the school which refuses to agree with the
general outlines of the theory.
These critics are the successors, though not by any
means merely the followers, of the late Dean Burgon.
Their views can be seen at length in the various
publications of the late Mr. E. Miller, Prebendary of
Chichester. The points which are attacked are two : —
(1) The preference shown to the text of KB., and the
little attention given to the later mss.
(2) The theory of a 'Syrian revision.'
As to (1) it is said that AV.H. ignore the probability
that the late mss. represent lost originals, more in
number and equal in age to the archetypes of KB., and
an attempt is made to support this theory by an appeal
to Patristic evidence as supporting 'Syrian' readings.
The reply which advocates of AV.H. make (and to the
present writer it appears valid) is that the only Syrian
readings which are supported by early Patristic evidence
are Neutral and Syrian, or AVestern and Syrian, not
distinctively Syrian, so that the argument in itself only
proves, what all admit, that AVestern readings are very
early, and that many of them were adopted by the
Syrian text.
As to (2), it is said that we have no right to imagine
a Syrian revision in the third century in the complete
absence of any reference to it in the writers of the time.
This is the most important argument which there is
against the theory of AV.H., and it is widely used
in Germany. But it derives its strength largely from
a mistaken point of view. W.H.'s theory does not

HISTORY OF MODERN CRITICISM 69
depend on the fact of a Syrian revision, but on those
textual phenomena which have been described, and
which show that the later text is an eclectic one, made
up of readings sometimes Neutral, sometimes AVestern.
These phenomena are facts, which any one who wishes
can verify by working through Tischendorfs editio
major vm.™ and noting how the Patristic evidence dates
the variants. Taking Origen as the earliest full authority
for the Neutral text, and Irenaeus for the Western,
with Chrysostom for the late text, it is at once obvious
that though Chrysostom and Origen often unite in differ
ing from Irenaeus, and Chrys. and Iren. in differing from
Orig. , yet Chrys. does not differ from them both at once.
And this is almost demonstrative proof that his text, char
acteristically representative of the later Fathers' versions
and mss, , is an eclectic one.
How can you explain an eclectic text, except by a re
vision ? No one has answered this question as yet. But
if there be any answer, it might be adopted without
upsetting W.H. 'a views in the least. The fact of the
'Syrian revision' is merely the deduction which W.H.
drew from the facts. If any one can draw any other
deduction, well and good. But the facts will not be
altered, and they prove that the later text is definitely an
eclectic one, posterior in date, as shown by Patristic
evidence, both to the Neutral and AVestern texts. Yet,
recognising this, it is noticeable that this attack on the
Syrian revision theory has had three great effects : —
(1) It has brought home to us the scantiness of the
evidence which we possess for the earliest periods, especi
ally in the East. For when one is asked why there is no
historical evidence of the revision, it is natural to ask
where we are likely to find such evidence, and the
answer seems to be that there is practically no Syriac
literature of an historical character dating from the third
century. (2) Changes of text can be shown to have occurred at
other times with so little notice, that it is the merest
accident that we have any record of them. For instance,
Theodoret, in the fifth century, replaced the Diatessaron
by some other text, probably the Peshitto, without

70 THE TEXT OF THE NEW TESTAMENT
apparently attracting any very great attention (cf.
p. 37).
(3) Most important of all, the suspicion has gained
ground that the B. or Neutral text is itself not im
probably due to a revision, or, at all events, is not so pure
as W.H. thought. While the local connection of KB.
with Caesarea has been strongly brought out, its textual
connection with Alexandria and Origen has grown more
evident, and at least two elements in modern research
suggest that the KB. text is merely an early form of the
Alexandrian text. These are : —
(a) The growing conviction of Egyptologists that the
most primitive Egyptian version is not the Bohairic, but
rather the Sahidic, which is probably more Western than
Neutral. O) The equally growing conviction that the Western
text has an even greater antiquity and more extended
prevalence than AV. H. imagined.
The suspicion raised in this way is unfavourable to the
idea that the Neutral is always right, and will, almost
certainly, gain further weight when it is more gener
ally recognised that (as is pointed out on p. 81) the
Sinaitic Syriac, the oldest representation of the Oriental
branch of the Western text, suggests that there is a con
siderable class of Neutral interpolations (or Western
non-interpolations as W.H. call them) which have
affected every text except the Old Syriac, just as there is
a similar class which have affected every text except the
Old Latin and its ally Codex Bezae.
Such are the chief attacks and modifications which
W. H. 's theory has suffered.
II. Besides them there is also an important develop
ment of criticism which has not attacked W.H.'s general
view, but has occupied itself with the problem, which
W. H. left untouched, of the origin of the AVestern text.
This is, however, so important and complex a problem
that it demands a separate chapter for its discussion.

HISTORY OF MODERN CRITICISM 71
Note A.
The Gospels and Acta.
The full grouping of authorities, according to W.H.'s
scheme, for the Gospels would be approximately this ; the
authorities in brackets have a mixed text.1
Neutral text.— KB[LTsCA in S. Mark NPQRZ]
[1-209, 33] Boh [Sah Syr""]
[Origen Clem""].
It may be noted that even K boh are not free from
mixture, though there is so much less that it would give
a wrong impression to include them in brackets. The
element of mixture in Sah. Orig. Clem"'1, Meth. is pro
bably entirely Pre-Syrian.
Western text. — D. [ferr. grp.] Lat.™' [vg].
Syr""h- Syr0"- Syr""- [Syr"1""""' arm. sah]
Iren.1" Iren.1*'- Tert. Cyprian [Clem*1™]
[and almost all Latin literature].
There are, of course, other Greek mss. which have
traces, but they are scarcely sufficiently marked to be
included here.
Alexandrian text.— [KCLASX 33, boh sah ClemM"- Orig.]
There is no pure Alexandrian text; it is mostly mixed
with a Neutral base. L is perhaps the best ms. of it.
Syrian text. — EGHKMS and in almost all mss. versions
not mentioned above, also mixed with the other elements in
[CAX 33, 1-209, ferr. grp. Lat.-ital. vg. Syr"""'' Syrhl" arm.]
In the Acts —
Neutral text— KB. [61] boh [sah].
Western text.— BR 137. Lat."'- esp. d. h. gig. p.e.
[Syr"""'"'8 Syr1""'] Cyp. Aug. Vigil.
Alexandrian text. — [KC 61, boh sah Orig.].
Syrian text. — Everything else.
1 The abbreviations used are explained in the note on ' Tischen
dorf a System of Notation,' p. 92.

72 THE TEXT OF THE NEW TESTAMENT
The Pauline Epistles.
In the Pauline Epistles a grouping can be traced
somewhat similar to that which exists in the Gospels
and Acts, but the characteristics of the texts in each
case are less marked. The Western text especially has
far less addition and omission. AVe get:
Neutral.— KB[AC] boh [Orig.].
Western.— DEFG[B] Old Lat., early Lat. Fathers.
Alexandrian. — If anywhere in [AC. Orig.].
And also a Caesarean group, KCH Euthal.1
It is noticeable that in the Pauline Epistles B has a
distinctly AVestern element in places, so that if it be
found in combination with DEFG against KAC, the text
of the latter is sometimes to be regarded as the Neutral.

The Catholic Epistles.
The evidence for the text of these epistles is not sufficient
to allow any definite grouping. It is only possible to
say that here also KB seem to have the best text, and the
Latin to represent a rather different type. One ms.,
ff=theCorbey S. James, is noticeable as perhaps repre
senting a local version used in the neighbourhood of
Aquileia, while m. has an ' African ' text.
1 Vide Sanday and Headlam's 'Epistle to the Romans.'

CHAPTER VI
THE WESTERN TEXT
Although not perhaps universally received, the cheory
of AVestcott and Hort is certainly the basis of most
modern textual criticism. But the problem which
especially exercises the minds of critics is the origin
of the Western text. It is widely felt that until some
probable theory can be reached, which will explain
the curious phenomena found in that group of mss., our
views on the text of the New Testament as a whole,
however probable, can only be tentative.
First of all, then, what is the attestation of the Western
text? It is found in a widely spread family of mss.,
almost all of them demonstrably containing an early text.
The chief members of this family are : —
D. Iren. The African Latin (k. e. Cyprian).
The European Latin (a. b. d. i., etc., Irentot-).
The Old Syriac (Syr0"', Syr"'"-, Aphraates, Ephraem)
with its derivatives, the Arabic Tatian and the
Armenian version.
These are the primary authorities, but they often
receive support from authorities which contain as it were
a residuum of Western readings, such especially as the
marginal additions of the Harklean Syriac, the Ferrar
group, and the Sahidic version. There are also cases
where the AVestern reading was adopted by the makers
of the Syrian revision, and is found in the mass of
manuscripts. In this way there is scarcely any ms.
extant, with the possible exception of B in the Gospels,
which does not afford some support to some AVestern
rending. But of course their support is merely interesting

74 THE TEXT OF THE NEW TESTAMENT
as throwing a light on the later history of the Western
text, not on its origin.
Taking then the combination D Old Lat. Old Syr. as
the typical attestation for Western readings, there are
three preliminary questions which must be discussed
before it is possible to say anything satisfactory about
the ultimate origin of the text : —
(1) What are the characteristic features of the text?
(2) AVhat is the probable date at which we first find
traces of it ?
(3) Is it a distinct whole, or can we divide it into strata ?
(1) What are the characteristic features of the text?
Using the Neutral text as represented by AV.H. for a
standard of comparison, the main characteristics are
addition, omission, and paraphrastic rendering.
A few examples will illustrate this : —
Addition. — In Luke ix. 65. The Western text, as repre
sented by (D) e. Cypr. a. b. Syr.-cur. and the late mss.
which adopted it, reads — oiie oKSare noiov irvevparos tare 6
yap vios rov dvdpdtirov oiK rj\8ev diro\io-ai dX\a aSitrai.
In Matt. xx. 28. The Western text (D*. e. a. b. al.
Syr.-cur. hlms) adds — vpcls Se fr/reTre ix piKpov av£rjaai
Kai iK p.et£ovos tkarrov etvai elaepxopevoi de Kai napaK\jj8ivT(S
Seinrijo-ai jxr) dvaK\lvca8e fls rovs i£ix0VTas tokovs p4]irOTe
ivdo^orepos o"ov iiriXQn feat npoaeXBav 6 BenrvoKXrjratp citttj
0"ot*ETl Karoo X^Pn Kal Karaio'xvvBrja'ri iav 8e dvairioTjs fit
rov rfrrova tottov Kai tl<re\8n o-ov tjttiov ipei aoi 6 fSenrvo-
KkjjTtop Svvaye in avat, Kai iarai <roi roOro ypjjcri/iOJ'. with
some variants of an unimportant character among the
authorities. Omission. — In Luke xxii. 19f. The Western text, as
directly represented by D. a. and implied by the reading
of b. e. Syr.-cur., omits ro vwip vpaiv BiBopevov . . . iKxvvva-
pevov, thus making it appear that at the institution of
the Eucharist the iraTrjpiov preceded the apros.
In Matt, xxvii. 49. The Western texts, D. Lat. are the
real authorities which add weight to omission of aXXos 8e
\a$wv \6yxvv Zvv£ev airov ttjv wXcvpdv, Kai egrjXScv vSatp
Kai aipa, which is found in what are usually the best mss.
Paraphrase.— In Acts xvi. 12. The Western reading
(D) for irp&Trj is Kf(pa\rj.
In Luke xxiii. 53. D al. pauc. read iiriOrjKfv r§ pvypeU

THE WESTERN TEXT 75
\idov ov pSyis elKoo-i iKvKiov, which seems to be merely a
wild paraphrase, in a Homeric spirit.1
Such then are, roughly speaking, the chief charac
teristics of the text. Some would perhaps add that there
is a tendency to make one Gospel harmonise with another.
This is perhaps true, but it is inadvisable to lay much
stress on this point for the following reasons : —
(a) This tendency is not peculiar to Western texts.
All types of mss. seem to be affected by it more or less,
especially the later mss.
(/3) There is little room for doubt that the labours of
the higher critics of the Gospels have shown that there is
a common document lying behind at least much of the
common tradition of the three Gospels. Therefore this
is a vera causa which explains verbal agreements in the
Synoptic Gospels, and renders the question of harmonisa-
tion of readings an excessively delicate one.
It is often probable, then, that the Western text
harmonises, but it is scarcely fair to assume this as a
general characteristic.
(2) What is the probable date at which we first find
traces of the text 7
The answer to this question is a good example of the
use of Patristic quotations and the date of versions. So
far as Greek mss. go, we have only D of the sixth century
with any pretensions to an early date, although it is true
that sporadic readings of a Western type are found in K,
and among later mss. there is the Ferrar group. But we
have in Greek the quotations of Irenaeus2 in the second
century, and in Latin the Old Latin version, which we
know to be older than S. Jerome, and which the quota
tions of Cyprian and Tertullian take back to the second
century. And in taking back the Old Latin, these quo
tations also take back the AVestern text. For it is precisely
these Fathers, especially Cyprian, who appear to have
habitually used a Western text of the most pronounced
character, and to have used jio other. Therefore we can
i Dr Eendel Harris, in Codex Bezae, tried to show that it is
actually due to a Latin version written in verse of the style of the
Homeric centonists ; cf., too, his Homeric Centones.
2 Even when only extant in the Latin translation, Irenaeus is
primarily Greek evidence. ,

76 THE TEXT OF THE NEW TESTAMENT
say with absolute certainty that the Latin version used
in Africa in the second century was a purely AVestern
text. And we have in Irenaeus evidence for the use
in the Rhone valley, in the second century, of a Greek
text with much the same markedly Western character
as that which was perhaps copied in the same district in
the sixth century and now survives as Codex Bezae at
Cambridge. Turning to the East, we find much the same thing,
although the evidence is not so full. For there is an
unbroken chain consisting of the Old Syriac mss., the
quotations of Aphraates and Ephraem, with the connect
ing link of the Armenian version of Ephraem's commen
tary, and Tatian's Diatessaron, which in the same way
takes the AVestern text back to the second century.
Can we do more? There is one generation left in
which we can be fairly certain that our Gospels were
used ; and one still earlier of which we cannot speak
certainly. AVhat is the evidence of these two genera
tions? The former is represented chiefly by Justin
Martyr and Marcion.
The evidence of Marcion is only derived from the
quotations of Tertullian and Epiphanius, while it is
further discounted by the fact that it is known to have
been a deliberately 'edited' text. At the same time,
especially when Tertullian agrees with him, and there is
no question of his own doctrine affecting the reading,
there is much weight in Marcion's evidence, which on the
whole seems to point to the use of the Western text.
The same is true of Justin. There is certainly a
considerable number of Western readings proved to have
been read by him. But in his case matters are com
plicated by the question as to his quotations as a whole,
namely, whether he may not have used another document
as well as our Gospels, or else used a harmony, the exist
ence of which is perhaps made probable on independent
grounds by Tatian's text so far as we know it. It is a
fact of some significance that thus the earliest Syriac
shows us the AVestern text in the form of a harmony
made by Tatian, and that the earliest Greek evidence is
that of Tatian's old master, who is also under suspicion

THE WESTERN TEXT 77
of having used a harmony. But exactly what con
clusions ought to be made from this as to its bearing
on the AVestern text is not clear.
Having thus found traces of the Western text in the
age of Justin Martyr and Marcion, it only remains to
examine the remnants of the subapostolic literature
which we have. It is questionable whether this will ever
be a very profitable task. For (i.) there is a doubt in
almost all places as to the source from which the quotation
may be derived. We can never be quite sure that the
quotations are from the Gospels which we now use.
(ii. ) The text of the writers themselves is often corrupt,
and fails just at the critical point. So that it is almost
impossible to say that the subapostolic text is Western,
while, at the same time, it is certainly more daring to
say that it is not.
One instance must suffice : —
The quotations, apparent or real, in the Didache. — There
are three places in this very early book (perhaps even
belonging to the first century), which may possibly be
connected with a use of the Western text, (a) In 1, 2. the
Didache reads — iravra Be o<ra eav 6eXr)o-Qs pr) yiveaBai aoi,
Kai av aXXco pi) rroiei. This may be a perversion of Matt.
vii. 1 2, due to the influence of the Jewish saying attributed
to Hillel and found in Tobit iv. 15 (6 o-ii piaeis, AXXa pi)
iroiei), and with variants in other Jewish books, or it may
be due to the AVestern text of Acts xv. 20, 29, where D
Iren. Cyprian read Kai oaa pr) diXere iavTols yiveaBai, eripa
pi] iroielv, or again it is possible that the Western reading
is due to the use of the Didache. Different minds may
consider these possibilities as possessing different values,
but no one can say that any one of them is impossible.
(4) In Didache 1, 3. the text eiXoyeire tous Karapatpivov
vplv, Kai wpo<T€i>xeo8e vrrep tS>v ixBpav vpa>v, vncrTevtre Be
virep to>v 8io>k6vt(ov ipas, seems to be nearer S. Matthew
in the Western form than to anything else, but it is not
a clear instance, and how are we to explain the unique
vrjiTTeveTe ?
The idea of a kind of positive efficacy in fasting is found
in the AVestern text of Mark ix. 29, but no ms. has
anything of the kind in the Sermon on the Mount

78 THE TEXT OF THE NEW TESTAMENT
(c) It is at least more probable than not that Didache 9
refers to the Service of Holy Communion, and that
the writer places the Upros after the iroTr\piov. This at
once reminds us of the Western text of Luke, which, as
mentioned on p. 74, similarly transposes the usual order.
But the actual details of the passage have nothing to
point to one Gospel rather than another, and the wording
of the prayers is more Johannine than Lucan.
Similar results, seem to follow from an examination of
all the rest of the literature of that date. It is at
present impossible to affirm or deny the use of the
Western text. Therefore the result of this glance at the
evidence for the existence of the Western text in various
ages and places is this : that it is shown to have existed
in the earliest times of which we have any certain know
ledge — both in Syriac, Latin, and Greek speaking circles ;
in the East, in Africa, in Italy, and in Gaul. Can we go
further, and say that it was also found in the Nile valley ?
Perhaps we can. For the quotations of Clement of
Alexandria, as recently published in the Cambridge
Texts and Studies by Mr. Barnard, seem to show that he
also used a form of the Western text in Alexandria in
the second century. This would suggest that the use
of the Neutral text in Alexandria began at some date
between Clement and Origen, — a theory which finds
support in the 'view of the more modern Egyptologists,
who date the Sahidic version, with its AVestern readings,
earlier than the Bohairic, which is Neutral.
If this theory be true, we can say that the AVestern
text is everywhere found wherever we have any evidence
for the text of the second century in Patristic quotations.
But in thus generalising we run the risk of begging
the question when we talk of the Western text having so
extended a prevalence. Strictly speaking, all that we
can say is that Western readings are found. The
Western readings in Latin sources are not the same in
all cases as those in Syriac ones, and we have no right to
construct an hypothetical original text, containing
readings for which there is only the authority of one of
these versions, unless we are prepared to show that the
other has been corrected by a Neutral source.

THE AVESTERN TEXT 79
This naturally leads us to the consideration of the
last preliminary question : —
(3) Is the Western text a clearly denned whole, or can we
divide it up into groups or strata 1
The importance of this question is, that if the latter
alternative be shown to be probable there is a presump
tion of considerable strength that we have to deal, at
least in part, with successive layers of corruptions. And
on the whole the existence of groups and strata is fairly
clear. It is not a point which has been fully worked out at
present, but, as Mr. Burkitt has shown, the interpolations
of the Western text do seem to fall into three main
divisions : — (a) Latin interpolations, greater.
1/3) Latin interpolations, lesser.
(y) Syriac interpolations.
Of these, (a) comprises such passages as the ' pericope
adulterae,' John vii. 53-viii. 12, or the appearance of a
light at the resurrection, Mark xvi. 3.
These are especially characteristic of the African
Latin ; many of them are also found in the European
Latin ; comparatively few would probably find a place
in a critical reconstruction of the Old Syriac.
(/3) The second class contains small additions, as that of
Kai 6 (tvv'icov o-wiira in Mark iv. 9. In character they are
easily distinguishable from the greater interpolations.
They are shorter, and almost always suggested by the
context ; they very rarely add a new fact, or tell a new
story. The greater interpolations, on the other hand, are
usually bold additions, some of them strikingly original
and apparently primitive, which often seem to be due to
some tradition external to the general current of the
evangelical narrative.
As their character differs, so also does their attestation
for while the African Latin was seen to be the stronghold
of the greater interpolations, the European Latin is the
text which is especially characterised by the smaller ones.
Moreover, in this case also the oldest Syriac text would
seem to have omitted most of them.
(y) The Syriac interpolations are not in character

80 THE TEXT OF THE NEW TESTAMENT
unlike the 'greater' interpolations of the African Latin.
For their attestation we have to rely chiefly on the
Sinaitic Syriac. A characteristic pair are the preserva
tion of the tradition that Barabbas name was Jesus, and
the addition in Luke xxiii. : Woe to us, woe, etc.
The most useful list and statement of these three
classes of interpolations will be found in Mr. Burkitt's
Old Latin and the Itala (Texts and Studies, iv. 3).
We may perhaps put aside the smaller interpolations
of the European Latin, a manifestly later form of the
Latin text than that represented by the African version,
though it is an interesting question whether the omission
of the longer ones by the European is due to excision
or to the prevalence at the beginning of a different type
of text, which afterwards was contaminated by a set of
small corruptions which did not affect the African.
But however that may be, we certainly have the two
classes of greater Latin and Syriac interpolations.
At the same time it must be remembered that, as was
shown in Chapter III., there is evidence for the theory
that the Latin and Syriac versions once lived side by
side, and the common text which they exhibit certainly
does contain some of the most remarkable of Western
readings, such, for instance, as the addition to John iii. 6 :
'For God is a Spirit.'
And if we turn to the question of omissions we
find the same phenomenon. There are two distinct
groups : — (a) Latin omissions.
(6) Syriac omissions.
(a) The Latin omissions are curiously distributed : there
are no less than eight important omissions in Luke xxiv.
(all of which omissions W. H. recognise as correct, and
call non-interpolations). But these do not stand abso
lutely alone, e.g. the omission in Matt xxvii. 49 stands
on precisely the same evidence ; for the testimony of the
late mss., and therefore of the textus receptus, does not
affect the point materially.
Just as was the case with the interpolations, the
evidence of the Syriac version is usually against the
Latin omissions, though there is a residuum which is

THE WESTERN TEXT 81
attested by both, which may be called Latin or Syriac
indifferently. (b) The Syriac omissions as represented by the Sinaitic
Syriac are more numerous than the Latin, and more evenly
distributed ; but, of course, owing to the lamentable loss
of almost all early Syriac evidence, there is not the
wealth of attestation which is available for Latin read
ings. A full list is found in the introduction to Mrs.
Lewis's The Sinaitic Palimpsest retranscribed. Typical and
interesting examples are the omission of Matt. xii. 9 ;
Luke xxiii. 11, 12 ; John xiv. 10, 11.
Thus we get both in omission and interpolation the
same phenomena of a double line, Latin and Syriac, each
having its own characteristic readings, with a residuum
of important passages common to both.
On the whole, therefore, the answers to the three
questions raised, as preliminary to the discussion of the
problem of the origin of the Western text, are these : —
(1) The characteristic features of the text are addition,
omission, paraphrase.
(2) The text can be traced back to the earliest times of
which we have knowledge, and in every part of Christendom,
with the possible, but not probable, exception of Alexandria
and the Nile Valley.
(3) We can trace at least two strata in the Western
text, separated not by characteristics, but by attestation, one
represented by the Latin texts, the other by the Syriac.
There is a common residuum of readings which do not differ
in internal characteristics from those which are peculiar to
either branch.
The main question is, then, open for discussion. It is,
What is the origin of the text which presents these re
markable phenomena ?
The theories which have been suggested may be divided
into two groups : —
(1) Those which assert the primitiveness of the Western
text. (2) Those which regard it as a series of corrupt
accretions. As to the first group. Before looking at that form of
the theory which is most before the public at present,

82 THE TEXT OF THE NEW TESTAMENT
it will be well to notice the features of the AVestern text
which have impressed critics with its primitiveness.
These are, first, some of the interpolations in the Gospeis,
such as the story 1 of the man working on the Sabbath,
and a somewhat greater number of the interpolations in
the Acts, which seem to be of so striking a character that
they can scarcely be false. And secondly, the omissions,
more particularly the Latin omissions (the Syriac omis
sions have not been long enough known), which it is said
are incredible if the original text had not, made the same
omission. So far as the omissions go, many scholars have agreed
that intrinsic probability declares strongly in favour of
the Western text, and W.H.'s view,, wjth later develop
ments, will be found summarised on p. 85. But in its
entirety the view that the Western is the most primitive
form of the text has found few supporters. Bornemann,
it is true, did make an effort to explain all other variants
as corruptions of the AVestern text, but his views have
never obtained many followers, and he may be safely
disregarded. Of recent years, however, Professor Blass
has made an attempt to rehabilitate the Western text
without giving up the Neutral. The Gospel of S. Luke
and the Acts are the starting-point of his investigation.
Indeed, his theory as a whole" applies to those documents
exclusively. It is this. There are many places in the Lucan narra
tives where the intrinsic probability of either the AVestern
or Neutral reading is convincing; where, in fact, it is
inconceivable that any one having either reading before
him would deliberately alter it to the other. Each read
ing has all the marks of originality.
The only possible theory, says Professor Blass, is that
the author himself actually wrote both, or, in other
words, that we possess a first and second edition of the
writings of S. Luke.
The details of this theory are interesting. The history
1 Luke yi. 4, D. d. add : rjj airy i/pipy, Seaadpevis two ipya-
\bpevov nji <r_a8f3&Ttp elirev a&Ttp, "Avdpaire el piv olSas rl Totets
paK&pim et" a di paj oldas eVurardparo; Kai ¦mpa/Jdrijs el too
vipw.

Gospel -^

THE WESTERN TEXT 83
of the Lucan writings, according to the learned critic,
is as follows : —
Luke wrote the first edition of his Gospel from Caesarea
to Theophilus, who may have been a Roman official
somewhere in the neighbourhood. After writing this
he went to Rome, and then wrote a second edition of
the Gospel for the use of the local Church. This was
the Western text of the Gospel, which Blass calls the
Roman text. He also wrote at the same time for the
Romans his first edition of the Acts, and afterwards made
another copy and sent it to Theophilus, which was the
archetype of the Neutral text of Acts.
He suggests, then, the following arrangement for the
Lucan books : —
To Theophilus from Caesarea, \ Neutral text.
For Roman Church, . ./Western „
Acts -f^ *i'or ^oman Church, . .\Western „
\(4) To Theophilus from Rome, J" Neutral „
This is an ingenious theory ; it is not a priori im
possible ; it has, the weight of anything coming from so
great a scholar as Blass. It is not possible in the limits
of a small book either to do it justice, or to explain fully
the case against it.
It must suffice to point out that the main reason why
almost all scholars are inclined to reject it is that it
does not recognise the fact that there are strata in the
Western text.
This cannot he shown so clearly in the Acts as in the
Gospel. But in the latter the case in favour of strata
is overwhelming (vide p. 79 f. ). Now, granted that there
are strata, the deduction is this : —
The two great authorities for the earliest text of a
Western type are the African Latin and the Old Syriac.
The ordinary conclusion from this is that the Western
text represents a series of accretions from some source
which we cannot yet identify, and which is not a
homogeneous whole. It would be argued that so far as
the places where the Latin and Syriac disagree are con
cerned, we have in addition to the evidence for the
Neutral text, as a homogeneous whole, the evidence

84 THE TEXT OF THE NEW TESTAMENT
that the earliest Western text (i.e. the common arche
type of Latin and Syriac) agreed with the reading of the
Neutrals. The only way in which this argument could
be invalidated would be by showing that the Western
authority which agreed with the. Neutral text had been
corrected to a Neutral standard. But in the case of
the Sinaitic Syriac and the African Latin all the evidence
is against such a theory.
Failing this, it would be necessary to put into the
'second edition' only those places where the Western
evidence is complete. But this is just what Blass has
not done. He has put down as belonging to the primi
tive Western text all the passages for which there is a
shred of Western evidence.
Therefore it is felt that Blass's edition cannot be taken
to represent the earliest form of the Western text even
on his own theory.
The earliest form would be based on the concurrence
of the Latin and Syriac. Undoubtedly it would contain
many interesting variants, but it is doubtful whether
these would be so many, or so different in character from
those which would have to be acknowledged to be later
accretions, as to justify his view, though it might justify
the adoption of many Western readings in preference to
Neutral ones.
Therefore, although Professor Blass's work is stimulat
ing and useful in drawing attention to the early date
and valuable character of Western readings, perhaps even
their primitive originality, his theory of double editions
does not commend itself for acceptance.
Leaving, then, that class of theory which considers the
Western text as primitive, we find several views set
forward which seek to explain the phenomena on the
assumption of a later date. The theories of this kind
which hold the field at present are : —
(1) Dr. Rendel Harris — Latinisation.
(2) Dr. Chase — Syriacisation.
Dr. Ramsay — Revision by an Asiatic scribe.
Dr. Resch — Effect of other translations of the
supposed Hebrew original.
But before looking at these theories, it is necessary to

THE WESTERN TEXT 85
examine W.H.'s treatment of the AVestern text; for,
except in the case of Resch, their view is the starting-
point of all the other suggestions. Stated roughly, their
view is that the Western text can be explained, so far
as interpolations go, as a series of corruptions of the
Neutral text, none of them authentic in the sense of
belonging to the true text of the canonical writings, but
some of them possibly preserving early and original
traditions taken from some other source either written or
oral. They do, however, accept the Latin omissions (the
only ones known at the time) as authentic, on the ground
of transcriptional probability. The words omitted they
consider to be due to some element of corruption which
attacked the Neutral text after the Western had split off;
and, as in the case of the Western interpolations, they
reserve the possibility that the additions represent a true
though not an evangelical tradition.
This position is the starting-point of the four theories
which have to he discussed. They are, all of them,
attempts to work out on these lines, but more definitely,
the causes which have produced the Western variants.
(1) Dr. Rendel Harris. — In his study of Codex Bezae
the Cambridge critic endeavours to show that many of
the readings of this AVestern manuscript are due to easily
recognisable causes. He tabulates them accordingly very
completely, but it is impossible to do more than give a
short summary of the chief results of these ingenious
researches. He endeavours to show that most of the
Western interpolations in the Acts are due to an early
Montanist scribe. For instance, he points to the numer
ous instances in which the Western reading refers to the
indwelling of the Holy Spirit — a doctrine on which the
Montanists laid great stress, — such, for example, as the
addition in Acts xv. 29 to the injunctions to Gentile
converts, eS npdijaTe (pepopevot iv t<5 rSyuu nvevpart,1 or
Acts xix. 1, where the Western text adds BeXovros Be tov
UavXov Kara ttjv 18 lav j3ovXi]V iropeveaBai els lepoaoXvpa
emev aurai to Trvevpa aywv iiroaTpi(peiv els Tr)v 'Aaiav.
1 Even if this be a misplaced gloss, the explanation of the
glossator's point of view would be the same.

86 THE TEXT OF THE NEW TESTAMENT
Here, of course, the purpose of the glossator is to explain
S. Paul's change of plan, but Dr. Rendel Harris's promt
is that the method of the explanation is Montanistie.
Similarly, in Luke he traces many readings to the
influence of Marcion. For instance, he points out that
the Western text of Luke ix. 54, 55, which adds the words
its Kox'HXi'as iiroinaev' (cat einov ovk oifrate iroiov irvevparos
Hare, may be well compared with the fact that Marcion
in his dvTiBeo-eis uses this incident of Elijah to support
his theory that the God of the Old Testament is different
from the God who sent His Son into the world.
He tries to show that these interpolations, or at least
some of them, were made primarily in the Latin side of
a supposed bilingual original.
For instance, the Bezan text of Acts i. 2 adds the gloss
Kai iKeXevaev Kr/pvcro-eiv to eiayyekiov = et praecepit praedi-
care evangelium. Probably this is an explanation of the
word ivreiXapevos, which comes just before, and is trans
lated by ' praecepit ' ; in which case it is obvious that the
glossator is working on a Latin text, taking his cue, as
it were, from the 'praecepit,' repeating and explaining it.
Had the gloss originated in the Greek, we should expect
ivereiXaTo ; and had the Latin been a translation of
iKeXevcrev, we should have expected its almost invariable
equivalent 'jussit.'
In this way Dr. Rendel Harris suggested that almost
all the AVestern additions might be explained. And he
has also endeavoured in a pamphlet on the Diatessaron
of Tatian to throw a little further light on the Latin
omissions or 'non-interpolations.' He draws attention to
the presence both in Tatian and in the Curetonian of the
passages omitted by the Latin authorities, and suggests
that the explanation of the presence of the so-called non-
interpolations is that there was a Pre-Tatianic harmony
which included the words omitted, and that this affected
all texts except the non-interpolating Latins. Of course,
Dr. Rendel Harris was writing before the discovery of
the Sinaitic Syriac Codex ; and the light which that ms.
has thrown on the Syriac side of the AVestern text, and
number of omissions to which it testifies, would probably
make him reconsider this point- It would be necessary

THE AVESTERN TEXT 87
to judge whether the Syriac omissions are similar in
character to the Latin or not, and to consider whether
it would be conceivable that there was yet another har
mony to account for the Syriac omissions. The answer
would scarcely be affirmative.
On the whole, probably Dr. Rendel Harris himself
would not claim that his theory is a complete solution of
the problem. But the value of a book to students is
more often to be found in its suggestiveness than its com
pleteness, and from this point of view few books are
more valuable than Dr. Rendel Harris's. Apart from
this, the permanent element in it is perhaps its demon
stration of the multiplicity of strata in the Western text,
and the emphasis laid on the influence of versions and
harmonies. (2) Dr. Chase. — Dr. Chase's theory is similar to that of
Dr. Rendel Harris, in that he seeks a cause for the
AVestern text in the influence of versions. But he con
siders the Syriac version rather than the Latin to be
the originating' cause. That is to say, that just as Dr.
Rendel Harris traces the AVestern text to an original
Graeco-Latin bilingual, so Dr. Chas* traces it to an
original Graeco-Syriac. And just as Dr. Rendel Harris
traces his text to some centre of Latin Christianity, so
Dr. Chase traces his to Antioch.
The arguments by which Dr. Chase supports his view
are roughly these : —
(1) In certain readings Syriac idioms seem to be re
produced, e.g. in Luke ix. 16 the Bezan text is eiXoynb-ev
in avrovs, which is said to be a literal translation of the
ordinary Syriac construction.
(2) In other cases there are examples of forms of ex
pression characteristic of Syriac, e.g. in John xxi. 7 the
Bezan text is Xryet ... 6 Kvpios ioriv T)pa>v ; where the
addition of f)pa>v is accounted for by Dr. Chase, by the fact
that such is the usual Syriac form of Kvpios when used of
Christ. [This might well be true of Other languages ; it
is of English.]
(3) Sometimes two glosses in different mss. are appar
ently traceable to one original Syriac gloss, e.g. in Matt.
xxvi. 59 ff. D reads Kai oix efy>°" to f§ijs koi ttoXXoi

88 THE TEXT OF THE NEW TESTAMENT
irpoaijXBov yjrevBopdpTvpes Kai oix (vpov to e£r)s, where in
several Latin texts the second egrjs is represented by
culpam or some cognate expression. Dr. Chase explains
this as due to the influence of a Syriac word, which
means both 'against' and also 'after.'
Probably what is generally felt about most of this kind
of criticism is that it is a little too ingenious. It is easy
to believe that the Syriac text has left many marks on
the Greek, but not that this is a satisfactory explanation
of all the Western readings. The impartial observer is
inclined to set Dr. Rendel Harris against Dr. Chase, and
to consider that the theory of each is partially true and
explains some readings, while neither entirely solves the
whole problem.
Mention, however, should perhaps be made of that
part of Dr. Chase's theory which, building on a review
by Dr. Sanday, connects the Western text with Antioch.
If the bond between the Old Latin and Old Syriac texts
is close, we must look for the birthplace of the text in
some district which was acquainted with both languages.
But Latin was spoken in official circles all over the
world, while Syriac can scarcely have been well known
in the West. Therefore we seem forced to suppose an
Eastern origin for the Latin version ; and if so, Antioch
is, on the whole, more probable than anywhere else, for
the following reasons : —
(1) It is known to have been the home of an early and
vigorous Christianity.
(2) It was undoubtedly bilingual or multilingual.
(3) It was in close communication with the rest of the
world.
(4) It was of considerable importance in the Roman
world as a centre of government.
This is important, for there are some indications of
superior knowledge of Roman administration and official
language in the Western text, e.g. the knowledge shown
in Luke that the proper title of Pontius Pilate was «ri-
rpoiros, not f/yepav. If however it be conceded that
readings found in the Old Latin and Old Syriac'have a
greater claim to be considered the right readings than
those which are only found in Neutral authorities, most of
the arguments for an original connection between the

THE WESTERN TEXT 89
Old Latin and Old Syriac are removed, though it remains
probable that they influenced one another directly or
indirectly at a later period.
(3) Professor Ramsay.— In his two important books on
the Life and Work of S. Paul, The Church in the Roman
Empire, and /S. Paul the Traveller and Roman Citizen,
Professor Ramsay has made many suggestive remarks on
the Western text in the Acts. He inclines to the view
that the Western readings are due to a very early
glossator who had a particularly good knowledge of
Oriental geography and customs.
Many of the examples adduced by Professor Ramsay
are strikingly convincing when taken in connection
with the archaeological information which he is able to
supply. For instance, no commentator had ever seen
anything peculiar in the fact that the ordinary text1
makes S. Paul go from Caesarea to Jerusalem in one
stage. But Professor Ramsay is able to show that this
is an impossibility, and at the same time to point to the
AVestern text, which explains that S. Paul was accom-
i panied by Mnason, with whom he stayed the night at a
village on the road. Thus he shows that the Western
glossator had a knowledge of the conditions of travel in
those regions so intimate as to notice a point which has
escaped the notice of all other commentators. Such
examples— and Professor Ramsay gives a great number
of them — are very hard to resist. We feel that we must
admit the accuracy and primitiveness of the Western
text. What we doubt is (1) whether Professor Ramsay
is equally safe in confining I3ie knowledge of the glossator
to Oriental subjects. Does not, for instance, the refer
ence in Acts xxviii. 16 to the aTpaTojreSdpxqs, whom
Mommsen has identified with the ' Princeps pere-
grinorum,' fall into the same category? (2) Whether
such good work is really that of a glossator.
Of course, Professor Blass hails these results as con
firming his theory of two editions both equally primitive
and original, and both written by the same author. It
must be admitted that Professor Ramsay's arguments do
seem to point to the Western text as at least as primitive
1 Acts xxi. 16.

90 THE TEXT OF THE NEW TESTAMENT
as the ordinary type. But the arguments adduced when
discussing Professor Blass's theory hold good. There
fore what is needed is some theory which will uphold
the primitiveness to which Professor Ramsay's theory
points, and yet will avoid the difficulties to which Pro
fessor Blass's idea, that both types of text are due to the
same author, has been seen to be liable.
Professor Resch. — This German critic has published
in two large books in the Texte und Untersuchutigen
series a theory of his own about the Synoptic problem.
So far as this theory concerns textual criticism it may
be put shortly thus : — There was, according to Professor
Resch, an original Hebrew document which was the
source of a great part of our present Gospels. This was
extant in its primitive form after it had been used by
the compilers of our Gospels ; and we can reconstruct it
by a comparison of the ordinary texts of the present
Gospels with each other, and also by the variant forms
found in the early Fathers, and those in the Western
texts, which he regards as often due to various transla
tions of the Hebrew original which they knew. There
is no need to discuss this theory, which finds but few
adherents in its entirety, but it is interesting from the
point of view of the textual critic as a curious attempt
to reconcile those phenomena which have made Blass
regard the Western text as equally primitive with the
ordinary type, and Drs. Chase and Rendel Harris
attribute it largely to the influence of versions reacting
on the Greek. It is also, perhaps, more important for
drawing attention to the fact that the textual critic of
the Gospels at all events, and probably also of the Acts,
has to consider the questions raised by the higher
criticism, and ask whether some of the phenomena which
puzzle him may not be due to the disturbing influence of
the sources used by the compiler of the documents.
Such are the chief theories which have recently been
put before the public. None of them even claims to be
final, but all must be studied by any one who wishes
really to master the problems of textual criticism. Per
haps the general result is to make it probable that W.H.
(largely from lack of evidence) underestimated the

THE WESTERN TEXT 91
possibility that a consensus of the Old Latin and Old
Syriac may give us a really primitive text even when
opposed to the great uncials ; but even if that were to
he proved, and the text reconstructed on these lines, the
problem is not fully solved. We often have readings in
which either variant is possible, and neither is decisively
the better. What is to be said as to the origin of the
readings which are rejected ? That is the problem which
has to be faced. At present it has scarcely been touched,
End it would be out of place to say anything at length on
the point , but the present writer cannot help thinking
that the solution of the origin of the AVestern interpola
tions, or Neutral interpolations, is connected somehow
with the sources of the New Testament rather than with
its text. It is a remarkable fact that the prominent
features of the Western text exist in the Gospels and
Acts, which are based on documents of an earlier date,
but are to a large extent wanting in the Epistles, which
are free compositions unconnected with other writings.
It is therefore well to keep in mind the possibility that
we have cases in the text of the Gospels and Acts of
readings which are authentic in so far as they are part
of the 'source-document,' but unauthentic in the sense
that the compiler did not use them, and which owe their
presence in any text of the New Testament to the re
action of the sources on the text of the compilation.
It may also be well to say one more word to any one
who proposes to study the Western problems. Begin
with the Acts : — not because the material for criticism is
greater, but in spite of the fact that it is less. For the
AVestern readings in the Acts are easier to judge, because
they are bound from the nature of the book to deal more
frequently with questions of geographical and archaeo
logical detail which can be readily tested. The Gospels,
on the other hand, more usually supply AVestern read
ings which deal with sayings and facts which can only
be judged by the criterion of a priori probability. It is
therefore the correct method to study the Western read
ings in Acts first of all, and to form some kind of judgment
on them, and after this to turn to the Gospels and apply to
them the conclusions derived from the study of the Acts.

APPENDIX A
TISCHENDORF'S SYSTEM OF NOTATION
A kitowledok of the symbols and abbreviations commonly
employed in textual criticism is essential to every student of the
New Testament.
Some of these have been explained in the preceding chapters,
but for general convenience these have been repeated here,
together with all the others which are in general use. The
multiplication of these symbols is an ever-growing evil ; and as
no two writers seem to employ quite the same method, it is
impossible to escape occasional misunderstandings. But since
the basis of all other systems at the present day is that of
Tischendorf, ed. viii., the necessary foundation of study is an
acquaintance with his notation. The. best way will be to take
an example and go through its details.
In M' 5,44. d7H7raTe toi)s (xBpods ipuv. Tischendorf's note
is : —
T. e%6p upuiv c. N E 1 22 209 al4 a b ff1- g1- 2- k 1 vg sax fr cop
syr™' Thphil3. "¦ Or 4,324.329.351. item W68- M53- Dial a0 EusP" S89
Irint 210 0ypt6r al  s (Gb00) add evXoyeire rovs Karapaipevovs
vpas (D vpiv, Athen om) c. DEKLMSUAII al pier c f h go syr,ch
etP arm aeth Athen >«s»tu Clem605 (omnino propter sqq e M*
vdtr) EusPra8P 13-7 (omnino eM' vdtr) Const1.2- Chr : : cf L" |.
The first thing to notice in this mass of symbols is the 
and the | at the end. These are the sign posts  implies
that the list of authorities for the variant quoted has been
exhausted, and that the alternatives will now be given. The |
signifies the close of the whole passage under discussion. Had
there been a third variant, there would have been some more
 after 'cf L0' instead of the |.
It appears, therefore, that in this passage there are two
readings : —
(a.) dyairare roi/s ixQpolis Vfiwv,
(/3) dyairaTe rois exdpovs vpwv, eiXoyeire rois Karapup^vovs
vp&s. Of these (a) is Tischendorf's reading, and therefore he first
gives it and the authorities for it, introducing the latter with a
c=cum. (j8) is the reading of s (Gb00). This is a highly complex
abbreviation. The first part («) is itself a compound symbol. It
represents the text of Stephanus, followed by Elzevir, Scbulz, and
92

APPENDIX 93
Griesbach. Now, although the agreement of these three editions
with Stephanus is fairly consistent, there are many places where
one or the other differs. In these cases Tischendorf s practice is
to indicate the reading of the editor or editors who differ from
Stephanus by inserting a ' bracket ' after s, and using the follow
ing symbols: 5°— Elzevir, Sz— Sohulz, Gb=Griesbach, '=' thinks
it probable,' "=' thinks it very probable, °=' prefers the other
variant,' 00= ' strongly prefers the other variant. So that s (Gb00)
in this passage means, The reading of Stephanas, Elzevir, Schulz,
and Griesbach ; but Griesbach strongly prefers the other reading.'
There then follows the list of authorities in their proper places
for each reading. These authorities are always quoted in the
same order.
1. Greek uncials, quoted by capital letters.
2. Greek cursives, quoted by numbers as a rule.
3. The Latin authorities, with notice of the more important
mss. individually.
4. Other versions.
6. Patristio quotations with references.
To continue, therefore, the explanation of the note in question :
Tischendorf means that the shorter reading is supported by the
Greek uncial mss. J<B (for the names, etc., of the uncial mss., see
Appendix B) ; by the minuscule or cursive mss. 1. 22. 209. and
al 4, i.e. four others ; by the Old Latin mss. a. b. ff1. g1. ga. k. 1.
(names, etc., of these also are in Appendix B) ; and the Vulgate
(vg) ; also by the two secondary versions, sax. = the Anglo-Saxon,
and fr.=the Frankish, both of which are derivatives of the Latin ;
by the Coptid version, now called the Bohairic ; and by the
Curetonian Syriac.
It is also supported by Theophylact, Origen, and an anonymous
tract known as the ' Dialogus de recta fide ' (Dial. ), by Eusebius
in his commentary on the Psalms (EusP".), by the Latin transla
tion of Irenaeus (Iren'nli=Irenaei interpres), by three quotations
of the passage in Cyprian, and by others (al.). In the case of
each Father quoted the reference to the standard edition (usually
the Benedictine) is given in small figures, e.g. Orig 4>324= Origen,
vol. iv. p. 324.
In the same way, the longer reading adopted by Stephanus,
etc., is supported by the uncials DEKLMSUAII and most other
Greek mss. (al. pier.), by c. f. h. of the Old Latin, by the Gothic
version (go), the Peshitto (syr"oh, i.e. SchaafTs Synao v. p. 40),
and the Harklean Syriac (SyrP=Syra posterior), the Armenian,
and the Aethiopic. It is also supported by Clement of Alexandria,
Eusebius in his Praeparatio Evangelica, the 'Apostolic con
stitutions ' (const.), and Chrysostom. In the case of Clement and
Eusebius, Tischendorf indicates in brackets that there are reasons
for believing that the quotation is from M' not L°, and the
sign : : cf. L° at the end means that he thinks that the reading is
due to assimilation to the text of Luke.

APPENDIX B
SUMMARY OF GREEK AND LATIN MSS.
MSS. of the Gospels
*K(iv.)1cod. Sinaiticus, at S. Petersburg.
*A(v.) cod. Alexandrinus, in the British Museum.
*B (iv. j cod. Vaticanus, at Rome, in the Vatican.
*C (v.) cod. Ephraemi, a palimpsest at Paris.
*D (vi. j cod. Bezae, Graeco-Latin, at Cambridge.
E (viii.) cod. Basileensis, at Basel, in Switzerland.
F (ix.) cod. Boreeli at Utrecht, formerly belonging to John
Boreel (1629).
G(ix.-x.) cod. Seid'elii, in the British Museum, brought from the
East by Andrew Seidel early in the eighteenth
century.
H'(ix.-x.) cod. Seidelii, at Hamburg; its history is the same as G's.
I (V.-yi.) some palimpsest fragments at S. Petersburg.
K (ix. j cod. Cyprius, at Paris, brought from Cyprus in 1673.
*L (viii.) qod. Regius, at Paris.
M (ix.) cod. Campianus, at Paris, given to Louis xrv. by the
AbbcS Francois de Camps.
N(vi.) cod. Purpureus, at Patmos, S. Petersburg, Rome,
Vienna, and the British Museum.
P(vi.) fragm. Guelpherbytana i., palimpBest fragments at
"Wolfenbuttel.
Q(v.) fragm. Guelpherbytana n., pahmpsest fragments at
"Wolfenbuttel.
R (vi.) cod. Nitriensis, a palimpsest fragment in the British
Museum, brought in 1847 from the monastery of
S. Mary Deipara in the Nitrian desert.
S (dated 949) cod. Vaticanus n., at Rome.
T (v.) cod. Borgianus, in the Propaganda at Rome, a Graeco-
Sahidic ms.
Tb and T°(vi.) Graeco-Sahidio fragments at S. Petersburg.
Two1 (vi.) a Graeco-Sahidic fragment in the Bodleian at Oxford.
U(ix.-x.) cod. Nanianus. at Venice.
V(ix.) cod. Mosquensis, belonging to the Holy Synod at
Moscow.
W*bode' fragments at Paris, Naples, S. Gall, Trinity College
Cambridge, Christ Church Library Oxford, and the
Wake Library Oxford.
1 These figures indicate the century to which the ms. Is assumed.
* See also Chapter IL
M

APPENDIX 95
X (ir.-x.) cod. Monacensis, in the University Library, Munich.
Y(viii.) Fragment* Barberina, in the Barberini Library at
Rome.
Z(vi.) cod-. DuUinensis, a palimpsest at Trinity College,
Dublin,
I (dated 979), at Oxford and S. Petersburg, brought by Tischen
dorf from ' the East,' probably Sinai.
A(x.) cod. Sangallensis, a Graeco-Latin ms. at S. Gall in
Switzerland.
6 (vii. ?) fragments at Leipzig.
A (ix.) cod. Tischendorfianus, in the Bodleian, brought from
' the East ' by Tischendorf.
* 5 (viii.) cod. Zacynthius, a palimpsest in the library of the
British and Foreign Bible Society, London.
II (ix. ) cod. Petropolitanus, at S. Petersburg.
Besides these mss. known to Tischendorf, a few others have
since been lettered as follows : —
2 (vi.) cod. Rossanensis, a 'purple' ms. like N, at Rossano in
S. Italy.
S* ood. Sinopensis (vi.), a purple and gold ms. at Paris.
? (v.) cod. Beratinus, another 'purple' ms. at Berat in Mace
donia.
& (viii. ) cod. Laurensis, at the Laura on Mount Athos.
ii (x. j cod. Diony siacus, at the monastery of S. Dionysius on
Mount Athos.
3 (x.) cod. Andreensis, at the ' scete' of S. Andrew on Mount
Athos.
(ix.) cod. Macedoniensis, to which no letter has yet been
assigned, belonging to Mr. J. Bevan Braithwaite of
London.
And a few other small fragments.

MSS. of the Acts and Catholic Epistles.
KABCD. The same as in the Gospels.
*E (vii.) cod. Laudianus, a Latino-Greek ms., probably written
in Sardinia, used by Bede, and given by Laud to the
Bodleian.
H(ix.) ood. Mutinensis, at Modena.
K (ix.) cod. Mosquensis, at Moscow. Acts is missing.
L (ix.) cod. Angelicus, in the library of the Augustinians in
Rome.
P (ix. ) cod. Porphy nanus, a palimpsest at S. Petersburg, for
merly belonging to Bishop Porphyry.
S (ix.-x.) cod. Laurensis n., at the Laura, Mount Athos.
3 (v.) fragm. Patiriensia, palimpsest fragments formerly at
Rossano, now in the Vatican.

96 THE TEXT OF THE NEW TESTAMENT

MSS. of the Pauline Epistles.
}?ABC. As for the Gospels.
*D (vi.) cod. Claromontanus, a Graeco-Latin ms. at Paris, for
merly at Clermont, near Beauvais.
*E(ix.) cod. Sangermanensis, a Graeco-Latin ms. at S. Peters
burg, formerly at S. Germain des Prez, a copy
of D.
*F(ix.) cod. Augiensis, a Graeco-Latin ms. at Trinity College,
Cambridge, formerly in the monastery of Augia
Dives at Reichenau, on Lake Constance.
*G(x.) =A of the Gospels, cod. Boernerianus, at Dresden,
formerly belonged to C. F. Boerner of Leipsic.
Perhaps the Greek was copied from F.
*H (vi.) cod. Coislinianus, fragments at Paris, S. Petersburg,
and at the Laura, Mount Athos.
K.L.P. As of the Acts and Catholic Epistles.

MSS. of the Apocalypse.
KAC. As of the Gospels.
B (viii.) cod. Vaticanus in., at Rome.
P. As of the Acts.

The Old Latin MSS. The Gospels.
». (iv.) cod. Vercellensis, a 'purple' ms. at Vercelli, said to
have been written by Bishop Eusebius (370).
b. (iv.-v.) cod. Veronensis, at Verona.
c. (xii.) cod. Colbortinus, at Paris.
d. (vi.) The Latin of D, at Cambridge.
e. (iv.-v.) cod. Palatinus, at Vienna, formerly at Trent,
' purple ' ms.
f . (vi. ) cod. Brixianus, at Brescia.
fft (viii.-ix.) cod. Corbeiensis i., at S. Petersburg, formerly at
Corbey, near Amiens. Sometimes quoted as f. for
the Catholic Epistles.
ffj (vi.) Corbeiensis n., at Paris, formerly at Corbey.
gt (ix.) cod. Sangermanensis i., at Paris, formerly at S. Ger
main.
gj, (x.) cod. Sangermanensis n., at Paris, originally at Angers.
h. (iv.-v.) cod. Claromontanus, now in the Vatican.
i. (vii.) cod. Vindobonensis, at Vienna, formerly at Naples
j. (v.) cod. Sareannensis, a 'purple' ms. at Sarezzano. near
Tortona.

APPENDIX 97
k. (vi.) cod. Bobiensis, at Turin, formerly at Bobbio, and said
to have belonged to Columban.
1. (vii.) cod. Rhedigeranus, at Breslau.
m. ' Liber de divinis Scripturis sive speculum.' Attributed
to Augustine, it is really a group of mss., and was
edited by Mai from one at Rome, and more recently
from six mss. in the Vienna Corpus script, eccl.
latt.
n. (v.-vi.), o. (vii.), p. (vii. -viii.) fragmenta Sangallensia, at S.
Gall.
q. (vii. 1 cod Monacensis, at Munich, formerly at Freising.
r. (vii.) cod. Ussherianus, at Trinity College, Dublin.
s. (vi. ) fragmenta Bobiensia, at Milan, formerly at Bobbio.
t. (v.) fragmenta Bernensia, at Berne.
v. (vii. ) fragm. Vindobonensia, at Vienna.
a2 (v.-vi.) At Coire, in the Realisches Museum, part of the same
ms. as n.
S. (x.) The Latin of A of the Gospels.

In the Acts and Catholic Epistles.
d. m. As in the Gospels.
e. The Latin of cod. E.
g. (xiii.) cod. Gigas, a Bohemian ms. at Stockholm. "W.H.
call it holm, and it is often quoted as gig.
h. (vi.) cod. Floriacencis, also known as f. (Blass), a palim
psest fragment at Paris, formerly at Fleury.
b. (v.-vi.) fragmenta Bobiensia, at Vienna, formerly at Bobbio.
And in the Catholic Epistles alone —
f. or ff. =S1 of the Gospels.

In the Pauline Epistles.
m. As in the Gospels, etc. d.e.f.g. the Latin of D.E.F.CX
respectively.
gue. (vi.) cod. Guelpherbytanus, at "Wolfenbiittel.
r. (v.-vi.) cod. Frisingensis, at Munich.

In the Apocalypse.

m. Of the Gospels.
g. h. Of the Acts.

98 THE TEXT OF THE NEW TESTAMENT

Vulgate MSS.
As there are some hundreds of these, it is impossible to give a
full list. The best appear to be : —
am. (viii.) cod. Amiatinus, at Florence. Written in the North
of England, and sent by Abbot Ceolf rid to the Pope
in 715, and afterwards remained at Monte Amiata
(v. Codex Amiatinus, by H. J. "White, in Studia
Biblica, vol. ii.). Wordsworth and White's A.
cav, (ix.) cod. Cavensis, at La Cava, near Salerno, in Italy,
written in Spain. Wordsworth's C.
for. (vi.-vii.) cod. Forojuliensis, at Cividale, in Friuli, but part
at Prague, and quoted by Tischendorf as prag.
Wordsworth's J.
fuld. (vi.) cod. Fuldensis, written for Bishop Victor of Capua,
and corrected by him. The Gospels are arranged
continuously in the same order as was employed
in Tatian's Diatessaron (v. p. 34). Now at Fvdda
in Prussia. Wordsworth's F.
tol. (?viii.) cod. Toletanus, formerly at Seville, then at Toledo,
how at Madrid in the National Library. Words
worth's T.

APPENDIX C
BOOKS VALUABLE FOR A STUDY OF TEXTUAL CRITICISM
Novum Testamentum Graece, ed. viii. maj. Tischendorf.
Prolegomena to Tischendorf. Gregory.
Canon and Text of the New Testament. Gregory.
The New Testament in the Original Greek. Westcott and Hort.
Die Schriften des neuen Testaments in ihrer altesten erreichbaren
Textgestalt. H. von Soden.
Introduction to the Criticism of the New Testament, ed. iv.
Scrivener.
Codex Bezae. Scrivener.
Old Latin Biblical Texts. Wordsworth and others.
The Sinaitic Palimpsest retranscribed. A. S. Lewis.
The Old Syriac Gospels. A. S. Lewis.

APPENDIX

99

Evangelion Da-Mepharreshe. F. C. Burkitt.
Codex Bezae. Rendel Harris.
Euthaliana. Armitage Robinson.
The Old Latin and the Itala. F. C. Burkitt. In Texts
Clement of A lexandria's Biblical Text. P. Barnard. - and
S. Ephraim's Quotations from the Gospel. F. Craw- Studies.
ford Burkitt.
CodeX I and its allies. K. Lake.
Acta Apostolorum. Blass.
Evangelium secundum Lucam. Blass.
Four Lectures on the Western Text. Rendel Harris.
Stichometry. Rendel Harris.
The Annotators of Codex Bezae. Rendel Harris.
The Syro-Latin Text. Chase.
The Syriac Element in Codex Bezae. Chase.
Agrapha. Resch. \
Parallel Texte. Resch. In Texte und
Textkritische Studien. Bousset. j- Untersu-
Der Codex D in der Apostelgeschichte. Weiss, I chungen.
Das lat. N. T. in Afrilca. H. von Soden. '
Church in the Roma/n Empire. Ramsay.
S. Paul the Traveller. Ramsay.
The Traditional Text. Burgon and Miller.
Histoire de la Vulgate. Berger.
The Vulgate. Wordsworth and White.
Diatessaron of Tatian. Zahn,
Some Criticism of the Text of the New Testament. Salmon.
' Textual Criticism of the Greek Testament. Nestle.
Handbook to the Textual Criticism of the New Testament.
F. G. Kenyon.

APPENDIX D
THE NEW RESEARCHES AND THEORY
OF PROFESSOR VON SODEN
A sufficient amount of a new and probably epoch-making
book is now published, to render it desirable to give some account
of its main features, though the time is not yet come for any
attempt finally to estimate its value. This is H. von Soden's
Die Schriften des neuen Testaments in ihrer altesten erreichbaren

100 THE TEXT OF THE NEW TESTAMENT
Textgestalt. The first instalment of the first part appeared in
1902, the second part in 1906, the third in 1907, and the fourth
in 1910. These parts contain a new notation of mss., a partly new
and much more elaborate grouping of mss. into textual families,
and a new theory of textual history.
1. Von Soden's Notation of MSS. — The older system dis
tinguishing between uncials and minuscules is abandoned, and
an attempt made to give information, by the symbol employed,
as to the date and the contents of each ms. All known mss.
are first divided into three classes : — (1) 8- mss., containing the
whole New Testament (SiaS^K-n), with or without the exception of
the Apocalypse ; (2) c-mss., containing the Gospels (eiayyiXiov) ;
(3) a-Mss., containing Acts and Epistles, with or without the
Apocalypse (dir6a"roXos).
The mss. within each of these classes are assigned numbers,
in accordance with their date and contents, on the following
system : — 8- and a-Mss., up to the end of the ninth century, are
numbered 1 to 49 ; those of the tenth century 50 to 99 ; for the
following centuries numbers of three ciphers are taken, and the
cipher in the hundreds' place indicates the century : thus, 121
means a ms. of the eleventh century, 221 of the twelfth, 321 of
thirteenth, and so on. In 3-mss. an indication is given as to the
presence or absence of the Apocalypse by using 1 to 49 in each
hundred for mss. with the Apocalypse, 50 to 99 for those without
it : thus, 8-421 would be a fourteenth-century ms. containing the
whole of the New Testament ; 8-271 would be a twelfth-century
ms. containing all the New Testament except the Apocalypse.
Similarly, the contents of the o-mss. are indicated by the figures.
Numbers of three ciphers are reserved for mss. containing Acts,
Catholic Epistles, and Pauline Epistles, with or without the
Apocalypse — the absence or presence of which is shown in the
same way as in the 3-mss. For mss. containing less than this,
numbers of four ciphers are used, the left-hand cipher always
being one, the next cipher showing the date, and the two right-
hand ciphers indicating the precise contents: 00 to 19 mean
Acts and Catholic Epistles, 20 to 69 mean Pauline Epistles alone,
and 70 to 99 mean the Apocalypse alone. Thus, a-321 means
a thirteenth century ms., containing Acts, Catholio and Pauline
Epistles, and Apocalypse; a-489 means a fourteenth century
ms., containing Acts, Catholic and Pauline Epistles, but not the
Apocalypse ; a-1109 means an eleventh century ms., containing
Acts and Catholic Epistles; a-1221 a twelfth century ms., con
taining the Pauline Epistles ; and a-1372 means a thirteenth
century ms., containing the Apocalypse. The e-Mss. are so
numerous that a further development of the system is necessary.
1 to 99 is used for mss. up to the ninth century, and as this does
not supply sufficient numbers, more are gained by prefixing a 0,
and so gaining another hundred figures. 1000 to 1099 are mss.
of the tenth century, 100 to 199 of the eleventh, and so on.

APPENDIX 101
When this proves insufficient, a 1 is prefixed, and since in the
twelfth and later centuries even this is not enough, use is made
of the thousands, and 2000 to 2999 is rendered available for
the mss. of the twelfth century, 3000 to 3999 for those of the
thirteenth century, and so on : in each case the left-hand figure
indicates the century. Normally an e is to be fixed to these
numbers, but it is obvious that this is only necessary when the
class of the mss. would otherwise be ambiguous, and, in dealing
with the Gospels, von Soden adopts the practice of never insert
ing the e when speaking of e-Mss., and never omitting the 3
when referring to the S-mss. This seems to be the most con
venient method, and can obviously be applied equally well to
a-Mss. 2. Grouping of MSS.— Yon Soden divides the mss. of the
Gospels into three groups, which he calls ' K, H, I.'
The K 'Group. — This is, roughly speaking, W. H.'s Syrian (or
Antiochene) text. Von Soden subdivides it into K1, which is the
purest existing form of K, and is best represented by SVfi 461,
661, 655, and a new ms. (e-94), K1, which is K influenced by
readings of the Ferrar group, E?\ K*, K', and some other varia
tions, which represent more or less unimportant sub-families of K.
It is suggested that K can be traced back to the fourth century,
and that it is the recension made by Lucian.
The H Group. — This represents W. H.'s Neutral and Alex
andrian, between which von Soden does not seem to distinguish.
It is represented best by kBCLZA* 33, 892, 1241, 679, and a few
other fragments. Von Soden thinks that there is a close relation
ship between x and B, to which he ascribes an Egyptian origin,
but that the other mss. are all independent authorities, though
all corrupted by varying degrees of admixture with K. In recon
structing the text of if he does not seem to attach nearly as much
importance to B as has been usual. It is suggested that II is
the recension of Hesychius; it was used by the makers of the
Egyptian Versions, and in the main by Athanasius, Didymus,
and Cyril of Alexandria, but not by Origen or Clement.
The I Group. — This corresponds, more or less, to the Western
Text. The oldest extant form of it is best represented, according
to von Soden, by D 28, 372, 565, 700, and a new ms. (e-050), which
seems to be closely connected with 700, and a new ms. related to
28 and 565. Important sub-families of I are J, which is the
Ferrar Group, and IP, which is made up of Fam.1 22, and some
new mss. (see p. 21). Another sub-family of less importance is n,
composed of the 'Purple mss.' (NS*, etc.), which von Soden thinks
can be identified with the text of the great Cappadocian Fathers,
Basil and Gregory. Still less important are OB23> (all various
combinations of / and K, or sub-families of K), and By, but
the last has a certain historical interest if von Soden is right in
thinking that it represents the recension (/ corrected to a A"
standard) used, and perhaps made by Chrysostom. In recon-

102 THE TEXT OF THE NEW TESTAMENT
structing the original I text, von Soden pays more attention to
28, 565, 700, and less to D than former critics. D he thinks
has been largely corrupted by the influence of the Latin, Syriac,
and, perhaps, Sahidic Versions. He does not recognise these two
versions as authorities for I — they belong to an earlier stage of
the history of the text, and he thinks that I itself represents
a Caesarean recension, with which the names of Pamphilus and
Eusebius may be connected, especially as the quotations of
Eusebius seem to belong to the I type.
The grouping of the mss. in the Acts and Epistles is less com
plicated, but not essentially different. In the Acts, von Soden
distinguishes the same three recensions, — K, H, and I. K, which
contains the sub-families K', and By, is represented by
the mass of late mss. ; H by ^BAC and a few other mss. which
all have a more or less contaminated form of the Neutral and
Alexandrian Text ; / falls into three sub-families, I*, i1, 1°, of
which /" is the most important, and is represented by DE and
a few minuscules. The same grouping is in essentials found in
the Catholic Epistles, though the evidence is much smaller.
In the Apocalypse there are also three main types : K, found
in most mss., H found in the early uncials t{AC etc., and a third
type, of which the commentary of Andreas is the best represen
tative ; to this von Soden gives the symbol Av.
In the Pauline Epistles the three main families are the same : —
H, I, and K. K is the text of the mass of mss., H that of
tfBAC and their allies, I of DEFG and a few others.
3. Textual Theory.— AM known mss. belong either to I, H, or
K; and it can, so von Soden thinks, be shown that they together
represent an original text, called I-H-K, which was used by
Origen, and probably by Marcion, Justin, and Tatian, who used
it as the basis of the Diatessaron. Of these recensions K is the
worst, and i" the best, and I-H-K can be recovered in the main
(1) by eliminating readings due to the influence of parallel
passages in the Gospels; (2) where this rule is insufficient, by
rejecting the reading which seems to be an accommodation to
Matthew, — the popular Gospel in earlier times ; (3) in other
places by adopting the reading found in two out of the three
recensions. Thus, I-H-K is reconstructed, and the question is
then faced whether it is really the earliest which we can reach.
The difficulty is that the Old Latin and Old Syriac versions,
and the quotations of Tertullian, Irenaeus, and Clement of Alex
andria do not seem to support the readings of I-H-K. Yet von
Soden is apparently prepared to argue that in the end they do
really support it, for their readings can all be explained as the
corruption of I-H-K by the Diatessaron, which was written in
Greek, probably early in Tatian's career, and became widely
known. In the Acts and Epistles von Soden's theory is in essentials
the same as in the Gospels. He thinks that in the former, at all

APPENDIX

103

events, the influence of Tatian is probable, though his work is no
longer extant. In the Epistles it is, he suggests, probable that
Marcion's text has had a great influence ; nevertheless in all parts
of the N. T. an original I-H-K text can be recovered, which
approves itself on intrinsio grounds as better than the rival texts
of Tatian or Marcion.
It is obvious that such important contributions to the history
of the text, and such great innovations in theory, are certain to
give rise to a long discussion among textual critics, which it
would be improper to anticipate, but there can be no harm in
pointing out that the foci of this discussion are almost sure to be
the following questions : — (1) Is K really independent of I and
H, or were W. H. right in thinking that it is the result of a
recension based on Western and Neutral texts (/ and H) ?
(2) Is von Soden right in regarding D as the result of Latin and
Syriac influence working on /, and not as evidence for a Greek
text which agreed with the Old Syriac and Old Latin more than
with any Greek mss. ? (3) Is he right in making no distinction
between Neutral and Alexandrian mss. ? (4) Is he right in think
ing that the Diatessaron of Tatian exercised such a wide influence,
and that this influence is sufficient to account for the variations
from I-H-K, shown by the oldest versions and Fathers?
The following scheme, whioh of course omits many of the less
important sub-families, will serve to illustrate von Soden's general
theory of the relations between the different textual families.

r-H-K

Syr. Lat.

K* If K*-

0 2 S K"

(Alt corrupted bg Jfbat Independents
and in varying degrees)

INDEX

Acts i. 2, 86.
 xv. 20, 77.
 xv. 29, 85.
 xix. 1, 85.
 xxi. 16, 89.
 xxviii. 16, 89.
Additions to Western Text, 74.
Aethiopic Version, the, 46.
African Latin, 26.
Akhmimic, 42.
Alcala, 60.
Alcuin, 31.
Alexandria, quotations in Clement
of, 78.
Alexandrian Text, the, 66.
Alexandrinus, Codex, 15.
Amelineau, M., 43.
Ammonian Sections, the, 54.
Ammonius, 54.
' AvnXeySpeva, the, 39.
Antioch, 30, 88.
Antiochene Text, the, 64.
Antonia, 40.
Antony, S. , 43.
Aphraates, 37, 49, 76.
Apostolos, 51.
Apparatus Criticus, 10 ff.
Aquileia, 72.
Arabic Versions of N.T., 46.
104

Archetypes, reconstruction of, 6 ff.
Ariston, 46.
Armenian Version, the, 45, 73.
Athanasius, 53.
Augustine, 25, 50.
Barabbas, Jesus, 80.
 Name of, 80.
Barnabas, Epistle of, 13.
Barnard, Mr. P. M. , 78.
Bashmuric Version of N. T. , 43.
Bengel, 61.
Bentley, Dr. Richard, 61.
Beza, 60.
Blass, Professor, 82.
Bohairic Version, 42.
Borgianus, Codex, 18.
Bornemann, 82.
Burgon, Dean, 63.
Burkitt, Mr., 27, 79, 80.
Caesarea, 14 ff., 56, 70.
Canons, Eusebian, 54.
Capitulation, 52.
Cardinal Ximenes, 59.
Caro, Cardinal Hugo d S., 58.
Catherine's Monastery, S., 34.
Causes of Corruption, ff.
Centones, 75 n.

INDEX

105

Chapter Divisions, 52.
Characteristics of Western Text,
74-
Chase, Dr., 39, 84, 87 f.
Chrysostom, 50.
 quotations in, 64.
Clement of Alexandria, 49.
 quotations in, 78.
Clement viii., Pope, 31.
Clementine Vulgate, the, 31.
Codex Alexandrinus, 15.
 Bezae, 16 ff. , 76.
 in the Pauline Epistles,
72.
 Borgianus, 18.
 Ephraemi, 15 f.
 Fuldensis, 34.
 Pamphili, 45.
 Regius, 18.
 Sangallensis, 18.
 Sinaiticus, 4, 12 ff.
 ¦ Vaticanus, 13 ff.
 Zacynthius, 18.
Col. ii. 18, 3, 9.
Complutensian edition, 60.
 Polyglot, 31.
Conflations, 38.
Conjectural emendation, z, 9,
Coptic Dialects, 42.
 Versions, 42.
Corbey S. James, the, 72.
Correctoria, 31.
Corrsen, Dr. P., 32.
Corruption, Causes of, 3 ff.
Criticism, meaning of Higher, 1.
 meaning of Textual, 1.
Cross grouping, 7.
Cureton, Dr., 33.
Curetonian Syriac Version, 33 ff.
Cursives, n, 19 ff.

Cyprian, 48.
Cyril, 49.
Damasus, Pope, 30.
Diatessaron, the, 69.
Didache, i. a, 77.
 i- 3. 77-
 ix. 77.
 quotations in, 77.
Dittography, 3.
Dogmatic alterations, 5.
Ebionites, 35
Egyptian, Middle, 42 f.
 Versions, the, 44.
Elzevir, 60.
Emendation, conjectural,
Enaton, the, 40.
Ephraem, 49, 76.
Ephraemi, Codex, 15 f.
Epiphanius, 76.
Erasmus, 60.
European Latin, the, 27.
Eusebian Canons, the, 54.
Eusebius, 49.
Euthaliana, 19, 55, 56.
Euthalius, 45, 56.
 Studien, 55 ff.
Evangeliarium, 51.
Evidence, liturgical, 50 ff.
 Patristic, 65.
Fathers, quotations in, 47 ff.
Fayoumic, 42.
Ferrar group, 20, 29.
Froben, 60.
Fuldensis, Codex, 34.
Georgian Version, 46.
Gibson, Mrs., 34.

106 THE TEXT OF THE NEW TESTAMENT

Gothic Version, 46.
Gregory Nazianzen, 50.
Griesbach, 61 1.
Harkel, 39.
Harklean Version, 40.
Harmonies, 5.
Harmonistic tendency of Western
Text, 75.
Harmony, Pre-Tatianic, 86.
Harris, Dr. Rendel, 75, 84 ff.
Hexapla of Origen, 40.
Hierapolis, 39, 40.
" Hippolytus, 49.
Homeric Centones, 75 n.
Homoioteleuton, 3.
Horner, Mr., 43.
Hort, Westcott and, 63 ff.
Interpolations, Latin, 79.
 in Western Text, 79.
Intrinsic probability, 8.
Irenaeus, text used by, 76.
Itacism, 3.
Itala, Old Latin and the, 27, 80.
Italic Latin, 27.
James, the Corbey S., 72.
Jerome, MSS. used by, 24 ff. , 30.
Jerusalem Syriac, the, 40.
Jesus Barabbas, 80.
John vii. 53, 57.
 viii. 12, 79.
 xiv. 10, n, 81.
 xxi. 7, 87.
Justin Martyr, 29, 49, 76.
Justin's use of harmony, 76.
Ke<jm\ata majora, 53.
Karkaphensian Syriac, the, 41.

Kirchengeschichte, Zeitschriftfiir,
SS-
Lachmann, C, Edition of N.T.,
62.
Latin Versions of N.T., 22 ff.
Lewis Codex, 34 f.
Lucar, Cyril, 15.
mss. , Bilingual, 29.
 Latin, 26 ff.
Marcion, 6, 29, 49, 76.
Mark i. 32, 38.
 xvi. 3, 79.
 xvi. 9, 20, 57.
Maria Dei para, S., 33.
Martyr, Justin, 29, 49.
Matt. vii. 12, 77.
 xii. 9, 81.
 xiii. 54, 4-
 xxvi. 59 ff,
 xxvii. 16, 4.
Memphitic Version of N.T., 43
Mesrop, 45.
Methodius, 49.
Middle Egyptian, 42 ff.
Mill, Dr. John, 61.
Miller, Prebendary, 68.
Minuscules, 12, 19.
Mixed variants, 7.
Moesinger, 34.
Mommsen, 90.
Montanists, the, 85.
Mopsuestia, Theodore of, 50.
Nazianzen, Gregory, 50.
Neutral Text, the, 65 ff.
Nitria, mss. from, 33, 94.
Novatian, 48.

INDEX

107

Old Latin and the Itala, 27, 80.
Old Syriac, the, 33 ff.
Omissions in Luke xxiv. 80.
 in the Western Text, 74,'
80.
Origen, 49.
Pamphili, Codex, 45.
Pamphilus, 49.
Pandects, 12.
Paraphrase in the Western Text,
74>
Patristic evidence, 65.
 quotations, 47 ff.
Pauline Epistles, Codex B. in the,
72.
Pericope Adulterae, 79.
Peshitto Version of N.T., the,
33. 38 ff.
Philoxenian Version of N.T. , 39.
Philoxenus, 39.
Polycarp, Rural Bishop, 3g.
Pope Clement viii., 31.
 Damasus, 30.
 Sixtus v. , 31.
Pre-Tatianic harmony, 86.
Primasius, 50.
Quotations in Clement of Alex
andria, 78.
 in Chrysostom, 64.
Ramsay, Dr., 84, 89 ff.
Reconstruction of Archetypes,
6ff.
Regius, Codex, 18.
Rendel Harris, Dr. , 84, 73.
Resch, Dr., 84, 90 f.
Robinson, Dr. Armitage, 55.

S. Gall, mss. at, 18, 95.
Sahak, 45.
Sahidic Version, 42 f.
Sanday, Dr., 88.
Sangallensis, Codex, 18.
Scholz, 62.
Scrivener, Dr., 63.
Sinai,- 34.
Sinaiticus, Codex, 4, 12 ff.
Sinaitic Palimpsest retranscribed,
81.
Soden, von, 20 f. , 58,100 ff.
Stephanus, Robert, 31, 60.
Stichometry, 52, 57 ff.
Subapostolic Text, 77.
Syriac, the Karkaphensian, 41.
 interpolations, 79.
 the Jerusalem, 40.
 the Old, 33 ff.
 Versions, 33 ff.
 omissions, 80.
Syrian Text, the, 64.
Syro-Latin Text, 39.
Tatian, 33 f. , 49.
Tattam, Archdeacon, 33.
Tertullian, 48, 76.
Tendency of Western Text, har
monistic, 75.
Text, the Alexandrian, 66.
Text, the Antiochene, 64.
 interpolations in the
Western, 79.
 used by Irenaeus, 76.
 the Neutral, 65 ff.
 omissions in the Western,
80.
 origin of the Western, 81 ff.
 Subapostolic, 77.
 the Syrian, 64, 65.

108 THE TEXT OF THE NEW TESTAMENT

Text, Syro-Latin, 39.
 the Western, 65 ft, 70, 73, 91.
Texte und Untersuchungen, 90.
Texts and Studies, 19, 55, 78, 80.
Textual Criticism, meaning of, i.
Thebaic Version of N.T., 43.
Theodore of Mopsuestia, 5a
Theodoret, 37, 69.
Thomas of Harkel, 40.
Tischendorf, 63.
Tobit iv. 15, 77.
Transcriptional probability, 8.
Tregelles, 63.
Ulfilas, Bishop, 46.
Uncials, n ff.
Variants, mixed, 7.
Vatican Sections, 52.
Vaticanus, Codex, 13 ff.
Version, the Aethiopic, 46.
Version, Armenian, 45, 73.
 Georgian, 46.
 Gothic, 46.
Versions, the Egyptian, 41.
 Latin, 24 ff.
 Syriac, 33 ff.

Von Dobschiitz, E., 55.
Vulgate, the, 30 ff.
 Clementine, 31.
 editions of, 32.
 Wordsworth and White's,
32-
Westcott and Hort, 63 ff.
Western Text, additions in the,
74-
 characteristics of, 74.
 harmonistic tendency
of, 75-
 omissions in the, 74,
80.
 origin of the, 81 ff.
 paraphrase in, 74.
Wordsworth and White's Vul
gate, 32.
Xenaias, 39.
Ximenes, Cardinal, 59.
Zahn, 34.
Zacynthius, Codex, 18.
Zeitschrift fiir Kirchengeschichte,
55>

' LIBRARY