The Epistle to the Romans

TN the last number of this Journal (in. p. 51 sq.) Mr Hort criticised and condemned a theory which I had suggested in the preceding number (n. p. 264 sq.) to account for certain facts connected with the text of the Epistle to the Romans. The facts, it will be remembered, were mainly these; (1) One or more ancient writers used a copy of the Epistle containing only the first fourteen chapters, with or without the doxology which in the common text stands at the close of the whole (xvi. 25-27). (2) In the existing copies this doxology appears sometimes at the end of the xivth chapter, sometimes at the end of the xvith, sometimes in both places, while in some few instances it is omitted altogether. (3) At least one text omits eV 'Pwfir] in i. 7, 15. The theory, by which I sought to combine and explain these facts, was this; that St Paul at a later period of his life reissued the Epistle in a shorter form with a view to general circulation, omitting the last two chapters, obliterating the mention of Rome in the first chapter, and adding the doxology, which was no part of the original Epistle. Mr Hort impugns some of these assumed facts and explains away others. Having done this, he attacks the theory itself, and endeavours to show that it is untenable.

No one, who is really anxious to ascertain the truth, would object to such a criticism as Mr Hort's, even though it should lead to the rejection of a darling theory. I am especially obliged to him for the thoroughness with which he has applied the test of textual criticism to my hypothesis. And, if I venture, notwithstanding his arguments, to maintain that the facts themselves are stubborn and in some respects even stronger than I had supposed, and to uphold my theory as the most probable explanation of the facts, until a better is suggested, I trust that I am not blinded by partiality. At all events I will give my reasons as briefly as possible, taking the facts first and then proceeding to the theory.

I. The first and most important of the facts is the existence, in early times, of copies containing only fourteen chapters. Of this the indications are various, and (as it seems to me) conclusive.

(i) The statement of Origen respecting Marcion has been 'universally understood/ as Mr Hort himself allows (p. 330), to mean that this heretic struck out not only the paragraph containing the doxology, but the two last chapters also; 'Caput hoc [i.e. the paragraph containing the doxology] Marcion, a quo Scripturae evangelicae atque apostolicae interpolatae sunt, de hac epistola penitus abstulit; et non solum hoc, sed et ab eo loco ubi scriptum est Omne autem quod non ex fide peccatum est (xiv. 23) ad finem cuncta dissecuit. In aliis vero exemplaribus, id est, in his quae non sunt a Marcione temerata, hoc ipsum caput diverse positum invenimus.' An universal understanding may be wrong, but most frequently it is correct; and I cannot doubt that this is the case here. Mr Hort however adopts a reading of a Paris MS. (Reg. 1639) which has 'in eo loco' for ' ab eo loco, ' and himself alters ' hoc' into 'hie.' Thus he makes Origen say that Marcion cut out the doxology, not only at the end of the xivth chapter, but also at the end of the Epistle. Now my reply to this is threefold; (1) Though we may allow the general value of the readings in this MS., whose date however is not earlier than about the 12th century, yet its text is far from faultless, so that only a slight presumption is raised in favour of a reading from the fact of its being found there. In the present instance however the reading 'in eo loco' has no meaning, unless with Mr Hort we likewise change hoc into hie—an alteration for which there is no MS. authority. (2) Mr Hort's reading and interpretation destroy the force of individual expressions in the context. L. E. 23

'Usque ad finem cuncta dissecuit' is natural enough when applied to two whole chapters, but not to the doxology alone; and again in 'hoc ipsum caput '- the ipsum becomes meaningless, unless it is contrasted with some other portion. If the words be taken as they stand and interpreted in the ordinary way, the sequence commends itself; 'Caput hoc...non solum hoc sed...usque ad finem cuncta...hoc ipsum caput'; but it is entirely broken up if they are read and explained as Mr Hort wishes. (3) One who reads continuously not only the passage quoted above, but the whole paragraph of Origen as given by Mr Hort (see p. 330) or by myself (p. 288), will hardly fail, I think, to see how Mr Hort's interpretation involves and confuses the natural order of the topics.

When again Mr Hort supposes the statement of Jerome (on Ephes. iii. 5), that the doxology was found in plerisque codicibus, to have been derived from Origen's commentary on the same Epistle, I allow that this supposition is probable. But I do not see that Mr Hort's view gains strength thereby. Commenting on Ephes. iii. 5, Origen would be concerned only with the doxology in which ' the mystery' is mentioned, and he would be going out of his way, if he said anything about the omission of the xvth and xvith chapters, with which he was not in any way concerned. Moreover it must be observed that, when there is a question of a various reading, Jerome sometimes manipulates Origen's statements in such a manner as entirely to disfigure their meaning. Such is the case for instance with the opening verse of this very Epistle to the Ephesians, where Origen, having before him a text which omitted iv 'E<£eo-a>, interprets rot? otiaiv in an entirely lucid though highly artificial way, but Jerome, repeating his great predecessor's comment, holds language which can hardly be called intelligible.

As regards the statement of Tertullian, when arguing against Marcion (v. 14), that the threat of the tribunal Christi (Rom. xiv. 10) occurs in clausula of the Epistle, I agree with Mr Hort that the inference which supposes Tertullian to refer to a copy of the Epistle wanting the xvth and xvith chapters, though 'natural,' is not 'conclusive.' Let the fact that the inference is natural have no more than its proper weight. I should not have laid much stress on the expression, if it had stood alone; but in connexion with Origen's account of Marcion it cannot be overlooked.

(ii) For the negative argument that the last two chapters are nowhere quoted by certain early writers I claim a supplemental value. More than this it does not deserve. The fact however remains that neither Irenaeus nor Tertullian nor Cyprian (except in a very doubtful allusion) refers to them. I will only add that this omission occurs in Western writers1, whereas they are more than once quoted by Clement and Origen. The importance of this fact will appear hereafter.

(iii) I owe it to Mr Hort's candour that my attention was directed to the capitulations of the Latin Bibles, and the evidence derived thence seems to me to strengthen my case enormously. In my former article I had referred to Wetstein's note: 'Codex Latinus habet capitula Epistolae ad Romanos 51, desinit autem in caput xiv.; ex quo conficitur ista capitula ad editionem Marcionis fuisse accommodata'; and, misled with others by his careless expression desinit (where desinunt would have been clearer), I had naturally supposed that the MS. itself, to which he refers, ended with the xivth chapter, and accordingly remarked that' later critics had not been able to identify the MS. and thus verify the statement.' I have no doubt however that Mr Hort is right, and that Wetstein refers to such a phenomenon as the Codex Amiatinus exhibits, where (though the Epistle itself is complete) the capitulations end with the end of the xivth chapter, there or thereabouts.

1 The first distinct quotation by any may be trusted) cites nothing from

Western writer, so far as I can dis- these two chapters but the doxology.

cover, occurs in Victorinus c. Arium The 'very doubtful reference' in Cy

iii. p. 280 c, a treatise written about prian is given by Mr Hort, p. 336,

A.d. 365—where xvi. 20 is quoted. note 2. Even Hilary of Poitiers (if the index

I have since been investigating the subject1; and the results of this investigation seem to be sufficiently important to justify my taking up a few pages in recording them.

In fact, there is evidence of two distinct capitulations—both ending with the xivth chapter—the first very widely spread, the second only preserved in a single though very early MS.

Of the first of these, the Codex Amiatinus affords the oldest and best example. In this MS. the table of contents prefixed to the Epistle gives 51 sections, the 50th section being described 'De periculo contristante fratrem suum esca sua, et quod non sit regnum Dei esca et potus sed justitia et pax et gaudium in Spiritu Sancto,' and the 51st and last 'De mysterio domini ante passionem in silentio habito, post passionem vero ipsius revelato.' Corresponding to these, the sections are marked in the text, and agree with the descriptions in the table of contents as far as the 50th. The 50th is marked as beginning at xiv. 15, and here again the description is accurate; but the 51st commences with xv. 4, and has no connexion with the description. The description of the 51st in fact corresponds to the doxology (xvi. 25-27), and to nothing else in the remainder of the Epistle. The natural inference therefore is, that the capitulation was made for a copy of the Epistle, containing only fourteen chapters and the doxology; and that the scribe who first adapted it to a full copy with the sixteen chapters, not finding anything corresponding to the 51st section in the immediate context, extended the 50th section as far as the subject allowed him and made the 51st section include all the remainder of the Epistle. This solution, which Mr Hort allows to be certainly possible, seems to me to commend itself as in the highest degree probable.

This capitulation appears to have prevailed very widely. It is found in not less than seven MSS. enumerated by Card.

1 After I saw Mr Hort's article in non of the capitulations in the Codex

type, I began to look into the matter; Fuldensis. To this conversation he

and, before it was finally struck off, I refers in a note appended to his article

mentioned the remarkable phenome- (p. 351).

Tommasi (Thomasii Op. I. p. 388 sq. ed. Vezzosi), and dating from the age of Charles the Great downwards. It occurs again in the British Museum MS. Add. 10,546, an Alcuinian copy, generally called 'Charlemagne's Bible,' but really written in one of the succeeding reigns; in the important MS. Harl. 1772 belonging to the 8th century; in the Oxford Bodleian MS. Laud. Lot. 108 (E. 67) of the 9th century (in which however the number is expanded from 51 to 67 by a subdivision of one or more of the earlier sections); in the MS. B. 5. 2 of Trin. Coll., Cambridge, belonging to the 11th or 12th century1; and in the Cambridge University MS. Ee. 1. 9 written apparently late in the 13th century*. In Add. 10,546 the sections correspond in number and position with those of the Amiatinus, but the words are occasionally varied, e.g. de non contristando fratre for de periculo contristante fratrem suum. In Harl. 1772 the number of sections in the table of contents is reduced to 49 by combining §§ 43, 44, 45 in one section, while (except unimportant various readings) the words of the Amiatinus are strictly followed. In the text however the whole 51 sections are marked; of these the first 49 correspond to those of the Amiatinus, but the 50th commences not with the beginning of xiv. 15 Si enim propter, but with the middle Noli cibo (while on the margin in a later hand stands xlviiij. opposite Si enim propter), and the 51st not with xv. 4 Quaecumque enim, but with the middle of xiv. 22 Beatus qui (the Q of Quaecumque being however illuminated). And again in Cambr. Univ. Ee. 1. 9, where the number of sections is similarly reduced to 50, the beginning of the 50th and last section 'de mysterio etc.' stands at xv. 1 Debemus autem nos, i.e. at the precise point where it would have stood, if the MS. had contained only the doxology after the xivth chapter.

1 In the older Trin. Coll. MS. of 3 In the Cambr. Univ. MS. Ff. 4.40,

St Ptiul's Epistles B. 10. 5, of the 9th which came from the Library of Christ

century, the Epistle to the Romans Church, Canterbury, and was written

and part of the First to the Corinthians probably early in the 13th century,

are wanting. The Amiatinian capitu- though the Amiatinian capitulations

lations are given for the other Epi- are not given, I find this note 'Haec

sties. epistola capitula li. dicitur habuisse.'

These variations show the difficulty which was felt in adapting the end of the imperfect capitulation to the complete Epistle: and they answer any objection founded on the fact that in the Amiatinus itself the "last section does not commence at the exact place in the text which the hypothesis seems to require.

In more than one MS. however, which I have examined, this capitulation is completed. The British Museum MS. Add. 28,107 formerly belonged to the monastery of S. Remacle at Stavelot, and was written in the year 1097, 'ipso eodem anno quo versus hierusalem facta fuerat gentium plurimarum profectio,' as is stated at the end. The capitulation to the Epistle to the Romans gives 63 sections. Of these § 1-41 correspond with those of the Amiatinus; §§ 42, 43, 44, 45, are formed out of § 42 of the latter subdivided; and §§ 46-53 correspond to §§ 43-50 of the latter. Thus the heading of § 53 is 'Periculum contristantis fratrem suum esca sua etc.' There is nothing corresponding to § 51 of Amiatinus, which comprises the doxology, but § 54 (xiv. 19) is 'Quae pacis sunt sectanda et fratres propter escam minime judicandi,' and § 55 (xv. 4) 'De doctrina et consolatione scripturarum et quod unanimiter sit honorificandus deus et pater domini nostri jesu christi'; while the last section of all (§ 63), beginning at xvi. 21, runs ' Salutatio timothei et caeterorum etiam et ipsius pauli qui epistolam in domino se scripsisse dicit.' The compiler was vigilant enough to see that the section ' de mysterio etc.' of the capitulation before him did not correspond to anything which followed, and therefore ejected it, and supplied (though not very intelligently) the remaining sections which were required to complete the Epistle.

Another complete capitulation, founded on the Amiatinian, occurs in the British Museum MS., Reg. 1. E. viii., which belonged to Christ Church, Canterbury, and may have been written about the middle of the tenth century. This capitulation, which is very brief and very slovenly, comprises 29 sections. The last of these are as follows:

xxiiii. de redditione unicuique omnium debitore (tic).

xxv. de periculo contristante fratrem esca sua.

xxvi. de mysterio domini ante passionem in silentio habitat (sic}.

xxvii. post passionem domini ipsius mysterio revelatus.

xxviii. obsecratio pauli ad dominum ut liberetur ab infidelibus.

xxix. salutatio pauli ad fratres.

The retention and subdivision of the section comprising the doxology, where it has no meaning, is a curious phenomenon.

A third instance of completed capitulation is found in the MS. B. 5. 1 of Trin. Coll., Cambridge, belonging to the 12th century. Here the scribe has retained all the Amiatinian sections, including the doxology; but by combining two in the earlier part, he reduces them to 50 in number. Thus the 49th is 'de non contristando fratrem, etc.', and the 50th 'de mysterio domini, etc.' To these he adds two new sections, which are the same as those described in the last MS.:

li. obsecratio pauli ad dominum, etc.
lii. salutatio pauli ad fratres.

In the text the 49th section begins at xiv. 50, the 50th at xv. 4, the 51st at xv. 30, and the 52nd at xvi. 1. The inequality of scale in these superadded sections shows that they did not proceed from the same hand as the rest1.

These facts have been elicited by an examination of such MSS. as came conveniently within my reach*. Doubtless a wider investigation would produce more striking results.

1 The relation between the two MSS. which perhaps appears first in the Al

last described is curious. For, while cuinian copies.

other indications would suggest that - My examination has not extended

the capitulations of Brit. Mas. Reg. 1. beyond the British Museum MSS. to

E. viii. were derived from those of the 11th century (inclusive), and the

Trin. B. 9. 1, the former presents the MSS. in the Cambridge University and

older form of the Amiatinian 50th sec- Trinity College Libraries. The infor

tion' de periculo contristante fratrem,' mation respecting Bodl. Laud. Lot.

while the latter substitutes the amend- 108 I owe to Mr Coxe, the Librarian, ed form 'de non contristando fratrem,'

But I have seen enough to convince me that the Amiatinian capitulation, though originally framed, as will be seen hereafter, for a short copy of the Old Latin, yet maintained its ground as a common mode of dividing the Epistle, until it was at length superseded by the present division into 16 chapters in the latter half of the 13th century.

The second capitulation, of which I spoke, is found in the Codex Fuldensis which, like the Amiatinus, was written about the middle of the 6th century. The sections in the text correspond exactly with the Amiatinian. Not so in the table of contents. Of the latter Ranke remarks (Codex Fuldensis, p. xxiii, 1868): 'Quae epistolae ad Romanos praemissa sunt capitula duabus in partibus constant, quarum altera (i.-xxiii.), totius /ere epistolae argumentum in se continens, per se ipsa stare videtur, altera (xxiii.-li.) iis respondet quae iisdem sub numeris in cod. Amiatino proferuntur.' The words which I have italicised are not very exact. These 23 sections, which belong to a different capitulation from the remainder, reach to about the end of the fourteenth chapter, the last (§ xxiii.) being 'Quod fideles dei non debeant invicem judicare cum unusquisq. secundum regulas mandatorum ipse se debeat divino judicio praeparare ut ante tribunal dei sine confusione possit operum suorum praestare rationem.' The 24th Amiatinian section, which follows next, begins with ix. 1, so that six chapters (ix.-xiv.) are included twice. The natural inference is that the scribe, remembering that the text contained 51 sections and seeing that the table of contents gave less than half that number, applied himself to another source, and completed the headings of the remaining sections from the Amiatinian capitulation. Whether the capitulation from which §§ i.-xxiii. are taken contained the doxology or not, must remain doubtful. The analogy of the Amiatinian sections would suggest that it did. The 23 summaries peculiar to the Fuldensis are very broad and general; thus § xxii.'de mundanis potestatibus honorandis quia oportet oboediri his quib. ad mundanum regumen dominus tribuit potestate,' though including the whole of our 13th chapter, omits to take account of the last half, vv. 8-14; and in like manner in § xxiii. the doxology may not have been thought worthy of any special attention in this heading1.

Mr Hort indeed impugns the value of this Fuldensian capitulation on the ground that the headings 'are loaded with Augustinian or Anti-Pelagian phraseology, and cannot therefore be dated much before 400 at earliest' (p. 351, note). I have no wish to deny that there is force in this argument; which nevertheless does not seem to me conclusive. The strongest expressions in this direction are 'pro fide romanorum...deo apostolus gratias agit ut probetur fidem in deum muneris est divini,' and 'in Christo Jesu qui solus sic humana [humanam] naturam recepit ut eum contagia veteris originis non tenerent.' The African fathers were more or less Augustinian before Augustine's time, and (so far as I can see) might have held such language*.

On any showing however the Latin Bibles bear strong testimony to the existence of the shorter form of this Epistle at an early date. The alternative hypothesis, that these sections were determined by the lessons read in Churches, is devoid alike of evidence and of probability. With this single exception, the Amiatinian capitulation in the New Testament includes, I believe, the entire book in every case. It does not bear the slightest trace of being intended for lectionary purposes. Nor indeed is there any reason why the 15th chapter should be excluded from the lessons; for it is much more fit for public reading than many sections elsewhere, which are retained.

1 Besides the capitulations mention- In this last MS., though the table of

ed in the text, I have noticed one contents gives 18 chapters, the Epistle

other which is unconnected with either. itself is divided by marginal numbers

It contains 18 sections and includes into smaller sections, 125 in number,

the whole epistle. This capitulation - e.g. Cyprian Ep. 64, says ' Secun

is found in: dum Adam carnaliter natus, contagium

(1) Brit. Mus. Add. 11,892, a MS. mortis antiquae prima nativitate oonwhich belonged to the monastery of traxit.' Compare also TertulL de Anim. St Gall, and was written in the 9th 40, 41; and see Neander Hut. of century. Christian Dogmas, I. p. 185 sq. (Eng.

(2) Brit. Mus. Add. 24,142 'Monas- Trans.). Augustine's own dogmatic tern S. Huberti in Ardvenna,' sup- views on these points were enunciated posed to have been written about A.D. before Pelagius took up the subject: 900. ib. p. 347 sq.

Even the 16th chapter would be treated with exceptional rigour on this showing, for in other epistles the paragraphs containing the salutations are religiously recorded in the capitulation. Moreover, the oldest evidence which we possess on the subject exhibits lessons for Sundays and Festivals taken from the 15th chapter; and if so, a fortiori it would not be neglected in the daily lessons, supposing (which seems improbable) that daily lessons had been instituted at the time when this capitulation was made.

When my attention was first directed to the Amiatinian capitulation, I naturally inferred that it had belonged originally to the Old Latin and was later adapted to the Vulgate. A further examination has shown this inference to be correct. The capitulation preserves at least one crucial reading of the Old Latin. In § xlii . the words 'de tempore serviendo' show that its author for Tw Kvpitp BovXevomei read Tq> Kaipm BovKevovre< ; in xii. 11, a reading which Jerome especially quotes as condemning the Old Latin and justifying his own revision (Epist. 28, Op. L 133, ed. Vallarsi).

Thus, taking into account all the evidence, the statement of Origen respecting Marcion (confirmed by the incidental expression of Tertullian), the absence of quotations in several early fathers, and the capitulation (or capitulations) of the Latin Bibles, we have testimony various, cumulative, and (as it seems to me) irresistible, to the existence of shorter copies of the Epistle containing only fourteen chapters with or without the doxology in early times. Even though it be granted that Mr Hort has given a possible explanation (I cannot allow that his explanations are probable) of each of these facts singly on a different hypothesis, still the convergence of so many independent testimonies direct or indirect towards this one point must be regarded, if I mistake not, as conclusive.

II. However the evidence does not end here. The fact that in existing MSS. the doxology occurs in different places (see p. 352) is very intimately connected with the fact or class of facts considered under the first head. And here again I cannot help remarking that my position has this great advantage over Mr Hort's, that whereas I postulate only one unknown fact to explain all or most of the phenomena, he is obliged to postulate a distinct one to account for each several phenomenon in turn.

As regards the varying position of this doxology, Mr Hort's explanation supposes the following stages. (1) The original place was at the end of the Epistle. (2) It was afterwards attached to xiv. 23 for reading in Church. (3) 'Scribes accustomed to hear it in that connection in the public lessons would half mechanically introduce it into the text of St Paul' at this place. (4) It would then be struck off from the end of the Epistle, that the same doxology might not occur twice. Thus we arrive at the vulgar Greek text, which has it at the end of the xivth chapter only.

Now, waiving for the present the consideration of its original position, I wish to point out two great improbabilities involved in the other assumptions in this sequence. First. There is no such obvious connexion between the paragraph at the end of chapter xiv. and the doxology, as should lead to their being connected together1, if separated in their original position by two whole chapters, while on the other hand these intervening chapters present material for more than one excellent lesson. Bengel indeed suggests, as Mr Hort points out, that the severa sententia afiaprla iariv, with which chapter xiv. closes, would be deemed unfit for the end of a lesson and that this inauspicious termination was got rid of by tacking on the doxology. But how much more easily would the difficulty have been overcome by continuing the lesson a little further and closing with the 2nd or 4th or 6th verse of the next chapter.

1 In a note (p. 342) Mr Hort remarks and Constantinople and from which

that 'the Synaxaria, valeant quantum, the Synaxaria are taken, they would

give Rom. xiv. 19-23, plus the doxo- naturally read it here. I would add

jogy as the lesson' for the Saturday that the Synaxaria (see Scrivener's

before Quinquagesima. But sinoe the Introduction, p. 68 sq.) present no

doxology , occurs here in the vulgar parallel to the omission of two whole

Greek text which prevailed at Antioch chapters.

The instance which Mr Hort quotes (p. 343, note 1), Acts vi 8-vii. 2 combined with vii. 51-viii. 4, as a lesson for St Stephen's day, will hardly bear out his hypothesis, for there the combination is naturally suggested by the subject. Secondly. This solution requires us to believe that all the three steps numbered (2), (3), (4), had taken place before Origen's time, so that he can speak of some MSS. as having the doxology in the one place and some in the other, without suspecting how the variation had come to pass. This supposes such an early development of the lectionary as (I believe) there is no ground for assuming.

LTI. Lastly there are the phenomena in the first chapter to be considered. Here the important fact is, that in one extant MS. (Q) certainly, and in another (F) probably, the mention of Rome has been obliterated in two distinct passages. In i . 7 Mr Hort explains the omission by the fact that'a Western correction substitutes iv 0y071-17 ®eov for dyaTrrfrols ©eov,' so that the words would run eNpa>MHeN&rAUH, where the repetition of iv might occasion the omission of one of the two clauses, especially as the archetype of this MS. appears to have been written stichometrically and each iv might commence a new line. Thus the omission would be accidental. But apparently dissatisfied with this solution he offers a second suggestion, that the omission was intentional; for he adds, ' These two MSS. (F and G) have further a trick of omitting words that do not appear necessary to the sense,' and gives instances. The accidental omission I could understand, but the intentional (thus explained) seems hardly credible, for the words iv 'Pa>fiy are essential to an Epistle to the Romans. Of the omission in i. 15 he gives no direct explanation, except so far as it may be involved in the words ' we may be content to suspect that in these two verses like causes produced like results' (p. 347). I do not understand this, unless by like causes is meant the desire in both cases to obliterate a superfluous clause. I too maintain that' like causes produced like results,' but I cannot allow that the historical fact involved in the mention of Rome could be regarded as a superfluity in an Epistle to the Romans; and, if the omission was intentional in both cases, it must have been (so far as I can see) from the desire of obliterating the proper name, because the proper name was no longer applicable. The hypothesis, that a coincidence so remarkable as the omission of the same name in two distinct passages could have been purely accidental, seems to me to be the most improbable of all.

That the twin MSS. F, G, did not stand alone in this omission, appears from the marginal note in 47, on which Mr Hort has some remarks, p. 344. Whether to these authorities we should add the commentaries of Origen and the Ambrosian Hilary, must remain uncertain. I certainly should not have discovered the omission in them, if it had not occurred independently, and I am not prepared to say that Mr Hort's explanation (p. 345) of their language is not right. At the same time to my own mind the ' Benedictio quam dat dilectis Dei ad quos scribit' of Origen, and the ' Quamvis Romanis scribat, illis tamen scribere se significat qui in caritate Dei sunt' of Hilary, still leave the same impression; but probably they will strike others differently.

It will thus be seen that Mr Hort denies some of my facts, and impugns the significance of others. As the facts give him no trouble, it follows that the hypothesis, which has no other raison ditre but to explain them, should not find favour with him. But, if (as I think I have shown) the facts are even more cogent than they appeared at first, being reinforced by the Latin capitulations, an explanation is still demanded. I cannot indeed say that my hypothesis is free from objections. But a piiori improbabilities could be detected by the keen eye of criticism in the most certain events of history; and a theory, which is based on circumstantial evidence, cannot hope to escape objection on this ground. But, if no other hypothesis has been offered which does not involve more or greater improbabilities, and if some hypothesis is needed to account for the facts, I must still venture to claim a hearing for my own.

In Mr Hort's criticism of the theory itself, as distinct from the facts which evoked it, there are three points especially which call for a reply.

(i) I had assigned the doxology (xvi. 25-27) to the shorter recension of the Epistle, which I supposed to have been issued by St Paul himself at a later date, and had produced parallels to show that its style very closely resembles that of the Apostle's later Epistles. Mr Hort himself considers it to have been the termination of the original Epistle. His argument is threefold: (a) that it is appropriate; (6) that St Paul at the time entertained the ideas contained in it; (c) that it presents numberless close parallels of expression to the earlier Epistles.

(o) As regards its appropriateness, I entirely agree with him. I cannot indeed assent to Baur's opinion which he adopts, that the main drift of the Epistle is revealed in chapters ix-xi. The central idea, as I conceive it, is the comprehensive offer of righteousness to Jews and Gentiles impartially, following on the comprehensive failure of both alike before Christ's coming. After this idea has been developed, the objection arises that, however comprehensive may be the offer, the acceptance at all events is partial and one-sided; that while the Gentiles seem gladly to accept it, the Jews stand aloof; and that thus the promises of the Old Testament appear to be nullified, and indeed reversed. It is to meet the objection which thus starts up, that St Paul pierces the veil of the future and discerns the gathering of the Jews into the same fold whither the Gentiles have preceded them. Thus the result will be comprehensive, as the offer has been comprehensive. But however fit a consummation of the Apostle's teaching this prophetic announcement may be, it does not in itself contain the nucleus of that teaching.

To the whole body of the Epistle however, in which the comprehensive failure, the comprehensive grace, the comprehensive acceptance, have been set forth in succession, the doxology forms an eminently appropriate close. An outburst of thanksgiving for the revelation of this 'mystery' of the impartial Fatherhood of God in Christ is the proper sequel to the contents of the Epistle. This adaptation would not indeed be easily reconcileable with any other authorship than St Paul's; but if written by him, whether written early or late, we should expect it to be appropriate.

(6) And again I grant that its main idea—the impartiality and universality of God's grace as a truth revealed in Christ—was not foreign to St Paul's thoughts at this time, though it assumed a much greater prominence afterwards. Indeed it may be said that this idea necessarily flowed from his commission as the Apostle of the Gentiles.

(c) But, as regards the expression of the idea, I join issue with him. The general style seems to me to be cast essentially in the mould of the later Epistles. The diffusive syntax of the paragraph is exactly what we find, for instance, in the Epistle to the Ephesians. And, when we come to individual phrases, there is (if I mistake not) a very wide difference in point of closeness between Mr Hort's parallels with the earlier Epistles and mine with the later. Compare for example his parallel of Rom. xiv. 4 with mine of Eph. hi. 20 for Tq5 Bvvafiivto, or of Rom. iii. 29, 30 with mine of 1 Tim. i. 17 for fj,ov<p aofyat ®e<p. The only exceptions in favour of the earlier Epistles occur exactly where on my hypothesis we should expect to find them. The expression viraitorj TTiarea><; is repeated in this final doxology from the opening paragraph of the Epistle (i . 5), and the reference to the prophetic Scriptures also has a parallel in the same paragraph (i. 2). On my hypothesis the opening portion was read over and altered, when some years later the Epistle was issued by the Apostle in this second and shorter form; and it was therefore natural that the thanksgiving which was then appended, should embody not only thoughts but also expressions taken from the commencement, thus binding together the beginning and the end of the Epistle.

(ii) The character and condition of the text of the twin MSS., F and G, is one of the points on which Mr Hort lays most stress; and certainly, if his account of my theory were correct, I should find it difficult to answer him. Expressing my hypothesis in his own words, he represents me as holding (1) that 'the scribe of G copied i.-xiv. from one MS. and xv. xvi., from another,' and (2) that'the scribe of F copied in like manner from the same two MSS., though he left no mark of the transition from the one to the other' (p. 339). He then remarks that'If the first of these hypotheses were true we ought surely to find some evidence of it in the respective texts; whereas the closest study fails to detect a shadow of difference in the character of the readings before and after the blank space'; and that ' when F is taken into account, fresh embarrassments arise.' But I did not for a moment contemplate the scribes of F and Q each of them copying directly from these two MSS., containing respectively the shorter and the longer recension of the Epistle. I was well aware that the phenomena of these MSS. would not admit of such a supposition. And I venture also to think that my language, which Mr Hort himself quotes just before (p. 338), cannot be taken in this sense: 'The copyist of an earlier MS., from which it [G] has descended, transcribed a MS. of the abridged recension till the end of chapter xiv., and then took up a MS. of the original Epistle to the Romans'; 'Either their common prototype [i. e. of F and G] or a still earlier MS. from which it was copied, must have preserved the abridged recension.' This language was expressly intended by me to leave open the question, as to the length of the pedigree which connected F and G with the scribe who first combined the two recensions; and the idea of direct parentage, which Mr Hort has imposed upon me, never once entered my mind. Thus I left ample room for the development of the peculiarities of F and G. Only I assumed that the retention of the vacant space at the end of chapter xiv., which I took to indicate the end of the Epistle in one of the two original MSS., had survived this development. But though I still think that (taking it in connexion with all the other textual phenomena on which I dwelt) my account of this blank space is the most probable, yet this is only a subsidiary support to my view, and

I could abandon it without any material injury to the main hypothesis.

But let us enquire what Mr Hort's statement, that 'the closest study fails to detect a shadow of difference in the character of the readings before and after the blank' (p. 339), really amounts to, when considered in its bearing on my hypothesis.

The characteristics of F and G, which differentiate them from what we may call the standard text of St Paul's Epistles, as based on the coincidence of the best authorities, are twofold: (1) Those which they exhibit in common with the Western authorities, and more especially that type of Western authorities which appears in the Old Latin Version; and (2) Those which are peculiar to these two MSS.

To the first class, comprising those readings which must be referred to the Western type, belong the most important, as well as the most numerous, variations from the standard text, whether in the first fourteen or in the last two chapters of the Epistle. If the two MSS. (containing respectively the long and the short form), from which on my hypothesis the text of FG was ultimately derived, were both of them Western, as on all accounts we might probably conclude that they were, then we should expect to find these readings pervading the xvth and xvith chapters, as well as the earlier part of the Epistle. It is difficult to explain the origin and prevalence of the Western type of text at all; but this difficulty was not introduced by my hypothesis, nor do I see that it is increased thereby.

Speaking of the peculiar features of F and G, Mr Hort says, 'The partial adherence of D excepted, this character is unique among existing Greek MSS.' On this statement I should wish to make two remarks. (1) The expression partial seems to me inadequately to express the degree of coincidence between D on the one hand, and FG on the other. Certainly in the two last chapters of this Epistle, with which we are mainly concerned, by far the greater number of the important deviations from the standard text are shared by D in common with FG. (2) These L. E. 24

three are the only1 three Greek uncial MSS. which, whether on external or internal grounds, can be assigned to the Western family. Whatever distinctive features therefore they possess in common, it is reasonable to set down to the Western type of MSS. generally. The Old Latin Version (with the exception of a few fragments) is only known to us through these same MSS., which are bilingual; for other independent copies, which contain a more or less pure Old Latin text, have not been collated: and its phenomena entirely accord with this supposition. The remaining source of evidence—the early patristic quotations— does not offer any obstacle to this conclusion; and indeed in the last two chapters of the Epistle, this evidence, as has been mentioned, is entirely wanting. On the whole then, I think it may be said that the coincidence of D with F and G represents very fairly the Western text.

The second class of readings, those peculiar to F and G, are in the xvth and xvith chapters comparatively unimportant. The divergences of these twin MSS. from D may be taken as approximately representing their peculiarities, though in the course of the analysis it will be seen that in many cases these divergences are supported by other, and especially by Western, authorities3.

These are as follows:

xv. 1 ape<jKov [aptaKtiv]; 3 Ovk [pvx] > 7 vfias [D* rj/zar, but D** ifuxs with most authorities, including Western]; 11 «raiiwart [D traivto-atoa-av, but the Latin of D has Magnificate with many other authorities, and the variation is easily explained in a quotation from the LXX.]; 13 ripoij>optj<rai...ra<rt) \apa Kai fip-qvq [D 7rJp(l>o'ai...7raoT;f Xafxs Kat

1 I pass over E, which is now acknowledged (at least so far as regards the Greek) to be a direct copy of D, and therefore to have no independent value.

- I have not recorded either the accidental errors of G when these have been corrected at the time when the MS. was written, or the divergences of F from G. Mr Hort's view, that F

was copied directly from G, deserves consideration, and may prove true, though his arguments do not seem, quite conclusive. So far as it has any bearing on my hypothesis, it is rather favourable than otherwise. The converse proposition, that G is copied from F, could not be maintained for a moment.

fipijwjr, but B agrees with FG, inserting however »r before iraorj. The Old Latin has repleat...omni gaudio et pace]. 16 hjo-ov Xptorov [D Xptarov bprov, but the Latin of D has Jesu Chruti which also has the vast preponderance of authority in its favour]. 18 6 Xpurros [om. 6} 21 avayyti \avrjyytj\. 24 t\m{a [D adds yap with the preponderance of authorities, but the Latin of D omits it, and so do the Latin fathers]. 25 vvv [row]. 26 Maratdove? [MaKtdovrr]. 27 o^xiXrrai yap [om. yap, but the Latin of D and Anibrosiaster have it]; avrav turt v [tiaiv avrav]. 28 ow apa [om. apa. The Latin of G is Hoc ergo igitur ergo], vfias \yfiav]. 29 yivaaKio yap [D ot5a 8*, but the Latin of D has tcio enim, and other authorities, especially Latin Fathers, have the same conjunction]. 30 rfioo-tvxau [add vrtp tfiov, but several Latin authorities, including the Latin of D, omit the words]. 31 rpoo-Sticrot [»u7rpoa-9«rror. The Latin of D is acceptalis (sic)]. 32 avayj>vxa [amnfrvfrt). 33 vfiav [add. afitjv, but A and others omit it].

xvi. 1 vfiav [jfimv, but the Latin of D has vestram, and AP also have vfiav]. 2 rrapaarartis [rpoarartt]. 3 aortaa-dai [a<nrao-ao-6e. This blunder recurs]. 8 Au7rXum»> [Afir\iav, but the longer form occurs in the Latin of D]. 10 Aptcrroflo\ov [Apiorofiov\ov, but the Latin of D has Aristoboli and this form is found in B and elsewhere]. 11 o-vyyevri [D o-vyytvriv, but corrected by a later hand]. 14 xvpia om. with A. 15 hivvtav [D 1ovXiav, which is correct, but C* has lovviav]. O\vfirti&a [D O\vfirtav, but Latin authorities, including the Latin of D itself, have Olympiada or Olympiadem]. 17 rapaKa\a [D* tpurm, but corrected. The rest have rapaKa\a]. rapa [D* rtpi, but corrected]. 18 mpuo [ra Kvpia], Sov\tvaovotv [&ov\tvovaiv]. 23 6\m ai Tkkxtjo-uu [ojt Tiji norXrjo-iat. The Latin of DFG alike is universal ecclesix, which would cover both readings. Another reading is dXq i) rxKXijo-ui. The .<Eth. is said to have oXai ai «eitXij<riai with FG]. 24 om. Irjaov Xptorov.

This analysis of the readings in the last two chapters shows two things: (1) That in almost every point even of minor importance, in which the text of FG diverges from the correct standard, it agrees with the Western text as exhibited by D or by some other authority; and (2) that the exceptions, which thus form the peculiarities of FG, are in almost every instance trivial and are easily explained by carelessness or caprice in copying. Hence it follows: first, that the scribe, who (on my hypothesis) wrote the archetype of F and G, taking up an average copy of the Western text to supply the xvth and xvith chapters, would find a text substantially such as we actually have here; and secondly, that no long pedigree need have been interposed between this archetype and FG, in order to develope the phenomena which they exhibit in these chapters; but that the intervention of a single scribe, or two at most, would explain everything. If so, the argument from the character of the text cannot be considered a substantial objection to my view.

(in) Mr Hort advances another argument against my hypothesis based on the assumption that the textual phenomena on which my theory is built are gathered together from incongruous sources; and he even goes so far as to ask, 'How is it that every authority, which supports, or may be thought to support, some part of this combination [i.e. the Short Recension, involving (a) the omission of the word Rome in the first chapter, (b) the omission of the xvth and xvith chapters, (c) the presence of the doxology], contradicts some other part ?' (p. 347).

To this statement I demur. I allow indeed that all these phenomena do not coexist in any extant authority. If this had been the case, I should not have had to frame a hypothesis, for the existence of this Shorter Recension would have been an absolute fact. But that there is any contradiction in my authorities, which prejudices the hypothesis, I cannot allow.

This attack has led me to marshal my troops to better effect. I wish especially to call attention to the fact, that the authorities, on which I chiefly rely, have for the most part a close affinity to one another and that they belong to the Western type. The Latin capitulations derived, as I have shown, from the Old Version are essentially such. The copy or copies, to which they refer, presented two (b, c) out of the three phenomena, and (for anything we know) may have presented the third (a) also. The remarkable absence of quotations from the last two chapters in the earlier Latin Fathers points in the same direction. The MSS. FG, which are the only indisputable vouchers for (a), are essentially Western. Their relation to (b), (c), is a matter of dispute between Mr Hort and myself; but the fact that there is a great break in G at the end of the xivth chapter (however explained) cannot but be held to favour my hypothesis to a greater or less degree. The exception to the Western origin of the evidence is Marcion, who, being an Eastern, used a copy of this Epistle in which the two last chapters including the doxology were wanting. But even Marcion is known to have resided for many years in Rome; and if, during his sojourn in the West, he fell in with a copy of the Short Recension, he might have welcomed it gladly, as sparing him the superfluous use of his scissors, which would be required to eliminate such passages as xv. 8, 27.

Hitherto there is no incongruity in the sources from which my data are taken. But the position of the doxology in the several authorities still remains to be considered; and it is evidently here that Mr Hort considers the main 'contradiction' to lie. Though 'there is no lack of authorities of a sort for subjoining the doxology to xiv.,' he writes, yet 'they have no sort of genealogical affinity with the MS. that ignores Rome, or with Marcion.' Now to this I would reply that the capitulations of the Latin Bibles certainly have this affinity, and that (for all we know) the MSS. mentioned by Origen as placing the doxology in this position may have had it also. On the other hand his statement, so far as regards the extant MSS. and the patristic authorities generally, which exhibit it at the end of the xivth chapter, is indisputably true. They belong to the great Antiochene or Constantinopolitan family, which though by far the most numerous, is of inferior authority. On the contrary the place of the doxology in the extant Western authorities is at the end of the xvith chapter. But, allowing the fact, I cannot accept the inference. For suppose that a scribe had before him copies of the two recensions (according to my hypothesis), the one comprising the 14 chapters together with the doxology, the other including all the 16 chapters but omitting the doxology and ending with xvi. 23 Kovapros 6 a8eX<f>6<t. If he set himself to combine the two so as to omit nothing, is it not at least as likely that, when he arrived at the end of the xivth chapter, he would reserve the doxology for the end of the whole Epistle where it seemed to be required to finish off an abrupt conclusion, as that he would leave it at the end of the xivth chapter? The same motive which led others to transpose the benediction (77 K.tx.), which properly

stands at xvi. 20, to xvi. 24, might even more easily induce him to treat the doxology in a similar way, inasmuch as he would still leave it at the end of the Epistle as he found it, though the Epistle had been lengthened out by the two additional chapters. Thus the fact that the Western authorities place the doxology after ch. xvi. seems to me to prove nothing as to the want of affinity between the several authorities for my hypothesis.

But this investigation leads me to observe (and I think the observation is pertinent) how entirely this Western character of the authorities coincides with my hypothesis. I suggested that 'at some later period of his life, not improbably during one of his sojourns in Rome, it occurred to the Apostle to give to this letter a wider circulation'; and that for this purpose he made the alterations which resulted in the shorter edition, so that it was rendered 'available for general circulation, and perhaps was circulated to prepare the way for a personal visit in countries into which he had not yet penetrated' (p. 319). This hypothetical change is made in the West and for the West; and it cannot be considered a matter of indifference that to this same region we owe the authorities which suggested the hypothesis, though at the time when I propounded it I did not see the full significance of this fact.

With these remarks I will leave the theory. For a reply so thorough and so suggestive as Mr Hort's I can only feel grateful. It has led me to consolidate the different elements of my hypothesis, and, unless I am mistaken, to present a stronger front to attack. From criticisms of inferior merit I might have found less to dissent, but I certainly should have found less to learn.