On the End of the Epistle to the Romans, By Dr. Hort

DR LIGHTFOOT in this Journal (n. 264 ff.) has demolished M. Renan's ingenious theory about the composition of the Epistle to the Romans, and along with it some others of inferior merit. He proposes instead a simpler view, which one could wish to believe true, so admirably does it harmonize the most salient phenomena of the text, and so free is it from broad historical improbability. A close examination however reveals difficulties which I am constrained to think fatal.

Dr Lightfoot supposes that the letter originally addressed to the Romans was our present epistle as it stands in the Received Text and Authorized Version, wanting only the last four verses, i.e. the second Benediction (xvi. 24) and the Doxology (25-27); but that at a later time St Paul himself' made it available as a circular letter or general treatise' by cutting off the last two chapters, substituting the Doxology, and omitting the name of Rome in i. 7, 15. The direct evidence lies in three chapters, i. xiv. xvi., which I will consider separately and in inverse order.

I. The apparently triple ending of xvi. in the Received

Text, when taken as a whole, rests on absurdly small and

worthless evidence, three or four obscure cursives and the

inferior MSS. of the Latin Vulgate: it is a mere jumble of the

Latin and the late Greek traditions, which owes its place in the

printed text to Erasmus¹.

¹ His account of his own proceed- the truth as it could be known at that

ing is intelligible, while his careless- date than it would be now. 'Hanc

nessgro8sly misrepresents the evidence; partem usque ad Debemus autem qui

indeed his statement is further from dam codices omnino non habent, quidam in fine adjiciant epistolae. Nos, readings generally, explains this sin

quoniam id non videbatur ad hnno lo- gular collocation. D is not so purely

com pertinere, semovimns in finem Western as FG: Sedulius combines

hujus epistolae' (note on xiv. 23 in ed. the Old with the Hieronymic Latin,

prinoepsof 1516). 'Haec est pars quae In each case the Doxology most be a

in plerisqne Graecorum codioibus non later accretion. The Gothic has the

additur, in nonnullis alio additur loco, Benediction at 24 and (in xvi.) no

sicut indicavimus, in quibusdam adji- Doxology: the extant fragments fail to

citur in fine. Id quod et nos fecimus, shew whether the Benediction was at

praesertim assentientibus Latinis ex- 20 likewise.

emplaribus' (note on xvi. 25 IT.). * If, as is probable, the shifting of

If the Doxology be put out of sight, we are met by a still worse confusion of incongruous traditions; that is, the doubling of the Benediction (20 and 24). The great mass of early authorities of various groups concur in placing the Benediction at 20 only: so KABC 5 137 lat.vg (best MSS.) memph aeth Orig.ruf. The pure 'Western' group D*FG (with Sedulius and perhaps the Gothic version) places it only at 24¹, evidently from the feeling that it must be the close of the epistle. Minor shiftings and other like freedoms taken by the same group of authorities occur in almost every chapter of St Paul: two whole verses 1 Cor. xiv. 34 f. are pushed 5 verses forward by DFG 93 and some Latin Fathers: compare 1 Cor. xv. 26. The scribes of the fourth century, bringing together MSS. from different regions, here as in countless other instances heaped up without omission whatever they found, and so the Benediction was set down in both places. The compound reading appears first in the Greek commentators of the fifth century from the Syrian school, then in the Harclean Syriac (A.D. 508-616): in extant MSS. it is found only in L (= J) of the ninth century and the great mass of cursives. There is however a similar combination in a few respectable authorities who retain the Doxology and place the second Benediction after it (P 17, the vulgar Syriac and the Armenian versions, and the Ambrosian Hilary): and this implies the previous existence of MSS. which simply transposed the Benediction to their end of the epistle, as (D*)FG transposed it to theirs*.

¹ D* and Sedulius add the Doxology the Benediction and the dropping of

after the Benediction. The nature of the Doxology were simultaneous in the

both authorities, as evinced by their common source of D'FGSed., P17 etc differ merely in taking one step in- own, notwithstanding the first person

Thus the historical relations of the authorities clearly shew that, be the claims of the double Benediction as a ' harder reading' what they may, it is as a matter of fact the last term in a series of changes.

Thus far there is no reason to suppose that Dr Lightfoot would dissent. He places the Benediction at 20 and there alone, and gives what is doubtless the right explanation of the order in saying that 'v. 20 is the true close of the epistle, and the salutations from the amanuensis and other companions of St Paul were added irregularly as a sort of postscript, as was very likely to have been done, considering the circumstances under which St Paul's epistles were written' (p. 319). Whoever will read the chapter through as far as 24 according to this arrangement, will find everything straightforward and intelligible; while the nature of the postscript is such as might easily mislead a mechanical transcriber. The difficulty begins when we go on to 25-27. Supposing however that we had no evidence about these three verses except as to their presence or absence in this place¹,1 do not see why we need hesitate to take them as an ending to the postscript, just as 20 is the ending to the epistle proper*. Having once made that fresh start to introduce the salutations sent by present companions, St Paul might gladly seize the opportunity to close the whole by a solemn giving of glory to God, as his first ending had carried grace to men. Compare xi. 36 in connexion with xi. 32 and the adjoining verses; also v. 2; xv. 5, 6. Similar pauses of adoration occur elsewhere in the epistle; i. 25; (viii. 39;) ix. 5; xi. 36; xv. (13,) 33, where I believe 'Afir/v to be genuine: and it is to be observed that, when St Paul's own salutations to Christians at Rome were ended, he was not able to refrain (xvi. 17-20) from breaking out afresh into renewed exhortations to mutual peace through willing obedience to the common Lord.

stead of two: the writer of their com- used for the moment in 22 by Tertius

mon original was willing to transpose the amanuensis in sending his own

but not to omit. The two transposi- greeting. Otherwise i avvepyfa [fiov],

tions were however apparently hide- o! avyyeviU nov before the mention of

pendent of each other. Tertius would not be intelligible. The

¹ Their total omission will be con- subsequent 6£tvot /now «toJ rijt ixxiatas

sidered further on. is also the language of an apostle.

* The postscripts evidently St Paul's

As he had gone back to the perils and hopes of the Church after the one set of individual greetings, so we can imagine him joyfully returning to the yet higher sphere of God's universal purposes after the other set of individual greetings¹. Nay the parallelism between 17-20 and 25-27 is one of contrast as well as likeness. The fust passage gives vent to somewhat of the anxious dread which lurks behind many a phrase of xv. 14-33, especially 30, 31. If these were St Paul's last words to the Romans except the two sets of greetings and the Benediction of 20 b, the epistle might have appeared to end in a note of discord: at all events its exulting comprehensiveness would have died back into the rebuke and controversy proper for the Galatians. The sudden upward flight of the Doxology seems therefore to be almost demanded, to swallow up not only trivial individualities of salutation but also the temporary strifes of the Church.

But it is said that the Doxology differs too much in style from the rest of the epistle to form part of it. I used to suspect that it might be the ending to one of the forms of the encyclical epistle to the Ephesians, which was preserved from being lost to the Canon by being appended to St Paul's longest epistle. Dr Lightfoot (after Dean Alford) points out its resemblance to the Pastoral Epistles as well, and accordingly treats it as marked by the Apostle's later style generally. Before scrutinizing words and phrases, let us look at the subject.

¹ Dr Lightfoot says (p. 317) that the tion of the type' ^ K.t.\. But

Doxology ¹ has nothing in common none of his other epistles have a post

with the usual endings of St Paul's script, following a benediction in that

Epistles, which close with a benedic- form already given.

The starting-point is doubly personal; an anxiety about the stability of the converts addressed, such as tinges the hopefulness of the first and last words spoken to and about the Romans (i. 11; xvi. 17-20); and a bold lifting up of what friend and foe knew as the distinctive 'Gospel' of St Paul, (and that in its distinctive form of 'preaching,' and with its distinctive appeal to 'faith,') such as marks the time of the conflict with Judaism within the Church (i. 1, 5, 9, 16; xv. 16; x. 8, 14, 15). Here the pronouns 'you' and 'my' face each other with an emphasis which ^n such a context is hard to explain till we remember the presaging instinct with which St Paul saw in the meeting of himself and the Roman Christians, if indeed it was to be vouchsafed, the pledge and turning-point of victory (i . 10 ff.; xv. 29-32; cf. Acts xix. 21; xxviii. 31). Then comes the idea in which the Doxology culminates, the counsel of the far-seeing God, the Ruler of ages or periods, by which the mystery kept secret from ancient times is laid open in the Gospel for the knowledge and faith of all nations. This idea no doubt pervades the Epistle to the Ephesians, though with considerable enrichments. But is it foreign to St Paul's earlier thought? The second chapter of 1 Corinthians at once shews that it was not and explains why the fact is not obvious. St Paul is dealing there with converts who were in danger from pride of eloquence and wisdom (from i. 5 onward). For fear of this danger, he says (ii. 1 ff.), he himself kept back all excellency of speech or of wisdom when he came among them, and confined himself to the bare preaching of the Cross as alone fitted to their imperfect state. But for all that he desired them to know that he too had in reserve a wisdom which he spoke among the perfect. Its nature he briefly hints in words that closely resemble our Doxology (' We speak a wisdom of God in a mystery, that hidden wisdom which God fore-ordained before the ages unto the glory of us' etc. ii. 7), and then hastens to explain that, even after being laid open, it demands a spiritual power to discern it. The Churches to which he wrote about this time, at Corinth, in Galatia, at Rome, were not in a state to profit by an extended exposition of a belief which yet was strong in the Apostle's own mind, and so the traces of it in the early period are few. Later it filled a larger space in his thoughts, it acquired new extensions and associations, and he had occasion to write to Churches which by that time were capable of receiving it. But it is not really absent even from the Epistle to the Romans. Kindred thoughts find broken and obscure utterance in viii. 18-30. The belief itself is the hidden foundation of the three chapters (ix-xi.) in which God's dealings with Jew and Gentile are expounded, and comes perceptibly to light in their conclusion (xi. 33-36). Now it is precisely in these chapters, as F. C. Baur (Paidus 341 ff.) saw long ago, that the main drift of the epistle is most distinctly disclosed: all its various antitheses are so many subordinate aspects of the relation of Jew and Gentile which in this seeming episode is contemplated in its utmost generality as reaching from the one end of history to the other. The whole epistle could hardly have a fitter close than a Doxology embodying the faith from which its central chapters proceed. Here at last that faith might well be articulately expressed, though a wise economy compelled it to be latent as long as the Apostle was simply instructing the Romans. This Doxology is in fact a connecting link between the epistle at large and the earlier concentrated doxology of xi. 36. In both alike human sin and hindrance are triumphantly put out of sight¹: but here the eternal operation of Him 'from Whom, through Whom, and unto Whom are all things' is translated into the language of history.

An examination of single phrases is attempted in the following table, which includes some less obvious coincidences of thought3.

¹ They could not be left out in the latter part of the Epistle, when St Paul's own position and the dangers of the Romans had to be spoken of (xv. 14-33; xvi. 17-20). But for this very reason it was the more necessary that the ground conquered at the

end of xi. should be maintained at the final close of the Epistle. See p. 324.

³ References to the later epistles are in []: the chief passages are set out at length by Dr Lightfoot, p. 318.

els rama ra Bom. i. 5 above; xi. patsim; xv. pattim; xvi. 3 f. Yru

yvupur6ivros, pifw is similarly used Bom. ix. 22 f.; 1 Cor. xv. 1; ? Gal.

i. 11; as well as (often) in the later period.

H6vip atxp$ 6t$ Bom. iii. 29, 30 rj 'lovSatuv i dtos fiivuv; otyl Kal idrum;

val Kal idvuv, ttrep ttt 6 6eos 4t cr.X. [M6r<j> deif 1 Tim. i. 17 a kindred passage, which early caused ruv ahhntu to be inserted here after roAt aluvas, and in its turn received ao<t>$ hence in the fourth century: cf. 1 Tim. vi. 15; but also Jud. 4, 25; John v. 44 etc] ZoQla is predicated of Ood by St Paul with reference to the working out of a distant purpose by unexpected means: so Bom. xi. 33; 1 Cor. i. 21, ?30; ii. 7; [Eph. i. 8; iii. 10; Col. ii. 3.]

Sia "Irpov XpioToC Bom. v. 1 f.; xv. 6 f.; GaL i. 4 f.; [Eph. i. 5 f., 11-14;

[f]¹ iiS6(aehtoit iii. 21; Col. i. 27; 1 Tim. i. 11, 17.]

BlQvaf &njv.

A minute examination of the passages briefly indicated in this table will shew that the dominant thoughts of the Epistle, —the thoughts which inspired its beginning (L 1-17), its primary close (xv. 6-33), and its three characteristic chapters in which the old faith and revelation are invoked on behalf of the new,—are precisely those expressed in the final Doxology; and that the separate words and phrases of the Doxology are for the most part what have already occurred in the Epistle, while there are hardly any not to be found in epistles of the same or an earlier period*. If this be so, the obvious resemblances to parts of the later epistles lose all force as evidence of date. The Doxology and 1 Cor. ii. 6-10, a passage absolutely inseparable from its context, support each other in shewing that St Paul's late teaching was his early belief; while in each case there was an adequate motive for his exceptional transgression of the limits imposed on him by the present imperfection of his converts. The condensed and cumulative style, which he used more freely afterwards, arises naturally from the compression of varied thoughts and facts into a single idea in a single sentence under the impulse of eager feeling. Rom. i. 1-7; iii. 21-26; 2 These, i. 3-10 offer a true analogy: what distinguishes them is their articulation, which was hardly possible in a doxology.

¹ if is probably an intrusion, not- of which is preserved in 1 Cor. ii. 7; withstanding the presumption in favour x. 11. Ou the other hand vraKo^ (riot an irregular construction. areut), both phrase and sense, is pecu

* The only dear exception is xpirtx liar to the early epistles. alivioi (2 Tim. i. 9; Tit. i. 2), the idea

But we may go further. As is the Epistle to the Romans itself in relation to the monuments of St Paul's early teaching, gathering up, harmonizing, concluding, such is the Doxology in relation to the Epistle. It looks at once backwards and forwards. Springing from the keen sense of a present crisis, it gives old watchwords of action a place in the dawning vision of thought which the epistles from Rome were to expound, and anticipates in its style as in its ideas the habitual mood of the time when the crisis was victoriously ended, and the unity of the Church secured.

II. The course thus far has been smooth, because the chief textual difficulties have been out of sight. The end of the fourteenth chapter is a point at which various phenomena present themselves which nothing in the context would have led us to expect. Some of them (a) on the surface mark only an interruption of the Epistle. The Doxology is inserted either (1) here alone or (2) both here and in xvi. In (3) a single MS. G, one of the twin MSS. which alone omit the Doxology altogether, an empty space is left here, occupying half a line at the bottom of an otherwise full page and 5 lines of the next page. Secondly (/3) the whole of the two following chapters are supposed to have been omitted (1) by Marcion (on the authority of Origen), (2) perhaps by Tertullian and even Irenaeus, and (3) in the capitulation of an unknown Latin MS. mentioned by Wetstein. The variety of this evidence, if it stands proof, is a strong argument in favour of any theory which will account for all the particulars.

The testimony of Origen requires consideration first. We have it only in the greatly abridged version of Rufinus, a careless and licentious translator. This is not a passage with which he is likely to have consciously tampered; but there is no certainty that the language is Origen's own. Characteristic terms of expression as well as ideas may be recognized through Rufinus's Latin in almost every page; but none such are conspicuous here: rather the sentences are short and simple for Origen. The comment on the Doxology (after xvi. 23) begins thus. "Caput hoc 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] usque ad finem cuncta dissecuit. In aliis vero exemplaribus, id est in his quae non sunt a Marcione temerata, hoc ipsum caput diverse positum invenimus. In nonnullis etenim codicibus post eum locum quem supra diximus, statim cohaerens habetur 'Ei autem qui potens est vos confirmare.' Alii vero codices in fine id ut hunc¹ est positum continent. Sed jam veniamus ad capituli ipsius explanationem." As the text stands, it asserts plainly that Marcion removed from the Epistle both the Doxology and xv. xvi.; and that of the MSS. unaffected by Marcion's proceeding some had the Doxology after xiv., some after xvi.

So the passage has been universally understood. On the other hand for many years I have had a strong impression that the Benedictine text is wrong in three letters, and that on the removal of this tiny corruption the whole interpretation collapses. De la Rue's notes on this book often mention the readings of a certain Paris MS. (Reg. 1639). Wherever I have examined them, they have appeared usually to give the truest text against all other known authorities, and very seldom to be evidently wrong. In this place Reg. 1639 has in instead of ab. If the preceding hoc is likewise altered to hie, and so small a variation may easily have escaped notice, we get an entirely new and, I venture to think, more probable statement. Origen begins by saying merely that 'Marcion, the falsifier* of the Gospels and [St Paul's] Epistles, removed this paragraph completely from the Epistle.' Then it appears to strike him that some reader might know the Epistle in a copy which had the Doxology at the end of xiv. (if not there alone), and acquit Marcion as having at most only removed a superfluous repetition¹. He adds therefore explicitly 'And not only here but also' at xiv. 23 'he cut away* everything quite to the end.' Then, for fear the remark might not be understood by those who knew the Doxology only in xvi., he explains 'But in other copies, that is in those which have not been corrupted by Marcion, we have found this very paragraph differently placed' etc.

¹ [' hunc' is a misprint for ' nunc.'] terpolate, but properly to give a spuri¹ Interpolo in ancient Latin, it will ous look of newness to old things, and be remembered, does not mean to in- so generally to falsify.

Of these three statements the end of the second might be thought a mere repetition of the first, according to the corrected reading. But I think Origen wished to make it perfectly clear that Marcion's offence, as he understood it, was no mere erasure of an obnoxious phrase but utter excision of the entire paragraph. Nor is it unlikely that the Greek original contained intermediate digressive sentences which gave a resumptive force to the repetition. No one, I presume, would seriously find a difficulty in the words 'to the end' as inappropriate to the removal of the Doxology alone, in the case of MSS. in which it had stood at xiv. 23: their correctness in reference to its normal position would make them sufficiently descriptive for Origen's purpose.

¹ Reasons will be given further on Dissccuit would not be an unnatural for suspecting that the MSS. here no- rendering of repiiKo^ev or possibly noticed by Origen had the Doxology in pitrciuv, either of which would mean both places. At this point the differ- simply 'cut away.' Compare Epiph. ence is without importance. Haer. 309 d oi> Pavov Si r)i¥ ipxty Ari

³ This is not, it must be confessed, rt\tai [of St Luke's Gospel]..., dXXd Kox the natural meaning of the single word roO ri\ovt Kai riir lUaui v Iroma irepUdistecuit: but will the context on any KO\pe Tur rrjs Aidelas \lryuv K.t.\.: and view tolerate another? As regards the again dXXd raxx airrGw ireptriiuiuv, rmi. Doxology, abstulit is decisive. Is it Si dXXouiaas xtipi\iua. In the first senconceivable that Marcion only 'teparat- tence, so closely resembling Bufinus's ed' xv. xvi. from the rest of the Epistle, in form, Aroriiwu and re/xxirrtd must while still acknowledging their autho- be practically synonymous, for the rity, whether he joined them to another preceding sentence describes the Gosepistle or not? or that such an opera- pel as repiKtKOftpUi/ov Aro rip Apxrit by tion would be unrecorded? The diffi- Marcion. culty surely lies in the translation.

Hoc ipsum caput is perhaps a slightly stronger phrase than we might have expected: how far it represents the Greek, and, if supposed exact, how far a knowledge of the unabridged context would explain it, we need not try to conjecture: even as it stands, it has a certain force in binding together the first and second statements.

On the other hand the internal evidence for the truth of the corrected reading is substantial. The order of the sentences, which Rufinus is not likely to have changed, runs naturally upon this view. By the common reading Origen keeps till last the only fact specially concerning the passage on which he is commenting: his first two sentences might have been written with equal force and appropriateness on any group of verses in the two chapters. He begins with saying that Marcion removed this paragraph, three verses, and then condemns, as an aggravation of the main offence, his removal of 59 verses, of which these three are nothing more than the end. Why should he choose this particular place for the remark, if Marcion's operation was really on that extensive scale? Why not mention it at the proper place, xiv. 23? It may be urged that possibly he was forgetful there, as he is certainly silent about the Doxology, but gladly repaired his omission when the Doxology brought to mind by association the earlier critical point in the Epistle. Certainly it might be so. But in that case we should expect him to begin with the transposition of his immediate text, and having so been carried to xiv. 23 to append by way of digression an account of Marcion's proceeding. The reverse order, which we actually find, has no logical justification on the common interpretation, unless Origen himself saw in Marcion's supposed omission of xv. xvi. and in the transposition of the Doxology two facts connected by community of origin. That however is a step in criticism which there is not the slightest evidence that he took. He regarded Marcion's omission, whatever its extent, as an original and unprecedented act; and he gives no hint that the transposition or repetition in certain MSS. was a consequence of Marcion's mutilation: in other words the two facts were in his eyes two independent phenomena. How then came the one to suggest the other? If Marcion omitted two chapters, the sole point of contact is xiv. 23; and thus the transposition, which alone forms a bridge from xvi. 24 to xiv. 23, must have preceded the omission in Origen's account. If on the other hand Marcion cut out only what the scribes transposed, then no bridge is needed. The first and the last sentences refer alike to the same subject, the paragraph on which Origen is avowedly about to comment. The second sentence refers partly to this place, partly to the other; and likewise serves to anticipate an erroneous criticism of the first statement, which might occur to Origen's readers.

The commentary of Jerome on Eph. iii. 5 explains diffusely how St Paul could say that 'the mystery of Christ in other generations was not made known to the sons of men' notwithstanding the language of the prophets. At the outset he repudiates the doctrine juxta Montanum that the prophets spoke in ecstasy, not knowing what they said. Three columns further on he repeats 'Those who will have it that the prophets understood not what they said, and spoke as it were in ecstasy, bring to confirm their doctrine not only the present text, but also that which is found [in the epistle] to the Romans in most MSS., reading Now to Him, etc.' The inference is obvious, that the writer had seen or heard of MSS. which did not contain the Doxology. But who is the writer? Jerome in his preface tells us that he had partly followed the three books of Origen on this Epistle. Comparison of the Greek fragments proves how freely he drew on his great predecessor's ample stores; and any one familiar with Origen's style will recognize it in many places where the Greek is entirely lost. Throughout this long disquisition Origen's hand cannot be mistaken, though Jerome may have added or altered this or that sentence. The controversy with Montanistic doctrine belongs moreover to the third, not the fourth century¹.

¹ The dislike of the early Alexan- phecy' or inspiration is well known, drians to the Montanist theory of 'pro

The character of the MSS. hinted at as wanting the Doxology is sufficiently indicated in the two sentences which follow the refutation of the Montanists. 'And in like manner it is to be observed that the mystery of our faith cannot be revealed except through the Prophetic Scriptures and the coming of Christ. Let those therefore know who understand not the Prophets, and desire not to know, protesting that they are content with the Gospel alone' etc. This evident allusion to the Marcionists, the other great sect which threatened the Church in Origen's days, suggests the strong probability that the passages from his two commentaries relate to the same subject. What he calls 'most MSS.' here are identical with 'those copies which have not been corrupted by Marcion.' In the former case the Doxology is said to have been omitted¹: may we not infer, in the absence of evidence to the contrary, that this and this alone constituted Marcion's offence? Whatever the argument might be worth taken independently, it appears to me a striking corroboration of the result obtained thus far.

Tertullian's language is ambiguous. After confuting Marcion out of Galatians and 1, 2 Corinthians, he proceeds to Romans (adv. Marc. V. 13). Henceforth, he says, he will touch but briefly on what has come before him already, and pass over altogether what has come before him frequently. He is tired of arguing about the Law, and about God as a Judge, and so an Avenger, and so a Creator. Yet he must point out the plain references to justice and judgement which meet him at the beginning of the Epistle (i 16 ff.; ii. 2). It will be enough for him, he declares, to prove his point from Marcion's negligences and blindnesses, from the sayings which he left undisturbed*.

¹ The words are 'Qui volant Pro- logy been the mere conclusion of a phetas etc., cum praesenti testimonio large section omitted. illud qnoque quod ad Bomanos in pie- "He notices but one omission by risque codicibus invenitur ad confirma- Marcion in this epistle, that of c. iz. tionem sui dogmatis trahunt, legentes The limits are not given, but there is FA autem' etc. They do not formally little room for doubt. Eight other negative the omission of the two whole (short) omissions are recorded by Epichapters; but other language would phanius, who professes to furnish only surely have been chosen had the Doxo- a selection (Haer. 317 f.). It is singu lar that Epiplianius should pass over Gospel, as -/iKparriplaarai pfyrt ipxhv

the loss of three consecutive verses: (xov Mre iUaa fifre Tatw, tfiariov /St

but his silence would be far more Ppufiivov iirb noWuv <tritur ^jt^x« ^t&v

He then runs over the Epistle in 5 pages, just half what he had bestowed on the little epistle to the Galatians, passing over in silence some long spaces of text containing appropriate matter, as iii. 1-20 and x. 5-xi. 32. The ethical paragraph xii. 9-xiii. 10 tempts him to give examples of the anticipation of its teaching in the Old Testament, and he concludes with insisting on the harmony of Law and Gospel in inculcating love of neighbours. There apparently he intended to stop, the doctrinal part of the Epistle being ended, but his eye was caught by the words 'judgement-seat of Christ' at xiv. 10. He therefore adds (14 s. f.) rather awkwardly, with evident reference to what he had said on the beginning of the Epistle¹, 'Bene autem quod et in clausula tribunal Christi* comminatur, utique judicis et ultoris, utique creatoris, ilium certe constituens promerendum quem intentat timendum, etiamsi alium praedicaret.' And then he proceeds to another epistle. The absence of allusions to anything in xv. xvi. requires no explanation: it is hard to see what could have been cited except xv. 4, 8, 18, which are slight and contain nothing new in relation to Marcion, and the Doxology, which all agree to have been omitted by him. But in clausula certainly means 'in the close of the Epistle,' and it is a natural inference that such a phrase would not have been used if xv. xvi. had stood in Tertullian's MS., whether that was his own or one of Marcion's recension. Natural but not conclusive. The verse quoted is not in the actual close on any view; thirteen verses follow of xiv. But the force of the word must be estimated by the context.

astounding if two whole chapters were rptrrov.

missing. Nothing could be safely in- 1 So not long before he had said, not

ferred in any case from his employ- it is true of a book but of a passage

ment of the word &Kpunjpidfa as ap- (1 Cor. ix. 10-x. 11), 'Denique et in

plied to St Paul's epistles (rai airrCi v Si clausula praefationi [apostolus] re

ijKpurripiaaiiivuv <rvvi)0ut Tj airov /ipSi- spondet' (c. 7).

ovpyiq. 317 D): his wide use of it is ma- * The true reading is Tov Beos, but

nifest when he says (311 D) that the confusion with 2 Cor. v. 10 was easy.

Antithesis to the beginning of the Epistle, not by any means the very beginning but i. 16 ff., ii. 2, is the motive of the remark. If xiv. 10 is included in a section of the Epistle, however large, which can fairly be called in any sense its close, the point of the remark is saved. Now Tertullian had to all appearance virtually ended his comments at xiii. 10. What follows to the end, with the partial exception of xv. 3 f., 8-12, is either hortatory or personal. The business of the Epistle, so to speak, is over: to the eye of a rhetorician, accustomed to study the members of a speech, the remainder would all constitute the close. Tertullian uses the word more loosely still on another occasion, again for the sake of an antithesis. To reinforce his position that Christ's command to flee from city to city under persecution became obsolete when the apostles went forth to convert the Gentiles, he urges that St Paul, who at an early time had consented to escape in a basket, in the close of his ministry (in clausula officii) rebuked those who urged bim not to go up to Jerusalem lest he should suffer there (de Fuga in Pers. 6). Yet this incident (Acts xxi . 13) preceded the events at Jerusalem, the two years' imprisonment at Caesarea, the voyage and shipwreck, and the two years at Rome; to say nothing of later occurrences not told in the Acts.

It remains true that Tertullian does not cite any words out of xv. xvi. in other parts of his writings¹: nor does Irenaeus or perhaps Cyprian³. Negative facts of this kind are by no means to be contemned, but their value depends on the attendant circumstances. Seventeen verses only of the two chapters (xv. 1-13; xvi. 17-20) were likely to be quoted. Of these Origen once quotes one (setting aside the commentary), Clement three; while of others it so happens that Origen quotes five, Clement three, besides the Doxology.

¹ Semler and Oebler indicate 5 re- he means p. 283 (Ep. 65 § 3) 'nec ferences to xv. 4,14; xvi. 18: but they ante se religioni sed rentrt potius et are imaginary. quaestui profana cupiditate serviut';

³ Fell's index gives only xvi. 18 ' ven- a very doubtful reference, tri serviunt: E[pist.] 233.' Doubtless

Lastly Wetstein has a note at the end of xiy.: 'Codex Latinus habet capitula epistolae ad Romanos 51, desinit autem in caput xiv.; ex quo conficitur ista capitula ad editionem Marcionis fuisse accommodata.' 'Later critics,' says Dr Lightfoot [p. '289],'have not been able to identify the MS. and thus to verify the statement.' Their failure however matters little. The phenomenon here obscurely described is not peculiar to a single MS.: it belongs to what was probably a widely current Latin capitulation, found e.g. in the earliest (540-550) MSS. of the Vulgate, the Amiatinus and the Fuldensis. The sections or breves of Romans are 51, § 50 beginning at xiv. 15, and § 51 at xv. 4. In the table of contents before the Epistle § 50 is headed 'De periculo contristante [sic] fratrem suum esca sua, et quod non sit regnum Dei esca et potus sed justitia et pax et gaudium in Spiritu Sancto,' a fair description of the section; and § 51 'De mysterio Domini ante passionem in silentio habito post passionem vero ipsius revelato,' which in strictness applies only to the Doxology¹. If the marginal figures were lost, it would be a natural inference that § 50 ended with xiv., that § 51 consisted of the Doxology, and that xv. xvi. were absent from the MS. on which the capitulation was originally formed. But as on this view the table and the marginal figures contradict' each other, it seems hopeless to attempt to clear up the confusion while the origin of the capitulation remains unknown*. There is no Latin authority whatever for associating the Doxology with xiv. 23; so that it would be rash to assume the table of headings to be alone authentic, and the marginal figures to have been inserted at xv. 4 by a misunderstanding. Yet that is certainly a possible solution.

¹ Either Wetstein examined only the answer to ecclesiastical lessons. Othertable of headings, or he overlooked wise one might have thought that the the inconspicuous figures li. at xv. 4, Doxology was appended to xv. 13 or 33 a place where he would scarcely expect for public reading, and the rest of xv. them. This is the sole point of differ- xvi. neglected. Some sections are deence. scribed only by their end, as others sup>1 Internal evidence proves that the only by their beginning, sections cannot, in their present form,

Only it must be remembered that the table of headings, with all its obscurities, would stand as the sole direct piece of evidence for the omission of xv. xvi. by any authority.

One indirect testimony Dr Lightfoot finds in the space left after xiv. 23 in the single MS. G, as noticed above (p. 329). His inference [p. 316] is that 'the copyist of an earlier MS., from which it has descended, transcribed a MS. of the abridged recension [i.e wanting xv. xvi.] till the end of chapter xiv., and then took up a MS. of the original Epistle to the Romans to supply the lacking matter, omitting however the doxology as inappropriate to what had thus become the middle of the letter, and perhaps intending to give it a place afterwards, but abandoning his purpose. It is an instructive fact that in the allied MS. F no space is left after ch. xiv., but the text is written continuously.' 'Either their common prototype¹, or a still earlier MS. from which it was copied, must have preserved the abridged recension.'

¹ The above was written on the assumption that F and G were independently copied from the archetype, as all considerable writers on the subject except Wetstein had laid down on apparently sufficient grounds. A query by Dr Westcott has recently induced me to examine the matter anew, and so led me to the conclusion that the scribe of G alone used the archetype, and that F is a copy of G. The few verbal (not orthographical) variations of F that might have seemed to preserve the readings of the archetype crumble away on examination. F often interchanges vnett with iifuts, not seldom against all sense, and 6 times alters ir6 to ari! it omits the article 23 times, and perhaps once inserts it: it omits other words 16 times, and inserts them at most thrice (Bom. vii. 19 fieiau as in 15; ix. 31, with a special mark, Swaioaivris as in the line above; Gal. Jin. daqv; all from the Vulgate): and the remaining changes, I believe fourteen, of which most are favoured by the Vulgate, are all trivial and natural. On the other hand FG agree in countless blunders, evidently such and not traditional variants, which cannot all, to say the least, be set down to the archetype. Again the confusion of spellings has its uniformities. To take only the more frequent cases, F incessantly interchanges tit, o u, i v, S d (t); in almost every line FG together interchange i «, e ai, very rarely either MS. separately; and I have failed to detect any permutations approximately peculiar to G. Misspellings of the promiscuous sort swarm in FG together and in F separately; in G separately they are rare and always so simple as to be within the capacity of the scribe of F to correct. Precisely the same may be said of the divisions of words; F is free from no ontrageous portent found in G, but has to answer for many of its own. No one can believe that two scribes independently arrived at e.g. Hamrovatv txai iryturtiwov w \oyuv (both FG have u over -vor: F further divides frytur. eurov) for irrorvruatv Ixt vyiaivivtuv \iyuv: and the absence of division of words in the archetype is proved by the numerous self-corrections of the scribe of G, where he has added to the end of one word the first letters of the next, seen his error, and began the second word afresh with a space between. In these cases he sometimes has forgotten to put in the cancelling dots or line, and then the writer of F confidingly transcribes the whole. But usually he is careful to follow only corrected readings. In 1 Cor. xi. 31 f. ax-6 translated by a happens to be under the end of iavroit in G; and the stroke or accent which, as usual in G, caps a looks like a cancelling line to the final s: hence F reads iavrbv thongh the verb is &teKplroixlv. Other instances might be given of the dependence of F on accidents in 6. The relations of the Latin accompaniments (fg) are complicated, but tend to the same result. The body, so to speak, of g must have at least a double origin, from a pure Old Latin text and from one or more altered texts, either the true Vulgate or one of the intermediate revised texts or both. Where none of his materials represented the Greek literally enough, the scribe evidently devised new renderings of words and still oftener

changed their order. This is shown not only negatively by comparison with the mixed and fragmentary yet frequently copious evidence of all sorts as to variations in Latin MSS. and Fathers, but also positively by mistakes arising from the wrongly divided Greek words and the like. Sometimes g offers two or more alternative renderings, either all traditional or part traditional part original. The body of f is tolerably pure Vulgate, unequally but always imperfectly assimilated to the Greek with, I believe, the aid of no docnment except g, all the elements of which may be recognized. In 1 Cor. x., singled out by Mr Scrivener for its frequent departure from the Vulgate, out of the 46 variants 23 agree with d and 42 with g, while the remaining 4 consist of 2 blunders, one correction of an obvious blunder, and one interpretative change of tense. The concordance of evidence so various seems decisive against any claim of F to represent the archetype where it differs from G. Nothing however in the text of this article is substantially affected by the result except the sentences in brackets.

In other words (1) the scribe of G copied i.-xiv. from one MS. and xv. xvi. from another; and (2) 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. If the first of these hypothetical facts 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. The partial adherence of D excepted, this character is unique among existing Greek MSS.: that it should prevail equally in two MSS. accessible to the scribe of G is possible certainly, but not likely; and the hypothesis involves this further anomaly that the two originals, so singularly alike in the main, must have differed on the capital point, the omission of xv. xvi. [When F is taken into account, fresh embarrassments arise. Either the scribe of F copied one MS. throughout or he did not. If he did not, an exact repetition of the circumstances attending the writing of G is demanded, without such evidence as the blank is said to afford. If he did, what becomes of the primary original of G ?] The blank may, I believe, be easily explained by a simple process. The Greek text of F and G alike was copied from a single archetype wanting only the Doxology. [The scribe of F wrote down exactly what lay before him.] The scribe of G on arriving at xiv. 23 remembered the Doxology as occurring there in some other MS. that he had read (all extant MSS. but 9 have it there, 4 older, 5 younger), held faithfully to his archetype, but satisfied his conscience by leaving a space which might be filled up hereafter if needful. He did in fact only what the scribe of B had done four centuries before, when he left a blank column for the supplement to St Mark's Gospel (xvi. 9-20). It follows that FG attest the omission of the Doxology alone, while the blank in G vouches merely for the vulgar Greek text as it prevailed from the fourth century onwards.

That reading of the vulgar text however remains to be explained if possible, and remarkable without doubt it is. The intrusion of the Doxology after xiv. 23 appears in two forms: conjointly with its retention at the end in AP 5 17¹, and some Armenian MSS.: in this place alone in L (= J) and all Greek cursives but 8 (or 10), some MSS. known to Origen (above, p. 330), the Harclean Syriac and the Gothic² (with, it is said, two other late and obscure versions),

¹ There is a doubt about 2 or 3 presence of the Doxology after xiv.

others, and more will probably be would make the gap exactly equal in

found in due time: see also p. 341, note length to the adjoining leaves of the

1. The introduction at xiv. 23 by the Codex Carolinus, which alone has pre

second hand of the Latin text in the served the verses before and after,

trilingual 109 is doubtless due to an The 4 existing leaves of this MS. shew

imperfect assimilation to the Greek. that xi. 33-xv. 13 was written on 8

² The fragments of this version do leaves; and all the measures give the

not comprise xiv. 20-xv. 3. But the same length to a leaf within a line.

Chrysostom¹ and the Greek commentators who follow him, and perhaps Cyril and John of Damascus. Perplexities abound here. The first small group is select* though not trustworthy: by the analogy of other passages it indicates a reading of high antiquity, probably current at Alexandria, but a correction. Origen's MSS. being waived, the certain portion of the second group is practically rubbish: that is, it contains no authority of the slightest value hereabouts except as a rare adjunct to some primary authority left nearly in solitude. That some MSS. known to Origen should have attested a reading of the first group is exactly what might have been expected: their association with the second is passing strange. It suggests a doubt (more is not permissible) whether Origen after all did not speak of those MSS. which had the Doxology at xiv. 23 as having it also at the end. Rufinus's clumsy scissors may easily have shorn off the additional fact, especially as the antithesis became clearer in consequence: on this view the words about Marcion's doings 'not only here but also in that place etc.' would have increased force, though it must be allowed they do not require it. But another difficulty remains. We might have supposed the double position of the Doxology to be owing to the combination of texts from two sets of MSS., each of which had it in a different place and there alone; yet the character of the authorities inverts this order. In cases like this it is ultimately found safer to trust to the historical relations of the evidence than to any speculations about probability. But indeed here the only tolerable explanation that offers itself of the introduction of the Doxology at xiv. 23 in either group would point to the first group as exhibiting the earlier form of corruption.

¹ One Vatican MS. of Chrysostom that Chrysostom himself used only the

according to Mr Field (p. 547) has both vulgar Greek text,

text and commentary in both places, ² Though inferior to 17, 6 is a cur

and so might be added to the first sive of the first rank, group. But internal evidence proves

Changes in the Greek text of the New Testament, chiefly by interpolation, arising from the modifications required for Church lessons are common in' MSS., though they have rarely found their way into printed texts. The salutations in xvi. might easily be thought to disqualify the bulk of the chapter for public reading¹, especially at a time when but a few select lessons were taken from the whole Epistle*:

\

¹ The Greek 'Euthalian' capitulation found in divers MSS. (printed by Mill N. T. 418 and elsewhere) has for the heading of its § 18 rtpl [ri}i]

atut rrfl Xpiarou ivd-iKaKtat, of § 19 repl rfi s \titovpylat abrov trp lv iraro\y Kox Svaet, and nothing after. These must correspond to xv. 1-13,14-33. It follows that xvi. (but not xv.) is omitted, evidently because not publicly read in tome church. The latest sectional number (24) in P stands at xv. 14, doubtless for a similar reason. By a singular coincidence § 18 of the Vatican capitulation begins with xv. 1 as in the 'Euthalian' capitulation: but they do not coincide in the earlier chapters, and the Vatican sections proceed to the end, commencing § 19 at xv. 25, § 20 at xv. 30, and §21 at xvi. 17. Fritzsche (Rom. i. p. xlvii.) pleads that on the same grounds we might argue the exclusion of 1 Cor. xvi. from public reading, since no trace of its contents appears in the' Euthalian' capitulation for that epistle. Why not? The last sectional numeral (20) in the margin of P in 1 Cor. is at xv. 51. Thus again both independent capitulations equally agree with what the nature of the chapter renders intrinsically likely. The Capuan Lectiouary in the Fulda MS. of the Latin Vulgate takes no lesson from Bom. xv. xvi. except xv. 8-14 (for the Circumcision), and none from 1 Cor. xii.-xvi.

³ Dr Lightfoot (p. 312) refers to Eeiche as having shewn that xv. xvi. were not omitted in public reading. Beiche depends on Fritzsche and after him

Meyer, who argue (1) that the profound reverence of the early Christians must have saved every letter of the N. T. from being unheard in the churches; (2) that the lectionaries prove the whole epistle to have been actually read. But this continuous reading noted in the lectionaries belongs only to the Daily Lessons, which E. Banke (Herzog R. E. xi. 376 ff.) shews to be of late date, perhaps not earlier than the 12th century. The ancient lessons for Sundays and Saturdays are all more or less selected, continuous only in certain definite cases. The existing Synaxaria, valeant quantum, give Bom. xiv. 19-23 plus the Doxology as the lesson (an appropriate one) for Saturday before ' Tyrophagus' Sunday (Quinquagesima): see the tables in Scrivener Introd. 72; Scholz N.T. ii. 459; Matthaei Rom. xxiv. They have but two other lessons from this part of Bomans, xv. 1-7 for the 7th S. and xv. 30-33 for the Saturday before the 10th S. after Pentecost (Scrivener 69 f.; Scholz 458; Matthau ib.). All these arrangements however are probably Constantinopolitan, and originally derived from the 'use' of Antiooh. An Alexandrine Table of Lessons is preserved in a Vatican MS. (46 Paul, of Wetstein), and has been edited by Zacagni Coli. Hon. 712-722; but the first leaf, containing from Easter to the 3rd S. after Pentecost, is missing. In the part of the year where Bomans is chiefly read, xiii. 1-8, xv. 1-6, 13-19, 30-33 occur consecutively; but no other lesson from this Epistle after xiv. 11 appears anywhere. A few scattered lessons agree with those in the common Synaxaria, bnt the coincidences are such as might easily be accidental: the systems are independent throughout, though partly analogous. Saturday lessons are wanting, according to the custom of the early Alexandrine and Roman Churches (Socrat. v. 22), except in Lent. But as it is the long eightweek Lent of Inte Alexandrine usage, comparison as to 'Tyrophagus' Saturday is out of the question. All the Lenten Saturdays have in place of a definite lesson the single obscure formula 'Ex Tov i.iroarb\ov fIs ayiovs: the 4 lessons eU urelas iylur, Rom. v. 1-9; viii. 28-34; Heb. x. 32-38; xi. 33-xii. 2, can hardly be meant, as Zacagni seems to suppose; but the reference may be to a Menologium, or Table of Lessons for Holy-Days, not preserved in the MS.: the common Synaxaria have lessons from Hebrews on the Saturdays of their Lent. 'Tyrophagus' Sunday is one of the days of coincidence, the lesson being Rom. xiii. 11xiv. 4. In short nothing can be clearly made out, except the prevalence of variety of usage and the utmost freedom in the selection of lessons; that is, Fritzsche's and Meyer's arguments are found to have no support from facts.

and yet some church, for instance that of Alexandria, may have heen glad to rescue the striking Doxology at the end for congregational use by adding it to some neighbouring lesson¹. It could not well be used by itself, even if it were longer: it craved to follow some passage which in like manner craved a close. Many would find in the benedictions at xv. 13, 33 a reason against appending the Doxology in either place², while it would make an impressive termination to a lesson formed out of the latter verses of xiv. which when alone have both a harsh⁸ and an unfinished sound.

¹ The late Alexandrine lesson for St Stephen's Day begins Acts vi. 8 and

ends vii. 60. As the other lessons are all short, this must have been made up of two passages, the speech being omitted. A similar Old Latin lesson for St Stephen's Day has been printed by Ceriani (Mon. S. et P. i. n. 127 f.), combining vi. 8-vii. 2 with vii. 51-viii. 4. Ranke in Herzog R. E. x. 81 notices two Mozarabic lessons from Jeremiah, one of which omits 13 verses in the midst, and the other is a cento of 5 fragments.

² Gabler in Griesbach Opusc. ii. p. xx vi.

³ This is the ground taken by J. A. Bengel {App. Crit. 340 Burk), to whom we owe the first suggestion about Church Lessons. He says 'Videntur Graeci, ne lectio publica in severam sententiam Quicquid non est ex fide peccatum est desineret, hanc ei clausulam attexuisse. Conf. var. Matth. iii. 11.' His note on the omission of Kox rupl in this last place is worth quoting. 'Citra haec verba finierunt Graeci, v. gr. in Aug. 4 [the Lectionary numbered 24], lectionem ecclesiasticam, ne tristis esset clausula. Simili euphemismo et Judaei post ultimum eumque severum Iesaiae, Malachiae, Threnorum, et Koheleth versum rescribere penultimum solent: et Graeci nonnulli post ultimum Malachiae versum ponunt antepenultimmii. Etiam in liti:. [86] ri\ot primum post haec verba, deinde his erasis ante, notatum est."

Scribes accustomed to hear it in that connexion in the public lessons would half mechanically introduce it into the text of St Paul, just as they seem to have introduced a liturgical doxology after the Lord's Prayer into the text of St Matthew (vi. 13). Then in the course of time it would be seen that St Paul was not likely to have written the Doxology twice over in the same epistle, and it would be struck out in one place or the other; while familiar use would override any effort of critical judgement¹, and so the Doxology would vanish from the end of xvi., nothing in the context seeming to demand its retention. Such I conceive is the history of the position which the Doxology lipids in the vulgar Greek text, a position which it would probably retain in the Received Text and in the popular versions of Europe but for the confused impulse which led Erasmus in this instance to adhere to the Latin tradition.

III. In the two places of the first chapter (7, 15), where the name of Rome is mentioned, it disappears in the single MS. G. Some leaves are wanting at the beginning of F; doubtless if extant they would shew the same omission. At the first passage there is a note in the margin of 47 to the effect that 'he [or 'it': no nominative] mentions the phrase iv 'Ptofiy neither in the commentary nor in the text.' The subject may be some unknown commentator, but is more likely to be an 'ancient copy' of St Paul's Epistles which is expressly cited in a similar marginal note on vi. 24^s, and which like 47 itself may have been provided with a marginal catena or ' commentary'³.

¹ Yet ancient criticism, finding the Doxology between xiv. and xv., would probably see nothing to object to; while it would readily stumble at the apparent violation of epistolary correctness in xvi. 25 ff. The influence of MSS. like FG may also have helped to expel the final Doxology, while it would be powerless to displace the same words where imbedded firmly in the text.

¹ The reading there quoted from T6 ra\aiir ivrlypa<j>ov is both rare and excellent: the other marginal readings of 47 are of no interest, nor is there I believe any other reference to another authority. Cf. Griesbach Symi. Crit. i. 155 ff.

³ An uncial MS. with a catena, like H of St Luke, might be called 'the ancient copy' in the 11th or 12th century.

Dr Lightfoot thinks he sees a trace of the same omission in Origen's criticism as rendered by Rufinus, notwithstanding the presence of Romae in the text. But the context gives another turn to the language used. 'Benedictio autem pacis et gratiae, quam dat dilectis Dei ad quos scribit apostolus Paulus, puto quod non sit minor ea quae fuit benedictio in Sem et in Japheth, quoniam per Spiritum impleta est erga eos qui fuerant benedicti etc.' 'Ad quos scribit' is substituted for 'qui erant Romae' because the point is that St Paul's benedictions had not less dignity and effect than the sacred benedictions of the Old Testament; as Origen proceeds 'Non ego his omnibus inferiorem duco hanc Apostoli benedictionem, qua benedixit ecclesias Christi,' while any inference from the generality of 'ecclesias' is precluded by the further remark that 'haec Apostoli consuetudo scribendi non erga omnes ab eo servatur ecclesias,' and by the classification which follows. Still less can I recognize any sign of the omission in the Ambrosian Hilary's words 'Quamvis Romanis scribat, illis tamen scribere se significat, qui in caritate Dei sunt.' For he goes on 'Qui sunt hi nisi qui de Dei filio recte sentiunt? Isti sancti sunt et vocati dicuntur: sub lege enim agentes¹ male intelligunt Christum' etc. Every word becomes clear on comparison with a passage in the Prologue (25 AB) in which he contrasts the 'Romani' with the Judaizers who were equally at Rome (iv 'P&J/i??): the meaning is that St Paul writes not to all 'at Rome' indiscriminately, but to those at Rome who were in caritate Dei.' The true text in full is iraaiv Tok ovaiv ev Ptu/z?? dy<nrrjroi<; deov Kxtjtok ayiot?. A Western correction (D* lat. [the Greek lost] G, the 2 best MSS. of the Vulgate, apparently the Ambrosian Hilary, and perhaps Hilary of Poitiers) substitutes ev dydirrj deov for dyaTnyroi? deov, doubtless on account of the Kxrjtois following ('who...through the love of God are called to be saints'). The result is that ENPfiMH and ENArATTHOY were left contiguous, each beginning with ev.

¹ Not 'they agentes' but 'they who agunt.'

The loss of one or other out of a pair of such groups of letters is common in MSS. of any form, and would be peculiarly liable to occur in one written in columns of short lines, such as was assuredly the archetype of FG¹. These two MSS. have further a trick of omitting words that do not appear necessary to the sense, as might easily be the case with ev 'Po>/«7 here when the following words were changed:

¹ Hug pointed out [Einl. in N. T. i. 252 ff.) the evidence afforded by the frequent capitals in G that it was copied from a 'stichometrical' MS. resembling D, and perhaps older. In F many of the capitals are wanting, and probably even the scribe of G neglected a large proportion. It has not however been noticed, I believe, that the three equal chasms in the Greek text common to F and G measure for us the contents of each leaf of the archetype, about 20 lines of the 'Oxford Lloyd,' a convenient standard for reference. Now in these three places (1 Cor. iii. 8-16; vi. 7-14; Col. ii. 1-8) a leaf of D contains on the average 24 lines of Lloyd, Greek alone. If then the archetype of FG had like D a Latin column, we might form a fair impression of the general appearance by cutting off 2 lines from each page of D. If there was no Latin, each leaf of the archetype must have contained rather less than those of any extant Biblical MS.: the nearest approach would be to the purple and silver N (21 Lloyd lines) and the peculiar Z (23), apparently onoe a MS. of the same class. E of the Acts has indeed but 12 Greek Lloyd lines; but there is the Latin in addition. One exception might have been found in the lost archetype of a part of C. A fortunate displacement of text in the midst of a page of the Apocalypse (x. 9,10; vii. 17-viii. 4; xi. 3-12) proves, on accurate measurement and calculation, notwithstanding the loss of the preceding leaf, that the archetype hereabouts was made up of quires of 8 sheets, with 12 Lloyd lines to a leaf, while a leaf of C itself has 100 Lloyd lines. The outer sheet but one of a quire must have been somehow turned inside out before stitching, and so the scribe of C, copying on without thought, interchanged vii. 17-viii. 4 and x. 10-xi. 3. But it is possible, though unlikely, that the archetype of C was bilingual: the Grteco-thebaic fragments of T have 21 Greek Lloyd lines to a leaf, nearly double. The great primary Eastern MSS. of the 4th and 5th centuries, KABC (with 160, 148, 131, 100 Lloyd lines to a leaf respectively), owe I believe their stately appearance to the new impulse to exhibit together the settled and completed Canon of Scripture. Before Constantine the parchment copies were in all likelihood small and portable. Our two earliest MSS, K and B, seem to represent the older period in the narrowness of their columns, not in the ample structure of their pages, which may or may not have been suggested by a partly opened papyrus roll. During the time when most variations arose, narrow columns were assuredly general, to say the least. The date when 'stichometry' proper began is still unknown; the evidence which refers it to the middle of the 5th century is most precarious. And the example of E of the Acts shews on how different scales stichometrical arrangements might be made.

so eh trcurTtpiav i. 16; 17 etc <fivaea><; iiKpofivarta ii. 27; (ov iravTg>? iii. 9;) 'itjo-ov iii. 26; fiovov iv. 16; o 6dvaro<; v. 12; (rai ? iiridvulais avrov vi. 12;) ott e/ioi To Kukov irapdxeirai vii. 21; ei 8e Xpio-ro? iv vfilv viii. 10; viodeaiav viii. 23; &c. The omission in i. 7 might therefore be neglected without further thought but for the parallel omission of Tok iv 'P(ifirj in i. 15, the name of Rome being confined to these two passages in the Epistle. The coincidence would certainly be noteworthy if it were sustained by other documentary evidence, or if there were independent reasons for believing a recension of the Epistle to have existed in which the marks of a special destination were purposely obliterated. There is no such reason apart from the supposed removal of xv. xvi.: the hypothesis is suggested by the reading of G at i. 7, 15. We may therefore be content to suspect that in these two verses like causes produced like results.

All the phenomena of text alleged to prove a double recension have now been examined. The enigmatical Latin capitulation excepted, they have been found, if I mistake not, to be more naturally explicable by other causes. This result becomes clearer still when the hypothesis is examined as a whole. The second recension, it will be remembered, was said to consist of chapters i. to xiv., with the Doxology, and without the two namings of Rome. How is it then that every authority, which supports, or may be thought to support, some part of this combination, contradicts some other part? For the omission of xv. xvi. the one direct testimony, if such it be, is that of Marcion: and yet the one incontrovertible fact about him is that he omitted the Doxology. If G is to be added on the strength of the blank space after xiv., yet again it leaves out the Doxology. Once more there is no lack of authorities of a sort for subjoining the Doxology to xiv. We may waive the fact that they all retain xv. xvi. We cannot forget (1) that they all make mention of Rome at i. 7, 15; and (2) that they have no sort of genealogical affinity with the MS. that ignores Rome, or with Marcion. In few words, the authorities, which as a matter of fact contain the rude outlines of the first recension, supply the main data for constructing the second. Meanwhile neither recension is represented in the great mass of good authorities, Greek, Latin, Syriac, Egyptian, or other, on which the text of St Paul stands in ordinary cases. Both recensions, as wholes, are purely conjectural. If Rome and the transposed Benediction are set aside, the first recension is vouched for by FG (standing for a single archetype) alone of extant documents and by some traditional evidence. The second recension can be reached only through a hypothetical text which Marcion altered, and a hypothetical duplicate original of G.

Such being the relations of the textual evidence, little requires to be said on the intrinsic probability of the hypothesis. There is nothing in it that we need hesitate to accept if only the evidence were stronger. But it surely has not that kind of verisimilitude which would raise the feeling that it cannot but be true. The only analogous instance known to us is the encyclical epistle addressed to the Ephesians and other neighbouring churches. But that letter appears (1) to have been sent simultaneously to its different recipients; and (2) to have been general in form in the first instance, not a special appeal trimmed for general use. Analogy apart, it is difficult to imagine St Paul deliberately cutting out in after years the words that spoke of personal bonds to definite churches and believers, and the passionate hopes and fears which they had once called forth. If for any purpose he needed an impersonal treatise on the old subjects, he would surely have written it anew. Indeed the fitness of our Epistle, however altered, may well be doubted. Its catholicity springs from the marvellous balance that it holds between Jew and Gentile, which in its turn rises historically out of the equal or almost equal combination of the two bodies in the metropolitan Church, as Dr Lightfoot has justly insisted (p. 312 ff). Is it probable that the same characteristics would recur in the unlike 'countries into which he had not yet penetrated' (p. 319)? Even that single point of connexion disappears when we recall the pregnant paradox of his relation to the Romans, that, though he had not seen them, he knew them so well.

The inverse theory of several critics, that the original letter to the Romans ended with xiv. and, some add, with the Doxology, and that St Paul afterwards appended xv. xvi., escapes these difficulties to plunge into worse. Paley proves convincingly that xv. can belong only to the time when the body of the Epistle was written and can have been addressed only to the Romans: and there is cogent evidence which he has overlooked. Dr Lightfoot has shewn how much can fairly be elicited from xvi. to the same effect. The slight break moreover after xiv. is onesided, and on the wrong side. The opening words of xv. furnish a tolerable beginning: the last words of xiv. make a very bad end, even when the Doxology is allowed to follow.

When all is said, two facts have to be explained, the insertion of the Doxology after xiv., and its omission. The former has occupied us enough already: the latter now claims a few words. If the view taken in this paper be right, the omitting authorities are FG, Marcion, and certain MSS. twice noticed by Origen, once distinctly and both times implicitly, as having been corrupted by Marcion. The readings of D* and Sedulius, mixed authorities substantially akin to FG, likewise imply omission as antecedent. Origen accuses Marcion of wilful omission: is the charge just? There is analogy favourable to either answer. It is now equally certain that Marcion sometimes mutilated the text of his favourite apostle, and that some variations or omissions imputed to his pen were in fact simply the readings which he found already in his MS. The reference to 'prophetic Scriptures' in v. 26 might conceivably annoy him, though, as far as we know, he tolerated much of the same kind that was less likely to please him. But the removal of four words, an operation more in his manner, would have served every purpose. Though copies of his Apostolicon were seemingly current here and there in the Church, no extant document can be shewn to have been affected by any of his wilful alterations. Indeed ' copies corrupted by Marcion' need mean to us no more than 'copies agreeing in a certain reading with Marcion's copy': and Marcion's copy, prior to his own manipulations, appears by various signs to have had much in common with the authorities associated with him in the omission of the Doxology. On the whole it is reasonably certain that the omission is his only as having been transmitted by him, in other words that it is a genuine ancient reading.

Genuine: but right or wrong? The question cannot be answered off-hand. Not right merely because shewn to be as old as the first quarter of the second century: not wrong merely because the outward evidence for omission is small and at the same time virtually responsible for many impossible readings. Experience shews that authorities, rarely or never in the right when they alter or add, are often in the right when they omit. Such is preeminently the case with the Western group of which DFG form an important section. Yet the omissions of DFG without the accession of B, when examined together, are for the most part suspicious. Thus on the whole authority is in favour of the Doxology. Internal evidence is likewise not all on one side. So considerable an omission might be expected to proceed only from a strong and evident motive, such as cannot be decisively recognized here. On the other hand the singular and yet unobtrusive correspondence with those parts of the letter which best reveal its purpose is an argument hardly to be gainsayed without strong documentary testimony. Pure accident is not to be rejected from the imaginable causes of the loss. The last or outer column of a papyrus roll, the outer leaf of a parchment book, would be subject to peculiar risks, as every keeper of MSS. can avouch; and it is probable that an epistle as long as that to the Romans would often form a book to itself in early times¹.

¹ On the scale of the archetype of C usually of coarse thick parchment, the this epistle would occupy 90 leaves. delicate thin vellum of our great MSS. They would necessarily be small, and being a recognized mark of luxury; and would thus form a sufficient volume. The variety of order in the Pauline epistles in early times, of which there is good evidence, would be promoted by their separate use. On this view the language used by Constantine and Eusebius (V. Const, iv. 36 f.) about the new Imperial Bibles, 'sumptuously prepared,' with their quires of 3 or 4 sheets, has more force: Const an tine's word aupArtov ( = corpus), the technical term for a combination of single works, doubtless expresses the change from books and groups of books to the full Canon.

¹ Since this article has been in type, Dr Lightfoot has kindly pointed out to me an oversight in pp. 337 i., 347. In the Codex Fuldensis the table of headings to Romans agrees with that in the Codex Amiatinus etc only in the latter part, as Banke himself observes, p. xxiii. The first 23 headings belong to a totally different capitulation, and exhaust the Epistle down to xiv. 13. Then follows No. 24 of the other table, describing ix. 1-5; and so on. The previous or peculiar headings have no marks or divisions answering to them in the text itself. The scribe evidently saw that his tale of 51 sections could not be made up without borrowing

elsewhere, and he ventured to save appearances at the cost of sense. Whether he had actually reached the end of the first table or only saw it near at hand, is less clear. The headings are not so exactly descriptive as to forbid the inclusion of xiv. 14-23 in §23; and thus it is certainly possible that we have two complete and independent Latin capitulations in which xv. xvi. are omitted. More cannot be said till ancient capitulations generally have been properly investigated, and this demands a wide examination of MSS. Meanwhile it should be observed that (1) the Fulda headings have no trace of the Doxology; and (2) they are loaded with Augustinian or Anti-Pelagian phraseology, and cannot therefore be dated much before 400 at earliest.

The sectional numerals in P, I now likewise see, may possibly once have been continued after Bom. xv. 14; 1 Cor. xv. 51: some numerals have faded out of sight in almost every epistle, and in Bom. i.-x. all have vanished; cf. Tisohendorf M.S.I, v. p. xiv. But as the §1 of each epistle (10) except 1 Cor. begins after the salutation, analogy favours the view taken above (p. 342, n. 1).

Nor again dare we assume that the rash hands which shifted the Benediction would hesitate to let go the Doxology, in their zeal to give the Epistle a correct ending. Having once lost the vantage ground of possession from whatever cause, the Doxology would not easily recover it. Henceforth conservatism and criticism would be on the same side. Presently, when the Doxology had found a home after the fourteenth chapter, every motive for replacing it at the end of the Epistle was gone. We cannot wonder that the evidence for retaining it there, and leaving inviolate the continuity of the fourteenth and fifteenth chapters, is exclusively ancient and good¹.

F. J. A. HORT.