A.—The Historia Acephala.

————————————

This most important document was brought to light in 1738 by the Marchese F. Scipio Maffei († 1755), from a Latin ms. (uncial parchment) in the Chapter Library at Verona. It was reprinted from Maffei’s Osservazioni Letterarie in the Padua edition of Athanasius; also in 1769 by Gallandi (Bibl. Patr. v. 222), from which edition (the reprint in Migne, xxvi. 1443 sqq. being full of serious misprints) the following version has been made. The Latin text (including letters 46, 47, and a Letter of the Council of Sardica) is very imperfect, but the annalist is so careful in his reckonings, and so often repeats himself, that the careful reader can nearly always use the document to make good its own gaps or wrong readings. Beyond this (except the insertion of the consuls for 372, §17 ad fin.) the present editor has not ventured^37673767    The corrections were made before he could obtain the essay and text of Sievers (Zeitsch. Hist. Theol. 1868), where he now finds them nearly all anticipated. Sievers’ discussion has been carefully and gratefully used, but his text is defective, especially from the accidental omission of one of the key-clauses of the whole (§17). to go. The importance and value of the fragment must now be shewn.

The annalist evidently writes under the episcopate of Theophilus, to which he hurriedly brings down his chronology after the death of Athanasius (§19). At the fortieth anniversary of the episcopate of Athanasius, June 8, 368, he makes a pause (§17) in order to reckon up his dates. This passage is the key of the whole of his chronological data. He accounts for the period of forty years (thus placing the accession of Ath. at June 8, 328, in agreement with the Index), shewing how it is exactly made up by the periods of ‘exile’ and of ‘quiet’ previously mentioned. To ‘quiet’ he assigns ‘xxii years v months and x days,’ to ‘exile’ xvii years vi months xx days; total xl years. He then shews how the latter is made up by the several exiles he has chronicled. As the text stands we have the following sum:

Table A.

Now the exact result of the figures as they stand is 182 months, 9 days, i.e. 15 years 2 months and 9 days, or 2 years 4 months and 11 days too little. Moreover of the well-known ‘five exiles,’ only four are accounted for. An exile has thus dropped out, and an item of 2 years 4 months 11 days. Now this corresponds exactly with the interval from Epiphi 17 (July 11), 335 (departure for Tyre, Fest. Ind. viii), to Athyr 27 (Nov. 23), 337 496 (return to Alexandria F. I. x). The annalist then (followed apparently by Theodt. H. E. ii. 1) reckoned the first exile at the above figure. But what of the first figure in our table, xc months iii days? It again exactly coincides with the interval from Pharm. 21 (Apr. 16, Easter Monday), 339 to Paophi 24 (Oct. 21), 346, on which day (§1) Athan. returned from his second exile. This double coincidence cannot be an accident. It demonstrates beyond all dispute that the missing item of ‘ann. ii, mens. iv, d. xii’ has dropped out after ‘Treveris in Galliis,’ and that ‘mens. xc, dies iii’ relates to the second exile, so that, in §1 also, the annalist wrote not ‘annos vi’ but ‘annos vii menses vi dies iii,’ which he repeats §17 by its equivalent ‘mens. xc, d. iii,’ while words have dropped out in §1 to the effect of what is supplied in brackets. (Hefele, ii. 50, Eng. Tr., is therefore in error here).

I would add that the same obvious principle of correcting a clearly corrupt figure by the writer’s own subsequent reference to it, enables us also to correct the last figures of §2 by those of §5, to correct the items by the sum total of §§6, 7, and lastly to correct the corrupt readings ‘Gregorius’ for Georgius, and ‘Constans’ for Constantius, by the many uncorrupt places which shew that the annalist himself was perfectly aware of the right names.

In one passage alone (§13 ‘Athyr’ twice for Mechir, cf. Fest. Ind. viii) is conjecture really needed; but even here the consuls are correctly given, and support the right date.

We are now in a position to construct tables of ‘exiles’ and ‘quiet’ periods from the Historia as corrected by itself.

Table B. Exiles &c., of Athanasius.

N.B. In the above Table, (a) denotes dates or figures directly implied in the existing text, (b) those implied by it in combination with other sources, (c) those based on conjectural emendation of the existing text. All unmarked data are expressly given.

Table B shews the deliberate and careful calculation which runs through the system of our annalist. Once or twice he indulges in a round figure, exiles 1 and 5 are each a day too long by the Egyptian calendar, and this is set off by his apparently reckoning the fifth quiet period as two days too short. But the writer clearly knew his own mind. In fact, the one just ground on which we might distrust his chronology is its systematic character. He has a thorough scheme of his own, which he carries out to a nicety. Now such a chronology is not necessarily untrustworthy. Its consistency may be artificial; on the other hand, it may be due to accurate knowledge of the facts. Whether this is so or not must be ascertained partly from a writer’s known opportunities and capacity, partly from his agreement or discrepancy with other sources of knowledge. Now our annalist wrote in the time of Theophilus (385–412), and may therefore rank as a contemporary of Athanasius (cf. Prolegg. ch. v.) His opportunities therefore were excellent. As to his capacity, his work bears every trace of care and skill. He is no historian, nor a stylist, but as an annalist he understood what he was doing. As to agreement with other data, we remark to begin with that it was the publication of this fragment in the 18th century that first shed a ray of light on the Erebus and Chaos of the chronology of the Council of Sardica and its adjacent events; that it at once justified the critical genius of Montfaucon, Tillemont and others, against the objections with which their date for the death of Athanasius^37683768    But our annalist gives May 3, while Fest. Ind. gives May 2, the day solemnised in the Coptic Martyrologies (Mai, Script. Vett. vol. 4, part 2, pp. 29, 114), and doubtless the right one. Perhaps, if Athanasius died in the night of May 2–3, the former day might be chosen for his commemoration, while our annalist may still be literally exact. was assailed, and here again upset the confused chronological statements of the fifth-century historians in favour of the incidental evidence of many more primary authorities^37693769    See Tillem. viii. 719 sqq.. But most important of all is its confirmation by the evidence of the Festal Letters discovered in 1842, and especially by their Index, the so-called ‘Chronicon Athanasianum.’ It is evident at a glance that our annalist is quite independent of the Index, as he gives many details which it does not contain. But neither can the Index be a compilation from the annalist. Each writer had access to information not embodied in the other, and there is no positive evidence that either used the other in any way. When they agree, therefore, their evidence has the greatest possible weight. Their main heads of agreement are indicated in the Chronological Table, Prolegg. sub fin.

It remains to notice shortly the two digressions on the doings of Eudoxius and the Anomœans (§§2, 12 of Migne, paragraphs II, IX of Gallandi). Here the annalist is off his own ground, and evidently less well informed. In §2 we learn nothing of interest: but the ‘Ecthesis’ of the Anomœans in par. IX is of importance, and only too evidently authentic. It still awaits a critical examination, and it is not easy to give it its exact place in the history of the later Arianism. Apparently it belongs to the period 360–364, when the Anomœans were organising their schism (Gwatkin, pp. 226, 180) the names being those of the ultra-Arians condemned by the Homœans in 360 (Prolegg. ch. ii. §8 fin.).

The contrast between the vagueness of statement in these digressions, and the writer’s firmness of touch in dealing with Alexandrian affairs is most significant.

The fragment runs as follows: