[In the following excerpt, Blakeway argues that certain conclusions about the chronology of Archilochus are erroneous because they are based on the solar eclipse of 648 b.c. rather than the solar eclipse of 711 b.c.]

I. THE ASTRONOMICAL EVIDENCE

The external literary evidence for the chronology of Archilochus is as follows:

Cicero places him in the lifetime of Romulus. (Traditionally 753-716 b.c.)

Clement, arguing from the foundation date of Thasos, dates his fame from c. 700 b.c.

Eusebius, in the Praeparatio Evangelica, Tatian, Cyril place his floruit about the Twenty-third Olympiad, 688-685 b.c.

Nepos places his fame in the reign of Tullus Hostilius. (Traditionally 671-640 b.c.)

Eusebius, in Jerome's version of the Chronici Canones, notices him against the first year of the Twenty-ninth Olympiad, 664-663 b.c. (665-664 b.c. Armenian version.)

Herodotus makes him the contemporary of Gyges. (? c. 652 b.c. according to the Assyrian Chronology.)

Since the identification of the eclipse of Archilochus with that which was total in Paros on April 6th, 648 b.c., much of this evidence has been disregarded as valueless. With a fixed point so firmly established by an exact science, Hellenistic, Roman, and early Christian chronologies could be treated with the scant respect they deserved. The 648 b.c. date stood firm on a fourfold foundation; Astronomy, the inscription of Assurbanipal, Herodotus, and Archilochus himself. It was one of the few certain dates of the eighth and seventh centuries, and to it all reconstructions of that period had to conform.

This creed I found fairly satisfying until I was forced into scepticism by (i) a study of Greek Colonization of the West and the dates of the Sicilian colonies, (ii) a study of the so-called ‘Lelantine War’, (iii) the date of the colonization of Thasos as given by Xanthus and the close connexion of Archilochus with that colonization implied by Clement of Alexandria and Oenomaus of Gadara.

I shall return to these points later. For the moment it is only necessary to state that in all three inquiries the 648 b.c. date for Archilochus seemed to me inconsistent with the majority of the best evidence, and that all three seemed to demand that part of Archilochus' lifetime should fall within the eighth century.

For this last there was indeed fairly good literary evidence, and further, with the exception of the reference to the eclipse, there was nothing in the extant fragments of Archilochus inconsistent with such a date. The eclipse, however, was not to be denied, and for long I thought that the 648 b.c. date must be retained even at the cost of discarding some of the most reliable dates in Thucydides. It was only when I found that the area of totality of the eclipse of 711 b.c. had last been calculated before the computations of Fotheringham, Schoch, and Neugebauer that I considered the possibility of the eclipse of that year being the eclipse of Archilochus and appealed to Dr. Fotheringham¹ to make the necessary calculations in my behalf.

I quote from two letters written to me in the summer of 1933.

I

‘I have now computed with the latest elements, i.e. Schoch's last published elements with the correction to the Moon's node which he sent me in a letter shortly before his death, the total eclipse of the sun of 711 b.c. March 14th.

‘But I find that this last small correction does not affect the result. I have in fact worked through the computation for the elements and the details for Thasos, substituting the last published corrections wherever they affect it. So the final result is not dependent on any unpublished figures.

‘I find that the eclipse was total in Thasos. I have computed for 40° 40′ N. 24° 20′ E., but it would be total throughout Thasos. I find for the middle of totality 10:16 a.m. local solar time, which is rather nearer to mid-day than the time when the eclipse of 648 b.c. April 6th was total at Paros.

‘I find that the eclipse of 711 b.c. March 14th attained a magnitude of 11·46 digits at Paros.

‘Totality is reckoned at 12 digits. This means that 95.5 per cent. of the sun's diameter was eclipsed. The uneclipsed crescent would have a width amounting to 4.5 per cent. of the sun's diameter—a thin crescent.

‘So if you have other reasons for regarding 711 b.c. March 14 at Thasos as a possible time and place, the astronomy supports you.’

II

‘I find that the last printed corrections to the position of the Moon's node at which I arrived by so curious an accident, and which were issued by Schoch on a fly-sheet and afterwards published by Neugebauer, most fully in his Astronomische Chronologie (1929) I, 133, give as magnitude of the solar eclipse of 711 b.c. March 14 for Paros (37° 2′ N. 25° 11′ E.) 11·47 digits and for Thasos (40° 40′ N. 24° 20′ E.) 12·15 digits, where anything amounting to 12·00 or over is total. The small correction which Schoch communicated to me in a letter, but which he did not live to publish, changes these to 11·46 and 12·14 respectively. These elements were verified by Schoch over a long series of eclipses, but he did not live to publish the verification. The elements, as printed, were also verified by Neugebauer in Astronomische Nachrichten, Band 8, Nr. 2 (1930), pp. B 24-B 32.’

The eclipse of 648 b.c. was total at Paros, that of 711 b.c. total at Thasos and not far from total at Paros.

At the least there is nothing to choose between the two eclipses astronomically. In the words of Dr. Fotheringham, ‘It is rather distressing that in spite of all our refinements astronomy should still offer this dual solution’. At the most the eclipse of 711 b.c. approaches more nearly to the description of Archilochus than that of 648 b.c. To quote Dr. Fotheringham again, ‘I find for the middle of totality at Thasos in 711 b.c., 10.16 a.m. local solar time, which is rather nearer to mid-day than the time when the eclipse of 648 b.c. April 6th was total at Paros.’ If anything astronomy supports the 711 b.c. date. 648 b.c. can no longer be regarded as a certain and fixed point in the lifetime of Archilochus. Ultimately our choice between the two can only be determined by an examination of the extant evidence, internal and external, for the period of his life. The majority of this evidence seems to me to point to the earlier date, but I propose to examine it in detail before attempting a reconstruction. …

X. CONCLUSION

We are now in a position to put forward a tentative reconstruction of the chronology of Archilochus' life.

c. 740-730 b.c. Conjectural date of Archilochus' birth.

734 b.c. Colonization of Syracuse (Episode in the Colonization mentioned by Archilochus, Athenaeus, iv. 167 d).

c. 720 b.c. Colonization of Thasos, Dionysius (Clem. Alex. Strom. i, p. 333 b).

716 b.c. Traditional date of end of reign of Romulus (752-716 b.c.), Cicero, Tuscul. Quaest. i. i. 3, refers Archilochus to his reign. Archilochus would be aged 14-24 in 716 b.c.

711 b.c. Total Eclipse of the Sun throughout Thasos. Archilochus aged 19-29.

c. 708 b.c. Colonization of Thasos, Xanthus (Clem. Alex. Strom. i, p. 333 b). Archilochus aged 22-32.

Archilochus' share in the colonization stated by Oenomaus ap. Euseb. Ev. Praep. vi. 7, and implied by Clem. Alex. Strom. i, p. 333 b. Both authors used a lost poem of Archilochus?

c. 700 b.c. Floruit of Archilochus according to Clem. Alex. Strom. i, p. 333 b. Archilochus aged 30-40.

c. 700 b.c.? ‘Lelantine War’ mentioned by Archilochus with indication that it was in progress or about to take place. Fragment 3, Diehl.

c. 700-c. 690 b.c. Fall of Magnesia. Dated to soon after c. 700 b.c. by Clem. Alex. Strom. i, p. 333 b, and to well before the accession of Gyges by Pliny, N. H. xxxv. 8 and vii. 38. Fragment 19, Diehl.

688-685 b.c. Floruit of Archilochus. Euseb. Ev. Praep. x. 11. 4; Cyril, c. Jul. i. 12; Tatian, Ad Graecos 122. Archilochus aged 42/5-52/5.

687 b.c. Conjectural date of accession of Gyges.

Before c. 670 - c.660 b.c. Archilochus slain in battle.²

The date chosen for Archilochus' death needs some further explanation.

According to a tradition preserved in Heracleides, Plutarch, and Dio Chrysostom,² Archilochus was slain in battle. If this is accepted he cannot have been an old man, say more than sixty-five, at the time, and in accordance with the evidence we have so far considered his death would fall before the decade 670-660 b.c.

This involves the rejection of the date given for Archilochus by Cornelius Nepos,³ who refers him to the reign of Tullus Hostilius—671-640 b.c. according to the traditional chronology—and the mention of him in Jerome's version of Eusebius under the year 664-663 b.c. (665-664 b.c. Armenian Version). The rejection of these two authorities, Nepos especially, is undoubtedly a serious matter and not to be undertaken lightly, but it is a far smaller sacrifice of evidence than that which is involved in the acceptance of the 648 b.c. eclipse. If 648 b.c. is the date of the eclipse of Archilochus we must discard the evidence of Cicero, Tatian, Cyril, Oenomaus of Gadara (probably based on Archilochus himself), and Clement of Alexandria (based on Xanthus, Dionysius, and possibly Archilochus himself). We must assume that Archilochus' reference to Aithiops the Corinthian was to a story at least sixty to seventy years old, that Clement was mistaken in his dating of Archilochus' reference to the fall of Magnesia, and that Pliny's story of Candaules and Bularchus' picture of that event is a hopeless chronological confusion, that Archilochus played no part in the colonization of Thasos, that indeed he was not born at that time, and that his reference to the ‘Lelantine War’ is to a struggle in the second half of the seventh century.

The difficulties involved in such a wholesale sacrifice seem to me enormous—in spite of the fact that it has so frequently been made. As long as we were forced to accept the 648 b.c. eclipse as the eclipse of Archilochus this sacrifice was necessary. Now, thanks to Dr. Fotheringham, it is no longer so. The literary evidence must be examined again without that unconscious bias in favour of the 648 b.c. eclipse which has so far influenced it.⁴

It may be that I have stated the case for 711 b.c. too emphatically. If that is so I can only plead that the case for 648 b.c. is habitually overstated and much of the relevant evidence tacitly omitted. I ask for a further examination of the evidence in the light of Dr. Fotheringham's computation.

Notes

1. I am indebted to Dr. Fotheringham for the information which follows, which in itself contains whatever value this discussion may possess, and also for much kindness and interest in my approach to the problem.

2. Heracleides, Pol. viii; Plut. Ser. Num. Vind. 17; Dio Chrysostom, 33, p. 397 m.

3. Cornelius Nepos ap. Aul. Gell. xvii. 21. 6 ‘Postea Pythagoras Samius in Italiam venit, Tarquini filio regnum optinente, cui cognomentum Superbus fuit, isdemque temporibus occisus est Athenis ab Harmodio et Aristogitone Hipparchus, Pisistrati filius, Hippiae tyranni frater. Archilochum autem Nepos Cornelius tradit Tullio Hostilio Romae regnante iam tum fuisse poematis clarum et nobilem.’

   Nepos' statement is logically capable of two interpretations. Archilochus was already famous before 640 b.c. or he was already famous in 671 b.c. (or soon after). Historically the second interpretation is far more probable. The first interpretation would mean that Nepos, alone among ancient writers, placed the greater part of the life of Archilochus after 640 b.c. The context in Aulus Gellius gives no help. It merely indicates a time very much earlier than the last-mentioned event.

4. The ‘modernity’ of Archilochus, the individual character of his poetry and its stylistic development, may appeal to some people as evidence for the later rather than the earlier date. To this I can only reply that Archilochus was a genius whatever his date, and that, to me, an innovating literary genius at the end of the eighth and in the first quarter of the seventh centuries is no more than what I should expect from the efflorescence of Greek Art in Corinth, Crete, and Ionia at that time.