THE FOUR GOSPELS - A Study of Origins, The Manuscript Tradition, Sources, Authorship, & Dates. By B.H. Streeter, Hon.D.D.Edin. Fellow of Queen's College, Oxford; Canon of Hereford. First published by MacMillan & Co Limited 1924. - Prepared for katapi by Paul Ingram 2004.



Home | Diagram – Synoptic sources | synopsis | the MS evidence | the loss- a primitive one | the longer conclusion | the lost ending- a speculation | list of MSS | (pages 333-360)


the MS. evidence

Greek, Syriac, Armenian and Old Georgian evidence for complete omission of Mk.xvi.9-16.

Significance of the Shorter Conclusion and of the epithet κολοβoδάκτυλος.
Early evidence for the Longer Conclusion.  The "Freer logion'

— an addition found in W.

the Loss a primitive one

There is no difficulty in supposing either
(a) that the Gospel was never finished, or
(b) that the earliest copy was accidentally mutilated.

The view that Mark went out of circulation for a time so that only one damaged copy survived is incompatible with the evidence for its wide use in the first half of the second century.

There are also fatal objections to the theory that the original ending was deliberately suppressed.

The copies of Mark used by Matthew and Luke seem to have ended abruptly at the same point as our oldest MSS. If so, the loss must be primitive.

the longer conclusion

The note in a Xcent. Armenian codex attributing the Longer Conclusion to the Presbyter Ariston probably represents, not a genuine tradition, but an ingenious conjecture by some reader of Eusebius.

Considerations of textual criticism suggest a Roman origin for the Longer Conclusion.

The addition found in W favours the hypothesis that it was originally composed as a catechetical summary of Resurrection Appearances, not as a conclusion to the Gospel.

the lost ending—A speculation

Tentative suggestion that the Appearance to Mary Magdalene and that to Peter by the lake in the Fourth Gospel represent, directly or indirectly, the lost ending of Mark.

Two objections from the standpoint of textual criticism considered.

The possibility that the conclusion of the Apocryphal Gospel of Peter was derived from the Lost End of Mark.

Considerations which suggest that John xxi. represents a portion of the Lost Ending.

Considerations in support of the view that an Appearance to Mary Magdalene was also found in Mark.

The evidence available quite insufficient to establish an assured result. But the improbability that the earliest tradition of the Resurrection Appearances should have left no trace at all in the Gospels is so great that even a tentative hypothesis is worth consideration.



the MS. evidence

EUSEBIUS, c. 325, the most widely read scholar of Christian antiquity, states that in the oldest and best MSS. known to him the Gospel of Mark ended with the words

for they were afraid, xvi.8; 

and he did not include the succeeding twelve verses in his canons or tables of parallel passages. [The fact that the verses were ignored in the Eusebian canons is noted at the end of Codex 1, 1582, and other MSS.]   The Gospel ends at this point in B א, the two oldest and best MSS. known to us; and as, in view of the statement of Eusebius, we should expect, there is good evidence (cf. p. 88) that it was absent from the old text of Caesarea represented by fam. Θ.  The Gospel ends at the same point in Syr. S.; also in three of the oldest MSS. of the Armenian, which is additional evidence for the omission either in fam. Θ or in the Old Syriac.   A fourth very early Armenian contains the last twelve verses, but separated from the rest of the Gospel with a note "of the presbyter Ariston."  In the oldest MS. of the Georgian version, which is dated 897, the Gospel ends at xvi.8.  But the LONGER CONCLUSION (as the last twelve verses are usually styled) is added as a sort of Appendix to the Four Gospels after the end of John, having apparently been copied from another text. [The Adysh  Gospels  (Phototypic edition),  Moscow,  1916.  I owe this information to my friend Dr. B. P. Blake.]

What is known as the SHORTER CONCLUSION is found in L Ψ 579 and two uncial fragments, in the Sahidic and Aethiopic versions, and in the African Latin k ; also in the margin in one Greek cursive, in the Harclean Syriac, and in the oldest MS. of the Bohairic. It reads as follows:

And all that had been com­manded them they briefly reported to Peter and his company; and after these things Jesus himself appeared, and from the East to the West sent through them the sacred and incorruptible proclamation of eternal salvation.

In these Greek MSS. and most of the versions, but not in k , the Longer Conclusion (Mk.xvi.9-20) follows the Shorter, being usually introduced by the words,

This also is current,
ἐστὶ καὶ ταῦτα φερόμενα.

As the Shorter Conclusion is obviously an attempt by some early editor to heal the gaping wound, the MSS. and versions, which contain it, really afford additional evidence for a text that ended with

φοβοῦντο γάρ.

The distribution of the MSS. and versions, taken in connec­tion with the statement of Eusebius, compels us to assume that the Gospel ended here in the first copies that reached Africa, Alexandria, Caesarea, and Antioch. Since in all probability the African text originally came from Rome, the burden of proof lies on the person who would argue that it was not also missing from the most ancient Roman text. And this would explain the epithet κολοβoδάκτυλος (as if in English one were to say "docked") applied to Mark by the Roman theologian, Hippolytus, c. 200.  Hippolytus himself used a text of Mark that contained the last twelve verses and understands the epithet of its author; but its origin is more easily explained as originally applied to the book. Originally κολοβoδάκτυλος; was used of a man who cut off a thumb in order to escape military service.  Wordsworth and White suggest it may have come to mean "shirker," and that Hippolytus found the term applied to Mark by Marcion in order to discredit his Gospel, in allusion to the withdrawal from the work in Pamphylia which St. Paul so much resented, Acts xv.38.  But even so, Marcion's attack would have been twice as effective if the epithet carried a double entendre, the author a shirker, his Gospel a torso. At any rate the author of the Gospel cannot have originally meant to end it without the account of the Appearance to the Apostles in Galilee, which is twice prophesied in the text (Mk.xiv.28, xvi.7). Indeed, the words ἐφοβοῦντο γάρ in Greek may not even be the end of a sentence; they lead us to expect a clause beginning with μύ, "They were afraid, lest they should be thought mad," or some­thing to that effect.

The LONGER CONCLUSION, found in the majority of MSS. and in our printed texts, is not at all in the style of Mark; and, as will appear later, a close study of its contents makes it in the last degree improbable that Mark himself wrote it. But it must have been added at a very early date. Irenaeus, c. 185, ["In fine autem evangelii ait Marcus," Adv. Haer. iii. 10. 6.] quotes xvi.19 expressly as from "the end of Mark"; and the Longer Conclusion already stood in the text used by Tatian when compiling his Diatessaron c. 170; and there are possible, though not quite certain, reminiscences of it in Justin and in Hermas. Since B א were written in the fourth century, both the Longer and the Shorter Conclusions were already of great antiquity, and can hardly have been unknown to the scribes who wrote these MSS. and, for that matter, to a fairly long succession of MSS. from which they were copied. Incidentally I may be permitted to remark that an asceticism which could decline to accept either of these endings argues a fidelity to a text believed to be more ancient and more authentic, which materially increases our general confidence in the textual tradi­tion which these MSS. represent.

The discovery of W has added yet another to the previously known endings of the Gospel. After xvi.14 occurs the section (part of which is quoted by Jerome, as occurring in some MSS.),

And they replied saying, this age of lawlessness and unbelief is under Satan, who does not allow what is under the unclean spirits (emending two words of the Greek to correspond with Jerome's Latin) to comprehend the true power of God; therefore reveal thy righteousness. Already they were speaking to Christ; and Christ told them in addition that the limit of the years of the authority of Satan has been fulfilled, but other terrible things are at hand, even for the sinners on whose behalf I was delivered up to death, that they might turn to the truth and sin no more, in order that they may inherit the heavenly spiritual incorruptible glory of righteousness.


The Loss a primitive one

But how are we to account for the Gospel thus breaking off short? It is, of course, possible that Mark did not live to finish it. But, if he did, it would seem probable that the end of the roll on which it was written must have been torn off before any copies of it had got into circulation. Otherwise such a loss would have been repaired at once from another copy.

There is no difficulty in supposing that the original copy of Mark, especially if the Gospel was written for the Church of Rome about AD 65, almost immediately lost its conclusion. The two ends of a roll would always be the most exposed to damage; the beginning ran the greater risk, but, in a book rolled from both ends, the conclusion was not safe. How in the case of Mark the damage occurred it is useless to speculate.  At Rome in Nero's days a variety of "accidents" were by way of occurring to Christians and their possessions.  The author of Hebrews, writing to the Roman Church, alludes to the patient endurance of "spoiling of their goods."
[This is not undisputed, but it is the simplest explanation of the fact that while Hebrews profoundly influenced the theology of the Roman Church as early as 1 Clement (c. 96 AD), it was only quite late, on the authority of the East, that it was accepted at Rome as by Paul. Probably for many years the Roman tradition preserved the name of the real author.]
That the little library of the Church, kept in the house of some prominent adherent, should have suffered in some "pogrom" is highly credible. Curiously enough, there is evidence that copies of Romans were in circulation, which lacked the last two chapters, which looks as if one of the earliest copies of that Epistle, the one other document of which we can be quite sure that the Roman Church had a copy at this time, was similarly mutilated.

Professor Burkitt accounts for the disappearance of the original conclusion of Mark on a different hypothesis. 
[F. C. Burkitt, Two Lectures on the Gospels, p. 33 ff. (Macmillan, 1901).]
  Mark, he argues, contained nothing that interested the Early Church which was not included in either Matthew or Luke; hence for a generation or two, after those Gospels had been composed, it ceased to be copied. Later on, when, in face of the struggle with Gnosticism, a formal canon of accepted Gospels was under discussion, the Roman Church remembered that among its archives was an old copy of Mark, and insisted on this being included. But the end of the roll had been torn off, and there was no other copy in exist­ence from which to repair the loss. To this theory there are formidable objections.

(1) A world-wide circulation of Mark in the first century is implied by the use made of it by the authors of Matthew, Luke, and John, who must have written in Churches at a wide remove from one another in theological outlook, and probably also in geographical situation. In view of this, the total disappearance in the course of the next fifty years of all copies but one is not very likely.

(2) Since Mark was made use of in the Diatessaron of Tatian, c. 170, the supposed rediscovery of the Gospel must have taken place before this date. And it must have been some consider­able time before, for two reasons. First, the only point of compiling a Harmony of the Gospels at all was to meet the inconvenience, for purposes of practical teaching, of having four parallel, and in some points apparently conflicting, Lives of Christ. But the difficulty arising from there being four standard Lives must have been in existence long enough to be felt as a difficulty, before the remedy was looked for. Secondly, Tatian's copy of Mark contained the Longer Conclusion. But since the earliest copies of the rediscovered Gospel that reached Africa, Alexandria, and Syria did not contain this, there must, on Burkitt's theory, have been an interval of time after the re­discovery of Mark during which the Gospel circulated without this addition to its text. The supposed rediscovery, then, must have been some time before 170, at the latest 165. Hence the period of complete disuse (during which all copies but one had time to disappear) must have been the fifty years or so previous to this. But for this period we have evidence of a widespread interest in and use of the Gospel.

(a) I regard it as practically certain that Mark was known to Justin c. 155; also to Hermas either c. 140, or, as I think more likely, c. 100. Both these are evidence for use in Rome.

(b) For Asia we have the evidence of Papias.
He wrote rather late in life at a date which Harnack fixes as between 145-160; other scholars prefer an earlier date, 130-145.
[Irenaeus, AD 185, styles Papias

a hearer of John, a companion of Polycarp, and an ancient worthy (ἀρχαῖος ἀνύρ).

This distinctly favours the earlier date.
But the De Boor fragment (printed Lightfoot and Harmer, Apostolic Fathers, p. 518 f.) seems to imply that he looked back on the time of Hadrian, who died
AD 138. As, however, the statement containing the reference to Hadrian is attributed by Eusebius, iv. 3, to Quadratus, who, he expressly says, lived in the reign of Hadrian, it is probable that the fragmentist (who is undoubtedly indebted to Eusebius elsewhere) is really quoting, not Papias, but Eusebius—especially as the statement that a contemporary of Christ lived till the time of Hadrian is absurd, while it is by no means unlikely that one should have lived till the time (i.e. till the birth) of Quadratus, who may have been an old man in Hadrian's reign. See the discussion by Prof. J. V. Bartlet in Hastings' Dict, of Christ and the Gospels, ii. p. 311, col. 2.]
His quotation of the famous statement about Mark made by John the Elder has already been discussed (p. 17 f.), so I will only stay to point out that, whatever else it proves, it is convincing evidence of three facts. First, at the time when Papias wrote, Mark was regarded in Asia as a standard work about whose origin Christians in general were interested. Secondly, the same thing held good in Papias' youth; or why, when he was collecting what seemed to him the most valuable of the teachings of the elders, did he trouble to note what they said about this Gospel? Thirdly, if we accept the view maintained by some scholars that Papias had not actually met the Elder John, Papias was not the first to elicit the statement from the Elder, in which case the date at which the Gospel of Mark, and the degree of authority to be attached to it, was a matter of public interest is pushed back earlier than Papias' youth.

(c) Irenaeus [Adv. Haer. iii. 11. 7.] says that Mark was the Gospel quoted as their authority by those heretics who separate Jesus from Christ and say that Christ remained impassive while Jesus suffered.

This statement is borne out by the fact that the apocryphal Gospel of Peter, which was evidently written in order to promulgate views of that kind, exhibits a special preference for Mark. The date assigned to "Apocryphal Peter" by most scholars is 130-140. Personally I think that too early; but, on any hypothesis, the above statement of Irenaeus and the preference shown by "Peter" for Mark are evidence of the vogue of that Gospel in yet another circle in the middle of the second century.

The only conclusion that can be drawn from the facts is that the comparative neglect of Mark, of which there is plenty of evidence in later times, began after, not before, the universal acceptance of the Four Gospel Canon.

There is still less to be said for a hypothesis, at one time popular on the Continent, that the original end of Mark was deliberately suppressed and the Longer Conclusion substituted for it. This is supposed to have been done in Asia as part of the process of forming an official Four Gospel Canon in the latter part of the second century, the object of the suppression being to get rid of the discrepancy between Mark's account, in which the first Resurrection Appearance is in Galilee, and the Jerusalem tradition, followed by Luke and John.

The main objections to this theory are four.

(1) The idea that the Four Gospel Canon arose in Asia, or, indeed, that it came into existence as a result of any one official act at all, is one for which, so it seems to me, the evidence is non-existent.

(2) While the revisers were about it, why did they not suppress the end of Matthew as well, since, in the matter of the first Appearance being in Galilee, his account equally conflicts with that of Luke and John? Again, if they were out to remove discrepancies between the Gospels, why did they not begin the "cut" a verse earlier, so as to remove the contradiction between Mark's statement that the women "told no man," and the statement by Matthew and Luke that they at once went and told the disciples?

(3) How was it that revisers succeeded in getting the Churches of Africa, Alexandria, and Syria to accept at once the excision of the original ending, which spoils the Gospel, without accepting the substitute which is said to harmonise it with the others?

(4) The use made of Mark by the authors of the other three Gospels proves, I must repeat, that Mark was universally read at the end of the first century; and it continued to be so throughout the second. Hence the suppression of the ending of a Gospel so widely circulated—and that at such an inappropriate point, ἐφοβοῦντο γάρ—would only have been possible if there had existed, as in the modern Roman communion, a highly centralised organisation able to enforce world-wide uniformity. All our evidence as to the history of the Church during the first two centuries points to the lack of any such thing. Least of all was it exercised to secure uniformity in the text of the Gospels. We have actual MSS. written in the fifth century to show that even then there were still current three different endings of Mark (not counting the absence of an ending found in א B), viz. the Longer Conclusion A C D, the augmented Longer Conclusion W, the Shorter Conclusion k (L Ψ). If this variety was possible in the fifth century, after a hundred years of oecu­menical conference, the notion is absurd that a machine existed in the second century capable of securing a world-wide excision in the text.

Let us now ask whether the end of Mark may not have been already missing in the copies of the Gospel used by Matthew and Luke.

(1) The message of the Angel,

Go tell his disciples and Peter he goeth before you into Galilee:
there shall ye see him as he said unto you

(Mk.xvi.7), is clearly intended to refer back to the previously recorded prophecy of Christ,

Howbeit after I am raised up I will go before you into Galilee. (Mk.xiv.28).

Thus we are bound to infer that the lost conclusion of Mark contained an account of an Appearance to the Apostles in Galilee. Further, this must either have come after an Appearance to Peter separately, or it must have been an Appearance in which Peter was in some way especially singled out for notice, as he is in Jn.xxi.

Now Matthew follows the text of Mark all through the Passion story with great fidelity; if, then, the copy of Mark used by him had contained a conclusion of this sort we should expect to find it reproduced by Matthew.   But Matthew, though he records an Appearance to the Eleven in Galilee, does not especially mention the name of Peter in connection with it. Again, the most striking thing about the Gospel of Mark is the author's gift for telling a story in a vivid, picturesque, and realistic way.   Elsewhere, wherever Matthew is following Mark, he abbreviates slightly and occasionally omits a picturesque detail; nevertheless the account he gives is always a vividly realised and well-told story—full of detail, though not quite so full as the Marcan original.  But Matthew's account of the Resurrection Appearances—to the two Maries (Mt.xxviii.9-10) and subsequently to the Eleven (xxviii.16-20)—is extremely meagre and is conspicuously lacking in these usual characteristics.  Both, then, because Matthew does not mention Peter and because his narrative becomes exceptionally vague at the exact point where the authentic text of Mark now ends, we infer that his copy of Mark ended at that point.

(2) Luke, we have seen, based his account of the Passion and Resurrection mainly on his non-Marcan source; but he has omitted nothing of interest in the Passion story as found in Mark. He prefers his own source wherever it gives an equally elaborate or interesting version of any incident, but, where Mark contained something not occurring in his other source, he has added it in an appropriate context.  If, then, Mark contained a detailed description of an Appearance to Peter and to the Apostles in Galilee, it would have been, to say the least of it, a "strong" procedure to ignore completely this well-established tradition and represent all the Appearances as having taken place in or near Jerusalem.  But if in his copy of Mark, as in ours, there was no account of Appearances to ignore, that difficulty disappears.  There is a more important consideration.  
At the end of the Emmaus incident Luke has a reference to the Appearance to Peter.  The disciples are made to say,

The Lord is risen indeed and hath appeared to Simon. (Lk.xxiv.34).

That the first Appearance was to Peter is stated by Paul (1 Cor.xv.5) and is implied in Mark.  Luke's allusion makes it clear that he wished to bring out this fact.  He accepted the tradition; but it would seem as if he knew no more details about it than can be inferred from 1 Cor. and the existing text of Mark.  If he had found a detailed account of the Appearance to Peter in his copy of Mark he would surely have made some effort to adapt it to his story, even if he was puzzled by the Galilean tradition. We conclude, then, either that Mark did not live to finish his Gospel—at Rome in Nero's reign this might easily happen— or that the end of the Gospel was already lost when it was used by Matthew and Luke.

the longer conclusion

The note by an unknown scribe of an Armenian MS. of the tenth century, which suggests that the Longer Conclusion was the work of the 'Presbyter Ariston', has been taken rather too seriously in some quarters.  It is, of course, always possible that a genuine tradition may survive in some late MS. in an out-of-the-way district. But the principles of historical criticism, as ordinarily accepted, do not encourage us to begin by taking for granted that a statement is good evidence when it appears for the first time in a writer who, on the face of it, is far removed both in time and place from the facts he attests.

That such evidence here is in itself totally wanting will appear from the three following considerations.

(1) In three of the four tenth-century Armenian MSS. examined by Mr. Conybeare (than which I believe no older are known) the Gospel of Mark ends as in B א. This is a piece of evidence for the history of the text of some importance; for it shows that MSS., either Greek or Syriac, which lacked the Longer Conclusion were used by the original translators c. 400, or else by those who revised it at a slightly later date.  Thus in any case it is evidence of the circulation or superior repute of the shorter text in the Far East. The fourth MS., which (after a break, indicating that the scribe regarded what follows as a sort of Appendix) adds the Longer Conclusion, does so with the words "of the Presbyter Ariston" in the margin of the first line. This again is an important piece of evidence; it is prima facie evidence that the Longer Conclusion was a late introduction in Armenia, and that when first introduced it was not regarded as being by the pen of Mark himself.

The possibility is theoretically open that the Longer Ending, plus the note attributing it to Ariston, was in the earliest form of the Armenian, but that later scribes—feeling that, if it was not by Mark himself, it was not canonical—dropped it out. But against this are the facts (a) that from the fifth century onward the Longer Conclusion stood in the texts received in both the Greek and the Syriac Churches, both of which had considerable influence on Armenian Christianity.
(b) In other respects the Armenian text is closely related to that of fam. Θ and Syr. S., which omit the Longer Conclusion,
(c) Later Armenian MSS. included the Longer Conclusion without any note of doubt.

It appears, then, that all the influences known to have operated were in the direction, not of excluding, but of accepting this particular reading. In that case the words "of the Presbyter Ariston" must either have come in at the same time as the Longer Conclusion, or else it must represent a conjecture made at a still later date to account for the fact that the addition was not contained in the older text. Hence, whatever the origin of the tradition, it has no claim to be regarded as specially ancient; for if the oldest Armenian did not contain the Longer Conclusion it could not have contained a note about its authorship.

(2) The end of Mark was a problem much discussed. Eusebius alludes to it several times, Jerome more than once. It is treated in the Commentary attributed to Victor of Antioch found in many MSS.; and numerous MSS. have scholia dealing with it. It is thus extremely improbable that any authentic tradition as to the authorship of the Longer Conclusion which survived would have entirely escaped the notice of all these, to turn up in Armenia in the tenth century.

(3) The occurrence in a tenth-century Armenian MS. of interesting information about the authorship of the disputed ending of Mark is a phenomenon which cannot be estimated apart from the appearance in the Greek-speaking world a century earlier of much information of the same sort, quite obviously based on mere conjecture or on tradition of no value at all. From the IΧcent. MS. K, for example, we learn that he published the Gospel of Matthew in Jerusalem eight years after the Ascension.  From Y, of the same date, we gather that it was written by Matthew in Hebrew, but translated by John; later MSS. have a similar note, but substitute for the name John that of James, the brother of the Lord, or Bartholomew. A Vulgate Latin MS. of the same century tells us that Papias wrote the fourth Gospel at the dictation of the Apostle John. [Quoted in Tischendorf, i. p. 967 ad fin.] Surely, when we find a whole crop of this sort of thing springing up in Greek and Latin MSS., the burden of proof lies with anyone who wishes us to take seriously a piece of information of a precisely similar character which turns up in an isolated Armenian MS. a century later.

If anyone asks, why should conjecture light on the name of Ariston in particular—or Aristion, which seems to have been an interchangeable form of it—one may hazard a guess.  What was wanted was a name that would give to the Longer Conclusion the authority of an eyewitness.    But all names of Apostles were excluded, since an Apostle would be hardly likely to add an appendix to a Gospel written by one not an Apostle.    The Church history of Eusebius was everywhere read, and no pass­age would be more familiar than the one (cf. p. 18 above) on the origin of the Gospels in which he tells how Papias reports that he diligently sought for authentic traditions from Apostles and from Aristion and the elder John, disciples of the Lord. Here we have all the materials for a "brilliant conjecture."  Is not the Longer Conclusion of Mark one of the traditions derived by Papias from an eyewitness, Aristion, the disciple of the Lord?  Eusebius in the very same context alludes to the theory (of Dionysius of Alexandria) that the Apocalypse was the work of John the Elder.  In the Eastern Church in the third and fourth centuries there was a strong tendency to regard the Apocalypse as, at best, of sub-canonical authority, and it was not in the canon of the Syriac-speaking church on which the Armenian largely depended for its literature.  The Longer Conclusion of Mark was another such sub-canonical writing— what more natural than to surmise that it was the work of the second of the two "disciples of the Lord" mentioned by Papias. The conjecture is such a brilliant one that we might be tempted to accept it and believe that Papias had actually said so, did we not know how puzzled were Eusebius and others, who must have read Papias, about the authenticity of the Longer Conclusion.

But if the Longer Conclusion has really nothing to do with Papias' Aristion, where did it come from?
At any rate we have two facts to start from.

(1) It was in the text of Mark used at Rome before Tatian, 170; possibly in that used in Ephesus by the author of Epistula Apostolorum, c. 180, and certainly in the text of Irenaeus, 185, who was connected with both Rome and Ephesus. [Even in view of the other passages (cf. p. 70) which suggest that the Epistula used a Western text, the inference that its author knew the Longer Conclusion is highly precarious. The women go to the tomb "weeping and mourning," Epist. 9; Mary goes to the Apostles "as they mourned and wept," Mk.xvi.10.]

(2) It was not in the text used in Africa, Alexandria, Caesarea, and Antioch half a century later; and, to judge from the MS. evidence, did not establish itself there before the end of the fourth century.

It is thus a fair presumption that it originated either in Rome or in "Asia." The case for Asia depends, so far as I can see, on four pieces of evidence,
(a) The idea that the Longer Conclusion is connected with the Aristion mentioned by Papias. On that no more need be said.
(b) The theory that the original Conclusion was suppressed and this substituted for it at the time of forming of the Gospel Canon. The baselessness of this theory I have, I hope, sufficiently demonstrated above,
(c) Mk.xvi.18 reads,

if they shall drink any deadly thing,
it shall in no wise hurt them.

Papias, according to Eusebius [Euseb. H.E. iii. 39.], gave, perhaps on the authority of the daughters of Philip who lived at Hierapolis, a "wonderful story about Justus who was surnamed Barsabas, how that he drank a deadly poison, and yet, by the grace of God, suffered no inconvenience." It is inferred that a document, which apparently alludes to this incident, must have been written in Asia where the story was known. The inference is precarious. It is not suggested Justus ever left Palestine; but "all roads led to Rome," and if a story of a wonderful escape of an eminent Christian from poisoning was current it would probably be told in Rome before long,
(d) The Longer Conclusion is supposed to be dependent upon the Fourth Gospel, and, therefore, to have been written in the Church where that Gospel was earliest in circulation. But, if the author of the Longer Conclusion knew John, why did he ignore the Appearance to Thomas? Presumably he compiled his list of Appearances with an apologetic purpose; why then leave out the most "evidential" of them all? There is only one point of contact between John and the Longer Conclusion, and that is the mention of Mary Magdalene as the witness of the first Appearance. But Matthew also records the first Appearance as being to Mary Magdalene and Mary the mother of James and Jesus. This second Mary is otherwise an absolutely unknown figure.

It is a law of the evolution of tradition that names to which no incident of dramatic interest is attached tend, either to gather incidents round themselves, or else to drop out. Consider this series. Mark mentions three women at the tomb, Matthew mentions two, the Shorter Conclusion only one. That one was not only their leader, but also the only one about whom a fact of interest was known, "that seven devils were cast out of her." Besides this tendency in tradition, the author of the Longer Conclusion was evidently influenced by the desire to be brief. If he summarises in two verses the Emmaus story, which takes twenty-three verses in Luke, he obviously does not want to fill up space with mere names. Obviously a knowledge of John is not required to explain the dropping of the other names. The Appearance to Mary may be derived either from Matthew or else from oral tradition, for it is the kind of thing in which oral tradition would be interested. Again, if we look more closely at the parallels we note that in the Longer Conclusion the disciples, when they heard from Mary that "he was alive and had been seen by her, disbelieved." There is not the slightest hint of this in John; Thomas, who does doubt, is not then present; but it is emphatically asserted in Luke in regard to the reception by the Apostles of the news of the three women that they had seen the angels who said He was risen. Thus, while the mention of Mary, the one supposed point of contact with John, can equally well be interpreted as a point of contact with Matthew, there is a notable absence of allusion to the most striking and, for the author's purpose, most useful story in John. But, since John was the characteristically Ephesian Gospel, we should expect to find quite a number of rather marked points of contact with it in a document of this sort emanating from Asia.

If, however, the case for Asia collapses, we may be content to accept the alternative view to which the textual evidence very decidedly points, and affirm that the Longer Conclusion was added in Rome.

With this in mind let us examine the document a little more closely. Everything in it, except the mention of the Appearance to Mary and the drinking poison, appears to be derived either from the Gospel of Luke or from the Acts. There are summary allusions to the disbelief by the Apostles in the women's message (Mk.xvi.11, cf. Lk.xxiv.11); to the walk from Emmaus; to the Appearance to the Eleven, with their hesitation to believe the reality of the thing they saw and the command to preach to all nations; to the Apostolic signs—casting out devils, speaking with tongues, the viper which clung to St. Paul in Malta but did not hurt him. Even the Appearance to Mary Magdalene, though not derived from Luke, identifies her by a formula taken from Lk.viii.2, "from whom he had cast out seven devils."

The natural inference that we should draw is that the Longer Conclusion was written in a Church where Luke and Acts had been long established, but where Matthew, if known at all, had only recently been accepted; and Harnack produces some reasons for the belief that Matthew was not accepted at Rome till about AD 120. [The Date of the Acts and the Synoptists, E.T., p. 134 n. (Williams and Norgate, 1911).] The Longer Conclusion opens in a way, which suggests that it was not originally intended, like the Shorter Conclusion, to heal a wound in the text of Mark. It reads as if it was originally a summary intended for catechetical purposes; later on the bright idea occurred to some one of adding it as a sort of appendix to his copy of Mark. In the first instance an interval of a blank line might be left to mark that it was not part of the authentic text, but in subsequent copies this blank line would soon disappear. The hypothesis that Mk.xvi.9-20 was originally a separate document has the additional advantage of making it somewhat easier to account for the supplement in the text of W (cf. p. 337 f.) known as the "Freer logion." A catechetical summary is a document which lends itself to expansion; the fact that a copy of it had been added to Mark would not at once put out of existence all other copies or prevent them suffering expansion. No doubt as soon as the addition became thoroughly established in the Roman text of Mark, it would cease to be copied as a separate document. But supposing that a hundred years later an old copy of it in the expanded version turned up. It would then be mistaken for a fragment of a very ancient MS. of Mark, and the fortunate discoverer would hasten to add to his copy of Mark—which, of course, he would suppose to be defective—the addition preserved in this ancient witness.

the lost ending—A speculation

If I venture a suggestion on this subject, it is with the dis­tinct proviso that what I write is intended to be read, not as "criticism," but merely as "scientific guessing." No harm is done by guessing of this sort, it may even have a certain interest, provided always that no one mistakes the speculations so reached for " assured results of criticism," and then proceeds to use them as premises from which further deductions may be drawn. The "scientific guess" in which I venture to indulge is that the lost end of Mark contained an Appearance to Mary Magdalene, followed by one to Peter and others when fishing on the Lake of Galilee, and that John derived his version of these incidents from the lost conclusion of Mark. I do not, of course, suggest that we have either of these stories exactly as they stood in Mark. Wherever John adopts a story from Mark, he does so with a considerably greater freedom in regard to language and details than do Matthew or Luke. Nevertheless, except where he is conflating material from Mark with another source, John does not seem substantially to alter the main facts and the general impression. This hypothesis I believe is worth working out in detail. But before doing this, two prima facie objections must be met.

(1) At first sight it may seem unlikely that the original ending of Mark should be preserved in Ephesus but lost in Rome. But we know that during Paul's imprisonment at Rome, Mark was contemplating a visit to Asia (Col.iv.10), and a little later (2 Tim.iv.11) Paul summons him to return to Rome, "for he is useful to me for ministering." The words occur in that portion of 2 Timothy that is most universally recognised as a fragment of a genuine letter, probably preserved at Ephesus. If so, Mark had been working in or near Ephesus; and, as the context implies, was regarded by Paul as a useful and effective worker very shortly before the date at which he wrote the Gospel. What, then, could be a more natural thing for Mark to do, the moment he has finished writing a Gospel for the Church of Rome, than to send a copy to the Church of Ephesus by the next Christian who was travelling that way on any kind of business? In that case the Ephesian copy would be the first ever made, and would have been made before the original was mutilated.

We may surmise that Mark's "usefulness for ministering" lay in some part in his command of anecdotes about the life and teaching of Christ. But he would have told these stories in Ephesus also. Hence the fact that Mark had worked in Asia makes it possible to suggest an alternative hypothesis that John is dependent, not on the lost written end of Mark, but upon Mark's account of the Resurrection Appearances, which survived in Asia in the form of oral tradition. This hypothesis, however, is for practical purposes so nearly equivalent to the one I have propounded above that it may, so far as arguments for or against it are concerned, be treated as a minor variation of the same hypothesis.

(2) If the original conclusion survived at Ephesus, how came that which now stands in our New Testament to take its place? The answer to this question may well be that the present ending was added so soon that it had time to become part of the text accepted at Rome before the date—some time in the latter half of the second century—when (as seems likely) the Churches of Rome and Ephesus exchanged notes on the Canon. At any rate we have already seen that it must have been added at a very early date. If the Longer Conclusion was composed as a separate document about 100-110, and had become firmly established in the text of Mark as read in Rome, say by AD 140, the original ending, even if preserved at Ephesus, would never be restored. Mark being the Roman Gospel, the Roman text of Mark would everywhere be regarded as the more authentic—except perhaps in Alexandria, which also claimed a special connection with its author. But the oldest Alexandrian text lacked an ending. Supposing the early Ephesian and the Roman text showed two different endings, this, if known to a scholar like Origen, would only confirm him, brought up as he was in the traditions of textual criticism current in Alexandria, in the belief that the mutilated text was the original. Moreover, as we have already seen (p. 69), the Textus Receptus was the text adopted in the great sees of Antioch and Constantinople, so that the old text of Ephesus was swamped at an early date and has left no trace on the MS. tradition. This would the more easily happen since, on our hypothesis, the same story, but told in a form more attractive to the Christian public, was contained in John. There would be no strong motive to keep alive what would seem a less interesting version of the same story.

The suggestion that the story of the Appearance and final charge to Peter on the Lake of Galilee in Jn.xxi. was derived, with some modification, from the lost ending of Mark has been commended by Harnack and others, on the ground mainly that the Apocryphal Gospel of Peter contained a version of the incident. The surviving fragment of the Apocryphal Gospel of Peter ends:

But we the twelve disciples of the Lord wept and were grieved: and each one grieving for that which was come to pass departed to his home. But I, Simon Peter, and Andrew my brother took our nets and went away to the sea; and there was with us Levi, the son of Alphaeus whom the Lord. ...

Here, unfortunately, the text of the fragment breaks off;
but evidently the words constitute the beginning of an account of an Appearance of the Lord by the Sea of Galilee. From the fragment of it that remains it is evident that the Gospel of Peter can in no way be regarded as an independent historical authority. It is written in the interest of the theory already combated in the Asian document 1 John [1 Jn.v.6,

not with water only,
but with the water and the blood

(i.e. a real death)—a reply to the position of Cerinthus, or some predecessor.] —that the Divine Christ departed from the human Jesus and was taken up into Heaven before the latter died on the Cross. Professor C. H. Turner [J.T.S., Jan. 1913.] in a brilliant, and on the whole convincing, article argues that the author of the Gospel of Peter was familiar with our Gospel of John. It does not, however, necessarily follow that he derived this particular incident from Jn.xxi. If so, why does he put first an Appearance that John distinctly affirms to be the third? Also, why are the names of disciples mentioned as present so different—Andrew and Levi as against Thomas, Nathaniel, the sons of Zebedee, and two others? The mixture of resemblance and difference is accounted for more easily if John and Peter are divergent versions of a third source than if either is dependent on the other. But it is clear that the main source used by the author of Apocryphal Peter was Mark. And since "Peter" claims to be written by Mark's master, we should expect it to concur with Mark, except where the author desired to supplement the traditional narrative with doctrinal modifications or legendary embellishments of his own. It is noticeable that Levi the son of Alphaeus is mentioned in "Peter"; the description of him, "son of Alphaeus," only occurs in Mark.  Also the account of the visit of the women to the tomb follows Mark rather closely. In particular the author preserves the detail " Then the women feared and fled," which corresponds to the last words of the true text of Mark, but are directly contradictory to the statements of both Matthew and Luke. Hence the hypothesis that in the paragraph that immediately follows, he is also dependent on Mark cannot be called improbable.

In support of the view that John xxi. represents either the lost end of Mark or an oral tradition more or less its equivalent, five considerations may be alleged.

(1) The lost ending of Mark must have contained an Appear­ance to Apostles in Galilee, which either followed an Appearance to Peter, or was itself one in which Peter figured in some con­spicuous way.

(2) If the story in Jn.xxi. had stood alone in a separate document, without the note (xxi.14) stating that this was the third Appearance, we should have inferred that the Appearance described was meant to be understood as the first. We seem to see (xxi.2-3) a group of disciples sitting dejected and inert after their disillusioned flight to Galilee, and Peter, always the one with the most initiative, rousing himself to the resolution to go back to the old and ordinary life, "I go a-fishing." The others follow. Jesus is seen on the bank. They do not know Him. They seem to be taken by surprise, which is strange if previous Appearances had already convinced them He was alive. An incident and conversation follow of which the general significance is a second call of Peter to be a fisher of men. His late denial of his Master is wiped out by a reaffirmation of devotion, and he is given the commission "Feed my sheep" and so made the leader in the Christian mission.

(3) The addition of a miraculous draught of fishes in the story of the original call of Peter in Luke v.4-7, and the addition, in Matthew xiv.29-31, to the story of the Walking on the Water of the incident of Peter leaving the boat to meet the Lord, are best explained as fragments of a story like that of Jn.xxi. current in oral tradition. If so, they are independent evidence that the story was in circulation at a very early date.

(4) Like everything else in the Fourth Gospel this story has been remoulded in the light of the experience and outlook of the author and the present needs of the Church, but it is certainly the kind of ending one would have expected Mark to give his Gospel. In particular, I much doubt whether Peter's denial would have been so emphasised in the Gospel unless as a foil to a subsequent story, the point of which lay in its cancelling a former weakness of the Apostle. Again, the fact that Peter, with whatever hesitation, did ultimately come out definitely on the side of the Gentile mission, and that in doing so he felt that he was carrying out his Master's real intention, must, I think, have somehow been adumbrated in a Gospel written for the Church of Rome.

(5) A critical analysis (cf. Chap. XIV.) of the Fourth Gospel suggests that the two main sources that John elsewhere combines are the written documents Mark and Luke (or Proto-Luke). It is, therefore, likely that in this passage also the traditions which he is combining are derived from the same two sources. Since, then, the Appearance to the Apostles in Jerusalem belongs to the Lucan tradition, those to Mary Magdalene and to Peter while fishing by the Sea of Galilee, with the final commission of Christ "Feed my sheep," may well have stood in the conclusion of Mark's Gospel as it was read in Ephesus about AD 90. Indeed we may surmise that one reason why the last chapter of John (which is obviously a kind of Appendix) was added was to harmonise that Gospel with the Marcan tradition of the Resurrection Appearances, while affirming that the Appearance by the Lake was the third, not, as previously related in Mark, the first, of the Appearances to any of the Twelve.

The suggestion that the Appearance by the Sea of Galilee was preceded by an Appearance to Mary Magdalene, something like that recorded by John, has not, so far as I am aware, been put forward before.  In its behalf I advance the following considerations.

(1) We shall see (p. 408 ff.) that, apart from the Appearance to Mary, John shows no trace of dependence on Matthew.  The hypothesis that the Appearance to Mary originally stood in Mark enables us to explain the occurrence in both Mt.xxviii.10 and Jn.xx.17 of an Appearance to Mary with the description of the disciples as "my brethren" which is not paralleled elsewhere.

(2) Again, the Appearance to Mary as described by John is entirely in the manner of the vivid and dramatic storytelling for which Mark is famed. Mark is one of those people who simply cannot tell a story badly—witness the tale of the daughter of Herodias and John the Baptist, the appeal of which to the artistic imagination every picture gallery in Europe proves. If ever he finished his Gospel, the Resurrection scenes would have been visualised in every detail. And there is no scene in the Fourth Gospel—again I call the painters in as evidence —more vividly pictured than that of Mary Magdalene in the Garden.

(3) John, as already observed, seems to follow alternately, or to conflate, two main sources, Mark and Luke (or a source of Luke). Since the Appearance to Mary is not found in Luke, it was probably absent from his non-Marcan source—and Luke's copy of Mark, we have seen, ended at xvi. 8. John, then, could not have derived the story either from Luke's source or from our Third Gospel. Whence, then, did John derive it? Of course he might have got it from Matthew; but, apart from this incident, John shows no definite knowledge of Matthew, still less any inclination to follow him. Much the simplest hypothesis is that John derived the Appearance to Mary from Mark (or an oral tradition representing what Mark would have contained), especially as an incident which turned the " fear " of the woman into joy would have formed a most appropriate con­tinuation of what remains of his broken text.

(4) Consider the situation at Rome if, after some police raid or riot, the end of the Church copy of Mark was found to have been torn off. Its general purport would have been known; many would remember roughly what it had contained, and the loss might have been replaced from memory. But this would have been inaccurate, and hopes may have been entertained that another copy might turn up. In the meantime there would remain a tradition, growing more vague in course of time, that the lost ending had contained an Appearance to Mary in Jerusalem, followed by an Appearance to the Apostles in Galilee. Now this is what we find in Matthew. The end of Matthew is exactly the kind of conclusion we should expect if the first man who took a copy of the mutilated Gospel to Antioch had written down on the back of the last sheet his recollections of the substance of what he had been told at Rome the lost conclusion had once contained.

(5) The view that oral tradition at Rome, ultimately dependent on the lost end of Mark, represented the first Appearance as being to Mary, would (equally with dependence on Matthew) account for the opening of the Longer Conclusion of Mark,

He appeared first to Mary Magdalene,
from whom he had cast out seven devils.

The seven devils are derived from Luke (viii.2); and as we have already noted, apart from the Appearance to Mary, all other details in the Longer Conclusion which occur in the New Testament at all are to be found in Luke and Acts. For though there is a point of contact with Matthew in the command to preach and baptize—Luke also, it should be noted, has the command to preach to all nations— there is in the actual language used nothing in common in the parallel Mk.xvi.15-16 = Mt.xxviii.19-20 but the single and inevitable word "baptism."

(6) From Paul's account of the Resurrection Appearance (1 Cor.xv.5) one would naturally infer that the first Appearance was to Peter. Luke's narrative confirms this impression. How, then, are we to explain the emphatic statement in the Longer Conclusion that the first Appearance was to Mary? I suggest that there was ancient tradition at Rome to this effect so firmly established that it could hold its own against the prima facie evidence of Paul. In that case the Longer Conclusion of Mark is best understood as the attempt to harmonise the old Roman tradition of a first Appearance to Mary Magdalene with the newly authenticated information that the Lucan writings had brought to the Church. Its addition to the text of Mark would not only help to preserve this tradition, but would be almost necessary, if the old Roman Gospel of Mark was to maintain its existence side by side with the longer and more interesting, but more recent, Gospel of Luke.

At any rate the preference in three of our Gospels, as we have them, of a tradition apparently contradicting a written statement of Paul does require an explanation. We have definite evidence that 1 Corinthians was the epistle which was most widely read in Christendom in the Sub-Apostolic Age. The critic is bound to produce a hypothesis to explain why, in despite of this evidence that the first Appearance was to Peter, a tradition prevailed in three different Gospels, representing presumably three different Churches, which assigns the supreme privilege of being the first to see the risen Lord, not to the Prince of the Apostles, but to a woman, of whom nothing is known save that seven devils were cast out of her. A tradition established so early in different Churches (most probably in Antioch, Ephesus, and Rome) must have gone back to great antiquity and have been regarded as authenticated by irrefutable authority. But if it originally stood in Mark, which in a point like this may be supposed to rest on Peter's own reminiscences, then there was the authority of Peter himself that he had in this matter been forestalled by a woman.

But why, it may be objected, if the Appearance to Mary originally stood in Mark, is it omitted in the Apocryphal Gospel of Peter, which we have assumed is here dependent on Mark? Apocryphal Peter, in order to vindicate its doctrinal curiosities, is particularly concerned to emphasise the fiction of Apostolic authorship. This is shown by the intrusion of the words "I, Simon Peter," which would be wholly unnecessary in a Gospel known to be written by an Apostle, and is, at any rate in this particular context, a most inappropriate repetition of a claim to authorship, which must have been stated before. But clearly an author who feels it so necessary to emphasise the ego of Simon at this point cannot afford to let the first Appearance of Christ be to anybody else. There is a further reason. Apocryphal Peter is a second-century work. Celsus, the great second-century opponent of Christianity, pours much scorn on the belief in the Resurrection on the ground that it originated in the fancy of a neurotic woman. There was an apologetic reason for the omission.

Such cogency as the foregoing arguments possess is largely dependent on the correctness of the analysis of the sources of John essayed in a later chapter. And, even if the correctness of that analysis be assumed, they fall far short of proof. Yet the view that the earliest account of the Resurrection Appearances has disappeared without leaving a trace is in itself so improbable that I have thought it worthwhile to outline a hypothesis, which makes it possible to affirm the contrary, even though from the nature of the evidence it can be no more than an interesting speculation.