• ILIHBIRAISBr •

DIVINITY SCHOOL
TROWBRIDGE LIBRARY

PHILOLOGY OF THE GOSPELS.

PHILOLOGY
OF THE GOSPELS.

BY
FRIEDRICH BLASS, De.Phil.,
D.Th., Hon. LL.D., Dublin,
PROFESSOR OF CLASSICAL PHILOLOGY IN THE
UNIVERSITY OF HALLE-WITTENBERG.

MACMILLAN AND CO., Limited.
NEW TOBK : THE MACMILLAN COMPANY.
1898.
All rights reserved.

GLASGOW I PRINTED AT THE UNIVERSITY PRESS SY
ROBERT MACLEH05E AND CO.

10
MY DEAR FRIEND THE
REV. J. P. MAHAFFY, D.D., Hon.D.C.L.,
WHO IS THE FIRST AUTHOR OF THIS
BOOK, IT IS AFFECTIONATELY
DEDICATED.

PREFACE.
The present book is due to the initiative of my
friend Prof. J. P. Mahaffy, who proposed to Messrs.
Macmillan that they should ask me to write it.
Prof. Mahaffy and Mr. George Macmillan revised
either the manuscript or the proof-sheets, and the
ms. had been previously submitted to four young
American friends of mine, namely, Dr. J. Carter, of
New- York, Mr. Charles Downing, of Clinton, Mr. W.
Prentice, of New-York, and Mr. H. Sill, of New- York.
F. BLASS.
Halle, March, 1898.

CONTENTS.

CHAPTER I.
PAGE
St. Luke's Gospel distinctly a Literary Work, 1

CHAPTER II.
The Proem of St. Luke, 7
CHAPTER III.
When did St. Luke Write? 21
CHAPTER IV.
Arguments for and against the Early Date of St.
Luke's Gospel, 35
CHAPTER V.
Importance and Method of Textual Criticism in the
New Testament, 53
CHAPTER VI.
Textual Condition of the Gospels : Matthew, Luke, 74

viii CONTENTS.
CHAPTER VII. PAGE
The Question of the Double Text in St. Luke's Gospel
and in the acts, "6
CHAPTER VIII.
The Proofs for Two Distinct Texts in the Acts, - 113
CHAPTER IX.
The Double Text in St. Luke's Gospel, 138
CHAPTER X.
Some other Textual Difficulties in St. Luke's Gospel, 165
CHAPTER XL
Textual Condition and Original Separate Forms of
Mark's Gospel, 190
CHAPTER XII.
Textual Condition of St. John's Gospel, 219

INDICES.
To Persons, Subjects, and Greek Words, 244
II. To Passages of the N.T. Discussed, 248
Addendum, 250

CHAPTER I.
ST. LUKE'S GOSPEL DISTINCTLY A LITERARY
WORK.
Any disquisition concerning the origin and connection
of the four Gospels containing the life and teaching of
our Lord ought to start from the Gospel of St. Luke,
although it confessedly is not the first written Gospel,
perhaps not even the earliest of the extant Gospels.
The reason is, that only this one is a literary pro
duction in the stricter sense of the word, opening
with an elaborate dedication like other literary pro- i
ductions of that age, as also of our own. A shorter
form of dedication appears in the continuation or
second part of Luke's work, viz., in the Acts, the
preface or proem being replaced by a simple address
to the person to whom both the first and second
parts are inscribed. Now, we may easily find plenty
of similar instances in Greek and Eoman authors,
not indeed in the very largest works, which, whilst
they usually have a long proem, are addressed to
general readers, but in those on a smaller scale, or in

2 PHILOLOGY OF THE GOSPELS.
those divided into separate parts, like the Biographies
of Plutarch. This celebrated and well-known work
has neither a proper beginning nor a definite close,
but seems to have been an unconnected series of pairs
of biographies dedicated to a certain Eoman grandee
of the name of Sossius Senecio. The Biography of
Theseus, which is the first one in our editions, although
by no means the first written, begins with a long
proem justifying the undertaking of this biography and
of the comparison between these two men (Theseus
and Eomulus) ; the address : Sossius Senecio, is put
immediately after the first words, and so it is in other
cases where the original beginning has been preserved
intact. In other works of literature or science we
have prefatory epistles, an even more developed form
of dedication than in the case of St. Luke's Gospel.
The two first books of the celebrated Conica of Apol-
lonius are inscribed in this way to a certain Eudemus,
the fourth to a certain Attalus.
But the most noteworthy instance to be compared
with St. Luke's Gospel seems to be a medical work.
One of our best Orientalists, the late Professor Paul
de Lagarde, has written a short note on the proem
of Dioscorides in comparison with that of Luke, in
order to show nothing less than that Luke has
modelled his own proem on that of the well-known
author on medical plants.1 As a philologist, Professor de
xPaul de Lagarde in Psalterium juxta hebr. Hieronymi, p. 165 i.

ST. LUKE'S GOSPEL A LITERARY WORK. 3
Lagarde does not entertain so many theological doubts
as others on St. Luke's authorship, nor on his being iden
tical with " Lucas the Physician," mentioned by St. Paul
(Col. iv. 14) ; he suggests, therefore, in support of this
identity, that Luke was in possession of some medical
works, especially of the Materia Meclica by Dioscorides,
who was a native of Cilicia (whilst Luke himself was
of the neighbouring Antioch), and, moreover, was at
that time a quite recent author. He proceeds next
to compare the two proems, and finds that Dioscorides'
proem is the model, and Luke's the copy, and not even
a good copy. At this point our doubts begin. The
main sentence in Dioscorides' proem is this : " Since
others have written on the same matter badly, I shall
try to write on it better " ; whilst Luke says : " Since
others have written on the same matter, I, too, may
do it." The train of thought therefore is quite
different in the two proems : why, then, should there
be imitation on the one part ? You will say, because
the words agree, and that agreement cannot be ex
plained otherwise. But in reality not even the words
are the same, except the first in Dioscorides (ttoWoov),
which is (in a different inflexion, however) the second
in Luke ; and, if we look a little further on, we
may find in the medical author (on page 2) such
striking test-words as irapaSiSovai and aKpiftecrrepov,
which are employed similarly, although not identi
cally, by Luke. So I again ask : Why should there be

4 PHILOLOGY OP THE GOSPELS.
imitation on Luke's part ? Because, says de Lagarde,
the argumentation goes the wrong way : if many had
already done it, then there was no need for a fresh
author, but Luke might refer Theophilus to the books
of the eye-witnesses (avToirrai), which were much
more trustworthy than his own could possibly be.
Here is, indeed, a strange mistake on the part of
the learned writer: the eye-witnesses of whom Luke
speaks had not written any books at all, and it
merely marks the modesty of our author that he
does not choose to disparage his predecessors, but only
asserts of himself that he has got very full information,
which is, I should think, a quite sufficient reason
for writing. There is another still : Theophilus, whom
he addresses, is evidently supposed not to possess any
of the books formerly written. Now, Luke might
send him these ; but possibly they were not even
written in Greek, but in Aramaic ; and, besides, they
did not contain so much information as Luke himself
could give ; so he prefers to write a new book. The
result is that we must, with regret perhaps, dismiss
the ingeniously contrived argument for the author's
identity with " Lucas the Physician," and rely on the
old ones, which are indeed, in my opinion, quite worth
relying on.
We might at once go on to examine the rest of
this interesting proem, but we are still detained by
the address we were speaking of. A literary work^

ST. LUKE S GOSPEL A LITERAEY WORK. 5
which is, like this, inscribed to a definite person
cannot come from an indefinite author, but must havej
borne from the first a distinct author's name. We
are not bound to inquire whether this rule holds
good for all modern works without exception, since
it does for all ancient ones as far as we know. But"1
in what way would the name of the author come in ?
An author might indeed begin with his own name,
either like Herodotus and Thucydides, in whose works
the first sentences are very like an enlarged title,
or, in a more modern Way, by prefacing his work
with a dedicatory epistle, the inscription of which,
according to the Greek and Eoman style, must run
thus : " N. N. to N. N. greeting." As we find
neither in Luke, nor in Dioscorides, nor in many
other writings, such an epistle, the only place left
for recording the author's name was the title of the
book. Long ago the title had got a separate and
independent existence, and for this same reason was
liable to get lost, as it is lost in Luke's case. The
title of the Gospel was approximately : Aovku 'Avtlo-
ve'fo? irpos QeocpiXov Xoyos Trpwros, and of the Acts
A. • 'A. irpbs 6. X. Sevrepos. No extant manuscript
gives these titles, because St. Luke's works had been
merged into the collection of sacred books of the
Christian Church, and especially his Gospel into the
collection of canonical Gospels, whose inscriptions run
all in the same strain : " According to Matthew, Mark,

6 PHILOLOGY OF THE GOSPELS.
'.Luke, John." If there had been selected but one
canonical Gospel, even these distinctions would have
disappeared, as they have in the case of the Marcionite
canon. Because Marcion recognized but one Gospel,
that of Luke, and but one apostle, namely, St. Paul:
one part of his canon was simply superscribed
Gospel and the other simply Apostle. It has been
wrongly inferred from this inscription " Gospel " that
Luke's name was not known to Marcion ; we might
just as well infer from the other inscription that he
did not know the name of Paul. But the case being
that the Church recognized four separate Gospels, the
names of the authors have been preserved, that is
to say, the most valuable part of Luke's original
superscription. We may well ask in what way the
Church got at the three other names ; but in Luke's
case it was most easy to pick it up. There are
classical writings whose authors have become unknown
by the loss of the title, as is the case with the well-
known rhetorical treatise on the " Sublime " (wrongly
ascribed to Longinus), which, like Luke's works, is
addressed to a single person (Postumius Terentianus),
and must of course originally have had an author's
r~name in the title ; the Acts also would have become
anonymous, and are really so (the inscription being
simply ILpd^eis rwv a-wocrroXwv), but that, with cer
tainty, the first part of the whole work preserved the
, author's name.

CHAPTER II.
THE PEOEM OF ST. LUKE.
The proem of Luke's Gospel gives so much valuable
information, both on the author's work and on the
preceding work of others, that it well deserves a most
careful examination, the more so as it has been from
very ancient times seriously misunderstood in more
than one point. We shall examine it, in the first
instance, from the literary point of view, and next
with regard to its contents.
The opening sentence of this Gospel is a very re
markable specimen of fine and well-balanced structure,
and at the same time of well-chosen vocabulary. It
has no parallel, in these respects taken together, either
in the Gospel itself or in the Acts ; it is unnecessary
to add that the other three Gospels are far from
exhibiting anything similar. Luke has tried to give
to his work (after the Pindaric phrase) a nzpocrwiTov
rijXavyes, and has succeeded very well. This, too,
is an additional proof of the literary character of the

8 PHILOLOGY OF THE GOSPELS.
work, in contrast to the writings of the other three.
He employs, in this proem, grander words and better
Greek than he generally does, or than others do;
for he was here free to give a sentence wholly his
own, which did not form a part of the sacred narra
tive, and would not lose its becoming, simple, dignity
by touches of literary refinement. Of course I cannot
explain this without entering into details which may
seem to us minute and trivial, though they were not
regarded in that light by the ancients. The very first
word is eireiSrjiTep, never occurring again in the New
Testament ; it was evidently employed instead of
e-ireiSr) on account of its length and grandeur. Then
comes -ire'ifKrjpocpoprjfj.ivoiv, instead of which he might
have written irXripwOevrwv, which gives the same sense
(on which we shall speak below) ; but again the former
was grander and more sonorous. The first word in
verse 2 is in the common text KaOws, but the Cam
bridge codex, which will occupy us more hereafter,
and two quotations in Eusebius1 give, instead of this
KaOws, the apparently identical KaOd. What is the
difference ? Simply this, that KaOws is vulgar, and
KaOd is Attic. As a native of Antioch, and of Greek
(or Hellenized) extraction, Luke had necessarily under
gone discipline from a grammarian, and that gram
marian had taught him : Do not write KaOws, for
that word has no classic authority; but write KaOo,
1 Euseb. Hist. Eccl. iii. 4. 6 ; Demonstr. Evang. , p. 120.

THE PROEM OF ST. LUKE. 9
which has the authority of Thucydides, or KaOd which
is equally good.1 Luke's days were those of reigning
Atticism, the general tendency of the literary world
being to look backwards to the classic period of the
language and literature, when both language and taste
were pure and not infected by barbarous influences,
which had, from Alexander's time, substituted in
literature the taste and style of Caria and Phrygia
for that of Athens. St. Paul, too, when he was
called to speak before King Agrippa, and Queen
Berenice, and the Praeses (or Procurator) Festus, and
the most distinguished society of Caesarea and of
the whole province, took care (if we trust, as we
ought to do, Luke's account in Acts xxvi.) not
to employ vulgar inflections of the verbs, but to say
'luacriv irdvres 'lovSaioi, not otSacriv. In his epistles,
he constantly has o'lSa/xev, -are, -acriv ; but his
schoolmaster at Tarsus had warned him against such
vulgarisms : " 'tcr/jiev, 'icrre, 'icracrw" he must have said,
'' are the true forms which you must employ if you
care to be considered a cultivated speaker or writer."
Likewise in Luke the next word is irapeSocrav. The
colloquial forms of the time, and for that reason
those of all New Testament writers, were eSwKafiev,
eSwKare, eSwKav, but schoolmasters (then as now)
drilled their pupils to make use of the correct Attic
1 Cf. Rutherford, The New Phrynichus, p. 495.

10 PHILOLOGY OF THE GOSPELS.
forms (although, in this case, the incorrect ones are
not altogether alien to Attic poetry, or to Attic
prose of the fourth century).
Moreover, it is easy to show that the structure
of this opening sentence is extremely well balanced.
It consists, as we say, of a protasis and an apodosis,
which are pretty nearly equal in length; but according
to the rhetorical doctrine of the ancients we may
divide it into six members, or cola, in this way :
'^iTreiSijirep iroXXoi eire^elprjcrav | dvara^acrOai oirjyrjcriv
rwv ireirXripocpoprip.evwv ev r}/j.iv irpay/marwv | KaOa
irapeootrav r)fuv oi ij' ap^rjs avroirrai Kai iiirrjperai
yevo/xevoi rou Xoyov II eSo^e kcx/j.oI iraprjKoXovOrjKori
avwOev irdcriv cxKpifiws KaQefrjs trot ypayyai Kpartcrre
QeocpiXe | 'tva eiriyvws irepi wv KarrjyriOrjS Xoywv rr\v
dcrcpdXeiav. Three of these members belong to the
protasis, three to the apodosis ; the sense of each of
the former members stands in corresponding relation
to the sense of the latter, inasmuch as iroXXol stands
in opposition to ica/nol, avard^acrOai Sirjyrjcriv is
parallel to ypa^rai K.r.X.; and, lastly, the tradition
of the eye-witnesses, which is treated in the third
member, is again referred to in the sixth with its
dcrcpaXeiav. There is even an opposition between
irapeSocrav (3), and Karrf^Oris (6) ; but of this we
must speak hereafter, as it belongs to the interpre
tation of the proem. Our author abandons this

THE PROEM OF ST. LUKE. 11
elaborate style suddenly, abruptly, the very first
words of the following narrative being not original
Greek, but transparent Hebrew, and he never returns
to his first style, not even in the beginning of the
Acts, which was modelled, as it seems, on the
beginning of the corresponding second part of one
of Luke's authorities.
I may add that there has been preserved a much
shorter form of the same proem, probably, or possibly,
belonging to a second edition, or, to speak more
correctly, to a second authentic copy of Luke's
Gospel, where the dedication to Theophilus had
been suppressed. A contemporary of Chrysostom,
Severianus, bishop of Gabala in Syria, some of
whose sermons have been preserved among those of
Chrysostom, gives a comparison between the begin
nings of the four Gospels, and states that of Luke
in this form : 'EireiSijirep iroXXol . . . irpayfj.drwv,
eSoPe Ka/xol iraprjKoXovQtjKori rois irdcriv air' ap^rjs
ypd^ai, KaOws irapeSocrav r)fuv ol dir dpyrjs avroirrat
ko.1 iiirrjperai yevo/mevoi rov Xoyov.1 Here the six
members have been reduced to four, by leaving out
the sixth, and retaining of the fifth nothing but
1Chrysost. opp. Vol. xii. 567. The editions give not ira.pr>Kohov-
0i>k6ti but 7rapaico\ov8T)K6Ti, and the editors record irapaKoXovdrjaai
as a various reading in the mss. If we adopt this reading, we
must of course strike out ypfapat, which might indeed be dispensed
with.

12 PHILOLOGY OF THE GOSPELS.
ypd^rai (which cannot form a member by itself);
but in order not to destroy the balance between the
protasis and the apodosis, the third member has
been (quite neatly) transposed from the former into
the latter.
Now, what is the meaning of the proem ? and
what is the true translation ? The latter question
is by no means superfluous, as there have been
many misinterpretations, from very ancient times.
There is no doubt that Hieronymus (Jerome), the
author of the Latin Vulgate, well understood both
the colloquial and the literary Greek of his own
time; but in the course of more than three centuries
the literary language had undergone some change,
and as the proem contains not vulgar but choice
Greek, misunderstandings on the part of a writer so
much later may well be expected. Still Jerome has
understood the proem much more correctly than the
great Eusebius himself, and we must confess that
the author had not sufficiently provided against
misunderstandings, purposing, as we have seen, to
make a irpocrwirov rrjXavyes, by means of words not
vulgar, but grand and sonorous. Jerome's translation
of the first verse runs thus : Quoniam quidem, multi
conati sunt ordinare narrationem quae in nobis com-
pletae sunt rerum. UeirXrjpocpoprjfj.evwv = completae is
right; avard£acrOai = ordinare is wrong; we shall speak

THE PROEM OF ST. LUKE. 13
of the former first. Eusebius in his History of the
Church (iii. 24. 15) makes the following paraphrase
of Luke's words : airiyr/cnv iroiijcracrOai &v avros
ireirXrjpQcpoprjTO Xoywv, which shows that he took
ireirXripocpoprjjj.evwv of conviction and ev tj/uiv of
Luke alone : " the things which have been brought
to conviction in me." This is strange English, but
rwv ireirX>jpocpopruj.evwv , if it is to have this meaning,
is even more strange Greek. "B./uv used of the
single person of the speaker, is by no means un
common in Greek authors, although it is somewhat
doubtful whether this idiom extends to New Testa
ment writers ; ifXrjpocpopelv, " to convince," ireirXrjpo-
cpoprmevos, " convinced," is Pauline (e.g. Eom. xiv. 5 ;
Col. iv. 12); but here we have "things convinced,"
instead of "things existing in the conviction of a
person," or " of which that person has been con
vinced." Would anybody understand " a narrative of
the things convinced in me " ? We must therefore
reject this interpretation, and adopt that of Jerome,
who takes irXrjpocpopeiv like irXijpovv, as we have in
Paul (2 Tim. iv. 5, 17), irXrjpocpopetv rr)v SiaKOvlav,
irX. ro KTjpvyfjia in quite the same sense as Luke says
(Acts xii. 25), irXrjpwcravres rr\v SiaKovlav. The
editor of Wilke's Clavis Novi Testamenti, the Jena
professor C. L. W. Grimm, has written a very
careful paper on Luke's proem, explaining the

14 PHILOLOGY OF THE GOSPELS.
single words and phrases by parallel passages, and
he compares the ireirXrjpocpoprj/j.evwv with ws o
eirXr/pwOrj ravra in Acts xix. 21.1 It follows, then,
that ev r)fj.iv in nobis must be " among us," and the
question arises to whom this " us " may refer. I
confidently answer, to the Christian community
existing in Judaea, most members of which were
a part of that population among which our Lord
had lived and died. The very important fact that
Luke includes himself in that number is to be
discussed later on. We proceed to avara^ao-Oai
ordinare (Jerome), which Eusebius renders simply
by iroirjaracrOai. Both the Latin and the Greek
writer confound dvara^acrOai and crvvra^acrOai, the
former being apparently unknown to them. Now,
it occurs, as Grimm shows, only twice in the
whole range of Greek literature, once in Plutarch
and once in Irenaeus, besides the passage in question.
Plutarch, in his treatise Be Sollertia Animalium
(Moral. 968 CD), gives a curious narrative of an
elephant, who was rather dull by nature, but at
the same time very eager to learn. This elephant
was being taught some tricks, in which he succeeded
much more poorly than his comrades ; but his
ambition made him rise at night and repeat by
himself those movements he was to learn. Now,
1 See Jahrbiicher fur deutsche Theologie, 1871, p. 38 ff.

THE PROEM OF ST. LUKE. 15
this is dvardrreo-Oai. The passage runs: cZcpOrj
WKros avros acp' eavrov irpbs rr/v creXrjvtjv dvarar-
rofievos ra fxaQrifxara Kal fxeXerwv, " bringing together,
repeating from memory," and we see therefore that
in this compound the preposition dva had by no
means lost its sense of repetition. Again, Irenaeus
(iii 21. 2, p. 534, ed. Stieren) gives the well-known
Jewish tradition of Ezra's restoring the sacred books
of the Old Testament, which had perished by the
flames in the capture of Jerusalem. This is ex
pressed by: rovs rwv irpoyeyovorwv irpocprjrwv irdvras
dvard^ao-Oai Xoyovs. Here we clearly see the wide
difference between crvvrd^acrQai and dvard^acrOai :
what those ancient writers had composed (crwerd-
^avro) Ezra restored from memory, dverd^aro. Now,
is it not indeed strange that Grimm, who gives
these two passages as the only instances where this
verb occurs, does not come to the conclusion that
it must have the same meaning in Luke ? As a
matter of fact, he identifies, like his predecessors,
dvard^acrOai and crvvrd^acrOai, giving, in this way,
one instance among thousands of the enormous power
of traditional teaching. In reality, Luke's meaning
must be this : " since many writers have undertaken
to restore from memory a narrative of the things
which have been fulfilled (or have come to pass)
among us." Perhaps you will say that any historical

16 PHILOLOGY OF THE GOSPELS.
writer writes, in a certain sense, from memory.
But we must look at the next clause : KaOd irape-
Socrav rjfj.lv ol . . . avroirrai k.t.X. The thing to
be achieved is the restoration of this irapdSocris,
which of course had been oral, and liable to pass
into oblivion, if it were not in time restored from
living memory. So the use of the verb in Luke is
much like that in Irenaeus, and the attitude and
problem of the earliest Gospel-writers is somewhat
( akin to those of Ezra in the legend.
In the second verse : KaOd irapiSocrav tjfiiv ol air'
dpyrjs avroirrai Kai virrjperai yevofievoi rov Xoyov,
the r)fuv recurs, and must of course be taken in the
same sense as before, denoting the Christian Church
of Judaea, which had received the instruction of the
eye-witnesses and first preachers of the " Word," that
is, of the gospel. There is a distinction drawn be
tween the writers (verse 1) and the eye-witnesses,
implying (as I have already observed) that none of
the written Gospels known to the author bore the
name of an apostle ; he cannot have known our fourth
Gospel, or the first in any form, and we must suppose
Luke's own work to be earlier than both of these.
The air' dpyrjs cannot refer to the birth of Christ, but
to His baptism only ; the passage in Acts i. 21 f.
is decisive in this respect (dp£d/j.evos dirb rod ftairrl-
crfiaros 'Iwdvov verse 22), and the beginning of St. Mark

THE PROEM OF ST. LUKE. 17
shows that the ordinary narrative of the Gospel started
from this point. So far as the apostles were called to
be witnesses of the things they had seen and heard,
they could not possibly begin their narrative at an
earlier period in the life of Christ; and the oldest
written Gospels, which were mere reproductions and
collections of their oral teaching, of course covered the
same space of time. From yevdfxevoi it has been
wrongly inferred that in the author's time those eye
witnesses already belonged to a past age, whilst in
reality the past tense refers only to% their quality as
eye-witnesses and as first teachers.
In verse 3, which contains the apodosis, there is still
another word liable to be misunderstood, iraprjKoXov-
OtjKori. Jerome renders this clause by " assecuto omnia
diligenter," but Eusebius (iii. 24. 15, and more clearly
iii. 4. 6) understands the verb in the sense of " having
followed," and consequently takes irdaiv as a mascu
line, referring it to the avroirrai Kai virrjperai. Like
wise Epiphanius (Against Heresies, 51, 7 ; ii. 458 ff.),
in a free quotation, makes Luke say, iraprjKoXovOrjKori
rois avroirrais . . . yevo/nevois. Now, although it is
quite possible that Luke had at some time seen one
or more of the Twelve, it would be a gross exaggera
tion if he asserted of himself that he had been from
the beginning (dvwOev) a constant follower of all the
apostles; and besides it does not make very good

18 PHILOLOGY OF THE GOSPELS.
Greek to combine this irapaKoXovOeiv with aKpipws.
But Polybius and other Hellenistic authors employ
the verb in the sense of studying, and there can be
no doubt that Luke's use is the same. "AvwOev means
much the same as air' dpyrjs ; but here it seems to
imply something more than dir dpyrjs in the preceding
verse, since the relation between this dvwOev and the
following narrative, which starts at the earliest possible
point, cannot well be denied. K.aOe£>js, belonging to
ypd^rai, might seem to contain a criticism on Luke's
predecessors; but I do not think there is any trace
of criticism in the whole proem. We know from a
passage of Papias (Euseb. Hist. Eccl. iii. 39. 15), that
Mark's Gospel had been criticised from very ancient
times as not giving the events in their historical
order ; but Luke could not possibly introduce a better
order, nor is there anything like orderly chronological
narrative in the middle and largest part of his own
Gospel; so it would be absurd if he promised to do
what he neither could do nor has done. But Papias
does not employ the adverb KaOe^rjs : he says ov /mivroi
raj-ei, and KaOe^fjs seems to me to have quite a different
meaning, referring to the uninterrupted series of a
complex narrative. Thus, in Acts xi. 4, we are
told that Peter i^erlOero avrois (to his fellow-apostles)
KaQegfis, or, according to the western text, iravra KaOegfjs
(as we have in the proem irdcriv side by side with

THE PROEM OF ST. LUKE. 19
KaOegjs), that is " completely/' not breaking the series
by any omission. The series was, in Peter's case,
given by the historical order of events ; but it might
be given also by a mere manner of narration, and it
cannot be Luke's meaning that Peter did not invert
the order of facts, but only that he did not omit any
important fact. On Kpdricrre it will suffice to say
that this epithet was the ordinary one in epistolary
and oratorical style, when the person addressed was
in a somewhat exalted position. So, in the Acts,
Paul says, Kparicrre cfrrjXi^, Kpdricrre ^rjcrre (xxiv. 3 ;
xxvi. 25), and in the dedication of books Kparicrre
occurs when the person addressed is something like
a patron, whilst cplXrare denotes familiarity.1 Who
Theophilus was nobody knows, nor is there any
reliance to be placed upon the tradition that he was
the first bishop of Antioch, although, as Luke him
self was a native of that town, we may conjecture,
as ancient writers might, that he addresses a fellow-
citizen. At any rate we may safely suppose that
Theophilus did not live in Judaea, and that he was,
like the author, of Greek extraction. He had been
previously instructed in Christianity, but not by any
authoritative person, so that he must desire a fuller
and more trustworthy knowledge than he had hitherto
1 See Otto in his edition of the Epistle to Diognetus, p. 79 ff. (p.
53 ff.).

20 PHILOLOGY OF THE GOSPELS.
received. For KarrjyfjOrjs does not denote more than
"you have been informed," not characterizing the
information as trustworthy; so, in Acts xxi. 21,
KarrjyrjOrjcrav is directly used of wrong information.
We have to acknowledge therefore a certain opposition
between KarrjyfiQrjs and acrcpaXeiav, and again between
this verb and irapeSocrav in verse 2 : irapaSiSovai is
always used of communicating one's own knowledge
to others as it is, without alteration. The Christian
community of Judaea, the r)fxels of verses 1,2, already
possessed that full and unadulterated knowledge, be
cause they had been taught by the eye-witnesses
themselves ; but Theophilus lived outside the circle
which the apostles had up to that time served with
their preaching. Now he is to obtain, by means of
Luke's Gospel, the same certainty on all particulars
which was enjoyed by the inhabitants of Judaea ; the
time had come when oral teaching was to be sup
planted by written teaching, and the perishable
impression produced upon a few hearers by the
preaching of the apostles was converted, by means
of letters, into an imperishable and widely diffused
treasure for mankind.

CHAPTER III.
WHEN DID ST. LUKE WEITE?
It is a most natural question at this point: When
did the change described at the close of the last
chapter take place ? It is easy to answer : some
time before Luke wrote ; for the proem itself mani
festly declares that Gospel-writing had begun in
Judaea, and was extended by Luke beyond that
limit. But then it will be asked again : When did
Luke write ? I answer : As soon as he could ; for
there was no reason whatever for postponing a work,
the usefulness of which was self-evident, and which
must be required in any Greek or Eoman town,
where the preachers of the gospel went and found
some willing hearers. People would ask the
preachers : Who was that Jesus ? What was His
teaching ? How did it come to pass that He was
so cruelly murdered by His countrymen ? and a
countless number of similar questions, which could

22 PHILOLOGY OF THE GOSPELS.
not be answered so easily by everybody as they may
/be nowadays.
Let us try to represent to ourselves the reality of
things, as it must have been. Paul and Barnabas
went together, for instance, to Antioch in Pisidia, and
began to preach there to both Jews and Gentiles.
They had been orally instructed by the eye-witnesses,
and were now instructing others in the same way,
successfully, as we see. The adherents they gathered
were eager to spread the gospel to others, and so it
rapidly went throughout the whole district (Acts xiii.
49). Of course Paul and Barnabas knew much fewer
of the particulars than Peter did, and again their
disciples knew less than Paul and Barnabas, and so the
store of knowledge went on diminishing by expansion.
That this was a very imperfect state of things, even
from the first, is self-evident, and in progress of
time it must have become more and more intolerable,
even if we assume that writing, as an aid for
memory, was in some measure employed. But, at
any rate, those writings were very scanty, and very
far from meeting the necessities which must arise
at every moment. There were, in Antioch or
Iconium, or in any of those towns where Christian
communities had been founded, plenty of books of
every kind. We see from the excavations recently
made in Oxyrhynchos in Egypt what a number of

WHEN DID ST. LUKE WHITE ? 23
books existed in a small provincial town, the
fragments of which are now coming to light.
Among the books in Antioch were the sacred books
of the Jews, and any Christian might read as often
as he would the predictions of the prophets about
that Jesus, whom he believed to be his only
Saviour. But on the fulfilment of those predictions,
on the life of his Saviour, on His teaching, on His
death and resurrection, not one of the many books of
his town gave him a syllable of information. Was
not this quite intolerable ? The neighbour or friend
of that Christian was anxious to learn something
about the Jewish man whom he adored ; what could
he answer to the manifold questions of his friend ?
" Go to our preachers, or elders, or prophets ; they
will teach you more than I can." The man went
there, and they taught him something more ; but
how much can that have been ? In the natural
and ordinary course of things, it may seem impossible
that a religious sect so scantily provided with means
of instruction, and so widely separated from its
original birthplace, would maintain itself, or even
increase, for any length of time; and if we must
assume that this particular case was not an ordinary
one, but quite exceptional, the necessity, never
theless, of fuller instruction would be eagerly felt,
and there would be in every town many more than

24 PHILOLOGY OF THE GOSPELS.
one Theophilus, who wanted to know "the certainty
of those things in which they had been instructed."
So I again say, Luke wrote as soon as he 'could.
When could he write? When he had provided
himself with full information. When was that time ?
After the first Gospels had been written, and after
Luke had come into contact with that remnant of
primitive Christianity which existed in Judaea after
the departure of the apostles : in other words,
after he had himself become one of the " we,"
among whom the original teaching of the apostles
was preserved. These are two means of determining
the time ; let us try to make as much as we can
out of each of them.
We have seen, in the preceding chapter, that
Gospel-writing was in the beginning a restoring, from
memory, of what the apostles had told in their
sermons, and what now could no longer be heard from
them, because they had finally left their native
country. As soon as that departure had taken place,
the necessity of restoring and preserving their teach
ing would be felt, and within one year, or two years
at most, that necessity might be provided for. So all
depends on fixing the time of that final departure.
We read in the Acts (i. 8) that the apostles had
received commission from their Lord to preach the
Gospel in Jerusalem, and in Judaea, and in Samaria,

WHEN DID ST. LUKE WEITE ? 2 5
and in every country of the earth. They acted ac
cordingly, beginning in Jerusalem, and after some time
proceeding into Judaea and Samaria (Acts viii.-x.);
as regards the rest of their doings Luke leaves us
without information, and from Paul we learn only
this, that Peter went as far as Antioch in Syria (Gal.
ii. 11). But indirectly the Acts inform us that at the
time of Paul's last visit to Jerusalem (Acts xxi.)
all the apostles had departed, because they are not
even mentioned in the narrative. "The next day,"
says Luke (xxi. 18), "Paul went in with us unto
James, and all the elders were there." Compare the
words in xv. 4, about Paul's former visit : " And when
they were come to Jerusalem, they were received of
the Church, and of the apostles and elders." There is,
moreover, an ancient tradition preserved by Clemens
and by Eusebius, that Jesus had instructed His dis
ciples to stay in Jerusalem twelve years, and after
that time to go out into the world.1 Now, it is well
known that the chronology of our Lord's life and of
His disciples' lives is very far from being satisfactorily
1 Clem. Al. Strom, vi. 5. 43 : St& tovt6 (fnjiriv 6 Herpos (an apo
cryphal writing bearing the name of Peter) elpr)K<:i>ai rbv nipiov tois
airoiTT6\ois' . . . fieTb, StliSeKa erij i^\9ere els rbv Kbafiov, pi tis (of the
Jews) eiirrf. oiK i)Koiio-a/j.ev. Euseb. Hist. Eccl. v. 18. 14, speaking
of Apollonius (a writer in Phrygia against Montanism) says : in 5k
iK irapadbcreus rbv airirrjpi. cpijiri -rrpoo-TCTaxivai tois aiiTou diroo-T&Kois iirl
StidcKd freo-i fiij xapur69fi>ai -ri)s 'IepoixraX^.

26 PHILOLOGY OF THE GOSPELS,
established, and I do not purpose to solve, or even to
discuss, these intricate questions in the present little
book. But, at any rate, the doubts do not extend to
more than four or five years, more or less; and if I
simply adopt, upon the whole, the chronology of the
ancients, the error, if there is any, cannot be said
to be important. The Chronicon Paschale places the
former of the two visits of Paul, of which we are
speaking, under the consulship of Asiaticus and Sil-
anus, that is to say, in the year 46 a.d.; for the later
visit we have Eusebius' computation, according to
which it falls in 54 A.D. The twelve years mentioned
in the tradition do not agree with those dates, but the
difference will not be very great if we place the de
parture of Peter and his fellow-apostles in 47 or 48.
For the visit of Peter to Antioch, mentioned by Paul
apparently in connection with his own visit to Jerusa
lem (Gal. ii), seems to me to be nothing less than the
first stage in a long journey, which conducted him
finally to Babylon. Everybody knows that passage in
the Epistle to the Galatians, where Paul relates that
he and the apostles at Jerusalem parted amicably,
under the agreement that he himself should go to the
Gentiles, and Peter and the others to the Jews. Every
body knows the absurd interpretation given by some
modern theologians, who add to the positive clauses of
the agreement the corresponding negative ones: accord-

WHEN DID ST. LUKE WRITE ? 27
ing to these expositors, Peter was under obligation not
to convert any Gentile, and likewise Paul not to con
vert any Jew. As the Acts contain numerous instances
of Paul's trying to convert Jews, in every place where
he went, they conclude that the narrative of the Acts
is utterly untrustworthy. Paul went to Corinth, and
found there a Jew, say Aquila, who showed some
inclination to adopt the religion of Christ, but first
wanted better information. Now, if we believe the
theologians in question, Paul was obliged to say to the
Jew : " I have no right to convert you ; you must go
to Peter." " Where is Peter ? " "I don't know, very
likely in Babylon." But, putting aside these absurdi
ties, we must try to understand the real meaning of
this agreement. As it was evidently impossible to
draw a line of separation between the Jews and Gen
tiles of one town, or of one district, and as Peter was
not to remain in Judaea, but obliged to go " to the end
of the world," there is but one partition possible : Paul
had to go to the West, where the Gentiles or, to speak
more correctly, the Greek Gentiles formed the
majority and the Jews the minority, and Peter to the
East where there was a small minority of Greeks, and a
very large number of Jews mixed up with Gentiles who
spoke the same language with the Jews, viz., Aramaic.
There Peter was in his proper place, as Paul would not
have been; conversely, in Corinth or Ephesus or Eome,

28 PHILOLOGY OF THE GOSPELS.
Paul was in his proper place, and Peter not. Now,
Babylon appears as the town from which Peter sent
out his (first) epistle, and Antioch lay on the route
from Jerusalem to Babylon. " But " — I shall answer
only one " But," and must refrain from entering
into the interminable field of controversy which sur
rounds Peter's epistle and Peter's person — " but the
real Babylon was at that time a deserted town, as
Strabo attests ; therefore the Babylon of the epistle
must be the apocalyptic Babylon, that is to say,
Eome." I answer, that Strabo attests nothing of the
kind, but only that, of the immense space contained
within the ancient walls of Babylon, which extended
for the length of about forty-five miles, far the larger
part lay waste, and that therefore one might well apply
to Babylon the line of a comic poet on Megalopolis in
Arcadia : eprjfiia fxeyaXtj 'crrlv r) fxeyaXij iroXis (Strabo,
p. 738). But still the large town had a great many
inhabitants, although the neighbouring Seleucia had
more. We might as well say the same of modern
Eome, when we stroll from the Capitolium to the
Colosseum and beyond, or to Mount Palatinus, or to
other parts which are enclosed by the ancient fortifica
tions. As for the epistle, it is sent (see i. 1) first to
Pontus, Galatia, Cappadocia, lastly to Asia Proper
and Bithynia, which agrees admirably with Babylon as
the starting-point ; but, if that starting-point had been

WHEN DID ST. LUKE WRITE ? 29
Eome, we may say with certainty that Asia
would come first in the address, and not remote
Pontus. _
But it is time to turn back to our own starting-
point. Peter then went away from Jerusalem in or
shortly after the year 46, and with him, or before
him, or shortly after him, the rest of the Twelve ;
and the necessity of a written Gospel, containing
and conserving the matter of Peter's past sermons,
immediately arose in Jerusalem itself, where that
necessity hitherto had least been felt. I say again :
if anybody prefers the year 48 to 46, or the year
50, or even 51-2 — as Theodor Zahn does in his
recent introduction to the New Testament — , I have
no mind to dispute that point at present, as it will
not affect the relative chronology of events, which
is more important to me than the absolute dates.
If Paul's visit to Jerusalem and the meeting of the
apostles and elders (Acts xv.) was in the winter
of 46-7, Paul's visit to Corinth may have been in
the spring of 49, and his first visit to Ephesus,
from which he immediately left for the East, in
the autumn of 50 ; if we follow Zahn, the dates
are respectively 51-2, 52 (November,— December), 54
(Whitsuntide). Now, after Paul had left Ephesus,
Apollos went there, a Jew born at Alexandria, who
"had been instructed in the way of the Lord (in

30 PHILOLOGY OF THE GOSPELS.
his native town, as the Western text inserts), and
being fervent in the spirit, he spoke and taught
diligently the doings of the Lord, knowing only the
baptism of John" (Acts xviii. 24 f.). Here we
meet with a Christian, who was at the same time
uncommonly well informed about the doings of Jesus,
and utterly ignorant about the rite of Christian
baptism, which the apostles had practised from the
beginning. Evidently he had had a teacher of the
same kind, who had not even baptized him after
the Christian rite, but, on the other hand, had
instructed him in a great many particulars generally
unknown to Christians. You see at once that such
a person is little less than an impossibility. Any
person who had been in contact with the eye
witnesses, and learned from them, could not but
have learned the Christian rite of baptizing, and
Apollos did not even know that rite. So there
seems to be here an enigma very hard to solve.
But what right have we to bring in the " person " ?
Are there not impersonal teachers, viz., books ?
Apollos had come into possession of a written
Gospel, and had been instructed by that, which
could neither baptize him, nor teach him anything
about that rite, if it did not mention it. Now,
this is not only possible, but is actually the case
in St. Mark, that is to say, in the genuine Mark,

WHEN DID ST. LUKE WRITE ? 31
which closes at xvi. 8. I do not infer that Apollos
had got Mark's Gospel, but only say, if Apollos
possessed a copy of a Gospel, either that of Mark
or of any other with a similar conclusion, the whole
difliculty is quite easily and satisfactorily explained.1
We find therefore in. the year 50 (or, according to
Zahn, 54) a written Gospel not only in existence,
but already rather widely spread. Is this possible,
and is it consistent with our former statements ?
Let us see, and let us not take the dates, but the
1It has been argued against this solution, that the word
KaT7fxnfub>tx, used of Apollos (Acts xviii. 25), implies oral and not
written instruction. I do not think that Ka.Tijxoiji.evos iK tov vbfwv
(Rom. ii. 18) is to be understood of a Jew who did not read in
the Law himself, but heard it read and explained by others ; and
the interpretation of " having been catechized in his youth," given
by Mr. A. Wright, in the Expository Times (Oct. 1897, p. 9 f.),
is in direct opposition with the present tense. But if KarisxelaBai
must stand always for oral instruction, much more must aKoieiv
do so. Now we find in Plato (Phaedr. 268 C) : iK fitflXlov iroBiv
aKoitras oterai larpbs yeyovivai, "having been informed by some book
he thinks he has become a physician." Here again, Mr. Wright
seems to me to misinterpret the words, by making them imply
that the man in question could not even read a book himself,
that is, that he was more ignorant than the sausage-dealer of
Aristophanes (see Knights v. 189). But there are more passages
of the same kind, in Plato and in later writers (see the Thesaurus
Or. linguae s.v. a-Koieiv) : for instance, Dionysius, Antiquit. Bom.
i. 48, has oi aKoiovres in the sense of "my readers." Well, you
will say, it was a general custom in antiquity to have a book read
to oneself by a slave or some other person. And is not Apollos
included in that same antiquity ? Let Karvx^^aj. be employed of
hearing even in the passage of the Acts: the book will still be there.

32 PHILOLOGY OF THE GOSPELS.
events only. Peter, having finally left Jerusalem,
went to Antioch, and found Paul and Barnabas
still there. This seems to indicate that that visit
of Peter happened shortly after Paul's visit to
Jerusalem ; for it is not to be , supposed that the
latter long delayed the communication of the apostles'
decree to the congregations of Cilicia and Lycaonia
(Acts xv. 40 ff. ; xvi. 4). A short time after Peter's
visit, Paul started on his own journey, which must
have taken him some time, even while it was still
confined to the different countries of Asia (Acts xv.
40 — xvi. 8). Meantime, the Gospel in question
might be written in Jerusalem. Paul went over to
Macedonia; a copy of the Gospel (for such copies
would be made at once) was brought to Alexandria.
Paul went to Corinth and made there a stay of one
year and six months, during which time he con
verted a great many Corinthians ; that same time
was more than sufficient for Apollos to study the
Gospel and to be converted by its means. Paul went
to Ephesus, and from there to Syria, and so on;
Apollos came to Ephesus. Why might not all
this really happen ? There are, at least, no chrono
logical objections to our suppositions. As for
years and dates, I only add that those of Zahn,
as given above, seem to me not to grant sufficient
space for Paul's journey from Antioch to Corinth,

WHEN DID ST. LUKE WRITE ? 33
and especially for the former part of it, from
Antioch to Troas ; whilst, if we make Paul start
early in 47, and come to Corinth early in 49, we
may assign about six months to the journey through
Macedonia, and eighteen to that through Asia. But,
as I said before, I do not lay much stress at present
on these differences of chronology.
We have sought and apparently found an answer
to the first of our problems, viz., when the first
written Gospels came into existence, and must now
try to find an answer to the second. When did
Luke become one of the " we " ? Here we meet
with no difliculty. Luke testifies himself (making
use of the first person of the plural) that he went
with Paul to Jerusalem in the spring of 54
(according to the chronology of the ancients), and
left Palestine, again as a companion of Paul, as
late as 56 (about August). During this period of
more than two years, he was a member of the
Christian community in Judaea, and at the end of
it he was fully competent to write to Theophilus,
let us say to Antioch : " the things which have come
to pass among us," the more so as he could not be
aware for what length of time he was still to remain
in Judaea. That these two years afforded very ample
time for collecting and studying written Gospels and
gathering oral information, and for writing a Gospel

34 PHILOLOGY OF THE GOSPELS.
of his own, it is almost unnecessary to state. So I
think that Luke finished his Gospel in the year 56,
if we follow Eusebius, or, if any one prefers Zahn's
chronology, in the year 60.

CHAPTER IV.
AEGUMENTS FOE AND AGAINST THE EAELY
DATE OF ST. LUKE'S GOSPEL.
I am well aware, that this early dating of Luke's
Gospel is in direct opposition to opinions most
current at the present day, according to which not
one of our Gospels was composed much earlier than
the destruction of Jerusalem (71), and Luke's Gospel
later than that event. Professor Harnack, in his
most recent publication, even while stating that now
the tide has turned, and that theology, after having
strayed in the darkness and led others into dark
ness (see Matt. xv. 14) for about fifty years, has
now got a better insight into things, and has come
to a truer appreciation of the real trustworthiness of
tradition, still puts Mark's Gospel between 65 and
70 a.d., Matthew's between 70 and 75, but Luke's
much later, about 78-93. Has that confessedly un
trustworthy guide of laymen, scientific theology, after

36 PHILOLOGY OF THE GOSPELS.
so many errors committed during fifty years, now of
a sudden become a trustworthy one ? Or have we
good reason to mistrust it, as much, or even more,
than we had before ? In ordinary life no sane
person would follow a guide who confessed to having
grossly misled him during the whole former part of
a journey. Evidently that guide was either utterly
ignorant of the way, or he had some views and
aims of his own, of which the traveller was unaware,
and he cannot be assumed now to have acquired a
full knowledge, or to have laid those views and aims
wholly aside. Nevertheless, let us examine what
reasons there may be for a later dating of Luke's
writings, after having first glanced at the second
part of them, viz., at the Acts.
As the second part of a work is later than the
first, an early date of the Acts implies an even
earlier one for the Gospel, and we might have started
from the Acts and arrived at the same conclusions
for the Gospel, if we had not chosen to go the opposite
way. For that the Acts were composed in Eome,
during the two years of Paul's first captivity which
^ the Acts attest (xxviii. 30), is an assumption made
as early as Jerome, who says of that book (de Viris
lllustribus, 7) : in urbe Roma compositus est, id quod
intellegitur ex eo, quod non ultra quartum annum
Neronis rerum gestarum memoriam Lucas vertexuit.

DATE OF ST. LUKE'S GOSPEL. 37
The fourth year of Nero is the year 5 8 ; more correctly
we should take the year 59, since Paul, if he started
for Eome in 56, did not arrive there before 57, and
the two years of the captivity extend from 57 (spring)
to 59 (spring). But the argument of Jerome is quite
clear : Luke would have continued his narrative and
not left it abruptly at this point, if he had known
anything more at the time he wrote; so the actual
end of the narrative must coincide with the time of i
finishing the book. The same argument is given more
explicitly by Euthalius, in his preface to Paul's
Epistles,1 and the scholars who lived after the
Eenaissance and the Eeformation of the Church have
adhered to it, not imagining that there would come a
time when this candid way of reasoning would be
abandoned for artificial ones, of which our ancients
could have no idea. One of these artificial reasonings!
is this : Luke had the intention of writing a third
treatise (rpiros Xoyos), which was to contain the
narrative from the end of Paul's first captivity down
to his second and to his death, and therefore he ter
minated his second part at the point where he actually
does, reserving the rest (which was of course, as
far as the main events were concerned, already known
to his readers) for the third part. I quite agree,
that under this supposition the facts find a sufficient
1 See Zacagni, Montim. vet., p. 531.

38 PHILOLOGY OF THE GOSPELS.
explanation ; but the supposition itself seems to me
i -not only unnecessary, but totally unfounded. If I
make a supposition not warranted by any testimony
(as there is no shadow of a trace of such a third part),
I must have and give good reasons for it ; otherwise
the supposition will be discarded as arbitrary. Now
there is one reason which may impose on those who
are not thoroughly acquainted with New Testament
grammar, namely this: Luke says in Acts i 1, that
the Gospel has been his irpwros Xoyos, and irpwros
implies more than two, like primus in Latin. So it
does in old Greek ; but in New Testament Greek the
distinction between the comparative and the super
lative, of which this is one instance, has been altogether
abandoned, and as fxe'iXwv stands for fj-eyirrros, where
three things are compared, so irpwros regularly occu
pies the place of irporepos, where there are but
'two (like English " first " and " second"). It must
be said, moreover, that the close of the Acts,
even if there was nothing to follow, is by no means
abrupt, since Luke has followed the course of the
Gospel from Jerusalem to Eome, from the Jews to
the Gentiles, and might well end his work with its
successful preaching in the centre of the world. But,
on the other hand, if he had known at that time the
final release of Paul, of whose former trials he has
told us so minutely, he would not have omitted to

DATE OF ST. LUKE'S GOSPEL. 39
indicate that much desired result, even if he chose not
to enter into details, which he reserved for another
volume. You will say, why did he not await the
result, before closing his book ? Because (I should
answer) in the case of a speedy release of Paul the
Eomans, for whom he partially wrote, would learn
it as soon as himself, and the Christians in other parts
of the world, especially Theophilus, could not remain
ignorant of so important a fact, but must hear of it
before Luke's book would reach them ; so there was
no necessity to await further events, the exact time of
which was quite uncertain. Besides, we must always
bear in mind that the course of the Gospel was for the
author (as for any Christian) the primary thing, and
the fortunes of persons the secondary thing, even those
of the principal persons, not to speak of subordinate
agents like himself, whose doings he does not stoop to
tell at all. But as I said before, if he had possessed
further knowledge of the events, he would not have
withheld a few words about them, as the main person
of his narrative was therein so eminently concerned.
If therefore the Acts were finished in the year 59, it
clearly follows that the Gospel must have been written
at the time we have already stated.
And now for the theologians, who assert that
Luke's Gospel must be later than the destruction of
Jerusalem. The reason is, of course, that the cata-

40 PHILOLOGY OF THE GOSPELS.
strophe of the holy city is told in the Gospel ; that
is to say, that Luke makes Jesus foretell it, which
is quite sufficient, for coarse and vulgar reasoning,
to lead to the conclusion that the author wrote after
the event. Omne vaticinium post eventum. My
readers must not be afraid that I am going to enter
into a theological discussion of this axiom, since
philology and history are quite sufficient to deal
with it, as far as is needed here. But theologians
may perhaps cry out, that I am treating them unfairly,
ascribing to them that coarse and vulgar reasoning
which they will pathetically disclaim. And yet you
will find in Professor Weiss's Introduction that St.
Matthew's Gospel is to be put later than 71, because
of the verse, xxii. 7 : " But when the king heard
thereof, he was wroth : and he sent forth his armies,
and destroyed those murderers, and burned up their
city." But Professor Harnack is not shocked by this
prediction, nor by any other in Matthew or in Mark ;
so his argument cannot be that vulgar one : Omne
vaticinium post eventum. Very well. I do not intend
to treat anybody unfairly, and the question is not
one of persons, except as they exemplify the general
tendency of a school or of an age. Now, it is un
deniable that the axiom : Omne, etc., underlies the
work of that theological school, which has, as Harnack
says, misled the world for fifty years, and so we must

DATE OF ST. LUKE'S GOSPEL. 41
deal with it, and prove the antithesis : Non omne
vaticinium post eventum, which will be proved, if I
adduce one certain instance of a vaticinium ante
eventum. I might go to the Old Testament, and
take Micah's prophecy of the destruction of Jeru
salem, which is quoted in Jeremy ; but as I need
give only one instance, I prefer modern history to
the Old Testament, and Jerome Savonarola to Micah.
That prophet — for such he claimed to be — was burnt
in 1497 ; his sermons were printed partly in his
lifetime, partly shortly after his death, and prove
that there have been and may be prophecies not only
spoken or even written before the event, but actually *
printed before it. Accidentally, you will say, the
event corresponded to the prophecy. But that is
not my point, whether it was accidental, or the pro
phet had really foreseen the event ; for in the case
of the prophecies recorded by Luke you may raise
the same controversy if you like. Most probably
you will not like to do so, but will try to substitute
the author for the prophet in order to get rid of the
prophecy. But whether of Luke or of Christ, the
prophecies are these : Luke xix. 43 f. : " For the days
shall come upon thee, that thine enemies shall cast
a trench about thee, and compass thee round," and
so on (which prophecy is found in Luke alone) ; and
xxi 20 ff. : "And when ye shall see Jerusalem

V

42 PHILOLOGY OF THE GOSPELS.
compassed with armies, then know that the desolation
thereof is nigh," and so on (which prophecy is re
lated in a somewhat different form by Matthew and
Mark). I do not think that either the former or
the latter of these foretellings is very distinct, since
there are neither names given, nor peculiar circum
stances indicated ; only the common order of events
is described : before the destruction the capture,
before the capture the siege and the circumvallation ;
together with the destruction, the killing of the
inhabitants, and the leading away of others into
captivity, and then the Gentiles' taking possession
of the vacant soil ; in all these particulars there is
nothing but the idea of destruction of the holy city
prepared and developed. There is still another pro
phecy : that the temple would be utterly destroyed,
so that no stone should be left upon another; but
as this one is common to the three synoptic Gospels,
and our critics have absolved Matthew and Mark of
the crime of falsified foretellings, we shall leave it,
although it is indeed much more peculiar, out of the
present discussion. On the other hand, Savonarola
foretold, as early as 1496, the capture of Eome,
which happened in 1527, and those sermons of 1496
were printed in 1497. I must not weary my readers,1
1 1 have treated this subject more copiously in a paper printed
in Neue Kirchliche Zeitschrift, 1896, 964 ff. See Villari, la Storia di
G. Sav., 2nd ed. in 2 vols., Florence, 1887-8.

DATE OF ST. LUKE'S GOSPEL. 43
else I might give many passages which agree as
tonishingly well with the account given by contem
poraries of that memorable event, for Savonarola
entered into particulars, and such particulars as were
indeed very hard to foresee. Especially remarkable
is this, that he extends the devastation to the churches
of Eome, which, in any ordinary capture by a Catholic-
army, would have been spared, but in this case were
not at all respected, because a great part of the con- '
quering army consisted of German Lutherans, for
whom the Eoman Catholic churches were rather
objects of hatred and contempt than of veneration.
Now Lutheranism did not exist in 1496. Among
Savonarola's prophecies we find this one : " Eome,
thy churches will be made stables for the horses,
which they will place therein." In striking accord
ance with this, Guicciardini, one of the eye-witnesses
of the capture, says in his narrative : " You might
see the sumptuous palaces of the cardinals, the sacred
churches of St. Peter and St. Paul, and the other holy
places, which were formerly full of plenary indulgences
and of venerable relics, now reduced to stables of
horses, and instead of hypocritical ceremonies and of
wanton music, you might hear in them the pawing
and neighing of horses."1 I think it quite unnecessary
1See C. Milanesi, II Sacco di Roma del MDXXVII (Flor. 1867),
where the original documents, and among them the narrative of
Luigi Guicciardini, have been reproduced.

44 PHILOLOGY OF THE GOSPELS.
to dwell any more on this topic, the general possibility
of vaticinia ante eventum having now been fully
established. But there have also been written vaticinia post
eventum, and many more than ante eventum. Of
course this is quite true, and on the score of general
probability the prophecies in Luke might seem more
likely to belong to the former class than to the
latter. My object, however, is not to prove a par
ticular, but to disprove an universal proposition ;
it would be foolish to enter into any proof that
Luke's prophecies were really written ante eventum.
But the objection itself, you will say, is wrongly
presented : Professor Harnack, for instance, might
put it in this form. " We have the same prophecy
in the earlier Gospels in a somewhat veiled and
obscure shape (which is true with trie important
exception of the prophecy concerning the temple) ;
on the other hand, Luke gives it in an explicit
and developed form. The real words of Christ
were most probably like those in Matthew and
Mark, but after the event the words were shaped
differently in order to fit them to the event." By
Luke ? That I deny, and very decidedly. Luke
has given what he found, without adulteration ; we
can easily distinguish a compiler of given materials
from one who moulds them artificially, and we clearly

DATE OF ST. LUKE'S GOSPEL. 45
see in Luke everywhere the raw material, whilst
Matthew,, that is to say the composer of the Gospel
bearing Matthew's name, is much more liable to the
charge of artificial shaping. Then, you will say, the
author whose writing Luke used, or if not that
author himself, the oral tradition which was his
source, brought the prophecies into this developed
form, whilst Matthew and Mark must represent the
original words of Christ. Why ? Because Mark is
certainly earlier than Luke. But if there was, as you
supposed, another written source for Luke, besides-
Mark, how will you prove that that source too was
later than Mark ? Because the developed form is
always later than the undeveloped one. And thus
we are gliding into an interminable and hopeless
discussion of general probabilities.
But I hope it will be granted on all sides, that
Luke really had, for his 21st chapter and those which
precede and follow it, a different source from Matthew
and Mark. Now, according to Matthew (xxiv. 15) and
Mark (xiii. 14), Christ says to His disciples: "When
ye therefore shall see the abomination of desolation,
spoken of by Daniel the prophet, stand in the holy
place (whoso readeth, let him understand), then let
them who be in Judaea flee to . the mountains," and
so on. The passage of Daniel referred to is this
(ix. 26 f.): "And the people of the prince that shall

46 PHILOLOGY OF THE GOSPELS.
come shall destroy the city and the sanctuary ; and
the end thereof shall be with a flood . . . and he
shall confirm the covenant with many for one week,
and in the midst of the week he shall cause the
sacrifice and the oblation to cease, and for the over
spreading of abominations he shall make it desolate,
even until the consummation," etc. Surely this is
more explicit and developed than anything in Luke,
and the whole difference between him and the
others amounts to this, that they make Christ refer
His hearers to an Old Testament prophecy, whilst
Luke makes Him give Himself the contents of that
same prophecy. For instance, in Luke He says
(xxi. 24) : 'lepovaaXrjfi. earai irarovfxevtj viro eQwov,
dypi ov irXripwOwcriv Kaipol eOvwv ; and in Matthew and
Mark He refers them to a passage where the Greek
words are : Kai 'ews rrjs crvvreXeias Kaipov crvvreXeia
SoOrjcrerai eirl rrjv epijfjwcriv. So the real difference is
not between undeveloped and developed, but between
veiled and open, which is quite another thing. We
may suppose that Christ really did speak in both ways,
first referring to Daniel's prophecy, and then declaring
it Himself; for it is self-evident, that the real speech
of Christ must have been much longer than we read
it now in any Gospel. Of these two parts, Matthew
and Mark give the first, leaving out the second, and
Luke gives the second, leaving out the first.

DATE OF ST. LUKE'S GOSPEL. 47
There remain these questions, why Luke should
have left out the first, and why the others the
second ? Now, Luke writing for Theophilus and
other Greeks, who were not very well acquainted
with the Old Testament, had good reason for leav
ing ou't the text of Daniel ; conversely, Matthew
and Mark being Jews and writing for Jews, had a
very good reason to leave out the declaration. That
declaration was extremely painful for them, and
would be so for their readers. We see from the
Acts, that the apostles and their disciples were daily
frequenting the temple, which had by no means
ceased to be a most holy place for them, and to
learn that that place and the whole city was to be
laid waste could not but pain them to the utmost.
Who would not fain cast a veil over things the
aspect of which gives pain ? And if we suppose,
as in my opinion we really must do, that the
Gospels were destined from the beginning to be read
in the weekly assemblies of the congregations, as
we have testimony to their having been read as
early as Justin's time (see Apolog. i 67), and that
they were intended to be a substitute for the former
sermons of the apostles, the contents of which they
reproduced, then it might also be for safety's sake,
and in order not to give unnecessary offence to the
Jews, that the open prophecy of the destruction of

48 PHILOLOGY OF THE GOSPELS.
the holy city was left out. Stephen had said
openly that the temple would be destroyed, and he
had become the first martyr.
There is still another objection commonly made to
Luke, viz., that he puts a distance of time between
the end of Jerusalem and that of the world, whilst
in Matthew and Mark these two events appear to
be closely connected. Matthew has (xxiv. 29):
"Immediately after the tribulation of those days shall
the sun be darkened," and so on, and Mark (xiii. 24),
a little less explicitly : " But in those days, after that
tribulation, the sun," etc. But in Luke (xxi. 24 {.)-,
the transition from the one prophecy to the other-
is made in this way : " Jerusalem shall be trodden
down of the Gentiles, until the times of the Gentiles1
be fulfilled. And there shall be signs in the sun,"
etc. Now, it is argued, that Matthew and Mark,
writing before or shortly after the destruction of
Jerusalem, still expected the end of the world to
be quite near at hand ; but Luke, writing later, had
lived to see some interval of time, and therefore took
care not to connect closely the two catastrophes. I
might answer, that this explanation of the difference >
is one out of many which are equally possible, the
more as Matthew and Mark themselves do not agree-
here in the words, and it is therefore impossible to
assert either that Christ really said "immediately,"

DATE OF ST. LUKE'S GOSPEL. 49
or that He said " in those days." But the most
important thing is, that Luke's insertion, " until
the times of the Gentiles be fulfilled," is nothing
more than an explanatory paraphrase of Daniel's -j
ecoy crvvreXelas Kaipov, and the passage of Daniel is
referred to in the other two. So there is in reality
much more agreement than at first there seemed to
be. As for the rest, we are fully at liberty to
trust Luke's account more than that of the other
two, who may have been partially influenced by
their being Jews. The destruction of the temple
and of the holy city was, in a Jew's eyes, next to the
end of the world (see Matt. xxiv. 3 ; Luke xxi. 7 /3) ;
and as they were to make a transition from the
one prophecy to the other, they made that transition
in conformance with their own ideas, but each of
them in a different way, so that we may suppose
their common source to have contained neither
" immediately " (a very common word in the two
first Gospels, much less so in Luke and John), nor
" in those days " (also very common in the three
first Gospels). But as for Luke, he can hardly be
said not to have partaken in the common belief of
the first age of Christians, that the return of our
Lord and the final judgment was near at hand, as
long as he says like the others, after the second
prophecy (xxi. 32) : " Verily 1 say unto you, This

50 PHILOLOGY OF THE GOSPELS.
generation shall not pass away till all be fulfilled."
It is true that there are here in Luke various
readings, which deserve at least a serious considera
tion. The Latin codex i has "it shall not pass
away from this generation until " etc., which makes
no great change of sense ; but the Latin e has
instead of " this generation," caelum istud, " this
heaven," and this reading leads directly on to the
form found in Marcion's Gospel : " Heaven and earth
shall not pass away, unless all be fulfilled. Earth
and heaven " (note the inverted order of words)
" shall pass away, but my words shall not pass
away." The heretic Marcion, indeed, does not
deserve implicit confidence ; but the Latin witness e
is above the suspicion of heretical tendencies, and
I am much inclined to regard Marcion's form as
the true one in Luke, whilst the other might very
easily come in by the way of assimilation to the
other Gospels, which has, as we shall see more fully
hereafter, much affected Luke's text. The sense
then will be this : " You may be quite sure that the
signs of which I have told you will come before
the end of things, and that the end will not come
abruptly, without those foregoing signs ; so I have
rightly instructed you to give heed to them, in
order that you may be well prepared for the final
catastrophe." The verses immediately preceding in

DATE OF ST. LUKE'S GOSPEL. 51
Luke (29-31) are common in substance to the
three synoptic Gospels ; but not quite so what comes
before them. Luke alone has verse 28:" And when
these things begin to come to pass, then look up,
and lift up your heads, for your redemption draweth
nigh," which is very well illustrated by the subsequent
parable (29 ff.), and stands in close relation with
the verse in question (32). So for internal as well
as for external reasons — since of two readings in
Luke, one of which exactly agrees with another
Gospel, whilst the other is a peculiar one, the former
is always liable to suspicion of assimilation — I do
prefer Marcion's reading of this verse and of the
next one, where the inversion " Earth and heaven "
after "Heaven and earth" (32) cannot easily be
deemed accidental.
But if this is true, you will say, that argument
of the theologians which I am combating will re
ceive additional strength. Be it so, the strength
will not even then be too great. Those epistles
of Paul, which are nearly contemporary to Luke's
Gospel, Colossians, Ephesians, Philippians, Timothy,
and Titus, while they are as full of references to
the day of the Lord as those preceding in time, do
not contain like them any clear expression of Paul's
hope to live to see that day. He seems to have
abandoned that hope which he formerly cherished,

52 PHILOLOGY OF THE GOSPELS.
and why might not Luke be of the same mind,
the more so if he did not find in the authors, whom
he follows in his Gospel, any distinct warrant for
these exaggerated hopes ? I do not think, indeed, that
Mark and Matthew deserve such exclusive confidence
in comparison with Luke, as many theologians are
inclined to give them. Just because Luke is the
later writer — compared with Mark, — he may well be
the more correct one. And here we may abandon
this subject, and pass to considerations of a different
order.

CHAPTER V.
IMPOETANCE AND METHOD OF TEXTUAL
CEITICISM IN THE NEW TESTAMENT.
No work of early literature which has been spared
by time and has survived to this day — with a few
exceptions not worth being mentioned here, — subsists
in the original writing of the author, but in the writing
of others, derived through a long series of successive
copies from that original writing. Moreover, we do not
ordinarily read it even in those written copies, but
in a printed book which has been made from the
copies, either directly or more often indirectly, being
the last in another long series of successive printed
editions. Nevertheless, we are used to regard that
printed book, say a book printed in 1897, as
substantially identical with the writing of the author.
Generally we may safely do so, for that identity
exists in the whole of the book and in most of its
particulars ; but if we come to care for any definite

54 PHILOLOGY OF THE GOSPELS.
particulars, and are to draw inferences from them,
the apparent identity may be, in a special case, a
gross deception, and the inferences will be utterly
wrong. Many and many a time, in cases con
cerning the Old or the New Testament, as well as
in other cases, that deception has been practised
upon readers by their book, and they have been
misled by it into drawing wrong conclusions and
adopting wrong opinions. We may indeed partly
provide against such deceptions, either by consulting
editions which come earlier in the series, or by
going back to the written manuscripts, and as far
back in the series of those manuscripts as we can.
But to speak in exact terms, no two different editions
are absolutely identical, nor are any two written
copies, nor is any edition or copy identical with the
original writing.
I was speaking just now of a well-known passage
in Luke : " Verily I say unto you, this generation
shall not pass away before all be fulfilled," that is
to say, before the final catastrophe of the world
come to pass. We know that in fact the generation
of which the Lord is speaking has passed away to
the last man, and many generations more, and still
the final catastrophe has not come. So it appears
as if our Lord had given us a false prediction. In
order to avoid this conclusion which must seriously

TEXTUAL CRITICISM IN THE NEW TESTAMENT. 55
shock us, we may have recourse to various alternatives.
We may, on the one hand, suspect that we have
given a wrong interpretation to the words (which
have been preserved in Greek). But upon con
sideration, this way will appear hardly practicable
in this case, although it may be so in others, the
instances of individual or even traditional misun
derstandings of a passage being very numerous. Or
we may question the accuracy of the account given
by the Gospel-writer. This way seems to be open in
every case ; but the more we get used to it, the more
we shall lose that confidence in the sacred writers,
which is so important for our Christian faith. There
is still one way more ; that of suspecting the accuracy
of the textual tradition, which has intervened between
the author and ourselves. This way proved to be
practical in this case of Luke's passage, at least for
me, and I (or we) have by its means quite got rid
of the difficulty, as far as it regards Luke; for
Matthew and Mark are still there, attesting the
false prediction. But the whole case has now
become somewhat different; for the prophecy is
now no more universally attested, and we may
correct the one extant tradition by means of the
other. It must be noted, however, that in this very
case even the theologian does not have easy access
to the sources of information. In ordinary editions

56 PHILOLOGY OF THE GOSPELS.
of the New Testament, he will find nothing at ' all,
and even in Tischendorf's large edition he will find
only the reading of Marcion, but not its partial
confirmation by the Latin codex e. He must needs
consult the separate edition of that codex made by
the same Tischendorf, that is to say, he must
consult editions of all the different Greek or Latin
or Syriac manuscripts, and seek for a different
reading, and be lucky if at last he finds it. But
how few can do this ?
We pass to another consideration which leads in
the same direction. The life of our Lord has of
course been the object of so-called scientific investi
gation, although, in reality, there are only some
particulars of it which come within the range of
such investigation, while the real import of it lies far
beyond. But now I am speaking just of these par
ticulars, one of which is, that our Lord, after He
was taken, was conducted either directly to the palace
of Caiaphas the high priest (so the synoptic Gospels),
or first to Annas the father-in-law of the high priest,
(so the Gospel of John). The latter testimony seems
to be more accurate and trustworthy ; but then, the
accuracy of the other narrative is rather seriously
affected, not because it omits an intermediate station
on the way, but because it refers the very important
transaction, the trial of our Lord and Peter's denial,

TEXTUAL CRITICISM IN THE NEW TESTAMENT. 57
to a wrong locality ; for John, as it seems, makes all
this take place in the house of Annas. I say, "as it
seems," for there is in reality no contradiction what
ever between the Gospels, but only a slight and quite
justifiable omission on the part of the three. Our
John is not identical with the real John, and it will
be quite clear even from a careful examination of the
text as it stands, that John can neither have meant
nor have written the commonly accepted account
with Annas' house as the scene of the trial. "They
led Him away to Annas first, for he was father-in-law
to Caiaphas, which was the high priest that same year.
Now Caiaphas was he, which gave counsel to the
Jews, that it was expedient that one man should die
for the people (see xi. 49 ff.). And Simon Peter
followed Jesus, and so did another disciple : that
disciple was known unto the high priest, and went
in with Jesus into the palace of the high priest"
(John xviii. 13 ff.). After having been distinctly
told that Caiaphas was the high priest that year,
and not Annas, we read that the other disciple went
in with Jesus into the palace of the high priest.
Whose palace, therefore ? Of course that of Caiaphas.
How has Jesus come there ? The writer, leaving
that serious omission unexplained and uncorrected,
goes on to speak not of Annas but of the high priest,
and to tell of Peter's being introduced into his palace

58 PHILOLOGY OF THE GOSPELS.
(ver. 1 8), and then of Peter's first denial, and next not
of the second one, but of the trial of Christ. After
that, he suddenly says (ver. 24): "Now Annas sent
Him bound to Caiaphas the high priest." Then he
returns to Peter, telling of his second and third de
nials, and from Peter again to Christ (ver. 2 8) :
" Then led they Jesus from Caiaphas into the hall of
judgment " (to Pilate). This narrative is so utterly
confused, that it is no wonder King James' trans
lators tried to correct it by interpretation, giving in
ver. 24 not "sent" but, "had sent." But the Greek
words give no warrant for this interpretation, and
even if it were possible, we could not withhold our
censure of the writer, as he would then have told a
simple story in the most awkward way. In reality,
the blame is to be cast upon the textual tradition and
not upon the author, and we may learn from this
quite evident case, that those written copies (not to
speak of editions), which we are accustomed to rely
upon, by no means deserve implicit trust. Which
copies, then, do deserve it ? No single copy at all,
but if anything the tradition taken as a whole,
with entire liberty to select in each individual case
that branch of the tradition for our guide which shall
seem to us to be in this case most trustworthy, even
if it is a heretical witness like Marcion. I deliber
ately say, "if anything," for there may be cases in

TEXTUAL CRITICISM IN THE NEW TESTAMENT. 59
which no branch 'of the extant tradition has preserved
the true reading. In this very case of John's ch. xviii.
until a few years ago there was no witness for it in
its totality, although there were a very few for a
part of the truth. But since Mrs. Agnes Smith-
Lewis discovered and published the Syriac manu
script of Mount Sinai, we read in this one witness 1
the following account : " They led Him away to
Annas first, the father-in-law to Caiaphas, who was
the high priest that same year (13). Now, Annas
sent Him bound to Caiaphas (24); Caiaphas was
he who gave counsel," etc. (14). Then comes the
mention of Simon Peter and of the other disciple,
and the statement that the latter went in with
Jesus into the palace (15). Next comes (19-23)
the story of the trial, and after that (16 ff.), "But
Peter stood at the door without," and then the
whole story of the three denials coherently, of
course without the repetition standing in our texts :
(18) "And Peter stood with them, and warmed
himself," and (25): "and Peter stood and warmed
himself." This is the narrative of a real author;
the other one is that of blundering scribes.1
*I may note that Prof. F. Spitta (Zur Geschichte und Litteratur
des Vrchristenthums, 1893, pp. 158 ff.) got at a part of the truth
in a purely conjectural way. He combines the two separated parts
of the story of Peter's denial, and establishes this order: vers. 12, 13,
19-23, 24, 14, 15-18, 256-28.

60 PHILOLOGY OF THE GOSPELS.
We shall speak later on of the condition in which
John's Gospel has come down to us ; for such gross
misplacements are far from being a general feature
in New Testament tradition. What I am now insisting
upon is the absolute necessity of textual criticism
for all studies connected with the ' New Testament.
It is not a certain amount of textual criticism,
which is required, but thorough and sound textual
criticism, as even with that we shall often remain
very far from the goal we want to attain.
Now it is a well-known fact that the number
of various readings collected from the manuscripts
of the New Testament by successive generations of
scholars is already, and has long been, astonishingly
great.1 Nevertheless many men in Germany, and in
England and elsewhere, are still engaged with praise
worthy zeal in increasing the number from fresh
sources. Nor can it be said that this zeal is altogether
misplaced, although it may be so in some cases. For
I do not think that every manuscript ought to be
thoroughly examined, because a great many of them
very soon show themselves to be quite worthless, and
it is even more strongly my opinion that not every
collation of manuscripts ought to be put into print.
On the other hand, we saw just now that there may still
be hidden in some corner of Europe, or Asia, or Africa,
1There are estimated now to be about 150,000.

TEXTUAL CRITICISM IN THE NEW TESTAMENT. 61
a veritable treasure of a manuscript, as that Sinai
manuscript was until recent years. Nor will those
treasures always present themselves in a very ancient
and musty shape. There is in London a Greek manu
script of the Gospels, written in the eleventh century —
that is to say comparatively very late — which was
bought at some auction in the year 1882; it was
examined by Dean Burgon first, then by W. H. Simcox,
who published his collation in a journal, and lastly
by H. C. Hoskier, who published his in a special book;
and the pains they took have been requited by 270
new readings, hitherto found in no manuscript. In
the passage of John's Gospel, of which we treated
above, one Greek manuscript, dating from the end
of the twelfth century, has preserved a part of the
true reading. We must therefore not look simply
to age, but to intrinsic value, which is to some extent
independent of age. For it is quite possible that an
ancient and valuable manuscript was carefully copied,
say in 1500, and afterwards perished ; the recent copy
in this case will possess by inheritance all the merits
of the lost archetype, side by side with which it would
of course be worthless.
This therefore is the first task, to collect all the
attainable materials for the restitution of the primitive
text. A second will be to sift the materials, for by
far the largest part of them is utterly worthless.

62 PHILOLOGY OF THE GOSPELS.
What, for instance, is the value of knowing, that in
a given passage a given manuscript has elire and not
elirev ? or that <5>apicraioi is spelt therein with et instead
of 1 ? For a philologist this may have some value,
in case the manuscript is very ancient ; the spelling
will teach him, that the i in this word is long. But
the apparatus criticus even of the largest edition ought
to be disencumbered of such minutiae, which may
be treated in a general way in the Prolegomena;
for the editor himself must know more than he gives
his readers. Moreover, as is nowadays the practice
in editions of Greek and Eoman profane writers, there
must be a sifting of the manuscripts, and all those
which have nothing good of their own must be dis
carded. Those endless lists of witnesses are not only
very cumbersome, but quite worthless, since the deci
sion on the correctness of a reading never depends on
numbers. In this way, we shall get a comparatively
short apparatus for most parts of the New Testament,
although there are some books where no sifting will
appreciably diminish the heap.
The condition, indeed, of the different writings
combined into the New Testament is far from being
equally good, or equally bad. There is, in the first
place, a wide difference between them as regards the
number of extant witnesses, which is exceedingly
large for the Gospels, but much smaller for the

TEXTUAL CRITICISM IN THE NEW TESTAMENT. 63
other parts, the Acts and Catholic Epistles (which
are usually contained in the same manuscripts), the
Pauline Epistles, the Apocalypse (for which the
number is smallest of all). Then the discrepancies
between the manuscripts are comparatively small in
number and importance in the case of all the
epistles, which appear to have been handed down
in as good and trustworthy form as any classical
author. Here therefore the third task of a critic,
that of discriminating and deciding between the
various readings, is no more difficult than in the
case of classical authors. I do not say that it is
easy, either for these parts of the New Testament
or for the classics ; the term easy may apply to a
great deal of patristic literature, which has come
down to us nearly as it was written by the authors.
But for the other writings contained in the New
Testament, viz., the Gospels, Acts, and Apocalypse,
the difficulties are much greater than they are, as
a rule, in the classics, and more especially in some
of the Gospels, where they reach an amount which
perhaps is nowhere attained in any other literature.
It is not only the great number of witnesses which
makes the difficulty, for that number is equally
great for Matthew and Luke, and nevertheless
textual criticism in Matthew is comparatively easy,
at least to one who has been occupied with the text

64 PHILOLOGY OF THE GOSPELS.
of Luke ; but it is chiefly the wide divergencies be
tween the witnesses which make one sometimes feel
burdened by a weight too great to be borne.
There is, indeed, a way to get rid of most of
those difficulties, that of discarding beforehand the
majority of the witnesses, not on the ground that
they are not independent (in which case the dis
carding is quite justified), but as untrustworthy. A
brief survey of the witnesses shows that, besides those
giving the common text as it was current in Byzantine
times, there are two main classes : one " Alexandrian,"
represented by the oldest uncials, Vaticanus B,
Sinaiticus K and Alexandrinus A, and by other
mss. besides; and one "Western," represented by the
old Latin versions and the Graeco-Latin Canta-
brigiensis D, etc., in very close agreement with
the old Syriac versions. Now, if we had to deal
only with the former class our task would be
extremely simplified ; because, although K and B and
A are far from giving an identical text, their
divergencies are small in comparison with those
given by the Western witnesses. The editors of
the New Testament in our century, from Lachmann
to Westcott and Hort, have indeed thought them
selves justified in relying chiefly upon Alexandrian
authority, and in neglecting more or less the Western
testimonies. The result would have been generally

TEXTUAL CRITICISM IN THE NEW TESTAMENT. 65
speaking worse if they had chosen the opposite way,
basing their text chiefly on Western authority and
neglecting Alexandrian. There is among the former
nothing to be compared with the latter for carefulness
and for beauty of writing, for instance, with the Vati
canus B, which is thought to have been written as
early as the fourth century, that is to say, two centuries
earlier than the Cantabrigiensis D. So if a general
option must be made between the two classes,
nobody will hesitate to give his suffrage to B and
to the Alexandrians. But such necessity is far from
being the case : we are only bound to individual
option in individual cases, and if that option may
be justly influenced by a general preference given
to the one class, we must always take heed not
to let that influence prevail over the individual con
siderations. In order to see how unjust it would
be wholly to neglect Western authority, let us
take the example of the Epistle to the Eomans.
There is no doubt that the Eoman Christians always
held in great honour that precious treasure, which
belonged in one sense to themselves exclusively:
and yet are we for that same epistle wholly to
disregard the testimony of Eoman or Western
witnesses, and to rely only upon Alexandria, where
the original of that epistle certainly never came ?
Moreover, let us compare the position of a judge

66 PHILOLOGY OF THE GOSPELS.
who has to decide an intricate and very complex
case, where there are many witnesses frequently
contradicting each other. For many of the parti
culars, let it be quite evident that some of the
witnesses have told a lie, and the others the
truth ; but in some few particulars let it be equally
evident that the truth is on the side of the former
witnesses, and the lie on that of the latter. Now,
would it not be quite absurd for the judge, as
regards the great bulk of particulars that might
still be in dispute, simply to adhere to the state
ments given by those witnesses who have been
convicted only of a few lies, and wholly to shut his
eyes to all other evidence ? On the contrary, he
' would say : All these witnesses are liars, nor does it
matter how often a witness has been convicted of a
lie, since everyone of them has been convicted of not
always telling the truth; so I must rely on the evi
dence given by the facts themselves, and not on the
witnesses. But, if the critic acts on these principles,
how will he be able to decide in everyone of the
innumerable cases put before him ? Is there always
an evidence given by the facts themselves? Certainly
not, and in these cases he will necessarily recur to
the evidence of those witnesses who seem to him
to be least untrustworthy, taking care, however, not
to admit such cases more than is necessary. If his

TEXTUAL CRITICISM IN THE NEW TESTAMENT. 67
general appreciation of the witnesses is right, he
may nevertheless decide wrongly in some cases, but
rightly in the majority of them, and he will have
fulfilled his task as well as is possible. It is an
ideal task for a critic, a task lying far above his
reach in the clouds, to restore the original form of
the writing throughout ; as things are, he will deserve
commendation, if he has approached that form even
by a small degree nearer than his predecessors.
There is still one more important question: Whether
conjectural emendation is justified for the New Tes
tament as well as for classical authors. Now, there
can hardly be any critic who will absolutely deny
that in a given case all manuscripts and versions
may present a wrong reading, or wrong readings,
and that the true one, lying not far behind the
corruption, may be found by conjecture. But as the
witnesses are for this book so very numerous, so early,
and so independent of each other, such cases will be
quite exceptional. Even if a right conjecture be
made by a critic, the probability is, that on closer
examination of the vast amount of existing external
evidence, that true reading will be detected among
that mass of testimonies, and the emendation will
prove to be an attested variant.
I may give a few instances from the Acts. In
xvi. 12 Philippi is described as being either irpwrrj

68 PHILOLOGY OF THE GOSPELS.
/j-eplSos rrjs MaKeSovlas iroXis (Vatic. B), or irpwrij
rrjs fxep. Mfflc. iroXis (Sin. K etc.), or irpwrij ms
fxep. rijs Mcoc. iroXis (majority of manuscripts). D
in this case gives an evident gloss : KecpaXtj rr/s
M.aKeS. iroXis, which is not even good Greek, but
coined after the Latin caput Macedonian The true
reading, viz., irpwrrjs /xeplSos rrjs Mar. iroXis (to
which B comes nearest of all, omitting but one cr),
was found by conjecture long ago by Pearce and
Clericus, who remembered that the Eomans, after the
conquest of Macedonia, had divided the country, as
Livy (45. 29) relates, into four districts, of which
the eastern part with Philippi and Amphipolis was
number one, the latter city not the former being
the capital of that district. I admitted irpwrijs into
my first edition of the Acts as a conjecture (which I
too had made independently) ; in my second edition,
which appeared a year later, it stands as an attested
reading. Whence did that attestation come ? I
found it first of all in a Provencal mediaeval version,
which gives en la primeira part de Macedonia ; next
the Paris Professor, Samuel Berger, called my atten
tion to a codex of the Latin Vulgate, with in prima
parte Mac.; lastly, Professor E. Nestle pointed out
that the oldest German versions also give that same
reading (zu dem ersten teyl zu M., or des ersten teyls
zu M.). But what kind of authority is all this ? A

TEXTUAL CRITICISM IN THE NEW TESTAMENT. 69
very slender one, you will say ; but when a corruption
has spread widely, as in this case, you must go to
the very remotest corner, if you wish to find the
true reading preserved.
Another case which I shall propose is different, in as
far that the conjecture has not really been made so far
as I know ; nevertheless, it might have been made by a
reflecting critic. In Acts vi. 9 the opponents of Stephen
are said to have been nves rwv e/c rtjs crwaywyrjs rrjs
Xeyofxevrjs AifSeprlvwv Kal TLvprjvalwv koi 'AXePavSpewv
Kai t<w diro JLiXiKias ko.1 'Acrias. Now, we are utterly
ignorant of a synagogue in Jerusalem bearing the
name of AifSeprlvwv, or the Freedmen, and there is
this additional difficulty, that the words Kal Kvpr/valwv
ko.1 'AXe^avSpewv seem to form a part of the same
appellation, although Cyrenians and Alexandrians
belong to definite towns, and freedmen existed every
where. I have tried in my commentary to disjoin
those words from AifHeprlvwv, and to bring them into
connection with ko.1 rwv dirb KiXiKias ko.1 'Acrias ; but
the right way lay quite in an opposite direction.
Mr. F. C. Conybeare and Mr. J. Eendel Harris directed
my attention, some time afterwards, to Armenian
versions of the Acts and of the Syriac commentaries
on that book, and in those sources I found the reading
Libyorum instead of Aifiepnvwv, a reading given
already by Tischendorf, but at the first disregarded

70 PHILOLOGY OF THE GOSPELS.
by me. Now, I saw at once that something like
AifSvwv would suit the context very well indeed,
as the Greek towns lying westwards from Cyrene
would come quite appropriately under that designation.
But can Aif3eprlvwv be a corruption of Aif3vwv ? Of
course not, nor does AtfSvwv seem to be the right
appellation for those Jews, as the Libyans were
nothing but barbarous tribes. But AifSvcrrivwv will
both suit the sense, designing them as inhabitants
of Libya, and come very near to the corrupted
AifSeprlvwv, there being but two letters different.
It is easy to establish that this form of the ad
jective from AlfSvs was a current one, from Catullus'
(60. 1) montibus Libystinis, and from the geographical
lexicon of Stephanus Byzantinus, and so on. This
therefore is the true reading, and the synagogue in
question bore the name of Ai(3vcrrlvwv km Kvprjvalwv
Kal 'AXe^avSpewv ; that is to say, of the African Jews
in the geographical order of their original dwelling-
places. In this case the true reading has been
preserved in the remote East, whilst in the former
it was from the remote West that the help came.
Of course, the fact that irpwrrj in one of these
passages, and AijSeprivwv in the other, is almost
universally attested, is not to be understood as
being the result of one great deliberate action, viz.,
of a revision of the text made at a definite time

TEXTUAL CRITICISM IN THE NEW TESTAMENT. 71
by definite men, and then imposed upon the whole
Christian Church. If such a revision had taken
place in the ancient Church, like those revisions
which have been made for instance at different
times in the English Church, we should certainly
hear of that fact from some of the numerous
ecclesiastical writers whose works have come down
to us. How, then, is that universal agreement in
blunders to be explained ? By the separate and
continuous agency of a common practice, which
consisted in collating copies with each other, and
when discrepancies were found, correcting the more
recent copy by the older one. The Jews, by
following this practice, have attained very nearly
absolute identity in their copies of the Old Testa
ment ; and the Christians, although they remained
far behind that extreme degree of accuracy and
scrupulosity, became nevertheless more and more
careful, lest their own sacred books might be
adulterated by corrupt readings. In the East this
care is to be seen much earlier than in the West,
where culture rapidly declined with the downfall of
the Eoman Empire; but as we are dealing with-
an originally Greek book, the Greek East comes
much more into our present consideration. The
care taken in revising and collating was by no
means confined to the sacred books, but extended

72 PHILOLOGY OF THE GOSPELS.
to classical writings as well. There can hardly be
a more serious mistake than when Byzantine scribes
are made responsible for the corruptions in classical
texts. Of course they committed blunders, partly
through ignorance and partly by inadvertence ; but
in most cases those blunders were removed by
correction, and we see from the remnants of ante-
Byzantine copies, which are coming to light in ever-
increasing number, that there is in most cases a
rather small difference between a copy, say of Plato,
of the fifteenth century and another of the fifth.
On the other hand, in the first centuries of our era
(and in those before, which do not come here within
our view), the work of revising and correcting was
in a great measure left to the buyer and owner of
a book, who might, if he liked, procure himself
another copy, and correct his own by means of that.
It is therefore not at all astonishing, that in the case
of the New Testament, as in other cases, the very
earliest copies may have been the most carelessly
made. Of course there was always a difference in the
matter of care or negligence between different contem
porary copies, and in a later age, when men had be
come more careful and when they saw the damage
already done, they recurred to those of the ancient
copies still extant, which seemed to them to be
the most careful and trustworthy, and used those

TEXTUAL CRITICISM IN THE NEW TESTAMENT. 73
copies as their standard for the correction of the
recent copies. Very likely their action was upon
the whole quite justified and extremely meritorious ;
but whilst a number of blunders were abolished in
that way, some few were at the same time per
petuated and became universal. For at the time of
which we speak, that from Constantine onwards, the
intercourse between the distant parts of the Church,
which had of course always existed, increased in
no small degree, that being the time when the
Oecumenical Council became an institution, and the
dogmatic controversies, which were to be decided by
the letter of the inspired books, required careful and
universally acknowledged copies of those books. So
we must suppose, that from that time on not many
fresh corruptions of the sacred text were allowed to
exist or to spread, and that the differences remaining
between the individual copies date from an earlier
time, and might be supported by the authority of
ancient manuscripts. But it is now time to close
these general considerations, and to treat more espe
cially of the textual condition of the Gospels.

CHAPTER VI.
TEXTUAL CONDITION OF THE GOSPELS.
MATTHEW— LUKE.
The union of the four acknowledged Gospels into
one Euangelium quadripertitum was effected, at
least ideally and theoretically, in the second century,
before Irenaeus, who tries to give even dogmatic
grounds for that number. It does not follow that
Irenaeus read them actually in a single volume,
the size of papyrus volumes or rolls being usually
not so large as to include more than one Gospel,
and papyrus rolls still prevailed in that age over
parchment books. Of course there had been a time
when the single Gospels had their quite separate
existence, and when Luke's Gospel, that is, Luke's
first book to Theophilus, still maintained its connection
with the Acts, that is Luke's second book to the
same Theophilus. A very curious trace of this
connection, as existing at least in the West for a

TEXTUAL CONDITION OF THE GOSPELS. 75
certain length of time, has been ' recently discovered.
The name of John has in Greek two spellings, one
with one N and another* with two. I do not doubt
that the former is the right one, and that from the
Hebrew Joohanan 'Iwdvrjs sprang up 'in the same
way as from Jonathan 'IwvdQrjs or 'IwvdOas, by
converting the an into an accusative termination,
and replacing the iV by an 2 for the nominative.
As there was another JV before the vowel, and not
a P or an I or JE, the termination in HS must
have seemed more regular than that in AS, just as
we have 'IwvdOr/s in Josephus. But in a later age
there crept in much irregularity in the doubling of
the liquid consonants; 'lwdvvrjs is parallel, for instance,
to AovkiXXios for Lucilius. Now the Vaticanus B,
than which there is no more trustworthy witness in
all matters of spelling, nearly always gives 'Iwdvrjs,
whilst the Cantabrigiensis D has both spellings, but
more frequently that with double V. The order of
the books contained in D is this : Matthew, John,
Luke, Mark, Acts. In the first two, the spelling with
double V has a very large predominance : twenty-four
instances to two in Matthew, and seventeen to seven
in John ; but in the third Gospel, that of Luke,
the writer of a sudden adopts the opposite principle,
writing 'Iwdvrjs twenty-seven times and 'Iwdwr/s but
once. This of itself would perhaps not be very

76 PHILOLOGY OF THE GOSPELS.
astonishing; but when he comes to Mark, he falls
back to his first spelling, there being in Mark
twenty-four instances of 'Iwdwr/s, and but two of
'Iwdvrjs. Last comes the second part of Luke, and
again he changes and spells 'Iwavr/s twenty-one times,
'Iwavvtjs but twice.1 That this definite inconsistency,
and at the same time ; consistency, cannot have been
effected by mere chance is quite evident, and there
is but one explanation for it. First, we must
acknowledge in the writer of D a degree of care
which hitherto seemed to be wholly alien from him,
the number of his blunders, more especially in the
spelling of the words, being exceedingly great; we
must now refer at least the greater part of these
blunders to the archetype from which the writer
copied. Again, if the archetype of D contained the
same books in like order, the same remark must
apply to the writer of that archetype. But perhaps
there was a different order in the archetype, with
Luke's Gospel coming last of the four. This too
is an order attested elsewhere, or to. speak more
accurately, there are different attested orders with
Luke as the last of all.2 But even in this case
the transition from one spelling to another requires
1 On this important discovery (made by E. Lippelt), see my edition
of St. Luke's Gospel, p. vi ff. Three of the number have been cor
rected by my friend Prof. J. Rendel Harris.
2 See E. Nestle's Introduction to the New Testament, p. 83 f .

TEXTUAL CONDITION OF THE GOSPELS. 77
a better explanation than that of chance, and the
supposition must necessarily be made, that the
archetype (or one of the archetypes) of D united
in itself parts coming from different sources, one
part of which consisted of the two books of Luke
written by one scribe. And from this we see that
there had been a time when there was a closer
connection between Luke's first and second parts,
than between Luke's Gospel and the other Gospels ;
a fact, besides, which must be credible independently
of all actual traces or testimonies.
In the first stage of its existence, therefore, every
Gospel was separate from other Gospels, and it is
self-evident that during that stage it cannot have
remained quite unaltered and unadulterated. That
very first age was also, as we have seen, that of
least care in transcribing, and there was besides
another reason for corruptions of a special kind.
Let us take the case of an individual Christian, or
of a single Christian community, possessing but one
copy of one Gospel. Now he or it might come into
contact with another individual or community in
possession of a different Gospel. Nothing would be
more natural than the wish to compare the two
Gospels, and to supplement, or even to correct, one
from the other. Let Luke's Gospel be the one, and
Matthew's the other: the former bore the name

78 PHILOLOGY OF THE GOSPELS.
of a later disciple, whose master himself, viz., Paul,
had not belonged to the eye-witnesses, and the other
bore the name of an apostle, one of the eye-witnesses.
Naturally Matthew must appear to be more trust
worthy than Luke, and accordingly the copy of Luke
underwent corrections from Matthew, in many of
the numerous cases where they differed from each
other; for instance, in the case of the Lord's Prayer.
On comparison with Mark, too, Luke might seem the
less trustworthy witness, because Mark's Gospel was
supposed to go back in substance to Peter himself,
that is to the chief eye-witness. But of course the
possessors of Matthew and Mark might also note
down in their copies the different account or
wording given by Luke. The final result of this
practice of comparing and correcting or supplementing
the Gospels (including even that of John, although
that of course had fewest points of similarity) is stated
by Jerome (in the preface to his Latin translation
of the Gospels) in the following way : "A great mass
of error has crept into our copies (the Latin copies,
but the case was of course quite the same in the
Greek originals of those copies), because the scribes
have supplied in one evangelist what another one
had more than he, deeming that to be wrongly wanting
in the former. Or, where the same sense was ren
dered in the different Gospels by different words,

TEXTUAL CONDITION OF THE GOSPELS. 79
the man who had read one evangelist first of the
four, thought that he must correct the others' by
that model. So it came to pass that in our copies
everything was mixed together, and that there are
found in Mark many things which belong to Matthew
or Luke, and again in Matthew many belonging to
John or Mark, and likewise in the rest many belonging
to some other of the four." Now, that this mixture
exists in the manuscripts (either Greek or Latin or
Syriac) which have come down to us may be seen
by a look into the apparatus of any large edition.
It would, moreover, be rather incautious to suppose,
that the cases of mixture are to be found in those
editions only under the text and never in the text.
Such implicit trust is not to be placed in any editor>
much less in any scribe ; for the idea of the scribes
having possessed the faculty of discerning the genuine
Luke or Mark from the corruptions coming from other
Gospels, would remind us of the legendary Indian
bird, which, out of a pot of milk and water, has the
miraculous gift of drinking the milk and leaving the
water. The case might be different if these inter
polations had sprung up about the date of our most
ancient manuscripts ; but since their origin is older by
centuries, it is next to impossible that any of our manu
scripts entirely escaped their influence. Besides this
kind of corruption of the text, which is the most

80 PHILOLOGY OF THE GOSPELS.
general, there must of course have come in, during the
same first period, a great many of the ordinary blunders,
and although many of these offered themselves to the
eye of an attentive reader, who might correct them at
once, others of these too remained unobserved. Upon
reflection, we must at once feel the absurdity of the
supposition (which is nevertheless frequently and
even generally made), that ancient readers and the
scribes perused the manuscript which they were read
ing or copying with the careful eye of a thoroughly
trained modern critic, attending not only to cases of
sheer nonsense, but also to those of awkward or sin
gular expression, and trying by correction to make
their text as smooth and pleasing as possible.
Coming now more closely to the object of our
examination, we may take first into our hands the
Gospel according to St. Matthew. A general con
sideration, which neglects some few exceptions,
shows us that the condition of the text is not
much worse than it is in the epistles. Editors have
to give generally for Matthew no larger apparatus
than for Paul, although the number of different
witnesses is far greater than for the latter. But even
those witnesses, which elsewhere exhibit a most peculiar
character, seem to forego that character as far as
Matthew is concerned. I may take for illustration
the Latin fragments of the two first Gospels, which

TEXTUAL CONDITION OF THE GOSPELS. 81
hail from Bobbio and are designed by k. Mark
according to k is very different from Mark according
to B ; but. Matthew according to k, and Matthew
according to B, are very similar. One exception,
however, must be made for the Syriac palimpsest from
Mount Sinai; there indeed we find in this Gospel
also not a few cases of unwarranted shortening,
which we may explain by the special condition and
origin of this codex, but must disregard here, as
leading far away from our theme. But as the other
witnesses, including D and Latin versions, exhibit
nearly the same Matthew, we may quite safely infer
that Western scribes also, and that too in the Gospels,
were by no means so utterly careless and lax as they
are ' deemed to be by some eminent modern critics.
Scribes who preserved 'Iwavr/s wherever they found
it (with a very few exceptions), and wherever they
found 'lwdvvrjs also preserved it, cannot possibly be
termed lax, but rather, if you like, unreflecting or
even thoughtless ; and so, as a matter of fact, scribes
generally were. Western scribes deserve therefore
some degree of confidence, as having rendered in sin
cerity of mind whatever they found or — by distraction
of mind — imagined that they found.
Nevertheless there are in Matthew some interpo
lations in the manuscripts (and consequently some
fewer in the printed texts), either of the kind

82 PHILOLOGY OF THE GOSPELS.
described above or of another kind. In viii. 5 we
read : " And when Jesus was entered into Capernaum,
there came unto Him a centurion," etc. But the
Sinai Syriac MS. and k give only : " After that
(fjerd ravra Se) there came unto Him," etc., omitting
the locality. Now, when we compare Luke (vii. 1),
where the same story is told in a partly different
way, we find " He entered into Capernaum," and
may justly suppose that this is one of those cases
described by Jerome, where one evangelist having
something less than another, the former has been
supplied from the latter. Moreover, instead of
eKarovrdpyr/s, centurio, the same Syriac witness alone
has yiXlapyos (the very Greek word in Syriac letters),
that is, tribunus (something like " colonel "), and I
am much inclined to regard centurio too as a wrong
assimilation, which has extended even to k. There
is a corresponding discrepancy between Matthew
(that is the pure text of Matthew) and Luke in the
further course of this narrative. According to Luke
(ver. 8) the centurio says of himself that he is " a
man set under authority " (dvOpwiros virb exovaiav
racrcr6fj.evos), as befitted a subordinate officer; but in
Matthew that " set " (racrcr6fj.evos) is an interpolation
not found in the majority of mss., and " under
authority" is to be combined with the following
words " having soldiers under me " (e^ow vir' e/xavrbv

TEXTUAL CONDITION OF THE GOSPELS. 83
crrpariwras), as befits an officer in command ; if indeed
we are not to go even further, and to read with the
Syr. Sin., " having authority and soldiers under me "
(exovaiav eywv Kal crrpariwras vir' efiavrov), which
is better Greek and quite unambiguous.1 i
But you will say that if we proceed in this way,
always giving preference to that reading which varies
from the narrative of another Gospel, we are augmenting
difficulties for the establishment of harmony between
the four Gospels. There was a time when theologians
defended the absolute infallibility of the inspired
writers, not only in matters of faith, but also in
matters of fact ; but afterwards there came a time
when at least a large part of those of my country
tried to establish the absolute fallibility of the same
writers. If we may credit Professor Harnack, that
time too has passed away ; but it cannot be a question
of returning to the former dogma, nor am I writing
for such over-sensitive readers as are apt to take
offence at any shade of alleged discrepancy between
two Gospels. Wherever there are discrepancies, the
reader may choose for himself that variant of the
story which will seem to him most just and most
expressive. Another kind of apparent interpolation, but from
1 An additional interpolation from Luke vii. 10 is found in some
mss. (among which the first hand of the Sinaitic N) in ver. 13.

84 PHILOLOGY OF THE GOSPELS.
Matthew himself, may be seen in the following
instance. In iii. 2 we read that John the Baptist
was preaching : " Eepent ye : for the kingdom of
heaven is at hand"; and in iv. 17 that Jesus preached
exactly in the same words. But the " baptism of
repentance" (Acts xiii. 24; xix. 4) more properly
(I do not say exclusively) belongs to John, and the
case being that not only k and the Syriac mss. of
Cureton and of Mrs. Lewis, but also, as it seems,
Eusebius (Demonstr. Evang., p. 438, cod. P), omit the
" repent ye " (fieravoelre) in the words of Christ, I
think we must read with these witnesses in Matthew
only : " to say that (oti) the kingdom of heaven is
at hand." The actual words of Christ may still
be questioned, because Mark too (i. 15) makes Him
say : " Eepent ye, and believe the gospel " ; but that
problem is clearly one of another order.1
While we had, in these two instances, cases of
what Dr. Hort calls " Western non-interpolation," 2
that is where some Western witnesses have remained
free from an interpolation found in the Eastern
authorities : there are, of course, other instances where
1 Similarly in xx. 16 the words, "For many be called, but few
chosen," are inserted in the majority of MSS., words which are
in their place in xxii. 14.
2 See also xxvii. 49, where our best mss. insert a sentence borrowed
from John xix. 34. Nestle, Einfuhrung in das Gr. Neue Testament,
p. 109.

TEXTUAL CONDITION OF THE GOSPELS. 85
the interpolation has infected only Western (and
Syriac) texts. The most conspicuous case in Matthew
is after xx. 28. There D and Latin and Syriac
witnesses insert a rather long passage, the greater
part of which is in sense very similar to Luke xiv.
8 ff., but wholly different in the words ; moreover,
many of the words, or combinations of the words, are
such as are never elsewhere met with in the whole New
Testament, viz., SeiirvoKXrjrwp, or en dvw, eri Karw, in
the sense ofthe comparatives dvwrepov (or -co), Karwrepw
Unquestionably this passage is alien to Matthew, and
must have been interpolated in a very early period
from an unknown source. There are in Matthew two
cases more of the same kind. In iii. 1 5 (baptism of
Christ) two old Latin witnesses insert the words : et
cum baptizaretur lumen ingens fulsit de aqua, ita ut
timerent omnes qui advenerant. Now, this is an apo
cryphal tradition, which is known and acknowledged
by Justin, and which, according to Epiphanius, was
found in the Gospel of the Ebionites. The other
passage has found its way into the textus receptus,
but the best Alexandrian witnesses and the testimony
of Jerome are against it : xvi. 2 ff., " When it is
evening, ye say, It will be fair weather : for the sky
is red. And in the morning, It will be foul weather
to-day: for the sky is red and lowering. Ye can
discern the face of the sky; but can ye not discern

86 PHILOLOGY OF THE GOSPELS.
the signs of the times?" The tradition in itself may
well be authentic, and it finds a close parallel in Luke
xii. 54 ff. ; but it does not stand here in its right
place, since the demand of the Pharisees, that Christ
would show them a sign from heaven (ver. 1), finds its
answer only in ver. 4 : " A wicked and adulterous
generation seeketh for a sign," etc., as in the corre
sponding narrative of Mark (viii. 11 ff.), whilst in
Luke the introduction is quite different.
A peculiar interest attaches to some passages of
the first chapter, where the various readings are
of great dogmatic importance. The close of the
genealogy of Christ (in ver. 16): "And Jacob begat
Joseph the husband of Mary, of whom was born
Jesus, who is called the Christ," is by two Greek
minuscules, and by Latin witnesses (among which
k and the Latin part of D, the Greek being lost),
given in this way : " And Jacob begat Joseph, to
whom being espoused (fxvrjcrrevOeicra, desponsata) the
Virgin Mary bore," etc. This, indeed, must be a
wilful alteration from dogmatic reasons : evidently
the expression " the husband (rbv dvSpa) of Mary "
had been shocking for those readers who believed
Mary's virginity to have been perpetual. The
reading of the Sinaitic Syriac comes very near to
this : " And Joseph, to whom was espoused the
Virgin Mary, begat Jesus," etc, We must bear in

TEXTUAL CONDITION OF THE GOSPELS. 87
mind that for " begat," and " was born," or " bore,"
the Greek verb used by Matthew (also according
to those two minuscules) is the same, and Syriac
too makes use of the same verb for both meanings,
only in different voices; so the difference in the reading
of the Syriac witness, especially when we retranslate it
into Greek, amounts only to the repetition of Joseph,
and to the changing of the participle " espoused," into
" was espoused." Materially this makes a thorough
alteration ; but the Syriac scribe is not to be
supposed to have known what he did,1 since he
contradicts himself by adding " the Virgin," and by
giving the detailed narrative in ver. 18 ff. quite in
the ordinary way, with one exception in ver. 25,
where again Latin testimony is concurrent. Instead
of ovk eylvwcTKev avrrjv 'ews oS ereKev rov viov avrris
rbv irpwroroKov, the Latin k has only : " And she
brought forth a son " ; and so the Sinaitic Syriac :
" And she bore to him a son," the word " first-born "
being omitted also by the two best Greek MSS. (B
and n), and by other witnesses besides. The pro
noun, "to him," which is not found in the Latin
codex, may seem to point in the same direction as
the Syriac reading in ver. 1 6 ; but the addition of a
1 Here I quite agree with Professor Zahn and with F. Graef e, who
treated this critical question very carefully in Theolog. Studien u.
Kritiken, 1898, p. 124 f. The Curetonian Syriac MS. has "was
espoused," but inserts " who (bore) " after " Mary."

88 PHILOLOGY OF THE GOSPELS.
pronoun to a verb is a very slight thing, and the
main importance of the various reading lies in the
omission of the ovk eyivwcrKev avrr/v ews ov, that is,
it lies quite in. another direction. Ovk eyivwcrKev
avrrjv ecoy, etc. It follows that afterwards eyivwcrKev
avrrjv, and the boast of perpetual virginity is de
stroyed for Mary. The " first-born " too was very
offensive to readers who adhered to such tenets,
although Jerome in his commentary on Matthew,
while acknowledging the reading, strongly denies that
the consequence of brothers to Christ having existed
is to be drawn from it. But as to the expression
" first-born," it has been not without good reason
argued that it may have crept in. from Luke ii. 7,
where it is universally attested. I am well aware
that the ovk eyivwcrKev, etc., too,, may seem to. have
been inserted, as there was possibly some reader
who was anxiously contending for the superhuman
generation of Christ. As a matter of fact there
are some passages in Luke where that tendency is
manifest. Luke was not allowed by some Western
readers to say (ii. 41): "And His parents went.
to Jerusalem," but they wrote : " And Joseph and
Mary," etc. ; or (ver. 48) : " Thy father and I have
sought Thee sorrowing," where they left nothing but
" we sought Thee sorrowing." But in Matthew, as
the variants in vers. 16 and 25 appear to stand

TEXTUAL CONDITION OF THE GOSPELS. 89
in close relation to each other, and as the tendency
of that in ver. 16 is unmistakable, the words ovk
eyivwcrKev K.r.X., are in my opinion to be retained
there. At any rate we clearly see that there have
been very ancient readers who did not shrink from
wilful alterations of the sacred text, if it did not
suit their dogmatic convictions, or if it might give
support to opposite tenets. Their reasoning was
either simply this : It is impossible that an inspired
writer should have written this which is incompatible
with truth ; or else : It is unsafe to allow people to
read these words as they stand, since somebody may
understand them in a wrong sense, which will lead
him to destruction. We shall find more of this kind
of alteration or mutilation in other Gospels. ~>
We may pass now from Matthew to the other
evangelists, whose textual condition is indeed very
different, both in relation to Matthew and to each
other. Setting aside for the present all cases of
mutual assimilation, we may state their textual
condition as follows. In Mark the words and
expressions are very frequently liable to doubt,
because of discord among the witnesses; there are
besides some cases of addition or omission, which
are not to be explained in any ordinary way.
John has some evident interpolations, e.g. that in
v, 3 f. (not to speak of the section from vii. 53 to

90 PHILOLOGY OF THE GOSPELS.
viii. 1 1, which case is quite peculiar) ; but besides
these a great many words and clauses are not
universally attested ; words and clauses, however,
which generally are without any importance for the
sense. There remains Luke, and in his writings,
including Acts, the discrepancies among the witnesses
become by far the greatest of all. It is true that
if we were allowed wholly to neglect Western
witnesses, the remaining discrepancies would not be
over-great, and we may allow ourselves to neglect
that part of the discrepant evidence for the present,
I in order to consider first what remains.
I do not intend to speak of minute matters,
but shall come at once to some very conspicuous
and well-known cases. In ix. 54 ff. we read in
ordinary texts : " And when His disciples James and
John saw this, they said, Lord, wilt Thou that we
command fire to come down from heaven, and
consume them, even as Elias did ? But He turned,
and rebuked them, and said, Ye know not what
manner of spirit ye are of" (or rather: Do you not
know, etc. ; the second " you " bears an emphasis,
as the pronoun is there in the Greek text). "For
the Son of man is not come to destroy men's lives,
but to save them." Now the words in italics are
omitted by a great number of witnesses, varying
however for the separate parts.

TEXTUAL CONDITION OF THE GOSPELS. 91
Again, in the narrative of the prayer in Geth
semane (xxii. 39 ff.), there is the same discrepancy
among the witnesses in giving or omitting these
verses (43 f.) : " And there appeared an angel unto
Him from heaven, strengthening Him. And being in
an agony He prayed more earnestly : and His sweat
was as it were great drops of blood falling down to
the ground."
Lastly, the words in xxiii 34: "Then said Jesus,
Father, forgive them ; for they know not what they
do," are far from being unanimously attested, there
being among the witnesses omitting them both the
Vatic. B and the Cantabrig. D. ^
So in these three cases, as it seems at first,
we have this alternative : either the clauses are
genuine, and we must seek for an explanation of
their omission, or they are spurious, and we must
seek for the reason of their insertion. But no
less a person than Bishop Lightfoot denies this
alternative. He says:1 "It seems impossible to believe
that these incidents are other than authentic, and
the solution will suggest itself that the evangelist
himself may have issued two separate editions.
This conjecture will be confirmed by observing that
1 See Lightfoot, Fresh Revision of English New Testament, p. 32
(3rd edition). I knew this book first from Dr. Salmon's quoting
it in his excellent Thoughts on the Textual Criticism of the New
Testament, p. 135,

92 PHILOLOGY OF THE GOSPELS.
in the second treatise of St. Luke (the Acts)
similar traces of two editions are seen, e.g. Acts
xxviii 16, 29." Now the question suddenly gets a
fresh and quite unexpected interest. Can it be that
we possess, either wholly or partly, two separate
editions of Luke's works ? That would indeed be a
very important enrichment of our store of authentic
knowledge. But three (or five) passages are a scanty
material for establishing a fact of this importance,
unless there are actually many more in the Acts
than Lightfoot expressly mentions. As a matter of
fact, there are many more in the Acts, and I myself
have written a great -deal on them ; but now we are
dealing with three passages of Luke, for which the
theory of two separate editions of Luke's works may
perhaps be the ultima ratio, as soon as we clearly
see that any other solution is impossible. But that
has not yet been proved. Moreover, the instances to
be taken from the Acts are dissimilar in this point,-
that the Cantabrigiensis D, in the case of the Acts,
enables us to propound such a theory, by constantly
presenting an enlarged form of the book ; while as
regards the Gospel, D seems to support the omissions
as much as witness against' them, agreeing with
Alexandrian authority in the third passage and in
V one-third of the first.
There is yet another marked point of difference.

TEXTUAL CONDITION OF THE GOSPELS. 93
Lightfoot very justly says that those passages of the
Acts " are entirely free from suspicion on the
ground that they were inserted to serve any pur
pose, doctrinal or devotional," and he might have
added, that there cannot be any more suspicion . that
they have been omitted for any such purpose. He
could not have said as much regarding the passages
of the Gospel. Not inserted for dogmatic purposes,
very well ; but not omitted for such purposes ? On
the contrary, the first passage might seem to sup
port Marcionite heresy, the second to ascribe to
Christ something unworthy of His divine Majesty,
and the third to involve a contradiction with other
verses. Let me explain the last case first ; the
second will not even require elucidation. When
Christ prays, of course what He prays is fulfilled.
Now He has prayed for those who hrought Him to
the Cross, that their sin might be pardoned ; con
sequently this sin has been pardoned, and if
pardoned, has not been avenged. Yet the same
Christ has said (ver. 28): "Daughters of Jerusalem,
weep for yourselves, and for your children. For,
behold, the days are coming," etc. Whether this
argument holds good, I need hot discuss, the less so
as I deny the minor premiss : Christ has not prayed
for the Sanhedrin, but for the soldiers who nailed
Him. to the Cross; for it is of these that the

94 PHILOLOGY OF THE GOSPELS.
evangelist is speaking both in the foregoing and in
the following words, and not of the high priests.
But I say thus much : the passage might be under
stood in this way, and so this argument might be
used. Again, in the first passage the disciples appeal to
the example given by Elias, and Christ rebukes
them by indicating that they are of another spirit
than that of the prophet, and consequently must not
act in the same way. Then there is a difference
of spirit between the Old and the New Testament,
and Marcionite heresy, maintaining a different God
for the former, might be justified from these verses.
In order to prevent such dangers, it seemed to some
orthodox man better to strike out that part of the
narrative which might give offence.
I If we accept this explanation, retaining of course
the questioned verses in our text, then, you may
say, Lightfoot's theory falls to the ground. But this
conclusion I deny : there may well be other support
for it. If the question of the Acts is to be decided
in Lightfoot's sense, that decision will contain for
the Gospel too a kind of presumption, compelling us
to search the state of things there very thoroughly
and carefully. Then we shall see that there is still
D with its associates, presenting an innumerable
host of- various readings, of omissions, of additions,

TEXTUAL CONDITION OF THE GOSPELS. 95
and claiming (which cannot in equity be denied),
that these variants be either simply accepted, or
explained in a satisfactory way, including Lightfoot's
theory. As a matter of fact, that theory ought not
to bear Lightfoot's name, although it was he who
propounded it first in England ; or mine own,
although I have written perhaps an hundred times
as much in support of it as Lightfoot has ; but that
of Joannes Clericus, who was by far the first pro-
pounder. But in order to get a better insight into
the genesis of this theory, we may go still further
back, and consider the origin and fate of the Codex
D (or Cantabrigiensis, or Bezae) more accurately
than we have hitherto done.

CHAPTER VII.
THE QUESTION OF THE DOUBLE TEXT IN
ST. LUKE'S GOSPEL AND IN THE ACTS.
The Graeco-Latin Codex ' D of the Gospels and Acts
may be justly termed the greatest literary treasure
which the University of Cambridge possesses. It
came into its possession as early as the year 1581,
being presented to the University by the celebrated
French reformer, Theodore de Beze, or Beza, who had
acquired it in 1562. Until then it had lain neglected
and ill-treated in a monastery in Lyons, which bore
the name of St. Irenaeus ; at last somebody had
carried it off from there and given or sold it to
Beza, in about the same mutilated condition which
it exhibits at present. So much Beza himself tells
us ; about the former fate of the MS. there is much
uncertainty,1 there being a trace of its having been
brought as early as 1546 by a Bishop of Clermont
1 See Scrivener's Introd. p. viii ff.

DOUBLE TEXT IN ST. LUKE AND IN THE ACTS. 97
(which is not far from Lyons) to the council of Trent
(Tridentum), when he proved by it that a certain
Latin reading in St. John's Gospel had for itself a
very old Greek authority; now, as a matter of fact,
D is the only Greek MS. we know of which gives the
passage in the form quoted by the Bishop.1 About
the same time Eobert Stephanus got " from Italy,"
as he says (it might seem from Trent), a collation
of a ms. which he in his edition designs by (3, and
which proves from the various readings given by him
to be identical with D. Between the sixteenth
century and the sixth, when the codex appears to
have been written, there is a great gap ; nevertheless
we may suppose it to belong originally to some place
in southern France, where, at any rate, the tradition
of a text of the Acts similar to that in D has con
tinued down to a very late age. Now, why did Beza
give it to the University of Cambridge ? This he
himself declared in the epistle accompanying his
present, stating that the precious codex would be
best preserved there, but "preserved," as he says,
" not published." Why not published too ? Because
he was afraid that the very numerous and important
variants, which the codex presented " especially in
1 John xxi. 22, iav airbv dika p.iveiv oSras Sus fyxo/Mi. The oStus
is given by D alone among Greek mss. ; it might seem to confirm the
more general Latin rendering sic (instead of si) volo eum manere, etc.
G

98 PHILOLOGY OF THE GOSPELS.
the Gospel according to St. Luke," might give offence
to persons who required for their faith a quite estab
lished and, so to say, infallible text of the Holy
Scriptures. Nevertheless, Beza himself has published
in his editions of the New Testament at least some
of the variants, and others came into general notice
by-and-by, until as late as 1793 the first edition
of D was made by Thorn. Kipling, in a careful and
sumptuous way for those times, although the later
edition by F. G. Scrivener (1864), if less sumptuous,
is more accurate and trustworthy, and now we are
to have an edition which is accurate in the utmost
degree, the codex being "lucis auxAlio in lucem emissus."
But before any of these editions, those variants in
Luke's works which had become known from D sug
gested the idea that Luke must have edited them
twice. Joannes Clericus (Jean Leclerc), born in Geneva
in 1657, but living in Holland and belonging to
the Dutch school of philologists contemporary with
the great Bentley, is said by the German theologian
Semler to have published, but under the assumed
name of Critobulus Hierapolitanus, the opinion that
Luke had made two editions of the Acts.1 It seems
1 See Semler, /. J. Wetstenii libelli ad crisin atque interpret. N. T.
(Halle, 1766), p. 8 : Clericus iam olim, nomine Critobuli Hierapolitani
usus, fere fuit in hac sententia, Lucam bis edidisse Actus. Nee
Hemsterhusius alienus fuit ab hac sententia, forte bis Apostolos
quaedam scripsisse.

DOUBLE TEXT IN ST. LUKE AND IN THE ACTS. 99
that the opinion, which he did not venture to give
under his own name, would have been liable in those
times to the charge of gross heresy, and this is indeed
very likely, since the Holy Ghost, who was believed
to be the real author of Holy Scripture, could not
possibly be supposed to have corrected Himself. But
for this overstrained dogma, I should think that more
scholars than Clericus and Hemsterhusius (whom
Semler relates to have been of a similar opinion,
viz., that the apostles had written some parts of
their work twice) must have hit on that very simple
solution. ,—
In cases of wide discordance among witnesses,
who seem to be upon the whole equally trustworthy,
it may very easily be supposed that the thing itself
which their evidence relates to is actually not the
same, and if this rule is followed in New Testament
criticism, we come at once to the hypothesis of morej
than one original text. In our day the charge of
heresy on account of such an opinion no longer
threatens us, though there may be other impediments
against the acceptance of a two-edition theory. Ac
cordingly, old Luke appears like one of ourselves,
who are used to publish the same book as many
times as possible, and if we are diligent, to intro
duce into it eacf^time more or less correction of,
style and matter. But what a distance is there

100 PHILOLOGY OF THE GOSPELS.
between us and Luke ! Of course there is a
world-wide distance between Luke and a modern
book of scientific theology ; yet in this respect
the distance between Luke's time and our own is
not so very great, since there exists in both
the notion of " edition " (in Greek ckSoctis), and the
possibility of more than one edition of one book,
that possibility having become actual in not a few
known and attested cases in antiquity. But if Luke
comes into question, we must indeed acknowledge
for him a somewhat wider distance. He is not to
be supposed to have given his book to a publisher,
as Cicero did to Atticus, in order that the publisher
might make, by means of his copyists, the required
number of copies, and send them to different
parts of the world. Nor is the term " edition,"
properly speaking, to be applied to the different
forms of Luke's work ; we ought to speak rather of
copies, which were privately made and privately
given to friends, and from which other copies would
be made for the use of friends of those friends, and
so on. So there is nothing too modern attributed to
the evangelist, but only a thing common in all ages
in which writing has been practised, and literary
works produced. One copy of the Gospel was that
sent to Theophilus ; but when Luke afterwards came
to Eome, he would of course be requested by the

DOUBLE TEXT IN ST. LUKE AND IN THE ACTS. 101
Eoman Christians, who heard of his having written
a Gospel, to give them, too, a copy of it, and he
would write out that copy in the course of perhaps
a month and give it them. That fresh copy would
not exactly agree with the former, for the writer
was entirely at liberty to shorten where he liked,
or to insert what he thought suitable for these
new readers, or to make improvements in style, or
what else he chose to do ; and he would naturally
desire to do something of that kind, as we usually
do, when we write the same essay a second time.
Likewise the Acts, which were written in Eome,
would be given to the Eomans first in one copy,
and afterwards sent or brought to Theophilus in
another different copy. That the copies in every
case were again copied, and in that way the work
became generally known in two different forms, is
obvious. It cannot, therefore, be denied that this
hypothesis of two editions or two original copies
sufficiently and amply explains the fact that there
are even now existing two different forms of each
of Luke's books. There is another fact which is
explained at the same time, namely, that Luke's
case, as I have stated it, is unique among New
Testament writers. Each of the epistles was written
once and not twice, and the Gospels of Matthew,
Mark, and John, however obscure their origin may

102 PHILOLOGY OF THE GOSPELS.
be, are not to be supposed to have been originally
written in more than one place, nor for more than
one definite circle of readers. Or, if this assertion
seems to be unwarranted, at least I may say that
only in Luke's case we can clearly see the reasons
for his writing different copies, since he had first
been a member of the Church of Antioch, with which
he of course continued to maintain relations, and
at a later time became a member of that of Eome.
Of course all these a, priori considerations are not
conclusive : we may admit that possibly there have
been different copies of Luke's work, which were
equally authentic, and consequently different texts
with equal authority ; but that the possibility has
come to actual existence, and that there are still
different texts of that description, remains to be
proved. Now, I have already intimated that the con
dition of the Gospel, and that of the Acts, similar
as they are, are still different in one essential point.
The similarity, to speak of that first, extends to
this, that of the two different copies of each one has
been written (as we supposed) for the Eomans, and
was consequently propagated chiefly in the West, and
the other was written for Theophilus, whom we
suppose to have lived in the East, and propagated
» there. But here at once this difference comes in ("which

DOUBLE TEXT IN ST. LUKE AND IN THE ACTS. 103
was found out, as I think, by Professor Nestle,
earlier than by myself), that the Eomans got the
later copy of the Gospel and the earlier copy of the
Acts, according to the dates of the two books, which
we have established in former chapters, and to the
course of the author's life. Now, which of the two
copies differing in date would be more correct in
style ? The latter. Which more prolix (which is
akin to a lack of correctness) ? The former. I at
least, from my own books and writings, should infer
this much, and I think it is in accordance with
general experience. As for the absolute extent of
the work, the later edition of a modern work is usually
the more extensive one, for which one reason is this,
that the subject-matter has become better known to
the author, or has been in a larger measure treated
by others, whose statements he has either to embody
or to combat. But in modern works the circle of
readers is the same for both editions ; and it would
be quite absurd to present to them the same book
in a shape not only not enriched, but on the con
trary materially reduced. In this respect Luke's
case, as we have supposed it, is of an opposite
description : the second copy was written for new
readers, who did not know the first. I should
think that in this case the second copy would not"
at all be enlarged, but rather abridged, the work

104 PHILOLOGY OF THE GOSPELS.
becoming somewhat tedious for the author, or at least
losing something of its freshness for him, so that
he was naturally disposed to omit many unessential
circumstances and details, which he formerly had
given. Consequently, we may suppose that the
Acts in the Eoman, or Western form, exhibits
a comparative prolixness in narration, which has
been removed, or lessened, in the copy sent to
Theophilus, that is in the form commonly known,
and that in the Gospel conversely we shall find in
the Eoman form many abridgments in comparison
with the other, which in this case too is the
form commonly known. As for correctness and
refinement of style (apart from succinctness, which
comes under the head already spoken of), it is clear
that Eoman Christians did not require it in a higher
degree than Theophilus, or Theophilus and his fellow-
Christians in the East in a higher degree than those
in the West ; but the author might give heed to
that, while writing the later copy, for his own sake,
and certainly he would correct errors committed in
the first copy.
f~ So in the Western form of the Gospel and of
the Acts, we may expect additions to the known
text in the Acts, and omissions in the Gospel, and
this is actually the case in D and its associates.
Now, additions may be tested, whether they are in

DOUBLE TEXT IN ST. LUKE AND IN THE ACTS. 105
harmony with the tenor of the narrative or not,
and whether their style is congruent with the bulk
of the text or shows differences ; but omissions of
course are not capable of any such testing. It
follows that the theory I am propounding is easier
to establish for the Acts than for the Gospel, and
so I will start from the Acts. But before entering
into my proofs, I must briefly discuss some prelim
inary topics.
Not only might it be asserted, but it has been
asserted, that the enlarged text of the Acts as
presented by D is the only original one, and that
all omissions found in the common text are due
to the negligence of copyists. This strong charge
against the common text was produced in Germany,
as early as 1848, by Pastor Fred. Aug. Bornemann,
who published an edition of the Acts exclusively
based upon D.1 His courage and independence of
judgment are very praiseworthy, and he was the
first to acknowledge the actual superiority of many
readings in D to those of the Eastern witnesses ;
but his charge against the latter has been, and
must be, universally rejected as wholly incredible.
There are some individual cases where the omission
in B, etc., of words given by D may be quite easily
explained by chance ; but this principle of explaining
1 Published in Grossenhain (Kingdom of Saxony) and London.

106 PHILOLOGY OF THE GOSPELS.
is not nearly sufficient for the great majority of
cases, or for the enormous number of them taken
as a whole. Therefore the solution given by
Bornemann is wholly inadequate, and I do not
think that it will ever again be taken up by any
scholar. On the other hand, to impute the divergencies to
the license taken by the scribe of D, or to that of its
archetype, is not only equally impracticable, but it
has become quite impossible. D does not stand
alone, but is supported by other witnesses, and
among these there are fathers of the Church of the
third and even of the second century, so that a large
number of variants and additions has been strictly
proved to go back to that time. This is chiefly
due to the researches of a very careful German
scholar, Peter Ctirssen, who has written a paper on
Cyprian's text of the Acts,1 showing that the text in
which Cyprian read the book was much like D; or, to
speak even more exactly, that the Latin version
Cyprian made use of was the same of which large
fragments have come down to us in some palimpsest
leaves from a French codex, called Floriacensis because
it came from the monastery of Fleury. These frag
ments have been published with great accuracy by
1 See P. Corssen, "Der cyprianische Text der Acta Ap.," in the
Program of the Gymnasium of Schoneberg-Berlin W., 1892.

DOUBLE TEXT IN ST. LUKE AND IN THE ACTS. 107
Professor Samuel Berger,1 and exhibit in their text
the nearest affinity to D. Corssen has also proved
that St. Augustin, at a certain period of his life (when
he wrote those of his treatises which are directed
against Manichaeism), used the same Latin version
of the Acts, which is the more valuable for us, as
he gives in those treatises the whole first chapter of
the Acts and a part of the second according to that
version, parts which are not contained in the Fleury
fragments. Irenaeus too had a text somewhat similar
to D, and he enables us to trace back this form of the
Acts (which I have denoted by /3 or by E(omana), in
opposition to a, or to A(ntiochena), as far as the end
of the second century. It is well known that we
have very little of Irenaeus in the original Greek, but
read the greatest part of his extant work in an old
and careful Latin version. TertuUian's quotations of
the Acts are extremely few, but nevertheless he too
is proved by them to have used a text of the same
condition as that of Irenaeus.
Thirdly, it is now an acknowledged fact (established
principally by Corssen), that in D we have not the
form /3 (or E) in its purity, but in a state of frequent
mixture and " conflation " with a (A). As I have
1 S. Berger, Le palimpsesle de Fleury, Paris, 1889. The first
editor of these fragments had been J. Belsheim {Appendix Epislo-
larum Paulinarum, etc., 1887).

108 PHILOLOGY OF THE GOSPELS.
already spoken at length of the general practice of
comparing manuscripts and correcting the one by
means of the other, it cannot be astonishing that in
the case of the Acts this same practice has given as
a result a conflation of the two texts a and (3 not
only in D, but also (of course in each ms. in a
distinct way) in other Greek and Latin mss., and
among them that used by Irenaeus, and those used
by Augustine at other periods and for other writings.
The thing is quite clearly to be seen in the actual
state of D. That MS. has undergone corrections by
many persons and at different times, but never (as
it seems) systematically, only sporadically, by which
the (S text has in many places been abolished,
and the a text introduced. We are justified in
supposing that the archetype of D was in the same
condition, there having been introduced, by correction,
many readings of the a text, which were of course
transcribed by the copyist of D. In mauy instances
the actual reading in D is an impossible mixture
of two texts, the correction having been but partly
made, or partly transcribed. In order to restore
| the /3 text to its integrity, as far as possible, we must
therefore have recourse to other witnesses, the more so
as no small part of the book, especially that from
xxii 29 to the end, is not extant at all in D, because
of the mutilations of that ms. Now, the fragments

DOUBLE TEXT IN ST. LUKE AND IN THE ACTS. 1 09
V
of Fleury, together with the quotations in Cyprian
and Augustin, help us a good deal, and that Latin
text seems to be nearly (not altogether) free from
conflation. A third witness for [3 is a Syriac version,
called the Philoxeniana, or Syrus posterior, in op
position to the older version, the Peshitta. This
version was collated in the year 616, by one Thomas
(who had been bishop of Hierapolis or Mabbogh, in
Syria, but having been exiled lived near Alexandria),
with some Greek Alexandrian mss., and the variants
found in those mss. have been added to the Syriac
text. The Greek ms. used by him for the Acts
was very similar to D, and we see by this fact that
the f3 text had not been strictly confined to the -West,
but had found its way also to the East, a fact which
is confirmed in the case of the Acts as well as in
that of Luke's Gospel by other evidence, and has in
itself nothing astonishing, if we regard the constant
commercial intercourse between Eome and Alexan
dria. This Greek MS. of the Acts, which has been
preserved to us in this indirect way, may be sup
posed to have been very ancient, since Thomas
would of course select for his work among the
many copies extant in Alexandria the very oldest,
as seeming to him for that reason the most trust
worthy. But it would be an unwarranted supposition
that this MS. exhibited the pure f3 text without any

110 PHILOLOGY OF THE GOSPELS.
conflation, as it is evident that Alexandria was least
of all the place where this form of the text would
f remain free from contact with the other.
I give a brief account of other witnesses : the
Graeco-Latin Laudianus (E), preserved in Oxford,
written towards the end of the sixth century, and very
likely used by Beda Venerabilis, but much inferior to
I) in purity ; the minuscule codex 137 in Milan, useful
especially for the last chapters where D is deficient ;
the Latin Gigas in Stockholm, giving a version which
was previously used by Lucifer, bishop of Cagliari in
Sardinia (t 371); another Latin ms. recently published
by Professor Berger, and partly by myself;1 lastly,
one of the old Egyptian versions, commonly called
Sahidica. The f3 text is iu all these witnesses much
more mixed with a than is the case in D or in the
Floriacensis ; but we are compelled to gather our
evidence from all sides, since even our best witnesses
are not wholly free from mixture, besides being so
sadly incomplete.
When we now come, after all these preliminaries,
to the testing of the /3 readings, we find the state of
things considerably changed. Instead of one single
1 See Sam. Berger, Un ancien texte Latin des Actes des Apdtres,
retrouvi dans un ms. provenant de Perpignan, Notices et extraits des
ms. t. xxxv. 1™ partie, Paris, 1895, and my second edition of the
Acts (Leipzig, 1896), p. xxv.

DOUBLE TEXT IN ST. LUKE AND IN THE ACTS. Ill
witness of the sixth century, originating probably
from Gaul, we have before us a coherent mass of
evidence, produced not only by the whole West but
also by a part of the East, and going back as far as
to the end of the second century. Of the two
contending texts the one has for itself the large
bulk of witnesses, while the other has still very few ;
but in point of authority the parties are pretty
well matched. There is Cyprian for j3 ; well, but
there is Origen for a. Irenaeus and Tertullian are
more favourable to [3, but Clement of Alexandria
stands on the other side. But, you will say, we must
not revoke the final decision given by the Church,
which has voted for a. That was because it wanted
to have one text and not two texts, and preferred
of course that accredited by those countries where
Christianity had sprung up ; we have in fact just
now mis-stated the question, speaking simply of two
contending parties. The real question is not, which
of the two is the original text ? but, has a a claim
to be the only original text ? We are not in the
situation of the ancient Church : we, or at least many
of us, would prefer to possess, if possible, two original
texts completing and explaining each other. So the
decision of the ancient Church counts for nothing,
and we are called to decide on a causa integra. We
must put aside all prejudices, and examine the

112 PHILOLOGY OF THE GOS-PELS.
question as if neither of the parties was beforehand
known to us. This may be difficult, but it is the
only way of coming to an equitable and unbiassed
I decision.

CHAPTER VIII.
THE PEOOFS FOE TWO DISTINCT TEXTS IN
THE ACTS.
The first test of [3 will be that of language, because
in this part of the question we have to deal with
facts which are in some degree incapable of being
differently apprehended, or differently stated. If the
claim of /3 is just, the language in the additions and
generally in the variants of (3 must be proved to be
Lucan. If this proof fails, the claims of [3 will be
hopeless; if it succeeds, they will have advanced a good
deal on the way to being satisfactorily established.
But you will perhaps object to- the question
being put in this way. Is there a definite style
which may be recognized as that of St. Luke ? There
is most certainly such a style, pervading both the
Gospel and the Acts, and recognizable everywhere.
Luke has, indeed, a very wide range in his choice
of words, and in almost every chapter he uses words
H

114 PHILOLOGY OF THE GOSPELS.
and expressions to which he does not recur again ;
but on the other hand, he has also his familiar words
and ways of speaking, and if anybody questions his
authorship of the whole Acts, that doubt will be
utterly removed (and has been removed) by the
test of the language. So we may safely place entire
confidence in that same test in the case of the
additions in [3.
I shall be as brief as possible in the following
analysis of the style of Acts [3, referring those readers
who want more information to my two editions of
the book and to my edition of Luke's Gospel.1
Here are some simple facts. (1) Number of words
occurring in Acts (3 and nowhere else in the New
Testament, 30. (2) Words restricted to Acts [3
and Luke's Gospel, 3. (3) Words in Acts [3 recur
ring in other parts of the New Testament, but not
in Luke (Gospel and Acts a), 20. (4) Words occur
ring in Acts (3 as well as in other parts of the
New Testament, and at the same time in Luke's
Gospel, 10. All the four categories have this in
common, that the words in question are alien to
Acts a; but it is evident that for proving the
spuriousness of {3 (2) and (4) are of no use, be-
1 See the index of the words occurring in the Acts at the close
of my larger edition, and the summary statistics in my edition of
Luke's Gospel on p. xxvii f.

PROOFS FOR TWO DISTINCT TEXTS IN THE ACTS. 115
cause those words are warranted by the Gospel.
In (1) and (3), taken together, the number amounts
to 50, among which are these astonishing enrich
ments of Luke's vocabulary : irefiirros- the fifth,
ef3So/j.os the seventh, SiaKovos the minister (while
Luke has elsewhere SiaKovw and SiaKovla), e^opKitw,
exorcise (while there is e^ppKicrrrjs elsewhere in
Luke). Of course all the fifty dVa£ Xeydfieva are
not so ridiculous. But now I shall compare the
instances, for each of the three first categories as
they are found in the whole Acts, according to a.
(1) 410, (2) 53, (3) 394. The fourth category,
which would contain the words used alike in Acts,
and Gospel, and elsewhere, cannot be taken into
consideration. So the relative proportions are :
410/30, 53/3, 394/20, much alike in the three
cases. Now, what proportion of sentences and
particles exclusively belongs to (3 as compared to
the rest of the Acts (which is common to a and [3) 1
As I have stated by a simple counting of the words
on three pages of my edition, about that of 86 to
530. Then we should expect much more than 50
aira£ Xeydfieva in (3. As a matter of fact I am sure
that there were more, but the peculiar readings of f3
are partly preserved to us not in Greek, but in Latin
or Syriac, and the words contained in these latter
portions are of course excluded from my statistics.

116 PHILOLOGY OF THE GOSPELS.
This is not the only way of testing by means of
the language, or the best either. The best would
be to take every separate addition or peculiar
reading in (3, and compare it, word for word, with
the other writings of Luke. But how can I do
that here ? I may give one instance as an illus
tration. Acts x. 25 D makes this insertion :
irpocreyyiCovros cVe rov Herpov els rrjv JLaicrdpeiav,
irpoopafiwv els rwv SovXwv Siecracprjcrev irapayeyovevai
avrov, and then goes on in this way : 6 Se KopvijXios
eKirrjSrjcras Kal (crvvavrrjcras K.r.X., as in a). Hpocr-
eyyiXeiv is new in Luke (one of the fifty words),
but he frequently uses eyyi(eiv, cf. x. 9, bSoiiropovvrwv
eKelvwv ko.1 rrj iroXei eyyiYovrwv, ix. 3, xxii. 6 ; see
further Gospel, xviii. 35, ev rw eyyi(eiv avrov els
'lepiyw ; xix. 29, xxiv. 28. Still you insist on the
irpos, as making a difference. Well, I may give,
you a reason for that too : the point of view of
the narrator, according to the words immediately
preceding, is Cornelius' house ; therefore the irpds
here comes in, while it is omitted where the place
approached is not that of the foregoing narrative,
which is the case in the other passages referred to.
ILpoSpa/iwv, see Gospel, xix. 4. Efy rwv SovXwv, like
Gospel, xv. 26, eva rwv iralSwv, etc.1 Aiecrdcprjcrev :
1 Weiss (see below) goes as far as to question the Lucan correct
ness of 5oi\uv, because in ver. 7 oIkctuv is used of the same persons.

PROOFS FOR TWO DISTINCT TEXTS IN THE ACTS. 117
another of the fifty words, occurring besides once or
twice in Matthew. Uapayeyovevai : irapayivecrdai is
found in Luke more than three times as often as
in the other New Testament writers altogether.
'EKirtjSrjcras, see xiv. 14. Are you content now?
Not yet, because of the Siao-acpeiv ? Well, I have
taken out a sentence containing ^T of all the aVa^
Xeyo/ieva ; of course I might have chosen many and
many other sentences without any aira£ Xeydfievov.
Then I shall give you an index of the compounds
with Sid occurring but once in Luke, and most of
them but once in the New Testament. *Aia/3dXXeiv
(the asterisk denotes air. Xey. in the New Testa
ment), Sia(3Xeireiv, * SiayvwpiXeivl (1. d. Gospel, ii. 17),
^Siayprjyopelv, SiaSeyeerQai, Siaipeiv,SiaKaQapi(eiv, * Sia-
KareXeyyecrOai, * SiaKOveiv, * SiaXeiireiv, * SiaXveiv 1 (1. d.
Acts v. 36), * SiafidyecrOai, ,r Stave fieiv, ^Siavevetv,
y'SiavvKrepeveiv, *Stavveiv, *SiairXetv, *8iairpay/ia-
revecrOai, * Siacreieiv, Siacrirdv, SiacrreXXecrOai, ""SiareXeiv,
* Siacpevyeiv, SiachOeipeiv (5c SiacpvXarreiv in an Old
Testament passage), * SiayXevd^eiv, * Siaywpi^ecrOai,
Sieyelpeiv, * SievOvfieicrOai, (* Sie£epyeo~Qai, 1. d. Acts
And yet Luke uses olKerris only twice (see Gospel, xvi. 13), but
dov\os (in the proper sense) 26 times. Likewise Weiss questions
the word immediately preceding (in ver. 24) irepii/i.evev, because it
stands without an object. Now, the rule of Luke's using this verb
with an object is based upon one instance, i. 4, which is the only
other instance of this verb in the New Testament.

118 PHILOLOGY OF THE GOSPELS.
xxviii. 3), * Siepwrav, Siopvcrcreiv. Does the dVa£
Xeyd/mevov Siaeracpeiv now seem to you sufficiently
warranted for Luke by these 2 8 other aira£ Xeydfieva
of the same kind (verbal compounds with Sid) occurring
in his writings ?
I stated above as one of the features in Luke's style,
that he frequently introduces new words, very often
without ever recurring to them ; if that distinctive
feature failed to be found in /3, this would be a
strong argument for my opponents, since the imitator
of course would not use any words but those found
in his model. As it is, I, for my part, give my
verdict for absolute identity of style between f3
and a ; whoever is still of a different opinion must
take the additions in /3 one by one, and then take
his concordance and compare, and he will come to
the same results — of course, only if he inquires
without prejudice ; in the opposite case, he may
come to any result he likes, and persuade himself
and perhaps others that it is next to impossible
that Luke should have written ir pocreyyi(eiv in this
one case, or that he should have used Siaeracpeiv
and never cracpeiv (which indeed does not exist), nor
cracprjs, nor cracprjveia, nor dcracpijs, etc.
Let us now see what is to be inferred from
these facts. In the first place, the spuriousness
of /3 cannot be shown in this way. Next, its

PROOFS FOR TWO DISTINCT TEXTS IN THE ACTS. 119
genuineness becomes at least highly probable, for
identity of language presupposes either a common
author, or a model and a very skilful and care
ful imitator, or two such imitators of the same
model. But it is extremely unlikely that there was
in any age a man who took Luke for his model
in order to interpolate Luke. Perhaps you will
say, that somebody who was thoroughly familiarized
with Luke's style would quite naturally and not
designedly give his explanations of the author's text
in the author's own language. It cannot be denied,
that in individual cases something like this might
happen ; for instance, we find in the section on the
adulteress, which certainly did not originally form a
part of John, a few peculiarities of John's style
introduced by copyists : eKacrros Se tcoj/ 'lovSalwv,
viii. 9, instead of ol Se aKovaavres, and ver. 6, rovro
Se eXeyov ireipdXpvres in most MSS., as it stands
in John vi. 6. But what might happen in a few
cases, cannot therefore be supposed to have happened
throughout a book ; the interpolator, if familiar with
Luke's language, did not cease to be still more
familiar with his own, and would give many more
samples of that than of the author's language.
And now for the second test, that of the matter
and of the facts contained in the additions. Can
it be rationally supposed that an interpolator rightly

120 PHILOLOGY OF THE GOSPELS.
understood all the text, without exception ? Certainly
not in the case of the Acts, which offer so many
difficulties to our understanding. We have seen
that no less a man than Eusebius thoroughly
misunderstood the proem in the Gospel. Secondly,
can it be supposed that the interpolator had any
knowledge of the story told, independent of the
Acts ? I do not say of the main points, but of
the minute detail, that which we have to deal
with here ? Evidently this is next to impossible,
and barely possible only if the interpolator was a
contemporary and a witness of the facts. Now
any ordinary interpolator would betray himself by
a number of misunderstandings (the book being of
such extent), and by a great many wrong state
ments, which might be refuted by other passages
and by our independent knowledge. I may safely
say that both Professor Eamsay and Peter Corssen,
who have used all their knowledge and skill in
order to detect such cases, have completely failed
in their attempt. As I have elsewhere answered
these highly esteemed scholars,1 I shall not repeat
things formerly said, but prefer to speak at once of
the most recent attempt, made by Professor Bernhard
1 See in my second edition of the Acts, on p. viii, and my edition
of Luke on p. xxiv ff. (against Corssen, who has given a very careful
review of my former edition of the Acts in Getting. Gelehrte
Anzeigen, 1896, p. 425-448).

PROOFS FOR TWO DISTINCT TEXTS IN THE ACTS. 121
Weiss, a veteran scholar in the whole province of New
Testament inquiry.1 It is not from the historical point
of view that Professor Weiss contends against D — for
he is still disposed to confine himself to that one wit
ness, waiving the rest of the host as much as possible
— but from that of textual criticism ; he imagines a
blundering interpolator, who perverted what he did
not understand. Let us take, as an instance of
Weiss's handling of the matter, the beginning of
xiv., and patiently enter for once into all the
minutiae. The English text is in ver. 1 ff . : " And
it came to pass in Iconium, that they (Paul and
Barnabas) went . . . into the synagogue of the
Jews, and so spake, that a great multitude, both of
the Jews and also of the Greeks believed. (2) But
the unbelieving Jews stirred up the Gentiles, and
made their minds evil affected against the brethren."
Then follows in ver. 3 : " Long time therefore abode
they speaking boldly in the Lord," etc. Now, I
say that this narrative wants coherence ; the result
of the action stated in ver. 2 is not given at all.
Turning to the Greek, we find that there are in
ver. 2 the aorists eirrjyeipav and eKaKwcrav, which
oppose the idea of an unsuccessful attempt (imperf.).
On the other hand, D and Syr. post, give a quite
clear and coherent narrative (ver. 2) : " But the
1 B. Weiss, Der Codex D in der Apostelgeschichte, Leipzig, 1897.

122 PHILOLOGY OF THE GOSPELS.
chiefs of the synagogue of the Jews and their
magistrates directed a persecution against the just,
and made the minds of the Gentiles evil affected
against the brethren ; but the Lord soon gave peace.
(3) Long time therefore," etc. I myself and Professor
Hilgenfeld, Provost Salmon and others, are quite
convinced of the superiority of /3 in this instance ;
and the only difficulty remaining is this, that Luke,
if he wrote /3 first and then a, must be said not
only not to have corrected his first form, but to
| have made it worse.
We shall have, of course, to revert to this ob
jection ; but let us first hear Weiss : " The things
which came to pass in Iconium were entangled by
the interpolator, because he did not understand
the dweiOijcravres 'lovSaioi in ver. 2." What ? Was
there ever a man who, after the statement in ver. 1,
did not understand ol Se direiOrjcravres in 2 ? " the
unbelieving Jews " after " a great multitude of
Jews believed " ? Are suppositions like this to be
considered justifiable ? " Nor did he understand
eirtjyeipav, which he supposed to want an object,
not seeing that that verb had an object in common
with the following eKaKwcrav, namely rds \J/-in/ay rwv
eOvwv." Here the words in a are really not easy to
understand. Even if eirrjyeipav rds ^vyds may be
compared with lxx. (1 Chron. v. 28), eirrjyeipev 6

PROOFS FOR TWO DISTINCT TEXTS IN THE ACTS. 123
Geo? rb irvevfia QaXwy pWiXeco? 'Acrcrovp, still there
is the important difference that in that passage there
comes a continuation : ml fierwKicrev rbv 'Tov/3ijv
( = " to do this "), while in the Acts that continuation
is wholly wanting, and there it is said as it were only
that " they made them," without the necessary
complement " do something." But as Weiss does
not seem to see any difficulty, we may pass on with
him. " So he makes the chiefs of the synagogue as
early as that time direct a persecution Kara rwv
StKaiwv ; besides he adds quite superfluously, rwv
'lovSalwv, which is but an ' echo ' of the 'lovSaioi in
ver. 1." By no means : as the Greeks ('EXXjJi/coj/)
had come in between 'lovSaioi and dpyicrvvdywyoi,
the addition was very necessary. But whence did
he get his chiefs of the synagogue ? whence his
dpyovres 1 Not a word on that. And yet the
accurate distinction between dpyicrvvdywyoi and
dpyovres, which has been shown by Professor Eamsay
to be quite correct for that time, while it would not
have been so for a later one,1 seems to demand some
explanation. Are we to suppose that the same person
was at once so stupid as not to understand " un-
1 See Ramsay in Church in the Roman Empire, p. 46, and in
Expositor, 1895, p. 212 ff. I may notice that D inserts rrjs o-vvayayijs
after apxovres (which has led Prof. R. into some error), but the
Syrus post, rightly omits the words. (See also F. H. Fisher, of
Pretoria, South Africa, in Expository Times, August, 1897, p. 524.)

124 PHILOLOGY OF THE GOSPELS.
believing Jews," and yet so fully supplied with
intimate knowledge as to give these details ? " In
this way the clause, ' and made the minds of the
Gentiles evil affected,' etc., becomes discordant with
the preceding one : for if the Jews themselves had
been able to direct a persecution against the Christians,
they did not want the assistance of the Gentiles."
I think my readers will. answer, that as the power
of the Jewish magistrates did not extend to Gentile
Christians, or to Paul and Barnabas, the assistance
of the Pagan populace and of the magistrates of the
town was on the contrary much wanted. Besides,
who are the SUatot in f3 ? who the dSeXcpol in /3 and a ?
" The SiKaioi are those Jews who had become believers,"
says Weiss. Quite right ; then rwv dSeXcpwv, which
word must be of a different meaning (else we should
have Kar avrwv), will embrace the whole congregation
of Christians (and of course the foreign apostles too),
and quite pertinently, as the hostility of Gentiles
would be at once more excited and more spontaneously
directed against their countrymen than against the
Christian Jews, who were to some extent under the
control of Jewish magistrates. "He makes this per
secution rapidly pass away by the Divine assistance ;
one does not, however, see how." But I do not see
how without some such thing the apostles could
have as much as remained in Iconium, not to say

PROOFS FOR TWO DISTINCT TEXTS IN THE ACTS. 125
" have been speaking boldly in the Lord " (ver. 3) ;
they would have been expelled by the combined
force of Jews and Gentiles, as they had been from
Antioch before. For, according to ver. 2, the hostility
of the Gentiles had been actually excited (aor.), and
of course against the foreigners and their Greek ad
herents, while the belief of any of the' Jews was for
them, as for Gallio (Acts xviii. 15), a matter of utter
unimportance. Professor Weiss continues in this way chapter
after chapter ; were I to follow him, I should only
weary my readers. Besides, I am not writing on
the Acts, but on the Gospels. So I may leave readers
to study this matter for themselves, giving them for
assistance a little book by another German professor,
Joh. .Belser, who, although a Eoman Catholic, never
theless, as an Englishman writes to me, "is very
free from prejudice, and full of good points."1 He
is free also from the dust of minute textual criticism,
which in Weiss's book is found from the very beginning,
overspreading everything else, and overclouding the
main problems, which, if they are to be rightly decided,
must be kept clear from minute encumbrances, like
the blunders of the copyists of D, or of those of its
predecessors. Let these blunders be stated in any
1 Joh. Belser, Beitrage zur Erklarung der Apostelgeschichte, Frei
burg im Breisgau (Herder), 1897.

126 PHILOLOGY OF THE GOSPELS.
number, you will by that means not even touch or
approach the problem ; and Weiss, by using that
method, has got nearly as far as to the middle of
his book before approaching it.
But the evident superiority of /3 which I have
stated in this passage, and which may equally be
found in many others — see Belser, or Hilgenfeld, or
H. Holtzmann, or Nestle, or Zockler, or Conybeare, or
Eendel Harris, or Provost Salmon1 — and the corre
sponding inferiority of a urges me to correct myself on
one point, which I never regarded as of importance,
but which in the general discussion has come into
much prominence.
In the beginning, the form I had given my solu
tion of the problem, was that I took /3 as it were
for Luke's rough draught, and a for the corrected
and final copy. It is evident that this is not at all
essential for the theory which requires merely one
older copy and one more recent, and besides it can
not be applied to the Gospel, where Theophilus' copy
is the older one, and the Eoman copy the later; and

1 See Hilgenfeld in Berl. Philolog. Wochenschrift, 1896, No. 43 ; H.
Holtzmann in Theol. Literaturzeitung, 1896, No. 3 ; 0. Zockler in
"Die Apostelg. als Gegenstand hoherer u. niederer Kritik," in
Greifswalder Studien (1895), p. 195 ff. ; F. Conybeare in American
Journal of Philology, xvii. 135 ff. ; R. Harris in Four Lectures on the
Western Text of the New Testament, 1894 ; G. Salmon, Introduction
to tlie Study ofthe Books ofthe New Testament, 7th edition, p. 592 ff.

PROOFS FOR TWO DISTINCT TEXTS IN THE ACTS. 127
consequently neither of them is a rough draught, for
such would never be sent to Theophilus. There
fore a better form of the solution is this, that Luke's
rough draught was kept by him for himself, and
that he made from this two successive copies of
each book, the older one following more closely the
first sketch, while the later departed from it with
more freedom and gave (which is the most prominent
feature) very frequently an abridged text. It is
by way of abridgment that this deterioration has
arisen, as in the passage I was just now speaking of.
Instead of apyicrvvaywyoi rwv 'lovSaiwv Kai ol dpyovres
was put direiOtjcravres 'lovSaioi, which is materially
the same, but in fewer words and in a less definite
form ; instead of eirrjyayov Siwy/ibv Kara rwv StKaiwv
he simply wrote emjyeipav, using the same word
which he had used a little before, xiii. 50 ; then he
left out the whole clause 6 Se Kvpios K.r.X., and be
cause of the omissions in 2 f. altered the words in
ver. 4 and those in 5 (as they are given by the
Syrian witness), writing in 4 ecrylcrOrj instead of rjv
ecryicr fievov, and so on. Likewise in x. 25 (the
passage of which we analyzed the words) the text
in a sprang out of that in /3 by mere abridgment.
Whether Luke is to be blamed, as a writer, because
of these omissions and abridgments, is a question
which I need not discuss here ; perhaps he is, from

128 PHILOLOGY OF THE GOSPELS.
our point of view : but he might answer with Euri
pides : rl 8' alcrypdv, rjv fir) roicri ypwfievois (to Theo
philus) Sokjj ? At the same time, by changing the form
of our hypothesis in this way, we get a full explana
tion of the converse cases, which, if much rarer, are
nevertheless not wholly absent even from the Acts,
viz., those, where a gives something more than /3.1
In these, Luke's archetype, or rough draught, has
been preserved in the later copy, as would of course
happen now and then, from respect to the peculiar
circle of readers addressed, or by mere chance, just
as in our own writing and copying.
What I have just said is also fitted to meet,
in some measure, the objections raised to my theory
by Mr. T. E. Page (see Classical Review, July, 1897).
On the whole, he treats it somewhat contemptuously,
saying that the question of the origin of the /3 variants
may occupy the attention of scholars " with ample
leisure." I am very much afraid he does not include
himself in that number. If he had given some
leisure hours to the individual problems he touched,
he would, e.g. have found in Provost Salmon's review
of my edition (see Hermathena, vol. ix. p. 235) a
very judicious appreciation of the passage in xxi. 16.
Eeaders of the common text are almost of necessity
led to the erroneous opinion that Mnason, who is
1 See my edition of Luke, pp. v f., xxvi, xxxiv f.

PROOFS FOE TWO DISTINCT TEXTS IN THE ACTS. 129
mentioned there, was living in Jerusalem, while
from the /3 text there comes the clear statement
that he lived in a village mid-way between Jerusalem
and Caesarea. " Anybody," says Mr, Page, " could
make these corrections." Certainly, anybody, who
was very attentive, and reflected a good deal, and
compared other passages, and had no small amount
of ingenuity : that is to say, not " anybody." The
most striking piece of reflection of that " somebody "
has been put into due light by Provost Salmon. As
Mnason is said to have been an old disciple, the
result is that, by giving him this dwelling-place,
a connection is established between this passage and
that in xi. 2, where Peter is said to have preached
the Gospel in the villages through which he passed
in going up from Caesarea to Jerusalem. " It is
a natural combination," says Salmon, " that Mnason
was one of his converts." And for that fine com
bination we are under obligation to somebody, who
must have been anybody except Luke. Why not
Luke ? Because Luke, if he had written first the
sentence as it stands in /3, never would have brought
it into that obscure form in a ? Then I answer,
that he did it for abridgment's sake ; for the unknown
village where Paul slept for a single night is a matter
of infinite unimportance, as Mr. Page says. No, I
do not say that, Mr. Page will answer : at whose

130 PHILOLOGY OF THE GOSPELS.
house he slept in a village is unimportant; but
Luke does tell us that he received kindly hospitality
from Mnason in Jerusalem, and the corrector has
quite misunderstood the real meaning. Impossible,
I reply, for it must " seem very strange to us that
Paul should have been dependent on a stranger
for entertainment at Jerusalem, a place where we
should have supposed he could have relied on the
hospitality of private friends " (Salmon), and that
in order to introduce Paul to Mnason, whose house
he might find himself, the disciples of Caesarea
should have accompanied him on that long journey.
The accompanying, on the other hand, was quite
natural, if Mnason lived in a village known to those
disciples, but not known to Paul.1 The aorist
^evicrOwfiev too, denoting an action -of some definite
time, makes very strongly against this interpretation
of the a text : if the sojourn in Jerusalem was
spoken of, we ought to have ^evi^wfieOa. But Luke,
Mr. Page might reply, makes a wrong use of the
aHere B. Weiss (p. 101) conies to the assistance of Mr. P.,
saying that it was not for Paul's sake, but for his large company
of uncircumcised men, that the accommodation in Jerusalem was
procured in this cumbrous way. But on Paul's former visit (Acts
xv. ) there had also been at least one uncircumcised man (Titus)
in his company, who had of course found accommodation, and
on that same occasion the equal rights of uncircumcised Christians
had been formally recognized by the whole congregation of
Jerusalem.

PROOFS FOR TWO DISTINCT TEXTS IN THE ACTS. 131
aorist also in xxviii. 14, rjXQafiev (a passage to
which my reviewer actually refers). Wrong only
in so far as the result is given beforehand, and
nevertheless the author, in the next verse, returns
to things which had occurred during the journey.
In xxi. 18 there is nothing of this kind, as the
author correctly says in 15: dvefiaivofiev els 'Iepoa-6-
Xvfia, and 17: yevofievwv Se rjfiwv els lepocroXv/ia ;
crvvrjXOov, which is used of the disciples in 16, seems
to imply that from the village they returned to
Caesarea. There are two passages in /3 which deserve our
particular notice, because of their bearing either
upon the author or on his former book. In xi. 27,
after the mention of the prophets who came from
Jerusalem to Antioch, /3 inserts : " and there was
a great joy. And we being assembled, one of them,
named Agabus," etc. Now, this we (crvvecrrpafifievwv
Se rjfiwv), which is also attested by St. Augustin,
clearly shows that the author was at that time
a member of the Church of Antioch, and as the
period is so very early, we can hardly doubt that
Luke really was a native of Antioch, which is the
tradition given by Eusebius (Hist. Eccl. iii. 4, 6),
and others. Eoman Christians cannot have been
ignorant of this fact, and the " we " was to them
perfectly clear. Besides this, we get by these three

132 PHILOLOGY OF THE GOSPELS.
words a fresh we-section, which is of the greatest
importance for showing that the occurrence of the
first person, which is found again in xvi., and in xx.,
etc., can by no means be employed (as theologians
have frequently employed it) for the purpose of
dissecting the Acts into parts originally independent
from each other.1
-— The other passage is that which opens the whole
book, and connects it with the Gospel. There D is
(as F. Graefe and P. Corssen have shown before me)
corrupt by way of conflation, and we are restricted,
in Weiss's words, to "Latin" witnesses ; in other
words, we have the testimony of St. Augustin given
in two places, and in the most trustworthy fprm,
the whole chapter having been copied by him. This
witness gives the text of ver. 2 as follows : in die
qua Apostolos elegit per spiritum sanctum et praecepit
praedicare euangelium. But, says Weiss, this com
bination of " to preach " (Krjpvcrcreiv) and " Gospel
(a combination which is here attested also by D and
other witnesses) is not Luke's habit ; he says (Gospel,
viii. 1), Krjpvcrcrwv kcu evayyeXiXpfievos. Now, how
many times does Luke make use of evayyeXiov ?

1 About this passage Weiss, who has reserved it for the end of his
book (p. Ill ff.), does not speak with any confidence, while he, at the
same time, judiciously rejects all theories of dissection by means of
the "we."

PBOOFS FOE TWO DISTINCT TEXTS IN THE ACTS. 133
Just twice, Acts xv. 7 (in Peter's speech),, and, xx. 24
(in that of Paul) ; besides him the combination exists
in Matthew, and in Mark, and in Paul, and Luke
might enrich his language by use of it, if need be,
from his teacher Paul. The main point is, that
instead of " until the day He was taken up," there has
been put in this text in die qua, the mention of the
ascension of Christ having been wholly suppressed.
As this is the proem of the work, we must consider
it very carefully, as we did in a former chapter in the
case of the proem of the Gospel. Luke is giving here
a summary of his former book ; in the a text, he out
lines its contents by giving the close, while in this (3
text he does so by giving the beginning, going on, in
the next verses, to sum up the further contents,
and inserting in ver. 4, "how" (et quomodo = ko.1 ws),
in order to indicate that this too was already related
in the Gospel, as it actually stands (Gospel, xxiv. 49).
At the same time he is giving here some fresh
details, especially the forty days in ver. 3, and
leads on, in this way, to the more detailed relation
of the ascension which is to follow. This text therefore
seems to be quite consistent with itself, but yet
we are told by Professor Weiss, that it is impossible,
because Jesus is stated to have begun His working
and teaching on the day (in die qua) He chose
the Twelve, a fact which is related by Luke as

134 PHILOLOGY OF THE GOSPELS.
late as in vi. 13 ff. As this objection is rather
serious, we must try to re-translate the " in die qua "
into Greek with more care than I did in my edition,
where I gave it in this Greek expression : ev fj
rjfiepa. But Augustin has not in qua die ; therefore
why not ev rjfiepa rj ? Very well, you will say,
but in this I do not see any real change. Never
theless, there is a change in the meaning. 'Ev fj
rjfiepa is the same as ev (eKeivrj) rrj rjfiepa fj, and
distinctly denotes some definite day ; but eV rjfiepa
fj, without the article, presents itself as being like
in sense to the identical lxx. phrase, as we have
it, for instance, in Jer. vii. 22: "For I spake
not unto your fathers, nor commanded them in
the day that (ev rjfiepa fj) I brought them out of
the land of Egypt, concerning burnt offerings or
sacrifices." Evidently the rjfiepa is applied there
not to a single definite day, but in a looser way
to a definite time extending over many days, and
so our Hebrew dictionaries teach us that the corre
sponding Hebrew word very frequently expresses
the general notion of time. Likewise, our dictionaries
of New Testament Greek tell us that the looser
Hebrew notion of " day " has introduced itself in
many passages, not only in the plural form, but
also in the singular, see John viii. 56 ; xiv. 20; xvi. 23,
26; Eph. vi. 13; 2 Cor. vi. 2 ; 2 Pet. iii. 18. So

PROOFS FOE TWO DISTINCT TEXTS IN THE ACTS. 135
Luke says, quite in accordance with the facts : All that
Jesus began to do and teach in that time when he chose
the twelve, and the " impossible " reading proves to
be quite possible and easy. It is true that Luke
elsewhere has either eV rjfiepa fj only of a single
but not definite day (Ev. xii. 46), or eV rjfiepais ah
of a definite time (ibid. i. 25); but these passages
also are both without parallel in Luke ; and it is very
clear that in that of the Acts he could not employ
the plural : ev rjfiepais als e£eXej~aro K.r.X., because
that event had been told in the Gospel as having
come to pass on a definite and very memorable
day, that of the great Sermon on the Mount.
It may be objected that I am gliding back into
the ordinary notion of " day," after having established
that here rjfiepa means "time." But the original
notion of the word could never be wholly absent
from the mind of the writer, and for that reason
a distinction between the plural and the singular
is maintained throughout : ev rjfiepa fj iirelSev dtpeXeiv
K.r.X. (words of Elizabeth, Ev. i 25), would have been
quite unnatural after : irepieKpvfSev eavrrjv firjvas irevre,
and so then the plural comes in.1 -^
Let us consider now the reading in a, ^Qv 'rjp^aro
'In Gosp. ix. 51 (see below), ras r>ixipas tt\s avaXrffispeids airov does
comprehend the time of the passion, the resurrection, and up to the
day of Ascension, which day, however, has no special importance
for this passage.

136 PHILOLOGY OF THE GOSPELS.
6 'Irjcrovs iroieiv re Kal SiSdcrKeiv, aypi rjs rjfiepas. The
combination of fjpparo with aypi may be justified,
in some measure, by other passages, but nevertheless
is rather harsh ; and Weiss thinks that' readers who
did not understand it have therefore altered " until "
to " in." And have left out, he ought to continue,
the Ascension; but he prefers, as we saw, to deal
with D alone, where (by means of conflation) the
Ascension is mentioned, and the rest of the evidence
is nothing to him. I say : readers of the New
Testament must have been accustomed to digest
harder things than this rjpparo aypi, and if they did
not understand it, the most simple way was to leave
out the " began." There comes next : evretXdfievos
rois airocrroXois Sia irvevfiaros dyiov ovs ePeXeParo
dveXrjficpOrj. The clause is in a very awkward way
complicated by the insertion of ovs ePeXeP^aro ; I
do not think that this can be the original hand of
Luke. But if we suppose the words as they are
in /3 to represent Luke's first writing, the thing
becomes quite transparent : he has encumbered the
clause in order to bring in the Ascension, without
leaving out the choice of the apostles. And why
did he want the Ascension to be mentioned? Because
the corresponding form of the Gospel (a, or A) closes
with the Ascension ; a of the Gospel and a of the
Acts must be made to agree.

PROOFS FOR TWO DISTINCT TEXTS IN THE ACTS, 137
Is there then in /3 of the Gospel any difference
on this point ? Assuredly there is, and now we
are at last returning to Luke's Gospel, on behalf of
which we have taken this long circuit through the
Acts.

CHAPTEE IX.
THE DOUBLE TEXT IN ST. LUKE'S GOSPEL.
It is a fact which has been the subject of much and
serious discussion, that at the close of Luke's Gospel
the Ascension of Christ is by no means attested by
all manuscripts. Besides the known text there is
another one, having very good attestation, and therefore
received by Tischendorf, as well as by Westcott and
Hort: (51) "And it came to pass, while He blessed
them, He was parted from them " (without " and
carried up into heaven"). " And they " (without
" worshipped Him, and ") returned to Jerusalem with
great joy," etc. If this text is to be accepted as
the only original one, and the omitted words are
to be rejected as interpolations, the fact of the
Ascension of our Lord, which we confess every
Sunday, will in the main rest exclusively upon the
testimony of the Acts, that of Mark being rejected
together with the whole close of his Gospel. But
this by the way : I am not writing on dogmatics.

THE DOUBLE TEXT IN ST. LUKE'S GOSPEL. 139
It has been argued, on the other hand, that Luke
gives another testimony for the same fact, by saying
as early as in ix. 51: " When the time was approach
ing that He should be received up " (eV tco crvfiirXr]-
povcrOai ras rjfiepas rrjs dvaXrjfi^ews avrov), and that
by mentioning the Ascension here, he clearly indicates
his intention of relating it at its proper place.
Moreover, as the reading in ver. 52:" they returned
to Jerusalem with great joy, and were continually
in the temple, praising (and blessing) God," is univer
sally attested, the words left out by a part of our
witnesses can hardly be supposed to have been
originally absent from the narrative. If the apostles
had seen their Lord carried up to heaven, there was
a reason both for their rejoicing and for their being
continually in the temple, that is to say, not expecting
any more appearances of Him ; but if the appearance
related had ended in like manner as the others, this
sequel becomes quite incomprehensible.
Why, then, are the words omitted by a part of
the witnesses? You ought to ask first: Who, then,\
are the witnesses upon whose evidence Tischen
dorf and Westcott-Hort omit tbe words ? There is,
in the first place, D, and together with D some Latin
versions, and St. Augustin, and the Syriac version1
'The Syriac witness gives, however, a. word like iwi)pffis instead
of Siitrris or arrio-ris.

1 40 PHILOLOGY OF THE GOSPELS.
of Mount Sinai, and lastly, but only as regards
the words in ver. 51, the first hand of the Greek
ms. from Mount Sinai (a). Now, the solution of
the enigma is quite near at hand : it is /3 of the
Gospel which omits these words, while a had them,
and a, in this case, is the earlier copy and represents
more truly Luke's original writing. The other question,
too, why the words are left out in (3, has already
almost found its answer : it was in order to fit the
close of the Gospel /3 to the beginning of the Acts
/3, and this was recognized ten years ago by the
German pastor, Ferdinand Graefe.1 Likewise as the
original beginning of the latter book was, in some
way, altered by the author, when writing for Theo
philus, in order that it might suit the close of that
form of the Gospel which had been sent to him :
so Luke, when he again copied his Gospel for the
use of the Eomans, altered- its close with respect
to the beginning of the Eoman Acts. It is true
that the last chapter of the Gospel /3 and the
first of the Acts /3 do not give a really continuous
narrative ; but the discordance, which consists in the
double relation of the Ascension, is at least veiled,
and does not strike the eye of an ordinary reader.
We must suppose therefore that during Luke's first
stay in Eome, at a time when he already contemplated
'See Theologische Studien u. Kritiken, 1888, iii. p. 522 ff.

THE DOUBLE TEXT IN ST. LUKE'S GOSPEL. 141
the writing of the Acts, or had even fixed the begin
ning of that book, he wrote out the copy /3 of
the Gospel. Why, then, did he not submit its last
chapter to a more thorough reconstruction ? I should
think, because he was depending on a written author
ity, and did not venture to alter overmuch the
account given by him on that authority. Like
wise he had a written authority for the first chapter
(or the first twelve chapters) of his Acts, and felt
himself bound to follow that authority very closely.
I am even inclined to think, that the rather rough
form, which the beginning of that book exhibits
even in (3, if we compare it with the extremely
refined proem of the Gospel, is due to his following
some written text, which furnished him with the
materials for the first part of the Acts. And here
I may venture to propound a conjecture. Suppose
that Mark was the author who had written a con
tinuation to his Gospel, and that this continuation
fell into Luke's hands at some time after he had
finished his own Gospel. I find that conjecture,
for instance, in Weiss's book on Mark, of course as
a conjecture, not as a certainty ; he thinks it pro
bable that Mark really had closed his Gospel at
xvi. 8, and afterwards wrote a continuation begin
ning with the appearances, that is to say, the first
actions of the risen Christ, and going on to tell

142 PHILOLOGY OF THE GOSPELS.
what the same Christ had done afterwards by means
of His apostles.
Supposing all this, we may well conceive the
beginning of Mark's second book to have been nearly
identical with that of the Acts, with this exception,
that the appearances did not come into the summary
of the earlier book, but were given in detail as a part
of the second. Of course this conjecture cannot be
tested or verified, except perhaps on one point.
Luke, as we have seen, refers his readers to the
memorable day when Jesus elected His apostles,
and to his description thereof in his Gospel, vi.
13 ff. Comparing that passage, we find that the
name of apostles is really there, but that the com
mission given to those " messengers," viz., to preach
the gospel, is not expressly stated. But when we
turn to the corresponding passage in Mark (iii. 13 ff),
there we actually find that commission (ver. 14), "that
He might send them forth to preach " (or " to preach
the gospel," after the text in D), and so the curious
fact is noticeable that Luke's summary suits Mark's
narrative even better than his own. We shall return
to that passage in Mark in a later chapter (see x.).
Leaving aside these conjectures for the present, and
returning to the last words of Luke's Gospel, I must
state that the omission of " and was carried," etc.
might be, not without some degree of probability,

THE DOUBLE TEXT IN ST. LUKE'S GOSPEL. 143
ascribed to some reader of Luke, who was offended
by the repetition in Acts i Of course that reader
must be sought for in the very first period, when
Gospel and Acts still formed as it were one whole ;
for in after- times, when Luke's Gospel had been
combined with the other three, and the Acts with
the Catholic epistles, no such offence could be taken.
But we have seen that there actually was a time
when Luke's works were not yet read separately.
So, if the omission of the words in Gospel, xxiv. 51 f.
were an isolated fact, we should require no other
explanation. But as the beginning of the Acts ex
hibits a corresponding phenomenon of double-reading,
and indeed in a„very complex way, which far exceeds
the abilities and the possible fancies of ordinary
readers, both of these phenomena are to be explained
on the same ground, which is sufficient for both ; and
Graefe himself, who formerly had applied to both the
hypothesis of an interpolation, now quite agrees with
me in referring the facts to Luke's own treatment of
his work.1
But my readers will still ask me this question : '
What does your supposition of a double authentic
form of the Gospel rest upon ? This one passage
in xxiv. 51 f. is a very narrow basis, and you
have expressly declined to adopt Bishop Lightfoot's
1 Theolog. Studien u. Kritiken, 1898, p. 137 f.

144 PHILOLOGY OF THE GOSPELS.
analogous solution for three other passages ; what
others have you then ? Of course I have plenty of
others to which I apply this solution ; but I am
bound to point out passages to which this solution
alone applies. As the difference between D and the
commonly adopted text, in the case of the Gospel,
consists mainly in omission, the proof for two
authentic texts must be much more difficult than in
the Acts. But to counterbalance this, the proofs for
the very early existence of the Western text are
more numerous, and go much further back in the
Gospel. I entreat my readers to allow me to give
\ these proofs first of all.
Now, it may be sufficiently established that Justin
Martyr, who lived in Eome about the middle of the
second century, used the /3 form of Luke's Gospel.
But as his quotations are few in number and free
in form, and as he always speaks of the Gospel in
general, never quoting by name either Luke or
Matthew,1 I shall waive this witness, since I have
another of the same time, Marcion, the well-
known founder of a heretical sect which lasted for
a very long time. This man, a native of Sinope
on the Black Sea, went to Eome about 138, and
1 E. Lippelt, an important discovery by whom I have mentioned
before, has directed his and my attention to this point also, and will
publish his results in due time.

THE DOUBLE TEXT IN ST. LUKE'S GOSPEL. 145
became there the chief of a separate sect about 144.
His notion of Christianity was that of an extreme
Paulinist, who decidedly rejected all that seemed to
retain a flavour of Judaism, not only therefore the
whole Old Testament, but also three Gospels out of
four, acknowledging only that written by 'Paul's dis
ciple Luke, and rejecting also the rest of the writings
in the New Testament, besides the Pauline epistles.
But even in those books which he kept very many
sections or verses were struck out by him, as tainted
by Judaism, and his Gospel, which of course bore no
distinctive name, being his only one, began in a very
abrupt way with Christ's coming down to Capernaum
(see Luke iv. 31). The New Testament of the Marcion-
ites has perished, but not without leaving considerable
traces in writings directed against this sect, among
which TertuUian's four books against Marcion occupy
the first place. It has been attempted, therefore,
partly to reconstruct Marcion's New Testament by
means of Tertullian and other writers, the most recent
and most complete and accurate reconstruction having
been made by Professor Theodor Zahn.1 Now it can
not be doubted that Marcion's text of Luke's Gospel
did not exhibit the form commonly known, but that
attested by other Western evidence, such as D and
1See Th. Zahn, Geschichte des N.T. Kanons, ii. 2, 1, p. 411 ff.
{Marcion's Neues Testament). K

146 PHILOLOGY OF THE GOSPELS.
Latin versions, and in this way a different Western
text of the Gospel is proved to have existed as early
as in the first half of the second century.
I might well be content, as regards antiquity of
evidence, with this one witness, the rest being of
course of a somewhat later age ; but I shall name
two more, in order to show by them how the Western
text spread more and more widely to the different
parts of the ancient world. Tatian, a Syrian by
birth, became a disciple of Justin, probably in Eome,
and afterwards went back to his own country, where
as it seems he gave to his countrymen the first
Gospel in their native tongue, the famous Diatessaron,
that is one Gospel made out of four. I am not
going to enter into the very intricate questions about
Tatian's Diatessaron, but shall content myself with
saying this much, that in the few unadulterated
remnants of that work there appear clear traces of
the Western text of Luke,1 and that the quadripartite
Syriac Gospel, as contained in the mss. discovered by
William Cureton and by Mrs. Lewis, counts as regards
a number of passages among the witnesses for the
Western text.
A third witness for the same is Clement of Alex
andria, although neither he, nor Tatian, nor perhaps
Marcion, seems to have possessed the Western text
1See on the Diatessaron, Zahn, ibid., p. 530 ff.

THE DOUBLE TEXT IN ST. LUKE'S GOSPEL. 147
while still in its pure form. Apart from the confla
tions with other Gospels, which as we have seen began
at a very early date, conflations with the Oriental
text must have been even more inevitable than in
the case of the Acts, because the Gospel was so
much more read and copied and collated than the
Acts. A general survey shows me that the /3 text
of the Gospel, while more widely spread than that
of the Acts, was in the same measure more conflated
with the other. Clement read the Acts in the a
text, but the Gdspel in a mixed one, and so our
chief mss. of the New Testament, the Vaticanus B
and the Sinaiticus N, which are either of Alexandrian
origin or intimately connected with Alexandrian texts,
follow a in the Acts, and exhibit a mixture of a
and /3 in the Gospel, more especially x in its first
uncorrected state, while in B there are but a few
passages showing Western influence. We saw how
ever before, that even of the Acts the /3 text had
spread also to Egypt, and that a MS. existing there
as late as at the beginning of the seventh century
became, by means of the later Syriac version, one
of our best witnesses for Acts (3.
I may now go on to test the combined evidence of
all these witnesses, and first as regards the omissions.
In viii. 43 we have in a: "And a woman having an
issue of blood twelve years, which had spent all her

148 PHILOLOGY OF THE GOSPELS.
living upon physicians, neither could be healed of
any, came behind Him," etc. D, which is supported
by an old Egyptian version, exhibits the whole relative
clause in five words : " whom no one could heal,"
while B and the Sinaitic Syriac version, together
with the Armenian (which is dependent on Syriac
authority), simply omits " had spent . . . physi
cians," so that we have "could not" instead of "neither
could." Can there be here any suspicion of inter
polation ? I do not think so (although Westcott
and Hort follow B here as elsewhere), for Mark (v.
26), while giving the same thing in substance, differs
entirely in the words, which is the ordinary relation
between the two Gospel-writers in matters which
they have in common. So there can be no doubt
that Luke originally wrote what we read in the bulk
of mss. Whence, then, the omission ? and whence
the even more abridged form in D ?
I Let us take another instance. Chap. x. 4 1 f., in the
well-known words of our Lord to Martha, there is a very
strong Western evidence for omitting the whole
bulk of them, leaving nothing but : " Martha, Martha,
Mary has chosen the good part, which shall not
be taken away from her." Now, can this in any
possibility be a case of (as Westcott and Hort express
it) Western non-interpolation ? Those editors them
selves do not think so, nor is it, on the other hand,

THE DOUBLE TEXT IN ST. LUKE'S GOSPEL. 149
in any way easy to account for the omission of the
words, if they are genuine, except by the theory
of two original texts, one longer and another more
abridged. Again, in xii. 19 the rich man of the parable,
according to the Western evidence, says nothing
but this : " And I will say to my soul : thou hast
much goods, be merry." This is quite sufficient
for the sense ; but everybody, I should think, would
prefer the longer form given in the other text :
" Soul, thou hast much goods laid up for many years ;
take thine ease, eat, drink, and be merry." Why
should we suppose ancient copyists or readers to have
had such a different taste, as arbitrarily to mutilate
this fine passage ? There was no dogmatic reason
for it whatever, nor any other save mere idleness,
which cannot rationally be thought to have formed
an essential factor in New Testament tradition.
Luke's case, of course, was different ; he did not
cherish his own work so much as others were bound
to do.
The account of the colt which Jesus made His
disciples bring Him from the village (xix. 29 ff.),
is by D given in this abridged form: (31) "And
if any man ask you . . . thus shall ye say unto
him : The Lord has need of him. (32) And they . . .
went their way . . . (34) and answered, The Lord

150 PHILOLOGY OF THE GOSPELS.
hath need of him. (35) And they brought the colt,
and cast their garments upon him," etc. In this last
verse we must consider the Greek words too :
a /3 (D)
Kat rjyayov avrbv 7robs Kat d-ya-yovres rbv ttioAov
tov 'Itjo-ovv, /cat eiripi'spaVTCs e7repi\j/av ra. i/idria ai)T<3v
avrwv Ta IfiaTia eirl tov eir' airbv koi eirej3ifSacrav
rrZXov eTrej3i/3aa'av rbv rbv 'fncrouv.
'Irjcrovv.
You see that 'Irjcrovv stands twice in a, and the
reason is, that after irpbs rbv 'Irjcrovv it was necessary
to bring in rdv irwXov, because eV avrbv would
have been ambiguous, and after that it was again
necessary to repeat 'Irjcrovv. On the other hand,
in |8 the unnecessary irpds rdv 'Irjcrovv has been
left out and the first clause given by the participle,
and in this way the whole sentence has been brought
into a smoother and more elegant form. Now, if such
transformations are to be ascribed to copyists or
readers, I am afraid we shall get a kind of copyists
or readers who are but the creation of our own fancy,
without having had or having now any existence
in reality. I, for my part, am not able to recognize
here anything but the license of an author, who is
handling his own work, and the skill of a writer.
Speaking generally, the case is the converse of what
we find in a passage of the Acts (xxv. 24 ff.). There

THE DOUBLE TEXT IN ST. LUKE'S GOSPEL. 151
Festus, according to the a text, is made to say thus
much : " But when I found that he had committed
nothing worthy of death, and that he himself hath
appealed to Augustus, I have determined to send
him." Of course in reality he said much more than
this, stating the case to his illustrious audience in
its full length, and in accordance with this we have
in /3 a much more detailed speech, but one in which
things are repeated which the author himself had
already told. He did quite well to avoid these
repetitions when he wrote a second copy. So, in
the account of the a Gospel, the things which came to
pass are first predicted and prescribed by Jesus,
and next represented for a second time in their
fulfilment and execution ; that repetition has been
avoided in f3, and yet not wholly, inasmuch as even
the words " and answered, The Lord hath need of him,"
cannot be deemed to be strictly necessary.
I might give a series of cases of a similar
character, the number of omissions in D being in
finite, but do not think that this would be in any
way profitable. So I refer my readers for further
instances to my edition of the /3 text, with which
I think they will be content, and will , be likely
to ask me this question : Are there then no con
verse cases of addition in D? There are indeed
such cases, just as in a of the Acts, but comparatively

152 PHILOLOGY OF THE GOSPELS.
few in number, and some others there are where
neither of the texts has either an addition or an
omission, but the words are different. Let me speak
of these first. There is a personage mentioned by
Luke, who may be unknown to some of my readers,
a man by name Chuzas, steward to Herod the tetrarch
and husband to Joanna, who was one of the women
accompanying Christ (see Luke viii. 3). The name,
of course an Aramaic one, does not occur anywhere
else. Now, if we scrutinize our Latin witnesses very
carefully, we find in I (an old Latin version of the
seventh century, existing in Breslau and published
by Professor Haase) instead of Chuzae, Cydiae. This
is a very ancient Greek name : there was one
Cydias a lyric poet, and another an Attic orator,
mentioned by Aristotle, and another a painter from
the island of Cythnus, and so on. How does the
Latin copyist come by that name ? By chance ?
Impossible. By correction ? Still more impossible.
I say he came by it in the simplest way in
the world, by tradition, which goes back to Luke
himself. That man had two names, one Aramaic
and one Greek, of somewhat similar sound, which
he had adopted as more convenient for the cultivated
and elevated circle in which he lived : just as other
Jews, as early as in the time of the Maccabees, trans
formed their name of Jesus into Jason, and as

THE DOUBLE TEXT IN ST. LUKE'S GOSPEL. 153
modern German Jews called Aaron prefer to call
themselves Arthur. Luke must originally have
written: "of Chuza, who was also called Cydias";
but, when copying first for readers in Syria and
Palestine, he left out the Greek name, and when
copying again for Eoman readers, he left out the
Aramaic one. There cannot be a more simple solu
tion of a puzzling problem, which, if you attempt in
any other way, you will find insoluble.
Now as regards the additions found in D, I think
it best to follow the same method of solution which
has proved practicable here, as far as it will go. The\
best known of the additions in D is that in vi. 5 :
" On the same day He saw a man working on the
Sabbath, and said unto him : Man, if thou knowest
what thou art doing, blessed art thou ; but if thou
dost not know it, thou art cursed, and a transgressor
of the law." I do not see here anything unsuitable
to our Lord ; but these words were likely to give
great offence even to Christian Jews, because the
spirit of the saying is quite that of Paul who, on
the one hand, gives us to understand that it is a
weakness in faith to esteem one day above another
(Eom. xiv. 5 f.), and on the other asserts that never
theless the man so esteeming is not allowed to act
against his own conscience (ibid. 23). It is, therefore,
quite credible that Luke preferred to leave out this

154 PHILOLOGY OF THE GOSPELS.
saying in the form of his Gospel destined for Oriental
congregations, a very considerable part of which
consisted of Jews, whilst in Eome there was no
cogent reason for omitting it.1 It is very remarkable
that for this addition D is the only witness ; there
is even no trace of Marcion's having known it, and
so it seems that it had very early disappeared from
Eoman copies, and was preserved only in Gallic
ones, which may have been derived (if I am allowed
to make a rather bold conjecture, without laying
any stress upon it) from an ancient copy carried
by Crescens (2 Tim. iv. 10) to Gaul. For that
there were written Gospels at that time within the
reach of an evangelist like Crescens, and that every
evangelist undertaking a mission in a new country
took care to provide himself with a written Gospel,
is for me a thing beyond question. But why did
1 The words are for the most part Luke's words : rrj airrj fi/iipa (cp.
e.g. iv Si Tais ij/iipats raiirais, in the same position, Acts vi. 1). ©ea-
adfievos (Gospel, v. 27; Acts xxi. 27, etc.). Tiva ipyatf/ievov (Gospel,
xiii. 14). Ty aa.fSfia.Tui (xiii. 14). YXrrev airtp "AvSpome (these three
words as in xii. 14 ; cf . v. 20 ; xxii. 58, 60 ; avBpunre is nowhere
found beside in Luke, Paul, and James ii. 20). Ei /liv (Acts xviii.
14 ; xix. 38 ; xxv. 11; never in Matthew, Mark, John). OlSas r'nroiets
(Gospel, xxiii. 34). ~M.aKa.ptos el (Gospel, xiv. 14, etc.). Ei Si jit) olSas,
imKarapaTos (only in Paul, Galatians iii. 10, 13, and perhaps John
vii. 49). Koi rrapafiaTTjs (only in Paul and James). Ylapafl. tov
vbfiov (Romans ii. 25, 27; James ii. 11). There is, however, one
difficulty, viz. , that we should expect el Si ovk olSas in New Testa
ment Greek, or else el Si /i$j (ye) without verb, cf. x. 6 ; xiii. 9.

THE DOUBLE TEXT IN ST. LUKE'S GOSPEL. 155
the Eomans strike out the verse in their copies ?
Because they found it was absent from the Oriental
mss., which must have appeared to them, as early
as in the second century, more trustworthy than
their own, or at least sufficient to render dubious
the authority of anything which did not exist in
them ; and of course the Eoman Church, at that
time as in later times, cared much for being reputed
thoroughly orthodox and untainted by untrustworthy i
doctrines or writings.
Neither are we quite devoid of instances of
additions which are much akin to this ; but here I
shall speak only of one addition, which is, it is true,
not preserved in D, nor in any Latin versions, but
only in some Greek minuscule mss. derived from a
common original, which probably was very ancient,
and existed somewhere in Southern Italy or Sicily. _j
These mss. have been carefully studied by my
learned and much esteemed friend Professor Eendel
Harris ; they are commonly called the Ferrariani,
because the affinity of some of them was first
observed by a Dublin professor, the late W. H. Ferrar.
The passage in question is a very long one, and
extremely well known, only not as a part of Luke's
Gospel, but as a part of John. I am speaking of
the section about the adulteress, John vii. 53— viii.
11, which is now generally recognized not to have

166 PHILOLOGY OF THE GOSPELS.
belonged originally to John's Gospel. The reasons
for this decision you may find in Westcott and Hort:
that it does not exist in the whole mass of ancient
and trustworthy Greek Oriental mss., including the
metrical paraphrase of the Gospel made by Nonnus ;
that it is wholly unknown to all the Greek fathers,
down to the time of Chrysostom and Cyril, nay,
down to the tenth or even twelfth century ; that it
is not unanimously assigned to this particular place
in John, but is placed by some witnesses at the close
of this Gospel, by others after vii. 36 or vii. 44 j1
lastly, that it neither has the well-known and very
definite style of John, nor does it suit the tenor of
the narrative into which it has been thrown. But
as there are two places assigned to it in documental
tradition, one in John, and another, as I have said, in
Luke, it by no means follows that it is altogether
spurious, if it is proved not to belong to one of
these places. I may state at once that, according to
Eusebius (Hist. Eccl. 3. 39, 17), either this or a
similar story was told by Papias, and found in the
Gospel according to the Hebrews (KaQ' 'E/3paiovs),
and that it has been conjectured that it came into
John from there; but it is only just to examine first
its claim to a place in Luke's Gospel, and in the case
of this also being proved unfounded, then to refer it
^ee Nestle, Einfuhrung in das Gr. N.T., p. 102.

THE DOUBLE TEXT IN ST. LUKE'S GOSPEL. 157
back to an apocryphal source. Now, which place in
Luke's Gospel is claimed by it ? The place is in
chap, xxi, after the great sermon on the., destruction
of the temple and the end of the world ; in the
Ferrariani it follows the last verse of that chapter,
and the connection is this: (ver. 38) "And all the
people came early in the morning to Him in the
temple, for to hear Him. ( = John vii. 53) And every
man went into his own house. ( = viii 1 f.) Jesus
went unto the Mount of Olives. And early in the
morning He came into the temple, and all the people
came unto Him," etc. Impossible, you will say; and
very justly. My young friend, E. Lippelt, who was
the first to call my attention to the Lucan character
of this section, and to this chapter as a fit place for
it, did not regard it as wholly impossible ; but never
theless I feel compelled to make two changes against
the authority of the Ferrariani : in the first place
wholly to cancel the introductory words, "And every
man went unto his own house," which are absent
from the Latin Corbeiensis (f2), and are, in my opinion
nothing but the link of connection added to the section
in order to adjust it to the place in John, and secondly
to place it two verses earlier. Now, the connection
becomes the following : " And (after the great sermon)
Jesus went unto the Mount of Olives. And early in
the morning He came again into the temple, and all

158 PHILOLOGY OF THE GOSPELS.
the people came unto Him," etc. (the rest of the
section, ending with Jesus' words to the woman).
" And (Luke xxi. 3 7 f.) in the day time He was
teaching (THf Se rds rjfiepas SiSdo-Kwv, imperfect of
custom) in the temple; and at night He went out, and
abode (epepyofievos rjvX'iXero, imperf.) in the mount
that is called the Mount of Olives. And all the people
came early in the morning (wpQpi(ev, imperf.) for to
hear Him." I venture to say that this connection is
so perfect that it cannot be the result of chance, but
must really go back to the author. There is first an
account of what came to pass on the next day, and
after that a general summary of what came to pass on
all of these days, given partly in the same words as
the beginning of the special account, but a little more
circumstantially, since a general custom deserved more
words than the occurrence of a single day. There is
an account somewhat akin to this in the f3 text of the
Acts, xv. 41 — xvi. 4 : " And he went through Syria
and Cilicia, confirming the brethren, and delivering the
decrees of the apostles and elders. After having gone
through these nations, he came to Derbe and to Lystra
(in Lycaonia)." Then (xviii. 1-3) about Timothy; and,
in conclusion (v. 4) : " And going through the cities,
they preached unto them with all confidence the Lord
Jesus Christ, delivering at the same time the decrees
of the apostles and elders who were in Jerusalem."

THE DOUBLE TEXT IN ST. LUKE'S GOSPEL. 159
Since, then, the place in Luke's Gospel claimed
by the section in question (according to the Ferrariani)
really seems to have been its original place, the second
question will be, whether its style may be pronounced
to be Luke's or not. Now, we may safely pro
nounce it so ; but I prefer not to give here the dis
cussion in detail, having already given it elsewhere,1
and the case being that everybody, who is in possession
of a concordance, may inquire into the matter for
himself. That in this long passage, which is besides
of a peculiar kind, there are some words not occurring
elsewhere in Luke, is in accordance with what we
find in any other piece of Luke's writing (see above
on p. 113 f. 118).
To sum up, there is external evidence for the
section's place having been in Luke's xxi,2 it really
fits perfectly well into that place, and its style is
quite that which is to be expected in the case of
Lucan origin. Is this all we may demand ? No ;
for everybody will ask : How, then, was it possible
that the section lost almost completely its original
place ? and how did it pass from Luke into John ?
Unless we see at least the possibility of these things,
we cannot restore it to its place in Luke. Has there
1See my edition of Luke on p. xlviii.
2 There is besides the evidence of a Greek evangelistary (nr. 435),
which gives this section as a part of Luke (iK tov Kard A.).

160 PHILOLOGY OF THE GOSPELS.
perhaps been offence taken at it by some very ancient
and highly authoritative Christian teachers ? But
although offence might be taken, this result, viz., that
this long piece was omitted from all the ancient
Oriental mss., nay, as regards Luke, from almost all
mss., and that it was unknown to all the ancient
fathers of the Greek Church, nay as a part of Luke, to
all the fathers of the Greek and Latin Church — never
can be explained in this way. There is but one
solution possible : the section must give up all claim
to the Oriental form of Luke, and content itself
with claiming a place only in the Eoman Luke.
Under this supposition a large part of the difficul
ties above stated vanish at once, and the rest cease
to be insurmountable. Suppose that early in the
second century the Eoman Church found that it
had in its Gospel a very remarkable narrative, which
was disclaimed by all Eastern copies and authorities.
To allege that this form of the Gospel had been given
to the Eoman Church by St. Luke himself, at a time
anterior to the Neronian persecution, by which the then
existing community of Christians had been almost
entirely destroyed, would be not so very easy, those
ancient facts being no longer upon record. On the
other hand, the charge of having an unauthenticated
Gospel must be avoided at any price. So the Church
of Eome, not finding in itself that strength of resistance

THE DOUBLE TEXT IN ST. LUKE'S GOSPEL. 161
which it possessed at a later time, banished from
its copies of Luke's Gospel all the seemingly un
warranted parts, and for that reason the section in
question is not known, as far as we can see, to any
Latin father before the fourth century. Perhaps
even Marcion ignored it, although the fact of its not
existing in Marcion's Gospel does not go far enough to
prove that : he might have his individual reasons for
omitting this section as well as a great many other
passages. Nevertheless, the section did not wholly
perish, but was permitted to exist in some copies
as a kind of appendix to Luke's Gospel, or (as now
in Westcott and Hort's edition) to the Gospels in
general; even the record of its original place must
have been preserved somewhere, by some adscript, for
instance, made to the end of the twenty-first chapter :
Here comes in the passage on the adulteress. For if
there existed in an archetype of the Ferrariani a remark
of this tenor, a copyist might be induced to take out
the section from St. John and put it in here in such a
way, that it is not quite at the right place, and yet
does not drop the additional link to John which it had
received at its beginning. So far we do not find any
serious difficulty ; but the other question still remains
unanswered : How did it come into John ? It existed
there, though only in the West, as early as in the
fourth century, as we gather from the attestations

162 PHILOLOGY OF THE GOSPELS.
by Ambrose, Augustine, Jerome ; we even see that it
was held in much esteem, and that its omission in a
part of the copies was ascribed to the supposition
of some jealous husbands having struck it out. Nor
did Jerome venture to banish it from his corrected
Latin version, which he certainly ought to have
done on his principle of going by the standard
of the Oriental Greek copies. So, having got a very
strong hold on its new seat in John, it finally gained
the East too, and is nowadays even better known
than a great many other stories in the New Testament.
But it remains quite unknown and undecided, who
brought the section into this place in John ; even
the reasons for such an act can be but imperfectly
guessed. If it formed, at some time in the third
century, a kind of appendix .to the Gospels, and
seemed to some authoritative person in the West,
whoever he was, to be both authentic and of a special
value, he would seek out for.it a fit place in one of the
four Gospels, and might imagine that Christ's not
judging the woman stood in a close relation to John
viii. 15, where He says: "Ye judge after the flesh;
I judge no man," wherefore he put in the section at
the next fitting place before that verse. And so
we may leave this interesting question, stating so
much, that as long as the chain of external and
internal evidence remains unbroken, by which it is

THE DOUBLE TEXT IN ST. LUKE's GOSPEL. 163
proved that the section about the adulteress is both
Lucan and absent from the Oriental Luke, we have in
it the firm proof for the existence of a different early
Eoman Luke, that is, for the existence of two authentic
forms of Luke.
There is still one question which may be asked:
Has Luke intentionally left out this piece in a ?
or did it come to him from some fresh source, after
he had written a? If I am to pronounce on such
a question, I should regard the second alternative
as rather improbable, and should prefer the first,
there evidently being the same reasons for leaving
out this section as for not giving the verse, vi. 5.
Christian Jews might have taken offence at this
position of Jesus regarding the Mosaic law, which
might seem to be practically abolished by Him. We
must bear in mind, that the woman's being stoned,
according to the law, was actually out of the question ;
by answering in the affirmative, Jesus would have
given the Jews an opportunity for denouncing Him
to the Eoman authority, which had deprived the
Jews of the right of putting anybody to death (see
John xviii. 31). But whoever did not bear that
in mind, and was, although a Christian, still zealous
for the law (see Acts xxi. 20), might be offended,
and Luke did not care to give unnecessary offence
by publishing this narrative in a country where there

164 PHILOLOGY OF THE GOSPELS.
were many thousands of Christians of that description
(Acts I.e.). In Eome, on the other hand, he might
feel himself quite at liberty and not bound to any
reticence.

CHAPTEE X.
SOME OTHEE TEXTUAL DIFFICULTIES IN
LUKE'S GOSPEL.
I shall not give other proofs for the double form
of Luke's Gospel ; but nevertheless we cannot leave
this Gospel yet, since there are many other passages
in it of high importance, of which some elucidation
may be given from textual criticism. We shall follow
the order of the book itself.
In ii. 7 there is a reading attested by Epiphanius
in his book against heresies (51. 9 ; ii. 460 f., Ddf).
Luke relates, says he, that the Child had been
wrapped in swaddling clothes, and lay in a manger
and in a cave (ev cpdrvy Kal [Iv] crirrjXaiw), because
there was no room in the inn. Now this cave
is even now generally known, not from Luke
or any other part of Holy Scripture, but from
tradition ; there is nowadays existing in Bethlehem
the cave and a splendid basilica, which has been

166 PHILOLOGY OF THE GOSPELS.
built over it. The tradition is attested as early
as Origen (against Celsus, 1. 51), and Justin (Dialogue
with Tryphon, lxxviii), unless the latter is directly
quoting from Luke ; his words are : " Since Joseph
had not in that village a house where he might
lodge, he found accommodation in a certain cave
belonging to the village." I hesitate to reject this
evidence, especially that given by Epiphanius, and
do believe that there were copies of Luke containing
that reading. It differs from the ordinary one in
two points more, viz., that Epiphanius gives " He lay,"
where the corresponding word in the common Greek
text is not "laid" (eOrjKe), but "laid Him down" (dve-
KXivev), and that in the next clause he omits " to them "
(avrois). Now these two readings are likewise found
in Latin witnesses and partly in Justin, and so we
may easily be induced to ascribe the whole of Epi
phanius' reading to the form p", although it cannot
be shown in what way the writer came by that
form, or by this part of it, whether directly or in
directly ; for it may well be that he is borrowing from
another writer. At any rate, this form of the pas
sage is quite in harmony with itself, and may seem
preferable to the ordinary one. It could not be said,
" they laid Him down in a manger and in a cave," but
"laid (the general and indefinite word) Him in a man
ger and in a cave " might be said ; moreover, to " inn,"

TEXTUAL DIFFICULTIES IN LUKE'S GOSPEL. 167
or generally, a "place where one could find lodgings"
(for the Greek word KaraXv/ia by no means excludes
private accommodation), the cave gives a much
better contrast than the manger, which (if the cave
is left out) may be supposed to have belonged to
the inn or to the private house. I scarcely need
to mention the well-known fact, that caves were
and are much used in the Orient for stalls, in
which case they must of course contain a manger.
So I should say that Luke's archetype is rendered
here by /3 and not by a, and that possibly the
preference given to the more expressive word " to
lean " (avaKXiveiv) had induced the author to leave
out the cave in a. Instead of " she wrapped up "
and "she laid," there is in the Latin codex e the
very commendable reading, "they wrapped . . . they
laid " ; the other one may be a mere corruption
occasioned by the preceding, " she brought forth "
(ereKev). Neither Epiphanius nor Justin tells against
this correction.
We proceed to the third chapter, where the baptism
of Christ is told, and in connection with it His
genealogy is given. Whence this connection ? Would
not the proper place for the genealogy be either in
the first or in the second chapter ? Editors are
accustomed to begin a fresh paragraph with the
genealogy, or to leave a blank space, signifying that

168 PHILOLOGY OF THE GOSPELS.
they find here a break and not a connection. But
if we want to understand this strange sequel, we
have only to look into D. Not that here again
a difference of the two forms is to be recognized ;
I am very far from pretending this solution to be
as it were a key which unlocks all doors. On the
contrary, there are very many cases where the reading
of D is a mere corruption, and other cases where
the reading of the a-witnesses is nothing but corrupt.
Now, if there are manifold other causes and kinds
of alteration, the one which most predominates in
the Gospels is their being jumbled with each other.
We must therefore be on our guard, whenever one
of two contending readings is at the same time uni
versally attested in another Gospel or other Gospels,
that we may not be deceived by the authority of
mss. which are by no means exempt from conflation.
In the passage in question (iii. 22), the words from
heaven are according to the great bulk of mss. :
" Thou art my beloved Son ; in Thee I am well
pleased." But, according to D and some Latin
witnesses (this evidence being supported by Justin
and by other fathers), the words are, " Thou art
my Son, to-day have I begotten Thee." If we
look into Matthew and Mark, we find in Matthew
(iii. 17) very nearly, and in Mark (i 11) exactly
the same reading as in the common text of Luke.

TEXTUAL DIFFICULTIES IN LUKE'S GOSPEL. 169
This must render us suspicious, and make us attend
very carefully to other individualities of Luke's passage,
in order to see which of the two readings is in
accordance with those. Now, the words following
in Luke are these : " And Jesus Himself began to
be about thirty years of age, being, as was supposed,
the son of Joseph, which was the son of Heli," and
so on. We have seen already that there is no
connection between this sequel and the preceding
words as they are commonly read ; but there is a
very clear connection if we take the words given
by D. The " to-day have I begotten Thee " stands {
in opposition to the thirty years, and the "Thou
art my Son," likewise to " being as Was supposed
the son of Joseph." This therefore is the genuine
reading, and the other one a product of assimilation \
to the other Gospels ; for I cannot believe that there
was ever a material discrepancy between the two
authentic forms, as there would be in this case if
we chose to regard one reading as that of a and
the other as that of /3. The words in the following
verse offer many and very great additional difficulties
which I cannot expound here ; I will only point out
that the "began to be" (23), dpydfievos, appears to
be a corruption from epydfievos, " when He came "
scil. to the baptism (epydfievos is given by the
minuscule codex 700, see above on p. 61 ; epydfievos

170 PHILOLOGY OF THE GOSPELS.
eirl rb ftdirricrfia, by Clement of Alexandria, and
together with the corruption dpydjuevos by Irenaeus),
and that the rest of the verse must, in my opinion,
be restored in this way, that the qualifying clause,
"as was supposed," refers both to the thirty years
and to the fatherhood of Joseph.1 Certainly a
disciple of Paul could not ignore the eternal pre-
existence of Christ, and the passage of the Psalm
(ii. 7), which is likewise referred to Christ by Paul
in the speech related in Acts xiii. 33, and by the
author of the Epistle to the Hebrews (i 5 ; v. 5),
might be interpreted not of a definite day, but of
the whole of eternity. But as these are purely
theological questions, I shall not venture to enter
in them further.
As for the genealogy immediately following, the
difficulties with reference to that in Matthew are
enormously great, and far surpass my ability to solve
them. It seems to me an unwarranted supposition that
the genealogy given by Luke is that of Mary and not
of Joseph, although, by the way, the Davidic descent
of Mary is also attested by the Western reading in
ii. 4, 5, which runs thus : " And Joseph also went up
unto the city of David, which is called Bethlehem, to
1 The text finally adopted by me is this : t]v Si 'I-rjo-ovs [ipxbf-evos]
lis irdv rptaKOVTa, ws ivofitfcro, vibs 'Icikttji^ k.t.X., "Jesus was, when
He came (to the baptism), about thirty years old, as was supposed,
and the son of Joseph," etc.

TEXTUAL DIFFICULTIES IN LUKE'S GOSPEL. 171
be taxed with Mary his wife,1 because they were of
the house and lineage of David." This reading gives
also a good reason for Mary's accompanying Joseph on
this journey, instead of remaining in Nazareth, which
would otherwise have been more convenient to her
condition. But as Luke expressly says, " the son of
Joseph, who was the son " etc., it is clear that at least
according to his own opinion, he did not give here the
forefathers of Mary. What I have to discuss in this
place is the very remarkable fact that, if we are to
credit D, the discrepancy between Matthew and Luke
is altogether non-existent, because D (but D alone in
this case) gives the same names of the forefathers as
Matthew does. The order of course is the inverse,
being the ascending, as we call it, while Matthew has
the descending order. But is not this a clear case of
assimilation of one Gospel to another ? After having
decided the critical question we were treating just
now, by giving the preference to the non-assimilated
reading, we are bound to abide by the same principle,
the more so as cases of assimilation are even more
frequent in D than in other mss., and as D stands
1 " The espoused wife " of the ordinary text is a very clear
corruption, due to an assimilation to i. 27 (where the case is quite
different), and to dogmatic prejudices like those which have
influenced the text of Matthew (see above on p. 86). At the time
of chap. ii. Joseph had already " taken his wife unto him "
(Matt. i. 24).

172 PHILOLOGY OF THE GOSPELS.
quite unsupported. It is true that there is a special
reason which may be alleged in favour of D, viz., that
the list of names is not quite identical with that in
Matthew. The latter is incomplete, the number of
names after David until the carrying away to Baby
lon having been reduced to thirteen, by the omis
sion of four kings ; now D in Luke has these four
kings, and the supposition may be made, that Luke
himself has given in his later edition this full list,
which he found not in Matthew but in Matthew's
authority, and which seemed to him more trustworthy
than his own formerly given.1 But if Luke used the
same author as Matthew, and had access to the full
list, he ought to give more names in another part too,
which must be deemed to be even more incomplete
than that from David to the Babylonian captivity.
Matthew divides his list into three parts, each of
fourteen names as he expressly states (ver. 17); the
second part, that from David downwards, contains now
but thirteen, very likely because of the omission of
one name (that of Joiakim or Eliakim) in the mss.2
1 A similar supposition has been made by Ferd. Graefe (see
above on p. 140) in Theolog. Studien und Kritiken, 1898, 123 ff.
He relies especially on the different spelling of the names in D
and in Matthew, e.g. Abia M., Abiud D in Luke. But Semitic
names and words in the New Testament are hardly ever without
variations, and what is to be compared with D of Luke is the
Western Matthew (D itself wanting in this part). Now the
Western Matthew (c, k, etc.) gives Abiuth, Abiud, or Abiu.
2 See Graefe, p. 124.

TEXTUAL DIFFICULTIES IN LUKE'S GOSPEL. 173
That part comprehends a space of about four hundred
years, and ought to contain, as we. have seen, seventeen
names ; the three others must have been left out by
Matthew himself. Now the third part (which in.
Luke comes first), comprehending about six hundred
years, ought to have about twenty-six names, and yet
it has but fourteen in D of Luke as in Matthew.
We are bound to conclude therefore that the four
names come from the Book of Kings, and their
insertion may be due to any reader acquainted with
that book ; there are even in Matthew various
readings giving this supplement. In this case there
fore the evidence of D must be rejected ; but there
is still another one in the same genealogy which is of
a different nature. The common text of Luke has
between Sala and Arphaxad (ver. 35 f.) one Cainan,
in accordance with the Greek Old Testament, but at
variance with the Hebrew, where this personage is
altogether wanting. Now D omits this Cainan. Is
this a correction made by means of the Hebrew text?
That seems to be very unlikely, since the readers of
Luke, with a very few exceptions (as Jerome), were
unacquainted with Hebrew. But the insertion of
Cainan into the common text may have been made
by means of the Greek Old Testament, just as in
the case of the four kings. So here we shall con
versely adopt the text of D, as in ver. 22, but

174 PHILOLOGY OF THE GOSPELS.
as we there did without any reference to the special
Eoman form.
In the beginning of the eleventh chapter, where the
prayer of our Lord is given, there are again very
serious variations in the reading, both in the intro
duction to the prayer and in the prayer itself. The
introduction as far as regards the disciples' words :
' " Lord, teach us to pray, as John also taught his
disciples," is the same everywhere. But in the
following words of our Lord D gives this amplified
text : " When ye pray, use not vain repetitions as the
rest do ; for some think that they shall be heard for
their much speaking." Why, this is again Matthew,
, you will say at once. Only that Matthew has not
" the rest," but " the heathen,"1 nor " some think,"
I but " they think." The expression, " as the rest do "
(ws ol Xoiirol), is besides well in accordance with Luke ;
we may compare xviii. 11, "I thank Thee that I am
not as the rest of men " (co? ol Xoiirol rwv dvOpwirwv),
while there is not any close parallel to it in the other
Gospels (John never uses Xonrds). It is even more
important that both the general tenor of this intro
duction, and in particular the words " as the rest do,"
are extremely well suited to all that precedes or
follows. Shortness is recommended ; now, our Lord's
prayer, according to Luke, is even shorter than accord-
1The Vatic. B gives not iBviKol but iiroKpiral, by assimilation to the
preceding verses (2, 5).

TEXTUAL DIFFICULTIES IN LUKE'S GOSPEL. 175
ing to Matthew. It is John the Baptist whose example 1
is alleged by the disciple ; John, then, and his disciples
must be included by the expression, "the rest," and the
fact of John's disciples making no fewer prayers than
the Pharisees is attested by Luke, and by Luke alone
(v. 33): "Why do the disciples of John fast often,
and make prayers, and likewise the disciples of the
Pharisees ; but Thine eat and drink ? " where the
addition " and make prayers " is absent from Mark
and Matthew. I do not think that a reading may
be safely discarded as worthless, when there is so much ¦.
to be said in favour of it. But if it does not come
from Matthew, whence does it come ? All the words,
with these few exceptions, are the same as in Matthew.
How shall we explain that ? Of course by the
assumption, that Luke and Matthew had some Greek
source in common ; that assumption must, in my
opinion, be made at any rate, because verbal agreement
between Luke and Matthew, if rare upon the whole,
exists nevertheless in some measure, and because that
Luke borrowed from Matthew is excluded by cogent
reasons, since he did not know any Gospel bearing
the name of an apostle (see above on p. 16). But
Matthew may have borrowed from Luke. If you
apply this to the passage in question, you have already
granted that D's reading is genuine Luke, and even
if you except this one case, you are sustaining a thesis

176 PHILOLOGY OF THE GOSPELS.
which you will find very hard to prove.1 Well, you
will say, I grant the common source, but how is it to
be explained that D alone has the words ? For there
is really no other witness for them but D, as in the
case of vi. 5, or in that of the genealogy. Since it
would be untenable to assume an omission in the
common text, nothing remains but to recognize here a
difference between the two forms. But f3 is the
shorter one, not a. True, if you except some cases where
Luke left out in a what might give offence to Oriental
readers. But this is not the case here. On the
contrary, might there not be Oriental readers still
adhering to John the Baptist ? In Eome there was
of course no one of that kind ; but for Palestine the
persistence of John's sect is well attested, and Luke,
when writing for those countries, might have his
reasons for considering the feelings of that sect.
But there is indeed a strange mixture of good and
bad in this codex Bezae (as there is, by the way, in
every " good " and ancient manuscript). Here it
alone has preserved something valuable, and in the
next verses it is in agreement with the mass of
untrustworthy witnesses, which exhibit the Lord's
1 The thesis has been sustained (in a very clever way) by G.
Schlager in Theol. Studien und Krit., 1896, 83 ff. See also my edition
of Luke's Gospel, xix. f. Schlager, too, does not always attend to
various readings. Matt. xxii. 35, vo/itKbs is an interpolation coming
from Luke.

TEXTUAL DIFFICULTIES IN LUKE'S GOSPEL. 177
Prayer not in the shorter form of genuine Luke,
but in the enlarged one of Matthew. I need not
dwell on a question decided long ago. Our best
witnesses give the prayer in Luke in this form :
" Father, hallowed be Thy name. Thy kingdom come.
Give us day by day our daily bread. And forgive us
our sins, for we also forgive everyone that is indebted
to us. And lead us not into temptation." Or, is not
even this genuine Luke? That it is not is asserted
by no less a personage than the well-known Gregory
of Nyssa, who expressly states that Luke has instead
of " Thy kingdom come," " Thy Holy Ghost come upon
us and make us clean." Nor does he stand alone and
unsupported : there is, in the first place, another
ecclesiastic writer, Maximus Confessor, attesting the
same; secondly, at least one of our own mss., the
minuscule codex 700 (see above on p. 61 and 169),
actually has this reading;1 thirdly, there is Marcion,
in whose Gospel, according to Tertullian, these words
stood side by side with " Thy kingdom come," the first
pait of the prayer, " Hallowed be Thy name," having
been suppressed. Lastly, D itself has preserved the
two words, " upon us " (ecp' rjfias), which stand in D
before eXO'erw. As this appears to be the result of
a conflation (see Professor Sanday in Westcott-Hort)
so also is Marcion's text ; in the copy used by
1 'EXSiroi rb irvevfia aov Tb dyiov tip' i)/ia.s Kal KaBapiaarta Tj/ias.
M

178 PHILOLOGY OF THE GOSPELS.
him the original Lucan words, " Hallowed be Thy
name," had given way to the adscript coming from
Matthew, " Thy kingdom come." For it is not likely
in any way that there were in Luke's original form
two adjacent clauses beginning with the same word
eXOerw. I believe therefore that Gregory and codex
700 give the true Lucan text, and that Luke's form of
the prayer differed even more from Matthew's than is
commonly thought, of course as well in the Oriental
copy as in the Eoman, so important a difference
between the two being wholly incredible. But which
form, or which author, is more deserving of our trust ?
Luke or Matthew ? Theologians perhaps will not
hesitate to give an answer ; but I deny the right of
putting the alternative, until another question is settled.
Are both of our authors speaking of the same thing ?
Certainly, it will be answered, for the introduction
is the same in both, if D's reading there is to be con
sidered genuine. Matthew cut off the question asked
by the disciple in order to insert the prayer into the
great sermon. I was indeed just now supposing a
Greek authority common to Matthew and Luke, and it
may be argued that Luke has preserved the true form
given by that authority. And Matthew, I shall
answer, has combined and welded that authority with
another authority, or else Luke has done the same.
What means are there to decide such questions ? As

TEXTUAL DIFFICULTIES IN LUKE'S GOSPEL. 179
long as scholars dream of one definite primitive Gospel,
in open contradiction to Luke's proem, they will both
raise and answer them ; but as soon as that un
warranted supposition is removed we get rid of a host
of inextricable questions. As for the welded character
of either account, it is to be noticed that the words
of the prayer in Luke and Matthew, even when there
is a substantial agreement between them, are but
partly the same, and that more comes into the ques
tion than abridgment or enlargement.
In the account of the Supper of our Lord (xxii. 7
ff.) the divergencies between the different witnesses for
Luke's text are uncommonly great, and its ordinary
form is indeed far from satisfactory. I have proposed
elsewhere x a very radical solution, striking out not
only vers. 19& and 20, but also ver. 19a, and by
these means establishing a narrative both clear and
consistent in itself, but not containing any longer
the institution of the Lord's Supper. Now, I hardly
need say, that I have not been determined by any
theological reasons. The audacity and presumption
of theologians — I speak chiefly of some German
theologians — is nowhere exhibited more scandalously
than here : they mount by the sole force of their
1 See Theolog. Studien und Kritiken, 1896, p. 733 ff. The subject
has been very carefully expounded in that same journal by F.
Graefe, 1896, p. 250 ff. ; 1898, p. 134 ff,

180 PHILOLOGY OF THE GOSPELS.
genius higher than St. Paul, into the very mind of
our Lord, and bring back revelations according to
which the Christian Church, nay even its founders,
the apostles, have been strangely deceived about
the real sense of this institution, or about its being
an institution at all. As for myself, I have no desire
to mount their Pegasus, who will very soon throw
his rider, a new Bellerophon, and am content to walk
on earth and to scrutinize the evidence of MSS., not in
order to get a sublime recognition of " spiritual things,"
which are far above the reach of the " natural
man" (1 Cor. ii. 14), that is both of philology and
of scientific theology, but to find out, if possible, what
Luke or another of our authorities has really written.
Now the vers. 19& and 20, which closely agree with
Paul (1 Cor. xi. 24 f.), are left out by D and other
witnesses; ver. 19a, which agrees with Mark, is not
left out by any witness; but its position not being the
same in all the witnesses, that verse too becomes
suspicious. Moreover, we must find a reason for the
whole mass of variations, that is to say an original
text, which invited, as it were, readers to alterations
and additions ; at the same time the required text
must be irreproachable from another point of view,
which we may suppose to have been that of the
author. The text of Westcott and Hort, who leave
out vers. 19b, 20, but retain 19a, does not answer to

TEXTUAL DIFFICULTIES IN LUKE'S GOSPEL. 181
the latter condition, since the institution of the cup
is wholly wanting ; Luke is not to be supposed to
have given a mutilated account, but either no account
at all, or a complete one. The latter alternative
being apparently excluded, if we are to explain the
variants — for the complete, and at the same time,
orderly text would not have given any offence — there
is but the former remaining, that is the supposition
that Luke chose to leave out what was commonly
known to his readers. As a matter of fact, this
point is not his only omission, since he does not
relate anywhere how the betrayer was disclosed by
Jesus, nor is he the only evangelist who omits the
institution of the Lord's Supper, since the case is the
same in John. But later readers did not understand
his intention, and either inserted both bread and cup
from St. Paul (vers. 19&, 20), or, wrongly imagining
that the cup was already there (in ver. 17 f.), they
inserted the bread from Mark. I entreat my readers
to read the narrative for themselves, leaving out ver.
19 f, and to ask themselves if this form is not quite
satisfactory under the single supposition, that it was
not Luke's intention to relate the institution of the
Lord's Supper. The sequence, " For I say unto you,"
etc. (ver. 18), and "But (irXrjv) behold, the hand of
him that betrayeth me," etc. (ver. 21), is much like
that in xviii. 8 : "I tell you that . . . Nevertheless

182 PHILOLOGY OF THE GOSPELS.
(irXrjv) when the Son of Man," etc., or in xix. 26 f . :
" For I say unto you, that . . . But (irXr)v) those mine
enemies," etc. Leaving the matter to the reader's
consideration, I add only this, that the insertion from
Paul must have been made in the very oldest times
by some man who possessed Paul's writings, but did
not possess either Matthew or Mark ; it is self-evident
that there actually was a time when many persons
and Christian communities were in that condition.
There is an expanded text in D and other witnesses
(among which for once Origen comes in) in ver. 24 ff.
of the same chapter ; as there is no material difference,
I recognize here the original form of the Eoman text.
The same text has something more on the healing of
the servant whose ear had been cut off by Peter (ver.
51). Then, in ver. 53 D gives a very remarkable
reading, by which the text is considerably improved.
Instead of : " But this is your hour and the power of
darkness " (rov o-Korovs), D gives : " But this is your
hour and your power, darkness " (rb o-Koros). The
words before these are : " When I was daily in the
temple, ye stretched forth no hands against me," and
it strikes one at once how much the text gains by
the clear and strong opposition coming out between
day and darkness. The ordinary reading must be due
to a copyist who had in mind Paul's words (Col. i. 13):
" Who hath delivered us from the power of darkness.'

TEXTUAL DIFFICULTIES IN LUKE'S GOSPEL. 183
In chap, xxiii. I notice first of all the apocryphal
addition to ver. 5, which is given by two Latin wit
nesses : " And He alienates from us our sons and our
wives ; for they do not baptize (or ' wash ') themselves
(cf. xi. 38), nor make themselves clean." Similarly
Marcion's Gospel had in ver. 2 : "We found this fellow
perverting the nation, and destroying the law and the
prophets (for these words there are Latin witnesses too;
cf. Matt. v. 17), and forbidding to give tribute, and
alienating the wives and the children, and saying," etc.
The verses 10, 11, 12, are skipped in the Syriac
palimpsest from Mount Sinai, and Professor Wellhausen1
thinks this omission to be well-founded, since in ver. 1 5
Pilate says to the Jews : " Herod (for I sent Him
[Jesus] to him) has not found," etc., telling them what
they were evidently still ignorant of; so the statement
in ver. 10 ("And the chief priests . . . accused
Him," before Herod) is excluded. But that reading in
ver. 15 is not the only one, nor is it, in my opinion,
the right one ; there is another : " for I sent you to
him," and a third, which I prefer to the other two :
" for he has sent Him back unto us." Was it possible,
that Pilate, declining to be judge in this case, sent the
accused to another tribunal, and did not send at the
same time His accusers ? The double text of Luke
is out of question here ; for D gives a special reading
'See Gotlinger Nachrichten, 1895, p. 9.

184 PHILOLOGY OF THE GOSPELS.
for ver. 12, materially unimportant, but very remark
able because of the words.1 So the omission of those
verses by the Syriac witness must be considered acci
dental. We pass on to ver. 44 f. : "And it was about the
sixth hour, and there was a darkness over all the earth
until the ninth hour. (45) And the sun was darkened,
and the veil of the temple was rent in the midst."
This again is one form of the text among many, and
by no means an irreproachable one : why " and the sun
was darkened " after " there was a darkness " ? On
the other hand, the Alexandrian reading in 45a, " the
sun having been eclipsed " (eKXnrdvros), is open to the
even more serious objection : How was it possible in
the very midst of the lunar month, which was the
time of the passover ? I think this is an erroneous
adscript (see Origen in Westcott-Hort), and go back to
the wider attested reading (which is also, in the main,
that of D), the case being that a very slight emendation,
and an attested one, puts things right. Strike out the
'' and " before " the sun," according to the attestation
of five Latin witnesses, and of D, which however has oe
after eaKorlaOrj, and insert the same particle with one
of these witnesses (the Vercellensis a) before " until,"
l"Ovres Si iv dvSla b UiXcLtos Kal 6 'HpoSSns iyivovro tpl\oi iv airrj
ttj -ij/iipa. This is a very good instance of the occasional refinement
of style found in fi. 'AvSla is new in Luke, but we have seen that
he is constantly introducing words not again used by him.

TEXTUAL DIFFICULTIES IN LUKE'S GOSPEL. 185
and you have a text as good as need be : " there
was a darkness over all the earth, and until the
ninth hour the sun was darkened." The latter part
of 45 is put by D alone after 46, which is the order
of events given by Matthew and Mark.
A very interesting addition is made by D, together
with the Latin c and the Sahidic version, to ver.
53 ; there begins besides with that verse a series of
variants extending over the beginning of the next
chapter. D's text in 53 is : " and laid it in a
sepulchre . . . and after it had been laid there,
he put unto the sepulchre a stone, which twenty
men could scarce roll." Let me first say that the
stone is there in Luke, as in the other Gospels, in
the beginning of the next chapter (ver. 2); it must
therefore seem strange that it was not mentioned
before, as in Matthew and Mark.1 In the text of D
we miss nothing, neither can it be said that that
text has received a supplement from another Gospel,
since the words, and not only these, are quite peculiar.
Even too peculiar, you will say. The addition in
D has indeed, at first sight, a somewhat strange
appearance ; my learned friend, Professor Eendel
Harris, was reminded by it of a passage in Virgil,
xIt is true that John too says nothing of the stone being put
on the sepulchre, and nevertheless mentions it afterwards (see xix.
42; xx. 1).

186 PHILOLOGY OF THE GOSPELS.
and there lies even nearer at hand a similar one
in Homer, on the stone which the Cyclops put into
his door : " a stone which not even twenty-two carts
might carry away" (XlOov S' eireQrjKe Ovprjcpiv di3pifiov
ovk av rov ye Sow ko.1 e'lKOcr' dfiapai ecrOXai rerpaKVKXoi
air' ovSeos dyXlcrcreiav, Odyss. ix. 240). There we
have the twenty or twenty-two (carts in this case),
and we have the verb eireQijKev as in Luke (3, while
Matthew and Mark use irpoo-KvXlw. But let this
come from Homer ; what is that against Luke's
authorship ? Must we not accept it for a certainty
that Luke, the physician of Antioch, had gone through
his Homer ? Nor are we devoid of other proofs for
this obvious fact : in Acts xxvii. 41, he says of
a sudden, eireKeiXav rrjv vavv, although neither he nor
any other New Testament writer elsewhere employs
the obsolete word rj vavs, instead of which rb irXoiov
was the common expression (occurring in this same
chapter no less than thirteen times), and eiri-
KeXXw, instead of eVo/ce'XXco, is altogether poetical.
But in Homer, in that same book of the Odyssey,
he had read thus, vrjas . . . eiriKeXcrai (148),
and again (546) vija eKeXcrafiev. But nevertheless
you will ask : Do you really mean to make Luke
embellish this narrative by a touch borrowed from a
heathen poet ? I do not say that the touch is from
Homer, but that the stone seems to have reminded

TEXTUAL DIFFICULTIES IN LUKE'S GOSPEL. 187
him of that door-stone in Homer, and that therefore
the description is similar, but nevertheless neither
unreal nor exaggerated. As he had been in Jeru
salem, he may be supposed to have gone and seen
the sacred places, and there the size of the stone
must have impressed him. Joseph, who was a rich
man, had as many workmen at his disposal as he
liked or thought necessary, and he chose a very
large stone in order that the sepulchre might not
be broken open and profaned by the hostile Jews.
Let us now see the continuation of the narrative in
D, ver. 54: "And that day was the preparation (or
' that of preparation '), and the Sabbath drew on." D
(the same thing in other and fewer words) : " And it
was the day before the Sabbath " (irpocra^j3drov,
cf. Mark xv. 42). Ver. 55, instead of "the women,"
" two women," that is Mary Magdalene, and Mary the
mother of James and Joses, see Mark xv. 47 (Matt.
xxvii. 61). I think Svo, "two," instead of at is right,
since it is attested by Eusebius also ; consequently in
xxiv. 1, the text of D, but also the common text (con
tradicted it is true by the Vaticanus B, etc.) has an
enlargement of the subject : " and certain others with
them." This enlargement would be cancelled as
useless as soon as Svo had vanished from the text.
The end of 55 is again abridged in D, and in 56,
" according to the commandment " is left out by the

188 PHILOLOGY OF THE GOSPELS.
("same ms. But in xxiv. 1 it has an enlarged form :
" Now upon the first day of the week, very early
in the morning, they went (rjpyovro, not rjXQov, came)
unto the sepulchre, bringing what they had pre
pared with certain others with them. And they
reasoned with themselves : Who shall roll away the
stone ? " "And when they arrived (eXOovcrai Se) they
found," etc. That addition comes from Mark xvi.
3, you will say. And yet Mark has not " reasoned,"
but " said " (eXeyov), nor rls apa, which is quite Lucan,
but simply rls ; and if " shall roll away the stone " is
in both, you must notice that the common form of
Luke has the same verb in the next verse. This, then,
is the special text of (3, while in a, Luke, consistent
with himself, had omitted to mention the stone
\ as well here as in xxiii. 53.
I shall state, but not solve, one difficulty more.
Ver. 12 of the common text : " Then arose Peter,
and ran unto the sepulchre," etc., is left out by recent
editors on the evidence of D and its Latin associates,
with which, as Tischendorf thinks, Eusebius agrees.
Comparing John (xx. 3 ff.) we find an agreement
strongly arguing the interpolation of the verse in
Luke. The verse does not leave an apparent gap
at its place; on the contrary ver. 13, "two of them,"
joins with 11 and not at all with 12, which refers to
Peter and John exclusively. So far all is very well ;

TEXTUAL DIFFICULTIES IN LUKE'S GOSPEL. 189
but when we come to ver. 24, we find there the same
fact mentioned : " And certain of them which were
with us went to the sepulchre," etc. Now, is this
text, that of D, and of Westcott-Hort and Tischendorf,
a consistent one ? Why has not Luke given this
important detail in its right place ? Then, ver. 24 too
is an interpolation ; its removal does not leave a gap
any more than 12. But neither have we evidence
for that omission, nor is the verse found in John. As
I have stated, I do not feel able to solve this diffi
culty; as an editor, I was bound to omit ver. 12
like Tischendorf and Westcott-Hort ; but doubts still
remain.

CHAPTEK XI.
TEXTUAL CONDITION AND OEIGINAL SEPAEATE
FOEMS OF MAEK'S GOSPEL.
After having said so much on Luke, I shall be com
paratively brief on the other two Gospels, that is on
Mark and John, for we saw already that on Matthew,
as regards the condition of his text, there is not much
to be said. This is indeed not true either of Mark or
of John ; but with regard to them there is so much
obscurity that I have not much to say about
either. To begin with the personality of Mark, it does
not lie more in the shadow than that of Luke, but even
less. His original name was John, and the Eoman
Mark a surname given in order to distinguish him
(which was indeed much required) from other Johns ;
his mother was called Mary (as perhaps were one half
of the women existing then in Judaea),1 and possessed a
1 In the Gospels we meet with five Marys, and besides them one
Elizabeth, one Anne, one Joanna, one Susanna, and one Martha,

TEXTUAL CONDITION OF MARK'S GOSPEL. 191
rather large house in Jerusalem (Acts xii. 12). It
has been quite plausibly surmised that the young man
having a linen cloth, who was in the company of Jesus
on the night in Gethsemane, and who is mentioned
by Mark alone (xiv. 51 f.), was nobody else but the
author of that Gospel himself. He, as well as his
mother, was a member of the early Church in
Jerusalem, and afterwards accompanied Barnabas and
Paul as an assistant on their first journey, but only
as far as Pamphylia, from whence he returned to
Jerusalem (Acts xiii. 5, 1 3). Barnabas supported him
notwithstanding, but not Paul, for which reason, when
they were to undertake their second journey, they
separated (ibid. xv. 36 ff.).
Paul, in his Epistle to the Colossians (iv. 10),
tells us that Mark was Barnabas' cousin; he himself
had at length forgiven him, and had him at that
time in his company, together with Luke (see also
Philemon, ver. 24). Where that was, whether in
Caesarea or in Eome, is a much discussed question.
Again, Paul says in his Second Epistle to Tirnothy
(iv. 11), which dates from his later captivity in
Eome : " Only Luke is with me. Take Mark, and
bring him with thee : for he is profitable to me
for the ministry." But Mark seems to have been
most of all with Peter, who calls him his son (1 Peter
v. 13), and this agrees with the well-known words

192 PHILOLOGY OF THE GOSPELS.
of Papias, who relates, on the authority of his " presby
ter " (apparently John, see the next chapter), that
Mark had been a follower and interpreter of Peter,
and that this enabled him to write his Gospel (see
above on p. 78). We are not warranted in giving
to the word interpreter (epfirjvevnjs) any other sense
than it commonly has, that of one who translates
the words spoken in one language into another,
and in this case the one language must be Aramaic,
and the other Greek. It is therefore to be supposed
that Peter, when travelling in a country of mixed
population, preached in Aramaic, and made Mark
translate his words into Greek. That sojourn at
Babylon and Peter's epistle fall into a late period ;
Papias' author may be supposed to speak chiefly
of the very earliest period of Gospel-preaching,
which was then confined to Judaea and the adjacent
regions, down to the time when Peter, having
escaped from Herod's prison, went into another
place (Acts xii. 17).1 After that time we find Mark
in Barnabas' company, and then again in Jerusalem
(whither Peter had meanwhile returned), and again,
after Peter's departure to Antioch (and Babylon,
1 1 briefly state that ip/njvevrris yevb/ievos means, "he had been"
(before writing his Gospel), and nothing more. Whether the
function of interpreter had come to a close by Peter's death (as
Harnack asserts), or by Peter's going away into another country
where an interpreter was not needed, is not stated.

TEXTUAL CONDITION OF MARK'S GOSPEL. 193
see above, p. 26), he joins his cousin. Tradition
assigns to him the role of being the first preacher
in Egypt, and the first bishop of Alexandria (see
Eusebius, Hist. Eccl. ii. 14, 1), which see he is said
to have occupied until the eighth year of Nero,
that is 62 a.d. (ibid. ii. 24). It appears, therefore,
that apart from the apostles themselves perhaps
no man was better able to write a Gospel than
Mark, and that he was also a very fit personage
to write the history of the earliest Church in Jeru
salem and Judaea. In a former chapter (see p. 141)
we hit upon the conjecture, that he actually did
write that history, and that Luke used his work for the
first part of his own Acts. So much is quite evident,
that for the story told in Acts xii. the authority
of Mark is claimed, as it were, by the narrator,
when he says (ver. 12) : " He (Peter) came to
the house of Mary the mother of John, whose surname
was Mark " ; and proceeds to say (ver. 17): " He
declared unto them (Mark is in all likelihood in
cluded) how the Lord had brought him out of the
prison." Now, I say that in case this conjecture
is right, Mark must have given the story of the
primitive Church in Aramaic, and not in Greek. For
Luke's authority was an Aramaic one, and if that
authority was a book by Mark, you see the con
sequence. But of course you will require proof
N

194 PHILOLOGY OF THE GOSPELS.
for the major premiss. I say, that the language
in the first twelve chapters of the Acts is markedly
different from that in the later chapters : in the
former Aramaicisms abound, in the latter they are
comparatively very scarce ; from these facts I argue
that the second part is an independent work by
Luke, but the former depends on an Aramaic source.
I think I may dispense with giving the proofs here
in detail, as the Acts are not the proper subject
of the present little book, and as I have given
them elsewhere j1 but at any rate, Professor Nestle's
discovery must be spoken of even here. There is
in the Acts (iii 14) a passage where the ordinary
text runs thus : " But ye denied the Holy One,"
whilst D (together with Irenaeus) has instead of
" denied " (rjpvrjcracrOe) the quite unintelligible reading
eftapvvare (something like " you troubled"). No
doubt " denied '' is right, and e/3apvvare utterly wrong;
but as the supposition of an ordinary corruption
is manifestly excluded, Professor Nestle is undoubtedly
right in supposing that Luke's Aramaic, or Hebrew
source, had here a word which might be easily
read and understood both ways, and that this word
was di"HB5, or (which is much alike in Semitic
writing) dfH35> the former meaning rjpvrjo-acrOe, and
the latter e/3apvvare.2 There was a misunderstanding
xSee my edition of Luke's Gospel, p. xxi ff.
2 Theolog. Studien u. Kritiken, 1896, p. 102 ff.

TEXTUAL CONDITION OF MARK'S GOSPEL. 195
of the Aramaic or Hebrew text, which crept into
Luke's first text, but has been corrected in the
second. But here comes a rather serious objection in the
shape of the question: Did Luke understand Aramaic?
As for Hebrew, that is quite out of question; but
his native town Antioch was situated in an Aramaic
country. Nevertheless the town itself was of Greek
origin, and it by no means follows that any Greek
native of that town must understand the language
spoken by the country-people around it, and perhaps
by a part of the lower classes in the town itself.
Even if he understood and spoke a little ~ Aramaic,
it does not follow that he could read Aramaic books.
Well, then I say that Luke, having got Mark's
book, which was written in Aramaic, got for himself
an interpreter of it, and that the blunder ef3apvvare
is due to that interpreter, and the correction
rjpvrjcrao-Oe is due to a person better informed, perhaps
to Mark himself, if he met with Luke either in
Eome or elsewhere. It is even easier to explain
the Aramaisms in Luke's first part, if he was using
an authority translated from the Aramaic into bad
Greek : copying from that authority, he corrected
the Greek, but not so thoroughly as to abolish all
traces of the Aramaic origin.
Let us now see what will follow from the

196 PHILOLOGY OF THE GOSPELS.
premisses we have hitherto stated. If Mark's second
part was written in Aramaic, then his first part,
that is the Gospel, must have been originally written
in the same language. Perhaps you will oppose
this by Papias, and Papias' authority, the presbyter.
Papias tells us, probably on the same authority,
that Matthew's Gospel was originally in Hebrew,
and that it had been translated or interpreted
(rjpfirjvevcre is again the word) differently according
to the different abilities of the interpreters (co? r)v
Swaros eKacrros). That is to say, Papias' presbyter
knew of different Greek forms of Matthew, besides
the Hebrew (or Aramaic) original, but in the case
of Mark, the interpreter of Peter, he knew only
one Greek form of that Gospel, and nothing at all
of an Aramaic original. Then, I say, he did not
know all that might be known, or even may be
known. Of course there is nothing remaining now
adays of an Aramaic Mark, but of the different
translations there are some traces still left in the
various readings given by our mss. and other wit
nesses. I shall take first a very clear case of triple
tradition. Mark xi. 13, in the story of the fig-tree
according to the Authorized Version, has : " He came,
if haply He might find anything thereon," which
renders this Greek text : r)X0ev el apa rt evprjcrei ev

TEXTUAL CONDITION OF MARK'S GOSPEL. 197
avrfj. But D (with the Latin b, c, etc.) exhibits
the same text in very different words : rjXQev ISeiv
eav ri ecrnv ev avrfj, "in order to see whether there
was anything thereon " ; and lastly, the valuable
minuscule codex 700, in accord with a quotation in
Origen and with three Latin witnesses, has at least
a different construction : rjXQev cby evprjcrwv ri ev avry,
" thinking that He would find something thereon."
I shall not allow you to decide this critical question
upon authority of MSS., but demand an impartial
hearing for the reasons which may be given in
favour of each of the three readings. Material
difference there is none, nor is there dependence on
Matthew (see Matthew xxi. 19 : rjXQev eir avrrjv,
nothing else) ; so we have only to examine the
language. (A) rjXQev el dpa. Why, this is Luke's
style, see Acts vii. 1 (D); viii. 22; xvii. 27. In Mark
there is no et dpa besides this passage (although
there is ISeiv el xv. 36), nor any apa except in
iv. 41, t/? dpa, and there too with an important
variant. (B) rjXQev ISeiv, see Mark v. 14, rjXQov
ISeiv rl ecrnv rb yeyovds. 'Eav . . . ecrnv is grossly
incorrect, but not altogether unparalleled in New
Testament Greek. (C) co? evprjcrwv : very good Attic,
but hardly known to New Testament writers, with
the exception of the Epistle to the Hebrews (xiii.
17). So we may give the preference to B, as

198 PHILOLOGY OF THE GOSPELS.
being most in accordance with Mark; but indeed
the whole of the case looks much like that described
by Papias : " everyone interpreted as he was able to
do " (rjpfirjvevcrev coy fjv Swaros eKaaros), in bad Greek
(B), or in better Greek (A), or in good Greek (C).
We may speak briefly of iv. 41, the passage we
were just now citing : instead of ris dpa ovros ecrnv,
on ko.1 6 avefios Kai rj QdXacrrra avrw viraKOvei, two Latin
witnesses simply give : 'ISe irws K.r.X. ("the wind," etc.),
and this is the ordinary style of Mark, see ii. 24;
xiii. 1 ; xv. 4, while the other reading is Lucan.
But there is a difference here : not only is this
reading Lucan in character, but actually Lucan (see
Luke viii. 2 5) in the very same story ; so the
decision will be that Mark's text here has suffered
by assimilation to the parallel passage in Luke, as
there are of course many cases of this character in
Mark as well as in other Gospels.
Again, in xv. 15, the ordinary text is: "And so
Pilate, willing to content the people, released Barabbas
unto them." But the words, /3ovX6fievos iroirjcrai rb
iKavov rw byXw or f3. rw byXw ro Ik. iroirjcrai, are
omitted by D and two Latin versions, which have
nothing but b Se H. direXvcrev avrois rbv Hapa(3f3av.
There is here no assimilation, except in so far as
Matthew too gives the bare fact without any reason
for it ; nevertheless you might try to explain the

TEXTUAL CONDITION OF MARK'S GOSPEL. 199
omission as occasioned by Matthew, and argue that,
in D, the next words are evidently assimilated to
that Gospel : rov Se 'Irjcrovv cppayeXXwcras irapeSwKev,
etc., instead of ko.1 irap. (irap. Se, B) rbv 'Irjcr. cppay.
On the other hand, it is a well-known fact that
verbal agreement between these two Gospels exists
to a very large extent. Then, the curious thing is
that only here Mark uses \3ovXecrQai instead of QeXeiv
(which was the ordinary word for "to will," while
/3ovXecrQai bears a somewhat learned character), and
that not even to kavov iroieiv occurs elsewhere in
the New Testament ; nevertheless, when we turn to
Luke (who at least in the Acts makes a rather
large use of f3ovXeo-Qai), we find there the similar
phrase rb Uavbv Xa(3eiv (Acts xvii. 9). So we may
recognize here again the forms A (in D) and B
(in the ordinary text), but in a different way, A
bearing an expanded character, and B an abridged one.
In ver. 1 9 of this chapter the same thing recurs :
" And they smote Him upon the head with a reed,
and did spit upon Him, and bowing their knees
worshipped Him." The Latin k (see above, p. 80 f.)
leaves out, "and did spit upon Him"; both D and
k leave out the last of the three clauses. So
Matthew, says Tischendorf. But the matter itself
is in Matthew, with a different word it is true, and
in a different place (yowirerrjo-avres, Matt, xxvii. 29,

200 PHILOLOGY OF THE GOSPELS.
in the place corresponding to Mark's ver. 18); why
then, if there is in D assimilation to Matthew, do
we not find in that MS. the corresponding addition ?
Moreover, the place of the clause in ver. 19 is
altogether wrong, the worshipping belonging to the
mockery told of in 18 and not to the cruelties
in 19, and lastly, the phraseology is again Lucan:
nOevai rd ydvara is a phrase used by Luke alone,
and by him five times, while Mark and Matthew
say yovvirereiv (not occurring elsewhere in the New
Testament), and Paul Kafiirreiv rd ydvara. Then
the clause must be spurious in Mark. Very well,
but it is not borrowed from Luke, although it bears
his character.
There are more passages of the same kind : in
ver. 24, where the soldiers are casting lots upon the
garments of Christ, the words rls rl dpy, which are
good Greek and may find a parallel in Luke (xix. 15,
AE, etc., Tischend.), are omitted by D and other West
ern witnesses, and a little before, in ver. 21, k offers
a reading, agreeing closely with Mark's manner :
("they compel Simon,") fuit autem nomen (a gross
blunder for pater) Alexandri et Bufi, et faciunt eum
crucem baiulare; cf. vii. 25 f. ("a certain woman
came,") nv Se rj yvvr) (rj yvvrj is omitted by b, etc.;
the reading of BD, etc., is rj Se yvvrj r)v) 'EXXrjvls
. . ., Kai rjpwra avrov K.r.X. Then the whole

TEXTUAL CONDITION OF MARK'S GOSPEL. 201
ver. 28: "And the scripture was fulfilled, which
saith, And He was numbered with the transgressors,"
which is omitted by N and B, and D and k, and
many other good witnesses, is in accordance with
Luke even more than with John : see for the Old
Testament passage, Luke xxii. 37, and for the form
of introduction, Acts i. 16: " This scripture must
needs have been (or 'be ') fulfilled." You see that
in this case the evidence against the apparently
Lucan form is not confined to the West, and if
you look further on, you find that the forms A
(apparently Lucan) and B (apparently original Mark)
are by no means connected in a definite way with
definite witnesses.
In chapter xiv. 4 the common reading is : "And
there were some that had indignation within them
selves." The words and constructions used are quite
in Mark's style. But D and Latin witnesses give
the reading : ol Se fiaQrjrai avrov Sieirovovvro. This
SiairoveicrQai (to feel pain or anger) is a very rare
word, occurring but twice in the New Testament,
namely in Acts iv. 2 ; xvi. 18. D's text therefore
is here A and not B, and likewise in xiv. 25, ov
firj irpocrOw irieiv, instead of (ovKeri) ov firj iriw. For
this Hebraizing use of the verb irpocrnOevai in the
sense of " again " is similarly found in Luke
xx. 11 f . ; Acts xii. 3 (where the middle voice

202 PHILOLOGY OF THE GOSPELS.
irpoo-rlOecrOai is employed). Lastly, in xvi. 1, even
this distinction between A and B fails to suit the
facts, inasmuch as both readings there bear the
character of A. "And when the Sabbath was past ''
(koi Siayevofievov rov cra/3/3aTou), a phrase exclusively
Lucan, see Acts xxv. 13; xxvii. 9. D has instead
of this and the following names of women nothing
but ko.1 iropevOeiaai, the subject remaining the same
as before. But iropeveo-Qai too, although seemingly
a quite common verb, is never used in Mark,
except three times in the spurious close of his
Gospel (xvi. 10, 12, 15), while Luke employs it
like Matthew and John. It is, besides, evident
that the text of Mark in this place has suffered
seriously : " And Mary Magdalene and Mary the
mother of Joses beheld where He was laid (xv. 47).
And when the Sabbath was past, Mary Magdalene,
and Mary the mother of James, and Salome, had
bought sweet spices," etc. Is Mary the mother of
Joses a different person from Mary the mother
of James ? By no means : a little before (xv. 40)
it was said : " And Mary the mother of James the
less and of Joses." A second difficulty : the " had
bought " of the English version is an attempt to
establish a harmony between Mark and Luke, who
tells us that they had bought the spices on Friday
evening (xxiii. 56); but the Greek text (rjydpacrav)

TEXTUAL CONDITION OF MARK'S GOSPEL. 203
gives no warrant at all for this translation, and com
pels us to translate " they bought," which is indeed
very strange, for then mention seems to be made
of a neutral time between the end of the Sabbath
(ver. 1) and the first day of the week (ver. 2). So the
repeated catalogue of names in xvi. 1 seems spurious,
and the " when the Sabbath was past " also. In fact,
when these additions are removed, we have quite a
clear narrative. But even if these additions are
rejected, they must be at least explained, and how
is that to be done ?
To use a simile : reading Mark (with due attention
given to the variants) reminds one of walking on
quicksand, that is where the surface is quicksand
while a little below there may be gravel or even
rock ; for the difference of readings mainly rests in
the expressions, and does not affect the sense. But
nevertheless we feel unsafe, and wonder in what way
such a condition of the text may have been produced.
It is true that the condition is nowhere else so bad
as in this 16 th chapter, where there is again in ver. 4
a wide difference between the witnesses, and the words
" for it (the stone) was very great," which in the com
mon text stand rather awkwardly, and which are besides
again of a strongly Lucan character (see Luke xviii.
23,r}v yap irXovaios acpdSpa; Mark never uses acpdSpa),
are placed by D, and Eusebius, and others much more

204 PHILOLOGY OF THE GOSPELS.
suitably at the end of the 3rd verse. The Latin k,
in this respect, goes with the majority, but it has a
long apocryphal addition in this place : " And suddenly,
about the third hour (that is at nine o'clock), there
was a darkness over all the earth, and angels descended
from the heavens, and rising in the glory of the living
God they mounted together with Him, and immediately
there was light again. And the women approached
the sepulchre," etc. As for the hour, we must notice
that in k (and D, etc.), there is no " very " before
"early" in ver. 2, and instead of "at the rising of
the sun" (araTe'AXoi/To? rov rjXlov, or even en avar.)
there is another well-attested reading, " after the rising
(dvaTe/\ai>TO?) of the sun." Here therefore the quick
sand seems to reach further below the surface. As
we are speaking of k, it is worth noticing that this
version at another place (xi. 4 ff.) exhibits a very
strong contraction, giving nothing but : " And they
went their way (4), and said even as Jesus had
commanded them" (6), much in the same way as D in
the corresponding place of Luke (see above on p. 149 f.).
You are looking, no doubt, for some clue which
may show you the way out of this wilderness. It is
of course impossible to hold the copyists responsible :
we should have to charge them all alike with taking
gross liberties, and yet should find, upon reflection,
no warrant for so unlikely a charge. Again, the

TEXTUAL CONDITION OF MARK'S GOSPEL. 205
theory of two texts is inapplicable to this problem,
at least in the sense it had in Luke: there the evidence
was for two texts, but one author, while here there
appear more authors than one. But one of the
authors seems to be Luke. Well, and then ? Did
Luke perhaps interpolate or revise Mark? No, but he
translated it, as the original Mark was in Aramaic,
or had it translated for his own use, and then revised
the translation. At a later time Luke's copy got into
circulation and was again copied, and those copies
went side by side with copies containing a translation
made by somebody else, and our text contains a number
of ' conflate ' readings, which arose by the constant
comparing and collating of the different copies. This
seems to be, at least, a possible solution; but before
adopting it, we must try to look more closely into the
relations which may have existed between Luke and
Mark. That they knew each other is certain, and they
lived together at some time in the company of Paul,
and what one of them had written, could not remain
unknown to the other. Moreover, it can hardly be
questioned that Mark wrote his Gospel before Luke,
and that this Gospel could not escape Luke's notice
even before they came into personal contact; it
follows, then, that Mark is one of Luke's author
ities. Now, if this be so, it must become evident

206 PHILOLOGY OF THE GOSPELS.
on comparing the two Gospels. They have been
brought to that test long ago, and the result is,
that Luke is rather generally regarded as depending,
in a large measure, on Mark's authority. In the first
place, the order of the narratives, as far as both
authors have them in common, is almost invariably
the same, and this argument is so much the stronger,
as that order is either artificial or accidental, certainly
not historical. There is on Mark the express testimony
of Papias' presbyter : " Mark has not related things in
their order, because he had not heard or followed the
Lord, but had heard, at a later time, Peter, who
delivered his teachings according to circumstances,
and by no means as one who was making an orderly
collection (avvrap^iv) of the Lord's sayings. So Mark
is not to be blamed for writing some things as they
came into his mind, his sole intention being to leave
out nothing of what he had heard." If this was the
origin of the order of events in Mark, it must be termed
accidental, and as the same order is found in Luke,
the reason cannot be that the same accidents happened
twice, but that Luke borrowed from Mark. It has
been suggested, it is true, that the order goes back to
the relations of Peter and other evangelists, who by
frequently relating the same things had at length
established for them a fixed order ; but I do not think
that this goes far enough to explain the constancy in

TEXTUAL CONDITION OF MARK'S GOSPEL. 207
the sequence of so many narratives, which would not
be given in one sermon, but in a long series of sermons.
In the second place, the matter in Luke, as regards a
large number of narratives, is the same as in Mark,
only that the latter gives more detail ; again and
again readers of Mark are struck by "autoptic
touches "* which are given by this narrator alone, and
for this reason one very soon finds it impossible to
derive Mark from Matthew or Luke, while the reverse
course lies wide open. Also, the exclusively Galilaean
character of the history of our Lord, as given by the
three synoptic Gospels, seems to indicate a common
source, which may be found in Peter's sermons related
by Mark, for it is very natural that Peter, when
preaching in Jerusalem, did not generally tell what
had come to pass in the holy city itself, but what had
come to pass far away in Galilee. But to prove the
hypothesis that Luke used Mark as his authority,
there is yet a third test, that of language, and that
test completely fails to give the expected result. Ver
bal coincidences of any importance between Luke and
Mark are very rare even in the common text, and that
common text is, as we have seen, not very trustworthy
in particulars in the case of Luke, and very untrust
worthy in the case of Mark. I may give one instance:
on the calling of Levi the publican we read in Luke
1 Salmon, Introduction (7th ed.), p. 138.

208 PHILOLOGY OF THE GOSPELS.
(v. 27 ff.): " He saw a publican, named Levi, sitting at
the receipt of custom : and He said unto him, Follow
me. And he left all, rose up, and followed Him."
Quite the same words, with the exception of the
"publican" and of "left all," are found in Mark
(ii. 14). But the words in Luke: "sitting at the
receipt of custom," are left out by the codex 700, and
upon reflection, we find that they are indeed rendered
superfluous by Luke's "publican"; likewise "rose up,"
which stands in close correspondence to " sitting," is
left out by the Sinaitic Syriac, and is likewise super
fluous after Luke's "left all." In this way Luke's
text having been purified, the verbal coincidence with
Mark has vanished. On reading on in this story, we
find the substantial agreement of both evangelists to be
complete, but the verbal agreement to be almost totally
wanting. Even in the concluding sentence : " They
that are whole have no need of the physician," etc.,
Luke replaces Mark's (and Matthew's) layvovres by
* r
vyiaivovres. You may say, of course, that even if Luke followed
the authority of our Greek Mark, he might never
theless change layvovres, which he never uses in this
sense, into vyiaivovres which he uses elsewhere, and
he alone of the evangelists. Besides, the difference
between Luke on the one side, and Matthew and
Mark (whose accounts have much in common) on the

TEXTUAL CONDITION OF MARK'S GOSPEL. 209
other, does not merely consist in the choice of words,
but also in the whole arrangement of the narrative ;
so Luke, even if he had another translation of Mark
before him, must have allowed himself much liberty
in its rendering. But the force of the argument
consists in this, that while material agreement between
the two extends over a rather large space, verbal
agreement is never found in any considerable measure.
And yet there is a great deal of such agreement not
only between Mark and Matthew, but also between
Luke and Matthew. Now we have seen in a former
chapter (see on p. 175) that Matthew is not likely to
have borrowed from Luke, but that more likely both
of them used a common Greek authority. If this is
true, it is proved that Luke did not shun verbal
copying from a Greek source ; and why then should
he have declined to do so when Mark was this
source? Was it because it was not quite good Greek,
as for example where Mark has (x. 25): Sid rrjs
rpv/naXids rfjs pacplSos ? Why then did Luke (xviii.
25) write Sid rprj/iaros fieXdvrjsi. Why, you may say,
did Matthew (xix. 24, Tischend.) write rpvirrj/uaros ?
It is perhaps difficult to answer, since nobody can tell
which of these words was the common one in Palestine
or elsewhere ; but, as I have said, the strength of the
argument lies in the constant occurrence of the
observation, and not in any particular instance of

210 PHILOLOGY OF THE GOSPELS.
disagreement. That the observation is generally true,
I have established elsewhere,1 and may refer my
readers to that place, if they do not choose to test
the thing themselves, taking care, however, not to let
themselves be deceived by assimilated readings.
I may now sum up my various arguments. Luke,
in the first part of his Acts, followed an author who
had written in Aramaic. Mark is very likely to be
the author who first published these stories ; he seems
therefore to be Luke's Aramaic authority. If Mark's
Acts were written in Aramaic, his Gospel originally
was written in Aramaic also. Secondly, the textual
condition of Mark's Gospel suggests the idea that
there existed a plurality of versions of a common
Aramaic original. But to speak more properly,
we should perhaps say not versions but redactions.
The discrepancies we found, e.g. in the beginning
of Mark's 16 th chapter, do not fall under the
description of various Greek renderings for the same
Aramaic words, and there are more cases like this.
Among the various readings in Mark we recognized
some as bearing distinctly the character of Lucan
style, without being borrowed from Luke. Lastly, it
appeared to us that the scarcity of verbal agreement
between Luke and our Mark strongly dissuades us
from the hypothesis of Luke's using this Mark ; while,
1 See my edition of Luke, p. xiv ff.

TEXTUAL CONDITION OF MARK'S GOSPEL. 211
on the other hand, the material agreement is such as
to render a dependence of Luke on Mark even more
than probable. The conclusion to be drawn is this,
that Luke used another Mark. And now we may
go further and suggest that Luke, before writing
his own Gospel, made a Greek redaction of that of
Mark, not for his own use only, but also for that
of Christians speaking Greek. Another translation
of Mark, or other translations, were made by other
persons, and one version among these was that which
eventually predominated, but the others have at least
left their traces.1
Or is all this rather wild conjecturing ? Is there
not a common and widely-spread opinion, that Mark
wrote for Eomans, therefore, of course in Greek ? I
ask, what are the proofs ? Because he bears a Eoman
name ? No, but because there are Latin words in Mark,
such as Kevrvplwv, instead of which the others employ
the Greek eKardvrapyos. I say this is a vulgarism,
1 Readings of Lucan character, besides those already mentioned,
are the following : iv. 19, Kal oi irepl Ta \oiird imdvpilai, words
which are omitted by D etc. (iirtffvfila and iiriffvfietv, irepl with
accusative, r\oiirbs, never in Mark, but all of them in Luke, cf.
Acts xix. 25, etc.); iii. 21, ol ypa/i/iaTels Kal o'i \oiirol, D, etc.,
see Luke xxiv. 9, etc. ; ibid. 22, oi airb 'lepoaoM/iaiv Karafiavres
(instead of these words the Latin e has et ceteri, and D's reading
in 21 seems originally to belong to 22), cf. Acts xxv. 7 ; x.
24, Toil's iren-oifldTas iiri xprjliaaiv (ACD etc.), see Luke xviii. 9; xiv.
58, N gives 8ri elirev, c, k, hie dixit, instead of ijfiets ¦nKoiaa/iev airov
\iyovros ; with the latter cf. Acts vi. II, 14.

212 PHILOLOGY OF THE GOSPELS.
not a Latinism. The Greek as well as the Aramaic
language of daily intercourse had received many such
words, and centurio is actually found in the Syriac
version. Any writer who cared for good Greek would
take the equivalent Greek word, but this so-called
Mark did not care for style, and this you may see
from the first page of the little book to the last.1
But there is still another argument : the two mites
(XeTrTa) of the widow are explained to be of the same
value as a Eoman quadrans (KoSpdvrrjs ; Mark xii. 42).
The supposition seems to be, that Eoman coins were
not known outside of Eome, and yet there is the
denarius in the four Gospels and in the Apocalypse,
and the as (daadpiov) in Matthew and Luke. But
at all events Mark was not writing for Palestinian
readers, who did not need an explanation of Xeirrd.
He was not translated for the use of such, I should
say, for of course this addition comes from the trans
lator. There is in Mark a much larger addition of
the same kind, vii. 3 f., on the washings of the Jews,
violently interrupting the construction, and of course
not in the Aramaic original, and perhaps not even in
the original translation, since it looks much like a
*An instance of stylistic refinement introduced by various read
ings is found in vi. 7, e D: Kal irpoaKa\ecr6\]i.evos rois St&SeKa d,ir&rrciXcv
airois dva Sio (instead of Siio Sio, which is an Aramaism), 8oi>s airois
i&valav tCiv irvev/idroiv twv aKaddpriav, TrapayycCXas auTots fxijSiv atpeiv
(instead of Xva . . . atptaaiv) els ttjv bSbv.

TEXTUAL CONDITION OF MARK'S GOSPEL. 213
scholion. If the condition of the actual Mark is such
as I think I have proved it to be, no individual
passages can be brought forward which will have any
weight against my hypothesis. Another feature in
the actual Mark are the Aramaic words : talitha kumi,
v. 41, ephphatha, vii. 34, which I take for relics of
the original, preserved by the translator.
On the other hand, if Luke — which must seem the
most hazardous part of the whole hypothesis- — had
made something like a Greek translation of Mark
before writing his own Gospel, and had set that
Mark in circulation, we understand much more
easily why there are so many omissions in Luke, He
might presuppose many things to be known to many,
or most, of his readers, and as he was himself com
posing a Gospel on a much larger scale, introducing
a great deal of fresh matter, he was well entitled to
leave out some part of that already known, lest his
book should grow to an immoderate size. Much stress
has been laid recently on the size of Luke's books, as
having been somehow prescribed to him by a literary
custom, or by the customary length of the papyrus-
rolls he used, and it is a fact that both of his
books are very nearly of the same size.1 I think that
there is some truth in this ; of course the size of
Luke's roll could not prevent him from giving in the
1 See Arnold Ruegg in Theol. Studien u. Kritiken, 1896, p. 94 ff.

214 PHILOLOGY OF THE GOSPELS.
Eoman copy the seven words which are left out there
(see above, p. 138), but that reason among others might
influence him in abridgments of greater importance.
Comparing the last chapters of Mark and of Luke, we
find the following important omissions in the latter.
He does not tell of (see above on p. 179 ff.) the in
stitution of our Lord's Supper, nor the anointing in
Bethany (Mark xiv. 3 ff), nor the discourses on the
betrayer of Jesus (ibid. 18 ff.), except very briefly and
in a general way (Luke xxii. 23), deflecting at once
to another discourse which is not given by Mark.
Next he omits all details of the trial before the San
hedrin (Mark 55 ff.), giving only the final and
decisive question. Also he omits the " Eloi, Eloi,"
etc. (Mark xv. 34), and the discourses made on it.
He must, of course, make his own Gospel independent
of the use of another Gospel side by side with it, and
must give, therefore, all that was necessary or desir
able for the understanding of the story, whether it
was already extant in Mark or not; but he was not
bound to give everything he knew. The same reasons
may account for the largest omission we find when
comparing the two Gospels, viz., that of the whole
series of stories beginning with Mark vi. 45, and ex
tending to viii. 13. There is in that part of Mark
the story of the second feeding of the multitude, and
Luke might easily dispense with giving that. But he

TEXTUAL CONDITION OF MARK'S GOSPEL. 215
actually leaps from the end of the story about the
first feeding to the end of that about the second,
leaving out all that comes between. This looks like
a piece of carelessness in using his authorities, which,
however, may be excused by the reflection that he
did not feel bound to give all, nor would he deprive
his readers of those stories which he did not give,
there being a book within their reach where they
might find them, the " Gospel according to Mark "
(evayyeXiov Kara M.apKov), as the title would be.
I think that if we still possessed this form of Mark
in its integrity, we should find there verbal agreement
with Luke to a very large extent : e.g. we should
find there "to preach the gospel," in iii. 14 ( = Acts
i. 2), as D gives (see above, p. 142), and the identical
stories told in similar words. That passage in Mark
iii. 14 ff. being of some importance, we shall allow
it to detain us a little more.
The form A (Lucan) is there preserved for some
length, differing greatly from B ; the witnesses for
A are, besides D, some old Latin versions : the
Vercellensis a, the Palatinus e, the Colbertinus c,
and others. Of course there is no want of con
flated readings in D and elsewhere ; but the purified
text A appears to be this: (14) "And He ordained
that twelve should be with Him, whom also He named
apostles ( = Luke vi. 13), (15) and gave them power

216 PHILOLOGY OF THE GOSPELS.
that they might heal sicknesses, and cast out devils,
and going about preach the gospel. (16) And Simon
(see i. 16) He surnamed Peter; (17) but collectively
He called them Baneregez, which is, The sons of
thunder. They were these : Simon and Andrew,
James and John (see i 19), (18) and Philip, and
Bartholomew, and Judas and Matthew, and Thomas
and James, the son of Alphaeus, and Simon the
Cananite (?), (19) and Judas Iscariot, who also be
trayed Him." x
The names are identical with those in Luke, with
one exception in the order ; the most important thing
is that the name of Sons of Thunder is transferred to
the whole body of the apostles, the thunder evidently
being compared to their voice of preaching. For that
same reason the preaching is mentioned after the heal-
1 1 give the two texts A and B in Greek (many peculiarities re
maining doubtful) :
A. B.
(14) Kal iiroiijaev tva Saiv Sti- (14) [Kai iiroiijaev StliSeKa] (to
SeKa fier* avrov, ovs Kal diroarbXovs be cancelled, see ver. 16) iva
tbvb/iaaev ( = L. vi. 13), (15) Kal tbaiv fier' avrov, Kal tva diroariXXij
ISukcv airois exovaiav Bepaireieiv airois Kwpvaaeiv, (15) Kal 2xei"
Tas vbaovs Kal iKfidXXeiv rd Sai- i%ovalav[depaireietvTasvbaovSKat]
jibvia (cp. L. ix. 1), Kal ire pie pxo- (om.t>;~Ba,l.)iKfidXXeivTa,Sai/ibvia,
fiivovs Kvpiaaetv Tb eiayyiXiov. (16) Kal iiroiijaev Tois StliSeKa (ko.1
(16) Kal iiri$i)Kev (rf) 'Slfiuvt ...S. om. AD al.),irpS>rov 'Sljiuva,
6vofia Hirpov, (17) koivQs 6" ixd- Kal iiridijKev Svojia rtf Xlfiwvi IHt-
Xeaev airois Bavvpeyi^ (700), 8 pov, (17) Kal 'laKafiov rbv tov Z.
iartv viol fipovrrjs. rjaav Si otVot Kal 'Itodvijv rbv dSeXtpbv airov, Kal
'Slfiav Kal 'AvSpias, 'laKwfios Kal iiridijKev airois 6vo/ia Boavwpyis,
'ludv-qs k.t.X. S iariv viol fipovrrjs' (18) Kal k.t.X.

TEXTUAL CONDITION OF MARK'S GOSPEL. 217
ing, that there might be a near connection between
the preaching and the common name.
I think I have now said enough on this subject,
as it is not my intention, nor ought it to be, to
solve the so-called synoptic problem, but only to
give such contributions for its solution as are afforded
to me by textual criticism. Upon the whole I am
afraid that textual criticism is apt to render the
problem even more complex, since it tends to split
a seeming unit, such as Mark's book, into a plurality
of books ; but this is, in my opinion, quite in accord
ance with the facts themselves. There was a time in
the Latin Church, when, as Jerome states, there existed
almost as many Latin versions of the Greek New Tes
tament as there were copies,1 each congregation having
not only its own copy, but in that copy a separate
version. Afterwards that plurality was gradually
more and more reduced, and now the Eoman Catholic
Church has one authorized Vulgate. I think there
was also a time in the old Christian Church when
there existed almost as many Greek Gospels as there
were Christian communities, not differing widely, per
haps, from each other in any individual case, but
still not wholly identical. Afterwards a gradual re
duction was effected, and now we are accustomed to
1 Preface to the Gospels (Tot enim sunt exemplaria paene quot
codices).

218 PHILOLOGY OF THE GOSPELS.
read only four Gospels, and each of them in one fixed
form and text. But Papias' presbyter still knew more
Matthews than one, and is it then astonishing that I
try to establish that there were, and in some measure
still are, two separate Lukes, and that in Mark the
case is even more complicated ? And, besides, all
those many "predecessors of Luke" of whom he speaks
existed, although, with the single exception of Mark,
we are not even able to guess who they may have
been.

CHAPTER XII.
TEXTUAL CONDITION OF JOHN'S GOSPEL.
There is but one Gospel now left for us to consider,
that of John. The problems it presents are suffi
cient to fill up many books ; but I shall speak in the
main of its textual condition, after having briefly
discussed, or touched upon, some important general
points. There is a tradition recorded by a Latin manuscript
of the ninth century, according to which Papias of
Hierapolis, whom we mentioned just now, was the
original copyist to wThom John dictated his Gospel.
Of course this cannot be true, but there seems to lie
behind the corrupt tradition some fact which was really
related by Papias. Now there is other testimony much
like this, coming from a Greek compilation of commen
taries (catena) on St. John, where the words are :
John dictated his Gospel to his disciple, Papias Eubio-
tus of Hierapolis, in order to give a necessary supple-

220 PHILOLOGY OF THE GOSPELS.
ment (irpbs dvairXrjpwaiv) to the information given by
those who before him had preached the gospel to all
the nations upon earth. This last clause reminds one
of a passage in Eusebius (Hist. Eccl. iii. 24, 7), which
is of this tenor : It is told (cpaa'iv) that after the three
other Gospels had already come into general notice,
John, while approving their contents as authentic, said
that he still missed in them the account of Christ's
first doings, and that this was his reason for writing
his own. Afterwards there comes in Eusebius the long
passage on Papias, and the excerpts from him refer
ring to Matthew and Mark ; as there are no excerpts
given regarding John, we may infer that Eusebius did
not find anything noteworthy in Papias that he had
not already related. But the name of John's copyist
would not seem to him very noteworthy, unless this
copyist were Papias himself, so that this form of the
tradition is disproved also by Eusebius' silence.
On the other hand the Catena, by presenting two
names for that copyist, offers the easy explanation that
in reality Papias had given the name of his own fellow-
citizen Eubiotus as that of the copyist, a name which
was blended by later ignorance with that of the author
of the notice. For Eubiotus (EvfSioros, wrongly spelt
-Iwros in the ms.) is a well-attested Greek name, and
on the other hand it is as impossible to take it for a
second name, or surname, of Papias, as to regard it as

TEXTUAL CONDITION OF JOHN'S GOSPEL. 221
an adjective. The adjective, of course, exists, but not
in common use, and it would require the article
before it ; a second name of Papias would also have
the article, like Aiwv 6 Kao-o-jo?, or EavXo? 6 Kai
TLavXos.1 I earnestly hope that the day will come when
this conjecture will be brought to the test, by the
recovery of Papias' work, either in a Syrian or
Armenian translation, or even in the original Greek
form, although there is more hope for fragments
of that, from Oxyrhynchus or some other ruins of
an ancient town, than for an entire manuscript.
Meanwhile scientific theologians do well to retrench
themselves in new positions, since the old ones, by
that discovery, might become hopelessly untenable.
I am sure that we should find in Papias, among
much rubbish of worthless tradition, a good deal
of trustworthy information on John, I mean the
apostle and not the presbyter. It is true that
Eusebius found in Papias two Johns mentioned,
as in the catalogue of authorities given by Papias
and preserved by Eusebius the name of John comes
twice, first among those presbyters, that is apostles,
whose death occurred before his own time, and
*It would not be impossible that the words in the Latin MS.,
disscripsit vero evangelium dictante Johanne recte, are a blundering
translation of Zypaxpe Si rb eiayy. birayopeiovros 'Xoidvov E«/3(otos.

222 PHILOLOGY OF THE GOSPELS.
secondly, together with one Aristion, as one of those
disciples of the Lord who in Papias' youth were
still alive ; Eusebius even affirms that he laid claim
to having heard these himself, although in the
quoted words of Papias, as they stand, this is not
stated. He calls the second John by the same title
which he had previously used for Peter and the
other apostles, viz., presbyter or elder, and gives
him that epithet in contradistinction to Aristion :
" what Aristion and the presbyter John, the disciples
of the Lord, relate " ; evidently then this John must
be the apostle or an apostle, if there were two.
But there were not two, and so the passage appears
to be utterly confused, unless we accept the con
jecture recently proposed by the Greifswald Professor,
Joh. Haussleiter, who strikes out the first mention
of John, taking it for an interpolation which had
crept into the text, before Eusebius' time, after the
name of John's brother James, with which his was
so frequently combined.1 Haussleiter urges that the
list is given by pairs: Andrew — Peter, Philip — Thomas,
James — Matthew (if John is left out), and that
Papias introduces the second, third, and fourth of
these names by " or what " (scil. " have related "),
while he drops the " what," leaving only the " or,"
1 Theolog. Literaturblatt, xvii. Jahrgang, 1896, nr. 39. See also
Provost Salmon's Introduction (7th ed.), p. 268 f.

TEXTUAL CONDITION OF JOHN'S GOSPEL. 223
before James and Matthew, as it was natural
towards the end of an enumeration; now John's
name comes in between these two with "or what,"
which is incongruous, and a clear trace of the
interpolation.1 If we adopt this emendation (as I
am much inclined to do), it does not follow of
course that the John who wrote the Apocalypse
was identical with the apostle. For although
Eusebius is not unwilling to ascribe that book to
the " presbyter " John, we must be aware that
Papias did riot mention the Apocalypse,2 and that
its author never styles himself "presbyter," nor
apostle, but simply "John," which indeed does not
go very far as a proof of identity. But the author
of the second and third Johannine epistles styles
himself "presbyter"; this then must be the apostle,
and why not ? 3 Moreover, the traditions introduced
1 The Greek words are : Ei Si irov Kal wapijKoXovdijKws ris rois
irpeafivripois &X6oi, rois rwv irpeafivripwv dviKptvov Xbyovs' rl 'AvSpias
i) rl Hirpos elirev i) rl QlXunros f) rl Qoifids 7) 'IdKwfivs i) rl 'Iwdviis
fl MarSaios ij ris 'irepos twv tov Kvpiov fuaBirrwv, i) rl 'Apiarlwv Kal
b irpeafivrepos Tu., ol rov Kvpiov fiaSijral, Xiyovaiv (this present tense
refers, of course, to the time of which Papias is narrating, and
not to the time when he wrote).
2 1 do not feel bound to credit Andrew of Caesarea (a writer of
the end of the fifth century), who adduces Papias among many
other witnesses for the authenticity of the Apocalypse. If there
had been in Papias a mention of that book, we should read it in
Eusebius, who expressly states that Justin cited the Apocal. (iv.
18, 8), and Theophilus of Antioch (iv. 24), and Apollonius (v. 18, 14).
3 See Salmon, 1. c, p. 270 ff.

224 PHILOLOGY OF THE GOSPELS.
by Papias with the formula, " This said the presbyter,"
must be referred to John the apostle, as e.g. that
about Mark's Gospel. I say again : Why not ? And
just so that about those vines of the Millennium, which
will have each ten thousand branches, and each branch
ten thousand twigs, and each twig ten thousand
bunches, and each bunch ten thousand grapes, and
each grape will produce twenty-five barrels of wine ;
for Irenaeus, who gives this quotation, expressly intro
duces it by the words, " as those Elders (Papias, etc.),
who had seen John the disciple of the Lord, relate,
that they heard from him, how the Lord taught about
those times, and said " ; then follow the words, and
after them Papias' written authority is adduced.
Why, this is mere rubbish, and cannot have been
taught by our Lord, or related by His disciple. I
fully agree with this criticism ; let us hope, then,
that when Papias comes to light, the formula, "This
said the presbyter," will not be found at the head
of this apocryphal teaching, although the words of
Irenaeus rather point to the opposite conclusion.
We see therefore that this is doubtful ground, and
hence the great divergencies between different scholars.
Whilst in Professor Haussleiter's eyes the " presbyter "
John is dead and buried, in those of Professor Harnack
he is, on the contrary, full of vigorous life ; Harnack
does not shrink from assigning to him, and not to

TEXTUAL CONDITION OF JOHN'S GOSPEL. 225
the apostle (who may have been, however, his teacher
and authority), the authorship of the Johannine writ
ings and especially of the Gospel.1 He insists chiefly
upon one passage of the Gospel, which gives, in his
opinion, the clear proof that the narrator and com
poser cannot be the eye-witness, but must be a different
person. There are in xix. 3 5 f. the words : " And
he that saw it bare record (fiefiaprvprjKev), and his
record is true : and he (e'/ceiVoy) knoweth that he saith
true, that ye might believe. For these things were
done, that the scripture should be fulfilled," etc.
There are, it is true, different interpretations given
of these words, and there are different readings also,
and more than that. Professor Harnack says : " It
would be quite unwarranted to regard the ver. 35
as an interpolation." Indeed ? He ought to be
aware that a very good warrant, the Latin e, and
besides an ancient MS. of the Latin Vulgate, omit the
ver. 35, e giving in ver. 36 Se instead of ydp ; for
Se there are even more witnesses, and among these
Nonnus in his metrical paraphrase of the Gospel, who
renders besides in ver. 3 5 the text : " He that saw
it bare record, and of him (eKelvov) we know that

1 Harnack, Chronologic der altchristl. Litteratur, i. 659 ff. Harnack
does even more than this. He makes his "John the Presbyter"
establish the canon of our four Gospels, including his own, and
impose that canon on the whole Church.
P

226 PHILOLOGY OF THE GOSPELS.
the record is true."1 This Nonnus was a gifted poet,
an Egyptian by birth, who lived about the beginning
of the fifth century, and wrote besides and before this
paraphrase a voluminous epic poem called AiowaiaKa.
In his paraphrase he is, of course, continually amplify
ing, but he takes great care not to omit any word of
the sacred text. Now, the passage in question reminds
one of the first epilogue to the Gospel in xx. 30 f . :
" But these (signs) are written, that ye might believe,"
etc., and of the second epilogue too, xxi. 24: "This
is the disciple which testifies (fiaprvpwv) of these
things, and wrote these things, and we know that
his testimony is true," the last clause " and we,"
etc., being omitted by Nonnus. Moreover, there
is to be compared v. 3 2 (in Christ's words) : " There
is another that beareth witness (fiaprvpwv) of me;
and I know that the witness which He witnesseth
of me is true," and in the Third Epistle of John
(of the authenticity of which I have no doubt), ver.
12: " Demetrius hath good report (Arjfirjrplw fiefiap-
rvprjrai) of all men, and of the truth itself: yea,
and we also bear record (fiaprvpovfiev) ; and ye know
(otSas) that our record is true." There is therefore
not the least doubt of the Johannine character of

1 Nonnus' words are : ivijp 6" bans birwirev iip iriarwaaro jiiBip jxap-
Tvplijv arlvaKrov dpiarovboio Si Kelvov tSjiev Sri £adiij Kal ir-rrrv/ios iirXero
tpwvi)' ravra Si irdvra iriXeaKev K.r.X.

TEXTUAL CONDITION OF JOHN'S GOSPEL. 227
the words in xix. 35, but we may question the
genuineness of the last clause, on Nonnus' authority,
as it may have crept in from the passage in xx. 31,
and likewise question, on the same authority, the
genuineness of "and we know," etc., in xxi. 24; for
that too may be an amplification due to xix. 35.
There is a difficulty removed in this way, for how
can a distinction be made between the author and
writer (b ypd^ras) of the narrative and the person
or persons narrating (rjfieis) ? But the same difficulty
is, on the same authority, translated into xix. 35,
where Nonnus gives • " we know " instead of " he
knows." The case there is, that an endless discussion
has been carried on about the eKelvos, the question
being, whether that pronoun may be employed of
the author himself. It is rather hard to affirm
this, and Zahn's view, who refers the pronoun to
Jesus, whilst it seems to be recommended by the
grammar, appears very strange from another point
of view. On the other hand, Nonnus' reading is,
as far as this sentence goes, quite unexception
able ; only there is again the distinction between
the author and the "we." Well, then, the whole
verse on the authority of e must be relegated from
the text to the margin, the ydp in ver. 36 being
of course replaced by Se. If we accept this, we
have to state that John's text has been commented

228 PHILOLOGY OF THE GOSPELS.
upon at a very ancient time by some disciples of
his, who copied his style; for the Johannine style,
as we have seen, is evidently there.
We may now leave this special question, in order to
take a general view of the condition of St. John's text.
Are there more instances of interpolation, or of com
menting upon a text, or whatever we may call it ? If
we are to regard as spurious, or at least dubious, all
those particles which have not an universal and unani
mous attestation, there is indeed a very great deal of
dubious matter in John. This is even in some
measure a recognized fact, certain passages having
been cancelled by recent editors upon the evidence of
the Alexandrian MSS. which omit them. The most
conspicuous instance is in chap, v., where some words
in ver. 3 and the whole of ver. 4 are now left out, on
the authority of the best mss. (among which B and
D). There can indeed be no doubt that the ver. 4
does not come from John, since the style is wholly
different,1 quite apart from all considerations with re
gard to the contents : " For an angel went down (but
according to the Alexandrinus A ' bathed,' eXovero), at
a certain season into (in) the pool, and troubled the
water : whosoever then first after the troubling of the
1 There is nowhere in the N.T. <jS Sijirore or otip Sijirorovv, nor
KarixeaBai voar/jiart, except in D, Luke iv. 38, Karix- irvperip, nor
vbaijfia itself.

TEXTUAL CONDITION OF JOHN'S GOSPEL. 229
water stepped in was made whole of whatsoever disease
he had." But the impotent man says to Jesus in ver.
7 : " Sir, I have no man, when the water is troubled, to
put me into the pool : but while I am coming, another
steppeth down before me." This is the universal
reading there, except that one MS. adds " and receives
healing," and another, " and I walk away in my in
firmity." After the removal of ver. 4, it is evident
that ver. 7 has become unintelligible. Turning to
Nonnus, we find in him no trace of ver. 3 or 4, but
instead of these he must have read something like
this : " where a sick man, as soon as he saw the water
being troubled spontaneously, bathed and was made
whole." I am sure that if we had a text like this,
no offence could be taken : if we suppose the pool
(KoXvfij3rj0pa) to have been rather narrow, there would
be no room for more than one man at a time, and
Nonnus indeed uses throughout the word dadfiivQos,
" bathing-tub," and, as we have seen, the singular of the
man. There is then in this passage not interpolation
only, but the genuine text has been replaced by a
spurious comment, of course at a very ancient time.
For even those witnesses who omit the comment, in
directly show that it had existed in one of their
archetypes, where it had been cancelled upon better
authority, but without giving the genuine words ex
hibited by that authority. Or else the narrative given

230 PHILOLOGY OF THE GOSPELS.
by John had been from the beginning incomplete in this
respect, and was afterwards supplemented by others in
different ways. You see there is the same quicksand as
in Mark. Going on in this narrative we find more
of it : ver. 9, '' and on that same day was the Sab
bath"; but D and e (Nonnus) simply: "And it was
the Sabbath"; ver. 10, " The Jews therefore said unto
him that was cured," but e without " unto . . .
cured," and the verb employed here for "cured" does not
exist elsewhere in John; "It is the Sabbath day"; but
this is not in Nonnus, or in the Sinaitic Syriac ; " it is
not lawful for thee to carry thy bed"; but Nonnus:
"who had bid him," etc., in close agreement with ver. 11,
where he renders this text : " And he answered them :
He that has made me whole, the same told me to take
up and walk." Conversely in ver. 12:" Then asked
they him (e, 'the Jews asked him, saying') : What man
is that which said unto thee, Take up thy bed, and
walk ?" Nonnus not only preserves the direct form
of speech (which is replaced by the indirect in N both
here and in 11), but also the words, " thy bed," which
are left out by NB and others. Ver. 13 : "And he
that was healed wist not who it was." Instead of
6 laQels, cf. iv. 47, there are these various readings :
b Se daQevwv, 6 Se reOepairevfievos, Kal 6 dvQpwiros (Syr.
Sinait., and apparently Nonnus), 6 Se (q). " For Jesus
had conveyed Himself away, a multitude being in that

TEXTUAL CONDITION OF JOHN'S GOSPEL. 231
place." The whole sentence is left out by Nonnus,
and the verb employed here (eKveveiv) is wholly foreign
to the New Testament. And was this a reason for the
man's not knowing Jesus ? Perhaps not, but if Jesus
had remained there the Jews would have recognized
Him. On the other hand, we do not even question
whether Jesus was there or not, if the sentence is left
out ; for the place where the Jews met with the man
who was carrying his bed was, of course, not Bethesda,
where the narrator has left Jesus.
Take another narrative, you will find the same
uncertainty of readings, and now and then a manifest
gloss, or interpolation. In chap. iv. 6 ff., the two
Syriac mss., those of Cureton and of Mrs. Lewis, make
a transposition of ver. 8 into ver. 6 : " Jesus therefore,
being wearied, sat thus on the well ; (ver. 8) for His
disciples were gone away into the city to buy meat.
(6) And as Jesus sat (there) it was about the sixth
hour, (7) and a woman cometh," etc. We have
seen on a former occasion (p. 56 ff.) a very good
transposition given by the Sinaitic Syriac ; in the pre
sent case both collocations of the sentences seem to
be equally possible. In the next verse (9) : " Then
saith the woman of Samaria unto Him, How is it that
Thou, being a Jew, askest drink of me, which am a
woman of Samaria? for the Jews have no dealings
with the Samaritans " (ov yap avyypwvrat 'lovSaioi

232 PHILOLOGY OF THE GOSPELS.
1,afiapirais), the last clause is left out by ND, etc.,
and here again the interpolation is evident, as avy-
yprjaQal nvi is foreign to New Testament Greek.
But also yvvaiKos "ZafiaplnSos ovarjs is superfluous, and
is left out by the Sinaitic Syraic. It is, however,
worth noticing that in the following discourse, as in
other discourses and speeches of Jesus, omissions are
much rarer. In ver. 14 e and Nonnus present a
shorter form : "And whosoever drinketh of the water
that I shall give him shall never thirst; but there
shall be in him a well of water springing up/' etc. I
do not think that the sentence loses anything by this
omission, as the words omitted (" the water that I
shall give him ") are already there in the former part.
Now, how are we to explain this condition of St.
John's text ? In the first place, we must define that
condition more exactly, and in contra- distinction to
that in Mark as well as to that in Luke. Is there
something like a double text ? By no means ; for the
witnesses for the longer and for the shorter form are
continually changing places ; for instance e, which
often bore testimony to an omission of words, is in
other places a witness for some enlargement. More
over, there are passages where all our witnesses give
a more or less prolix text, but in quite different ways.
In ii. 2 there are these readings : " And when they
wanted wine '' ; or, " And they had no wine, because

TEXTUAL CONDITION OF JOHN'S GOSPEL. 233
the wine of the marriage festival had been spent " ; or
(e) " And it came to pass, that because of the great
multitude of guests the wine was spent." None of
these readings commends itself by particular Johannine
words ; on the contrary, we may object to each of
them from this point of view, and what if we leave to
John nothing of all this ? The narrative will be some
what short, it is true, but will not cease to ¦ be in
telligible to everybody. So, the thing we find does
not look like a double form of text, but, as I stated
before, like an uncommented text on the one side (not
always preserved), and a text accompanied with dif
ferent comments on the other. This is not the case,
as we have seen, in Mark, or in Luke, where the
clauses omitted by D are sometimes very characteristic
and always in accordance with Luke's style, while in
John there is now and then a striking difference
between comments and text. Again, we did not find
anything in John which reminded us by its style of
another individual writer, as was the case in Mark.
Well, then, if this is the condition we actually have
before us, let us seek for an hypothetical explanation
which may suit it. The copyists of our MSS. and of
their archetypes are either to be condemned all alike,
or to be absolved all alike ; as the former is impossible
we shall do the latter. Then there remains only this
hypothesis. The archetype of St. John's Gospel,

234 PHILOLOGY OF THE GOSPELS.
written by Eubiotus, or by some one else, whoever
he may have been, and destined in the first place for
John's disciples who lived with him and for the
Christian community in which he lived, was very soon
copied for the use of distant disciples and communities;
and the copyists being themselves disciples, took the
liberty of enlarging the text here and there, of course
each in a different way, and this was the origin
of most variants. Whether the angel bathing in the
pool (v. 4) was added by a disciple of John, I am not
certain ; but so much is evident, that a man of " very
little understanding," like Papias (who is so charac
terized by Eusebius), was quite capable of commenting
in this way. On the other hand, xix. 35, which
also did not seem to us to be a part of the original
text, shows both understanding and familiarity with
John's style ; there must have been disciples of all
kinds, and so there are comments of all kinds.
Among individual passages of contested reading
there is none more deserving our attention than the
thirteenth verse of the first chapter : " Which were
born, not of blood, nor of the will of the flesh, nor of
the will of man, but of God." And yet very few of
my readers, as I should think, are aware that there is
any difference of reading here. But if we slight the
variants found in our mss. and versions (e.g. that
"which" is left out in D and others, being replaced in

TEXTUAL CONDITION OF JOHN'S GOSPEL. 235
English, of course, by " they "), we cannot slight the
evidence given by a witness of such antiquity and such
high standing as Tertullian, whose copy of John not
only did not contain the relative pronoun, but also had
" he was born " instead of " (they) were born." He
goes so far as to ascribe the other reading, " they
were born" (cyevvrjQrjaav), which he found with his-
adversaries the Valentinians (but without the pronoun
" which "), to their wilfully changing the text. If Ter
tullian is not sufficient, there is Irenaeus also attesting
the same reading — obscurely, it is true, as to the
omission of the pronoun, but very distinctly as to the
singular and to the reference to Christ — not to speak
of authors so late as Ambrose and Augustine. Is this
not enough? There is Justin, too (Dialog, ch. 63),
with a rather unmistakable allusion. You find the
whole case stated in the large work of Dr. Eesch, who
has the great merit of calling attention to this highly
important " variant," and decides in favour of it.1 But
let us judge, with all impartiality and fairness, on care
ful consideration, both of the external and of the
internal reasons. There are four readings : " He was
born," " who was born " (qui natus est, the Latin b, a
Verona MS. of the fourth or fifth century), " they
were born," and " who were born." The two readings
with the pronoun are suspicious, even for the reason
1 A. Resch, Aussercanonische Paralleltexte, iv. 57 ff.

236 PHILOLOGY OF THE GOSPELS.
that there are others without it; for copyists were
very apt to add pronouns and conjunctions, as they
have done a thousand times in John and elsewhere;
on the other hand, there is no apparent reason why
the pronoun should have been left out. That John's
style is asyndetic you will recognize at once,
wherever you open his book ; in this same proem
you have the conjunction only before vers. 5, 12
(where D and e, and Tertullian and Cyprian leave
it out), 14, 16, 17, while it is missing in 2, 3, 4,
etc. But if we leave out the pronoun here the case
will be decided ; for " they were born not of blood
. . ., but of God, and the Word was made flesh, and
dwelt among us," etc., is manifestly impossible. So,
if you will maintain the ordinary reading, you must
needs return to the " who," which still might be left
out accidentally. But even with the pronoun you
have the" ami7 "in 14 much against you: why are
these two sentences connected, although the former
refers to the time after Christ's appearance, and the
latter goes back to the appearance itself and at the
same time to the "Word" of vers. 1-3 ? There is a
manifest break in the series of ideas in ver. 14, and
here the author, after having spoken mostly in dis
connected sentences, puts in the " and."
But, you may say " he was born" must refer to the
Word, or to the Light, and with one of these subjects

TEXTUAL CONDITION OF JOHN'S GOSPEL. 237
the sentence becomes absurd. I deny the necessity :
look at the words immediately preceding: "But as many
as received Him, to them gave He power to become the
sons of God, to them that believe on His name." On
whose name ? On that of the Word, or of the Light ?
Of course not, but on the name of Jesus Christ. Well,
then you have there the subject for ver. 13. There is
still more to be said against the ordinary reading.
" Sons of God " in 1 2 evidently bears a spiritual mean
ing, and cannot be understood otherwise ; why then
this strong assertion that these spiritual sons are not
born of blood, nor of the will of the flesh, nor of the
will of man, but of God ? Of course, the reader must
say ; for this is indeed self-evident. Give back to the
sentence its reference to Christ, and the strong asser
tion is justified. But why has this reference been
destroyed by " were born " (eyevvrjOrjaav) ? Not by
malice, as Tertullian suggests, but by mere inadver
tence : "was born of God " was assimilated to the
preceding " to become the sons of God." Afterwards
the pronoun came in, either in the singular or in the
plural, because a sentence like this : " Not of blood,
nor of the will of the flesh, nor of the will of man,
but of God He was born," with its subject not indi
cated until as late as in the fifteenth Greek word,
was somewhat harsh and seemed to require the eluci
dation which was given by the pronoun.

238 PHILOLOGY OF THE GOSPELS.
There is a very well-known passage in ii. (ver. 4),
where Jesus says to His mother : " Woman, what have
I to do with thee ? mine hour is not yet come."
The Greek words are rl efiol Kal aoi, yvvai ; ovirw rjKei
tj wpa fiov. This " what have I to do with thee ? " or
literally, " what is there (in common) to me and you ? "
does not cease to vex readers, whatever may be said in
explanation of it, for the phrase used is the same which
the demons use in speaking to Jesus (see e.g. Matt.. viii.
29), and which is frequent both in the Old Testament
and in colloquial Greek of the time, quite in the mean
ing of our " Let me alone." But Nonnus, in his
paraphrasing of the words, makes a very slight
alteration, which wholly changes the sense : instead of
" and " (Kal) he has " or " (r)),1 and now this sense comes
out : What is that to me, or to you ? (cf. Paul, 1 Cor.
v. 12). Namely, that they have no wine does not
concern you at all, and it does not concern me yet :
for mine hour is not yet come. I am sure most
readers will be of my opinion and prefer r) to Kal,
although there is but one witness for the former, and
1 Ti ifiol, yivat, i)i aoi airy ; In case anybody should conjecture i)Si
instead of i)i, I notice that the index of words given in Scheindler's
edition of Nonnus does not contain r)Si at all ; besides the airy is
evidently incompatible with that alteration. By the way, I can
state by means of the same index, that N. renders ij instead of Kal
also in viii. 14 (f) irov; so BD, etc.), x. 10 f. (i) eiarj i) diroXiarj . . .
i) irepiaabv) ; xii. 38 (i) b fipaxlwv), 49 (i) rl XaXrjaw, as d and the
Coptic version), whether rightly or not, I do not decide.

TEXTUAL CONDITION OF JOHN'S GOSPEL. 239
that quite an obscure one. It is much to be regrettedj
that in this chapter we are destitute of the aid of the
Sinaitic Syrian, which, as my readers will remember,
afforded us excellent help for xviii., where we found a
very clear instance of inverted order (see above on
page 5 6 ff.). This too seems to be a special feature in
the textual condition of John ; carelessness in copying,
and the leaving out of sentences, which were afterwards
supplied in the margin, and from thence came again
into the text, but at a wrong place, may have been the
early causes of this damage. It seems to have taken
place now and then even on a larger scale : Prof. H.
Wendt1 has proposed a highly probable conjecture on
vii. 15-24, which he removes from their present place,
putting them at the end of v.
But as I am dealing strictly with MSS. evidence,
I prefer to give another instance which is no less
clear, although not of so great importance. Eead
attentively xxi. 7 f . : " Now when Simon Peter
heard that it was the Lord, he girt his fisher's coat
unto him (for he was naked), and did cast himself
into the sea. And the other disciples came in a
little ship, for they were not far from land, but as
it were two hundred cubits." Why, if they had
XH. Wendt, Die Lehre Jesu, i. 228 ff.: cf. also Bertling, in Studien
u. Kritiken, 1880, 351 ff. (before Wendt), and F. Spitta, Zur Geschichte
und Litter, des Urchristenthums, 199 ff. (after W.).

240 PHILOLOGY OF THE GOSPELS.
been very far from land they would much more
naturally have come in the ship, and not swimming.
Turn now to our Syriac witness from Sinai, which
gives this quite clear and coherent text : "... he
girt his fisher's coat unto him (the words 'for he
was naked ' are left out) and did cast himself x into
the sea, and came swimming (Nonnus too renders
this clause), for they were not far from land ('but
. . . cubits ' is left out). But the other," etc.
There can be no doubt that this is the right order,2
but you may be still wondering why it was inverted,
and what is to become of the words "for he was
naked," and of " but . . . cubits," and of the " and
came swimming." Of course it is impossible here to
know anything certainly : we may imagine an original
text having some authentic supplements in the margin
(as is very frequently the case with our own manu
scripts), and shall be able to explain by that hypothesis
the omissions (the supplements having been overlooked
by a copyist), either wholly or in part, and the wrong
order (the copyist having corrected his mistake, but
in a wrong way), and all we want to explain. At
any rate, the " but as it were two hundred cubits "
looks rather authentic, and the " for he was naked "
^''StfiaXev iavrbv common text, but D rJXaTO, and so Nonnus.
2 In Nonnus the clause "for they,'' etc., stands after aipovres . , .
IxSiwv, at the end of ver. 8.

TEXTUAL CONDITION OF JOHN'S GOSPEL. 241
(i.e. he had no undergarment on) seems to be indis
pensable, while " came swimming " may be a comment
made on the model of "the other disciples came in
a ship." Thus you leave much uncertainty, you will
say. Of course I do ; am I that Indian Bird (see
above on page 79) possessing the miraculous gift of
discerning, in any given case, between spurious and
genuine particles ? We may be able to do so in
some cases, and more easily when the object of our
doubts is of some extent; but there are only too
many things about which we cannot ourselves arrive
at a firm conviction, still less convince others.
Having spoken quite enough of variants and of
minute matters, I shall conclude with some remarks
on the words in v. 2 : " Now there is at Jerusalem
by the sheep market (rather gate ; the Greek text
has only iirl rfj irpofiariKfl) a pool, which is called
in the Hebrew tongue Bethesda,1 having five porches."
" There is " ? Then Jerusalem was still existing at
the time when this was written ; and even John's
Gospel, the latest of all, was written before the year
71. I am not at all afraid of this inference which
has actually been drawn, and before our century ;
but as it is of great importance, we must be very
cautious. There is a German pastor, 0. Wuttig,
who has recently published a book on this Gos-
111 Which . . . Bethesda" is missing in Nonnus.
Q

242 PHILOLOGY OF THE GOSPELS.
pel,1 trying to establish that it was written in Judaea,
and before any other Gospel, in spite not only of
common opinion, but also of tradition, since Euse
bius, and perhaps even Papias, nay even Papias'
presbyter, that is John the apostle, regarded it as
being destined to give supplements to other Gospels.
Of course I cannot enter into this subject here ; but
the objections to Wuttig's thesis seem to be very
serious. He too relies very much on the " is "
in v. 2, although it evidently does not go far
enough for him : I at least am quite convinced
that in the year 68, that of Nero's death, both
Mark and Luke were in existence and generally
known. But what he says on the " is " in question
is very judicious: he attaches to that present the
more importance, as John ordinarily uses the im
perfect even when speaking of localities (see iv. 6,
" Now Jacob's well was there "), and as this was and
is quite a common assimilation to the tense of the
narrative, not at all implying that the thing in question
has now ceased to exist. But if we use the present
tense amid the imperfects of our narrative, we certainly
imply that this is even now so. Moreover, there is
no passage in this Gospel where the destruction of
Jerusalem is alluded to, and the latest event mentioned,
1 Licent. 0. Wuttig, Das Johanneische Evang. und seine Abfas-
sungszeit, Leipzig (G. Bohme), 1897.

TEXTUAL CONDITION OF JOHN'S GOSPEL. 243
the martyrdom of Peter, is universally placed in Nero's
time. Nevertheless, if you are to suspend a hundred
weight, you must take a rope and not a thread. If
the " is " (eanv) is universally attested, there are so
many threads joining that they will form a rope ; if
not, I shall call the thing still a thread, and have no
mind to suspend the hundred-weight by it. Now,
which is the case ? Greek mss. are concordant for
" is," but Nonnus has yv, " was," and so have all the
Syriac versions, and the Armenian, and the two old
Egyptian versions. Still the evidence for "is" is far
stronger, as Nonnus does not count much here, his
stand-point being rather that of his own time, and the
" was " might be due to assimilation to the " was "
of the preceding verse; as an editor, I should certainly
give " eanv," and not r)v. But that is not the
question here. The threads do not make up the rope,
and so I leave this question like many others to my
readers.

INDICES.

I. TO PEESONS, SUBJECTS, AND GEEEK WOEDS.

A (A).
aKoieiv, 31. I.
Alexandrian text of the N.T.,
64 ff., 147, etc.
Alexandrinus A (N.T. MS.), 64,
228, etc.
Ambrose, 162, 235.
dvardrreadai, 14 ff.
Andrew of Caesarea, 223. 2.
avudev, 17 f.
Aorist, contradistinction to pre-
sent-imperf., 121, 125, 130 f.,
188.
Apollonius (author of Conica), 2.
Apollos, 29 ff.
dpa, 188, 197 f.
Aramaic words in Mark, 213.
Aristion, disciple of the Lord,
222.
Armenian versions of the N. T. ,
69, 148, 243.
Atticism (in Greek literature), 9.
Atticus (T. Pomponius), 100.
Augustin, 107 ff., 131 f., 134,
139, 162, 235.

B(B).
Baneregez (Boanerges), surname
of the Apostles, or of John
and James only, 216.
Belser, J., 125 f.
Berger, S., 68, 106 f., 110.
Bertling, 239. i.
Bethlehem, cave in, 165 f.
Beza (Th. de Beze), 96 ff.
Bobbio, Latin MS. of (k), 80 ff.,
84, 86 f., 199,201,204,211. i.
Bornemann, F. A., 105 f.
fioiXeaBai instead of BiXeiv, 199.
Breslau Latin MS. (I), 152.
Burgon, Dean, 61.
Byzantine scribes, carefulness
of, 72. C.
Cantabrigiensis D, 64, 75 f., 81,
86, 96 ff., etc.
Centurio, the word received in
vulgar Greek, 211 f . ; ditto in
Aramaic, 212.
Chronicon Paschale, 26.
Chuzas - Cydias, 152 f.

INDICES.

245

Cicero, 100.
Clement of Alexandria, 25, 111,
146 f., 170.
Clericus, J. (Jean Leclerc), 68,
95, 98 f.
Coins, Boman, 212.
Colbertinus, Latin MS. (c), 185,
197, 211. I, 215.
Conybeare, F. C, 69, 126.
Corbeiensis, Latin ms. (/2), 157.
Corssen, P., 106 f., 120, 132.
Crescens, 154.
Curetonian Syriac MS., 87. 1, 146,
231.
Cyprian, 106, 109, 111, 236.
D(A).
Daniel (ix. 26 f.), 45 f., 49.
de Lagarde, P., 2 ff.
Sid, compounds with, 117 f.
StairovelaBai, 201.
Siaaatpelv, 116 ff.
Diatessaron, 146.
Dioscorides (author of Materia
Medica), 3, 5.
SovXos, o'tKirns in Luke, 116. I.
E (E, H).
¦t) and Kal confounded in MSS. ,238.
iSwKav - gSoaav, 9 f.
Egyptian versions of the N.T.,
110, 148, 243 (see also Sahidic
version).
iKSoais "edition," 100.
¦ijfiels instead of lyti, 13.
rjjiipa = time, 134 f.
iireiSJjirep, 8.
iiriKiXXw instead of iiroK., 186.
Epiphanius, 17, 85, 165 ff.
ipinjvevrijS, 192.

Eubiotus of Hierapolis, 219 ff.
Eusebius, 8, 12 f., 14, 17, 25 f.,
34, 84, 120, 131, 156, 187 f.,
193, 203, 220 ff., 234.
Euthalius, 37.
Evangelium quadripartitum, 74
146.
Ezra, 15 f. F.
Ferrariani, mss. of the Gospels,
155 ff.
Fisher, F. H., 123. i.
Floriacensis, Latin MS. of Acts,
106 f., 110. G(T).
Galilaean character of Synoptic
Gospels, 207.
German versions of the N.T., 68.
Gigas, Latin MS. of the N.T. ,110.
v. d. Goltz, E., 250.
yovvirerelv, rtBivai rd ydvara, 200.
Graefe, Ferd., 87. I, 132, 140,
143, 172. I, 2, 179. i.
Gregory of Nyssa, 177.
Grimm, C. Z. W., 13 ff.
H.
Harnack, A. , 35, 40, 44, 83, 192.
I, 224 f.
Harris, J. Eendel, 69, 76. i, 126,
155, 185 f.
Haussleiter, 222, 224.
Hebrews, Gospel according to
the, 156.
Hemsterhusius, Tib., 99.
Hilgenfeld, A., 122, 126.
Holtzmann, H. , 126.
Homer known to Luke, 186 f.
Hort (see Westcott-Hort).
Hoskier, H. C, 61.

246

PHILOLOGY OF THE GOSPELS.

KD.
rb iKavbv iroieiv, Xafietv, 199.
Imperfect in contradistinction
to aorist, 130 f., 158, 188.
Irenaeus, 15 f., 96, 107 f., Ill,
170, 235.
'Iwdvijs - 'ladvvijs, 75 f., 81.
J.
Jerome, 12, 14, 17, 36 f., 78 f.,
85, 162, 217.
John the Presbyter ?, 222 ff.
Justin, 47, 85, 144, 146, 166 f.,
223. 2, 235.
K(K).
KaBd, KaBws, 8.
Ka$e£rjS, 18 f.
KarnxeiaBai, 20, 31. I.
Kipling, Th., 98.
Kpdnare, 19. L(A).
Latin mss. of Gospels, 50, 56, 85,
86, 139, etc.
Laudianus E of Acts, 110.
Lewis, Agnes Smith, 59, 146.
Atfieprlvoi - Aifivarivoi, 69 f .
Lightfoot, Bishop, 91 ff., 143.
Lippelt, E., 76. i, 144. i, 157.
Livy, 68.
\o«r6s, 174, 211. I.
[Longinus] irepl iltfrovs, 6.
Lucifer of Cagliari, 110.
M.
Marcion, 6, 50, 56, 93 f., 144 ff.,
154, 161, 177 f., 183.
Mary, very common name in
Judaea, 190.

Maximus Confessor, 177.
Minuscule codex of Acts (137),
110.
Minuscule codex of Gospels (700),
61, 169, 177, 197, 208.
N(N).
vavs instead of irXoiov, 186.
Nestle, E., 68, 76. 2, 103, 126,
194.
Nonnus, paraphrase of St. John's
Gospel, 156, 225 ff., 229 ff.,
238, 240. i, 2, 241. I, 243.
0(0).
otSa/iev - ta/iev, etc. , 9.
Origen, 111, 166, 182, 184, 197.
Oxyrhynchus, excavations of, 22,
221. P(n).
Page, J. E., 123 ff.
Palatinus (e), Latin MS. of Gos
pels, 50, 56, 167, 211. i, 212.
i, 215, 225, 230, 232 f., 236.
Papias, 18, 156, 192, 196, 198,
206, 218 ff., 234.
irapaylveaBat, 117.
irapaKoXovBelv , 17 f.
Participle of future with us, 197.
Pearce, 68.
irXijpotpopelv, 8, 12 ff.
Plutarch, 2, 14.
iropeieaBat, 202.
Present (conj., etc.) in contra
distinction to aorist, 130 f.
irpoaeyyl$etv, 116.
irpoartBivat, -eaBat, 201 f.
irpwros instead of wpbrepos, 38.
Provencal version of the N.T.,
68.

INDICES.

247

E.
Eamsay, Prof. W., 120, 123.
Eesch, A., 235.
Eoman Church, 155, 160 f.
Ruegg, A., 213. i.

Sahidic version of the N. Test.,
110, 185.
Salmon, Provost G., 91. i, 122,
126, 123 ff., 222. i.
Sanday, Prof., 177.
Savonarola, Jerome, 41 ff.
Schlager, G., 176. I.
Scrivener, F., 98.
Semitic words in the N. T. , 1 72. i .
Semler, 98 f.
Severianus of Gabala, 11.
Simcox, W. H., 61.
Sinaiticus (S) of the N.T., 64,
68, 83. I, 87, etc.
Size of papyrus rolls, 213 f.
Spitta, F., 59. I, 239. I.
Stephanus, E., 97.
Strabo, 28.
Syriac versions of the N.T. (Cure-
tonianus, Sinaiticus, Peshitta),
59,64, 82f.,84ff., 139f., 146f.,
148, 183 f., 208,230, 239 f.
Syrus posterior(Philoxenian ver
sion) in Acts, 109, 121, 123. I,
147.

T.
Tatian, 146.
Tertullian, 107, 111, 145, 177,
235 ff.
Tischendorf, 138 f., 188 f., 199.

Vaticanus (B) of the N.T., 64 f.,
68, 75, 81, 87, 91, 147 f., etc.
Vercellensis (a), Latin MS. of
Gospels, 184, 215.
Veronensis (6), Latin MS. of Gos
pels, 197, 235.
Virgil, 185. W.
Weiss, Bernh., 40, 116. 1, 120 ff,
130. I, 1321,136, 141.
Wellhausen, 183.
Wendt, H, 239.
Westcott-Hort, 64, 84, 138 f.,
148, 156,161, 177, 180 f., 184,
189.
Western text of Gospels and
Acts, 64 ff., 84 f., etc.
Wright, A., 31. I.
Wuttig, Lie. 0., 241 f.

Zahn, Th., 29, 31 ff., 87. I, 145,
227.
Zockler, O., 126.

II. TO PASSAGES OF THE N.T. DISCUSSED.

St. Matthew.

PACE

PAGE
xiv. 4,
201
i. 2-17,
171 ff.
xiv. 25,
201 f.
16,
86 f.
51 f.,
191
25,
87 ff.
58,
211. I
iii. 2,
84
xv. 15,
198 f.
15,
85
18 f.,
199 f.
viii. 5, 9,
82 f.
21,
200
13,
83. i
24,
200
xvi. 2 ff.,
85
28,
201
xix. 24,
209
39 (Kevrvpltav),
211 f.
xx. 16,
84. i
xvi. 1,
202 f.
28,
85
2,
204
xxii. 7,
40
4,
203 f.
35,
-
176. i
8,
141
xxiv. 15,
45
29,
48 f.
St. Luke.
xxvii. 49,
84. 2
i. 1-3,
25,
1-20 135
St.
Mark.
ii. 4f.,
170 f.
iii. 13 ff.,-
142
, 215 f .
7,
165 ff.
21 f., -
211. i
41, 48,
88
iv. 19,
211. i
iii. 22 f.,
167 ff.
41,
198
23 ff.,-
170 ff.
vi. 7,
212. i
35 f.,
173
vii. 3 f.,
212 f.
v. 27 f.,
207 f.
25 f., -
200
vi. 5 fi,
153 f.
x. 24,
211. i
13 ff.,-
134, 142
25,
209
vii. 1, 8,
82 f.
xi. 4ff.,
204
viii. 3,
152 f.
13,
196 ff.
43,
147 f.
xii. 42,
212
ix. 51,
135. :
xiii. 14,
45
54 ff.,-
90,94
24,
48 f.
x. 41 f.,
148 f.
INDICES.

249

PAGE
PAGE
xi. 1 f.,
174 ff.
vii. 53-viii.
H, ¦
89 f
, 155 ff.
xi. 2 f.,
176 ff.
viii. 6, 9,
-
119
xii. 19,
149
xviii. 13 ff.,-
56 ff.
46,
135
xix. 35 f.,
225 ff.
xviii. 25,
209
42, xx.
i.
185. i
xix. 29 ff.,-
149 f.
xx. 30 f.,
226 f.
43 f., -
41
xxi. 7f.,
239 ff.
xxi. 20 ff.,-
41 f., 46
22,
97
24 f., -
46, 48 f.
24,
226 f.
28 ff.,-
51
32,
49 ff, 54
i. 1,
2ff.,
Acts
38
132 ff.
37 f., -
157 f.
xxii. 7-23,
179 ff.
8,
24 f.
24 ff.,-
182
iii. 14,
194 f.
43 f.,
91
vi. 9,
69 f.
51,
182
x. 24,
116 f. i
53,
182
25,
116 ff.. 127
xxiii. 2, 5,
183
xi. 2/3,
129
10-12,
183 f.
4
18 f.
34,
91, 93 f.
27,
131 f.
44 f.,
184 f.
xii. 12, 17
193
53,
185 ff, 188
xiv. 1 ff.,
121 ff.. 127
54 f., -
187
xv. 20, 29
250
56,
187 f.
41 -xvi
4,
158
xxiv. 1,
188
xvi. 12,
67 ff.
12,24,
188 f.
xviii. 24 f.,-
30 ff.
51 ff.,-
-138 ff., 213 f.
xxi. 15 ff.,
128 ff.
St.
John.
xxv. 24 ff.,
150 f.
i. 13,
234 ff.
xxvi. 4,
9
ii. 2,
232 f.
xxvii. 41,
186
4,
238 f.
xxviii. 14,
131
iv. 6 ff.,
231 f.
16, 29,
92
14,
232
v. 2,
241 ff.
Galatians.
3f.,7,
89, 228 f., 234
ii. 11,
25, 26 ff.
9ff.,
230 f .
2 Timothy.
vii. 15-24,-
239
iv. 10, -
154
ADDENDUM to pp. 70 f. and 106 F.
Licent. E. von dee, Goltz, of Berlin, recently discovered, in
a monastery at Mount Athos, a manuscript of the Acts and
Epistles, not very ancient, it is true, but, at all events, derived
from an ancient and very valuable original. The margin contains
numerous scholia, in which many variants, especially in the
Pauline Epistles, are traced back to quotations by Eusebius,
Origen, Irenaeus, and Clement ; there is exhibited a degree of
carefulness even about small things, which we were by no means
prepared to meet with. But, at the same time, it is quite evident
that there existed no authoritative text of the sacred books. In the
Acts there is one important addition of D, which is acknowledged
both in the text and in a scholium : ch. XV. , 20 and 29, the words Kal
6aa av fiij BiXwaiv airois yeveaQai, iripots firj iroieiv (after tov a'tfiaros
in v. 20 and after iropveias in v. 29) stand in the text included by
asterisks (*), much in the same way as in the Syrus posterior of
Acts (see p. 109), or in the remnants of Origen's Hexapla. The
scholium attests that this was the reading given by Irenaeus in
his third book, and by Eusebius in his sixth and seventh book
against Porphyrias. Another scholium (to v. 29) gives the close of
the Epistle of the Apostles according to Irenaeus : 4£ wv StarnpovvTes
iavrois ev irpd^ere, tpepbfievoi iv dyiw irvei/iart (without 'ippwaBe), stating,
at the same time, that Kal rov ttviktov in v. 20 and Kal itviktwv in v.
29 was omitted by the same authority. We actually find this very
text ( = D) in the extant Latin version of Irenaeus. Now it becomes
more and more impossible to ignore Western variations, since it is
proved that they go back to such very early times, and that attention
was bestowed upon them by the ancient Greek Church itself. My
own larger edition of the Acts is nothing but an expansion of the
method followed by the Athous in those passages of ch. XV. , and I
should think that there were Greek manuscripts where the same
method was thoroughly followed, as it is in the Syrus posterior.
Lie. v. d. G. is preparing a publication of the scholia ; meanwhile
he has kindly allowed me to state so much here.

GLASGOW : PRINTED AT THE UNIVERSITY PRESS BY ROBERT MACLEHOSE AND CO.