THE APOSTOLIC FATHERS: by the late J.B.Lightfoot DD., DCL., LL.D. Lord Bishop of Durham. Edited and completed by J.R.Harmer MA. Fellow of Corpus Christi College, Cambridge, sometime Chaplain to the Bishop. First published Macmillan & Co. 1891. Prepared for katapi by Paul Ingram, 2014.
HOME | Introduction | Epistles: to the church in Ephesus | Magnesia | Tralles | Rome | Philadelphia | Smyrna; | to Polycarp.



THESE seven epistles were written in the early years of the second century, when the writer was on his way from Antioch to Rome, having been condemned to death and expecting to be thrown to the wild beasts in the amphitheatre on his arrival. They fall into two groups, written at two different halting-places on his way. The letters to the Ephesians, Magnesians, Trallians, and Romans, were sent from Smyrna, while Ignatius was staying there and was in personal communication with Polycarp the bishop. The three remaining letters, to the Philadelphians, to the Smyrnaeans, and to Polycarp, were written at a subsequent stage in his journey, at Alexandria Troas, where again he halted for a time, before crossing the sea for Europe. The place of writing in every case is determined from notices in the epistles themselves.

The order in which they are printed here is the order given by Eusebius (H.E. iii. 36). Whether he found them in this order in his manuscript, or whether he determined the places of writing (as we might determine them) from internal evidence and arranged the epistles accordingly, may be questioned. So arranged, they fall into two groups, according to the place of writing. The letters themselves however contain no indication of their chronological order in their respective groups; and, unless Eusebius simply followed his manuscript, he must have exercised his judgment in the sequence adopted in each group, e.g. Ephesians, Magnesians, Trallians, and Romans.

The two groups, besides having been written at different places, are separated from each other by another distinctive feature. All the epistles written from Smyrna are addressed to churches which he had not visited in person but knew only through their delegates. On the other hand all the epistles written from Troas are addressed to those, whether churches (as in the case of the Philadelphians and Smyrnaeans) or individuals (as in the case of Polycarp), with whom he had already held personal communication at some previous stage in his journey.

At some point in his journey (probably Laodicea on the Lycus), where there was a choice of roads, his guards selected the northern road through Philadelphia and Sardis to Smyrna. [See Map.] If they had taken the southern route instead, they would have passed in succession through Tralles, Magnesia, and Ephesus, before they reached their goal. It is probable that, at the point where the roads diverged, the Christian brethren sent messengers to the churches lying on the southern road, apprising them of the martyr's destination; so that these churches would despatch their respective delegates without delay, and thus they would arrive at Smyrna as soon as, or even before, Ignatius himself.

The first group then consists of letters to these three churches, whose delegates had thus met him at Smyrna, together with a fourth to the Roman Christians apprising them of his speedy arrival among them—this last probably having been called forth by some opportunity (such as was likely to occur at Smyrna) of communicating with the metropolis. The three are arranged in a topographical order (Ephesus, Magnesia, Tralles) according to the distances of these cities from Smyrna, which is taken as the starting-point.

The second group consists of a letter to the Philadelphians whom he had visited on his way to Smyrna, and another to the Smyrnaeans with whom he had stayed before going to Troas, together with a third to his friend Polycarp closing the series.

The order however in the Greek ms and in the versions (so far as it can be traced) is quite different, and disregards the places of writing. In these documents they stand in the following order:

  1. Smyrnaeans
  2. Polycarp
  3. Ephesians
  4. Magnesians
  5. Philadelphians
  6. Trallians
  7. Romans.

This sequence is consistent with the supposition that we have here the collection of the martyr's letters made at the time by Polycarp, who writing to the Philippians says 'The Epistles of Ignatius which were sent to us by him, and others as many as we had with us, we send to you, even as ye directed: they are subjoined to this letter' (ยง13). But though this order, which is given in the documents, has high claims for consideration as representing the earliest form of the collected epistles, I have substituted the chronological arrangement of Eusebius as more instructive for purposes of continuous reading.


Our documents are as follows.

1. The Manuscript of the Greek Original (G), the famous Medicean ms at Florence, from which Voss published the editio princeps in 1646. It is incomplete at the end, and does not contain the Epistle to the Romans. If this ms had been, as Turrianus described it, 'emendatissimus', we should have had no further trouble about the text. But since this is far from being the case, the secondary authorities are of the highest moment in settling the readings.

2. Among these the Latin Version (L) holds the first place, as being an extremely literal rendering of the original. The history of this version is especially interesting to Englishmen. It was discovered by Ussher in English libraries in two mss, one of which has been since lost, and was given to the world by him in 1644. It was certainly translated in England, probably by Robert Grosseteste, Bishop of Lincoln (c. a.d. 1250), or his immediate circle. It exhibits a much purer form of the text, being free from several corruptions and a few interpolations and omissions which disfigure the Greek. At the same time however it is clear, both from the contents of the collection and from other indications, that this version was translated from a Greek ms of the same type as the extant Greek ms; and therefore its value, as a check upon the readings of this ms, is limited. Whenever GL coincide, they must be regarded as one witness, not as two.

3. The Syriac Version (S) would therefore have been invaluable as an independent check, if we had possessed it entire, since it cannot have been made later than the fourth or fifth century, and would have exhibited the text much nearer to the fountain-head than either the Greek or the Latin. Unfortunately however only a few fragments (S1, S2, S3, S4) belonging to this version are preserved. But this defect is made up to a considerable extent in two ways. First. We have a rough Abridgment or Collection of Excerpts (Σ) from this Syriac Version for three epistles (Ephesians, Romans, Polycarp) together with a fragment of a fourth (Trallians), preserving whole sentences and even paragraphs in their original form or with only slight changes. Secondly. There is extant also an Armenian Version (A) of the whole, made from the Syriac (S). This last however has passed through so many vicissitudes, that it is often difficult to discern the original Greek reading underlying its tertiary text. It will thus be seen that ΑΣ have no Independent authority, where S is otherwise known, and that SAΣ must be regarded as one witness, not as three.

4. There is likewise extant a fragment of a Coptic Version (C), in the Sahidic (Thebaic) dialect of the Egyptian language, comprising the first six chapters of the Epistle to the Smyrnaeans, besides the end of the spurious Epistle to Hero. The date of this version is uncertain, though probably early; but the text appears to be quite independent of our other authorities, and it is therefore much to be regretted that so little is preserved.

5. Another and quite independent witness is the Greek Text of the Long Recension (g) of the Ignatian Epistles. This Long Recension consists of the seven genuine Epistles but interpolated throughout, together with six additional Epistles (Mary to Ignatius, Ignatius to Mary, to the Tarsians, to the Philippians, to the Antiochenes and to Hero). The Latin Version (l) of the Long Recension has no independent value, and is only important as assisting in determining the original form of this recension. The practice of treating it as an independent authority is altogether confusing. The text of the Long Recension, once launched into the world, had its own history, which should be kept quite distinct from that of the genuine Epistles of Ignatius. For the purpose of determining the text of the latter, we are only concerned with its original form.

The Long Recension was constructed by some unknown author, probably in the latter half of the fourth century, from the genuine Ignatian Epistles by interpolation, alteration, and omission. If therefore we can ascertain in any given passage the Greek text of the genuine epistles which this author had before him, we have traced the reading back to an earlier point in the stream than the direct Greek and Latin authorities, probably even than the Syriac Version. This however it is not always easy to do, by reason of llie freedom and capriciousness of the changes. No rule of universal application can be laid down. But the interpolator is obviously much more given to change at some times than at others; and, where the fit is upon him, no stress can be laid on minor variations. On the other hand, where he adheres pretty closely to the text of the genuine Ignatius, as for instance through great parts of the Epistles to Polycarp and to the Romans, the readings of this recension deserve every consideration.

Thus it will be seen that though this witness is highly important, because it cannot be suspected of collusion with other witnesses, yet it must be subject to careful cross-examination, before the truth underlying its statements can be ascertained.

6. Besides manuscripts and versions, we have a fair number of Quotations, of which the value will vary according to their age and independence.

From the above statement it will be seen that, though each authority separately may be regarded as more or less unsatisfactory, yet, as they are very various in kind, they act as checks one upon another, the one frequently supplying just that element of certainty which is lacking to the other, so that the result is fairly adequate. Thus A will often give what g withholds, and conversely. Moreover it will appear from what has been said that a combination of the secondary and capricious authorities must often decide a reading against the direct and primary. For instance, the combination Ag is, as a rule, decisive in favour of a reading, as against the more direct witnesses GL, notwithstanding that A singly, or g singly, is liable to any amount of aberration, though in different directions.

The foregoing account applies to six out of the seven letters. The text of the Epistle to the Romans has had a distinct history and is represented by separate authorities of its own. This epistle was at an early date incorporated into the Antiochene Acts of Martyrdom of Ignatius, and thus dissociated from the other six. In its new connexion, it was disseminated and translated separately. It so happens that the Greek mss which contain this epistle (the Colbertine, 18 S. Sab., and 519 Sin.) are even less satisfactory than the Greek ms of the other six (the Medicean); but on the other hand we have more than compensation for this inferiority in the fact that the Acts of Martyrdom (with the incorporated epistle) were translated independently both into Syriac (Sm) and into Armenian (Am); and these two versions, which are extant, furnish two additional authorities for the text. Moreover the Metaphrast, who compiled his Acts of Ignatius from this and another Martyrology, has retained the Epistle to the Romans in his text, though in an abridged and altered form.

From this account it will be seen that the authorities for the Epistle to the Romans fall into three classes.

(1) Those authorities, which contain the epistle as part of the Martyrology. These are the Greek (G), the Latin (L), the Syriac (Sm), and the Armenian (Am), besides the Metaphrast (M). These authorities however are of different values. When the epistle was first incorporated in the Acts of Martyrdom, it still preserved a comparatively pure form. When it has arrived at the stage in which it appears in the extant Greek MS (G), it is very corrupt. In this last form, among other corruptions, it exhibits interpolations and alterations which have been introduced from the Long Recension (g). The MS used by the Metaphrast exhibited a text essentially the same as that of G.

(2) The independent Syriac Version (S) of which only a few fragments remain, but which is represented, as before, by the Syriac Abridgment (Σ) and the Armenian Version (A).

(3) The Long Recension (g), which in great parts of this epistle keeps close to the text of the original Ignatius.


Though the principles on which a text of the Seven Epistles should be constructed are sufficiently obvious, they have been strangely overlooked.

The first period in the history of the text of the genuine Ignatius commences with the publication of the Latin Version by Ussher (1644), and of the Greek original by Isaac Voss (1646). The Greek of the Epistle to the Romans was first published by Ruinart (1689). The text of Voss was a very incorrect transcript of the Medicean ms, and in this respect subsequent collations have greatly improved on his editio princeps. But beyond this next to nothing was done to emend the Greek text. Though some very obvious corrections are suggested by the Latin Version, these were either neglected altogether by succeeding editors or were merely indicated by them in their notes without being introduced into the text. There was the same neglect also of the aid which might have been derived from the Long Recension. Moreover the practice of treating the several mss and the Latin Version of the Long Recension independently of one another and recording them co-ordinately with the Greek and Latin of the genuine Ignatius (instead of using them apart to ascertain the original form of the Long Recension, and then employing the text of this Recension, when thus ascertained, as a single authority) threw the criticism of the text into great confusion. Nor was any attention paid to the quotations, which in several instances have the highest value. Hence it happened that during this period which extended over two centuries from Voss to Hefele (ed. x, 1839; ed. 3, 1847) and Jacobson (ed. 1, 1838; ed. 3, 1847) inclusive, nothing or next to nothing (beyond the more accurate collation of the Medicean ms) was done for the Greek text.

The second period dates from the publication of the Oriental versions—the Syriac Abridgment with the Syriac Fragments by Cureton (1845, 1849), and the Armenian Version by Petermann (1849) « The editio princeps of the Armenian was published at Constantinople in 1783: but this version was practically unknown to scholars until Petermann's edition appeared. New materials of the highest value were thus placed in the hands of critics; but, notwithstanding the interest which the Ignatian question excited, nearly thirty years elapsed before any proper use was made of them. In some cases the failure was due, at least in part, to a false solution of the Ignatian question. The text of Bunsen (1847), Cureton (1849), and Lipsius (1859), which started from the assumption that the Syriac Abridgment represented the genuine Ignatius, must necessarily have foundered on this rock, even if the principles adopted had been sound in other respects. Petermann and Dressel (1857) however maintained the priority of the Seven Epistles of the Vossian text to the Three of the Curetonian; and so far they built upon the true basis. But Petermann contented himself with a casual emendation of the text here and there from the versions; while Dressel neglected them altogether. Jacobson (ed. 4, 1863) and Hefele (ed. 4, 1855) also, in their more recent editions which have appeared since the Oriental versions were rendered accessible, have been satisfied with recording some of the phenomena of these versions in their notes without applying them to the correction of the text, though they also were unhampered by the false theory which maintained the priority of the Curetonian Abridgment. It was reserved for the most recent editors, Zahn (1876), and Funk (1878), to make use of all the available materials and to reconstruct the text for the first time on sound and intelligible principles.

The text which I have given was constructed independently of both these editions, and before I had seen them, but the main principles are the same. Indeed these principles must be sufficiently obvious to those who have investigated the materials with any care. In the details however my views frequently differ from theirs, as must necessarily be the case with independent editors; and in some respects I have had the advantage of more complete or more accurate materials than were accessible to them.