Traces of the Eruption Outside Thera

Damage or dislocation for which there is evidence in northern Crete and in some of the Aegean islands in Late Minoan I A may be due to causes -earthquakes, shock-waves, tsunamis, tephra-fall - linked with the eruption, but for the moment at any rate this remains uncertain. The spiritual and psychological impact of the eruption, however, on peoples of surrounding areas must have been tremendous. Some Late Minoan I A votive deposits in Crete appear to reflect this.

Much useful evidence bearing on the question of possible effects of the Bronze Age eruption of Thera on surrounding areas has been collected in recent years, notably by the authors of Acta 1971. Cf. Ninkovich and Heezen 1965; Page 1970; Pomerance 1970; Vitaliano 1973.

The idea that the eruption had effects in Crete appears to have originated with Sir Arthur Evans. As a result of his work at Knossos in 1922 he concluded that the earthquake, which in his opinion caused havoc there at the end of the Middle Minoan III period, might have been connected with the eruption (Baikie 1926, 289 f.). But by the time he addressed the Society of Antiquaries in London in October 1922 Evans had realised that the "native wares" from the eruption horizon on Thera "were executed under a strong Cretan influence of the early part of the First Late Minoan Period", that is, of Late Minoan I A (Evans 1922, 328; 1928, 312 f. note 1). This had become clear to him from a study of the pottery found in the Thera excavations of the 1860s, much of which was actually published for the first time by Renaudin (1922) that year. In spite of this the idea of a connection between the eruption of Thera and the Middle Minoan III B destruction at Knossos was elaborated in the Rhind Lectures the following year by Hall (1928, 109, 163).

Evans also drew attention to the damage caused by a severe earthquake at Knossos in Late Minoan I A, the period to which he assigned the eruption of Thera; but for reasons which it would be interesting to know he does not appear to have been attracted by the idea of a possible connection between them. It was the late Professor Marinatos who in the 1930s began to revive old and develop new theories about damage in Crete as a result of the eruption. Marinatos excavated several Cretan sites destroyed in Late Minoan I, and concluded that their destruction was due to causes linked with the eruption of Thera. Inland sites like the settlement at Apodhoulou in the Amari valley he thought might have been wrecked by earthquakes connected with it, but sites on the northern coast of Crete like Amnisos he suggested could have been destroyed by the impact of tsunamis. Marinatos was the first Aegean archaeologist to realise the immense powers of destruction of tsunamis and the possibility that the eruption gave rise to them. He eventually published his views in a classic article in the English periodical Antiquity (Marinatos 1939).

In this article Marinatos assumed - wrongly, but in accordance with a belief prevalent at the time, although Evans had long since rejected it - that the Late Minoan I A and Late Minoan I B styles of Cretan pottery were contemporary; and he therefore equated the horizon of destruction in Crete at the end of the Late Minoan I period, in Late Minoan I B, with the eruption of Thera, which he dated like Evans c. 1500 B.C., that is in Late Minoan I A. Excavations in Crete since the Second World War have vindicated the opinion of Evans that the Late Minoan I A and Late Minoan I B styles reflect successive periods of time. It is now generally agreed that the pottery from the settlements buried by the eruption on Thera is assignable to Cretan Late Minoan I A, datable in rough terms c. 1500 B.C., while that from the well-marked horizon of fire-destruction in Crete belongs to Late Minoan I B, conventionally dated some fifty years later c. 1450 B.C. Attempts have been made to save the Late Minoan I B destructions for the eruption by arguing that it took place in two or more stages separated by intervals of years.

But this appears to be scientifically improbable, and in any case the character of the Late Minoan I B destructions in Crete makes it difficult to believe that they were due to natural causes, whether earthquakes, tsunamis, or shock-waves, or a combination of these.

1. The Cretan sites destroyed in Late Minoan I B were scattered throughout most parts of the island. Some were on the coast, others like Sklavokampos were tucked away in upland valleys.
2. In almost every case the destructions were accompanied by fire. Earthquakes may cause fires, but are unlikely to do so on such an universal scale. Tsunamis are of course even more improbable as agents of fire-destruction. It is true that at some of the sites involved the destruction was only partial. Thus at Knossos several houses in the town were destroyed in Late Minoan I B, while the palace seems to have largely escaped damage. At Mallia on the other hand the palace and the houses in Block Zeta immediately to the east of it were destroyed by fire, but the large house in Block Epsilon in the southern part of the town appears to have gone unscathed. Similarly at Mirtos in southern Crete the little "country house" on top of the hill was destroyed, although houses round it were spared.
3. Some ot the sites destroyed or abandoned in Late Minoan I B do not appear to have suffered damage in the kind of way that suggests natural causes. The case of the little town on Pseira is instructive in this regard (Seager 1910). The area of the town was deserted after the end of Late Minoan I B until late Roman times. The excavator, Seager, does not explicitly say so, but the evidence suggests that the town was destroyed by fire like neighbouring Mochlos and Gournia (Mountjoy 1972, 128 note 24). But when Seager excavated Pseira he was impressed by the remarkable state of preservation of the walls of the houses there. The exceptional state of preservation of the house walls is also clear from contemporary photographs. Such a degree of preservation is difficult to reconcile with any theory of destruction by natural causes such as earthquakes, shock-waves, or tsunamis.
4. Human bodies or treasures of gold and silver have rarely been found in the buildings and settlements destroyed in Late Minoan I B. Remains of human bodies have only so far been recorded at one Cretan site, Mochlos, but they were apparently found there in considerable numbers. The implication is that there was usually time for people to escape and to remove the most valuable objects from the buildings and settlements before they were destroyed.
5. Destruction levels assignable to the Late Minoan I B period have been recognised on some of the other islands, at Kastri on Kythera, Ayia Irini on Kea, and Phylakopi on Melos. But there is no evidence for a comparable horizon of destruction at this time on the Greek mainland.

In spite of assertions to the contrary (e.g. Page 1970, 12) this horizon of destruction in Crete in Late Minoan I B has all the appearances of a destruction caused by war. This is in accordance with the views of the excavators like Seager and Harriet Boyd who first explored Cretan sites destroyed in Late Minoan I B in the early 1900s. The evidence from these sites, virtually all of them ravaged by fire, provides in effect a model of what one might expect to find in the way of material traces of a war of conquest of that period in the Near East. It might almost serve as an archaeological commentary on a description like that in the Book of Joshua of the conquest of the Promised Land. The evidence for the presence of rulers from the Greek mainland at Knossos immediately afterwards is in harmony with the view that the horizon of fire-destruction in Crete in Late Minoan I B reflects a war of conquest. In any case both the archaeological and the scientific evidence appear to favour the idea that the eruption of Thera began and ended within a relatively short space of time, and within the limits of Cretan Late Minoan I A making the Late Minoan I B destruction horizon in Crete subsequent to it.

If the eruption of Thera took place during the Late Minoan I A period, can any traces of it be identified in surrounding areas? The kind of traces which it might be possible to detect in the archaeological record may be grouped under four headings:

1. Damage by earthquakes. It is always possible, although by no means certain, that earthquakes felt in surrounding areas including Crete were in some way linked with the eruption.
2. Damage by shock-waves at the time of the eruption. These are unlikely to have been responsible for damage at any great distance from Thera.
3. Damage at coastal sites by tsunamis. It is not certain, although quite possible, that tsunamis consequent upon the eruption afflicted some areas.
4. Damage and dislocation owing to the fall of tephra. Ninkovich and Heezen (1965, 443) estimated that the eruption deposited a layer of tephra over 10 cm. thick over all the Aegean islands around Thera, including the central and eastern parts of Crete. This tephra layer may have been thick enough to kill vegetation and impede agriculture for a period of months or even years. But in a mountainous island like Crete subject to heavy rains in the autumn and spring it is possible that the bulk of the tephra which fell was dispersed within a comparatively short space of time.

The areas closest to Thera and most likely to have been affected by the eruption are:

A)     the southern part of the Greek mainland;

B)     the islands of the Cyclades and other neighbouring islands; and

C)     Crete.

A - The Greek mainland.

No effects of the eruption appear to have been identified as yet in the archaeological record anywhere on the mainland. The lumps of pumice recovered at Nichoria in Messenia were evidently water-borne, and had been carried to the site by the inhabitants some time after the eruption (Rapp, Cooke & Henrickson 1973).

B - The islands.

At Ayia Irini on Kea there is a horizon of pottery including some complete vases assignable to the period of Cretan Late Minoan I A (Caskey 1972); but no major destruction of the site appears to have occurred then. Similarly at Phylakopi on Melos about 100 km. west of Thera City II came to an end during the Cretan period Middle Minoan III B and City III suffered a destruction of some kind at a time when imported Late Minoan I B vases were in use, but there seems to be no evidence for a major cataclysm in Late Minoan I A.

C - Crete.

There is evidence of extensive damage and dislocation in Late Minoan I A at some of the sites in the northern part of the island, which would have been most exposed to any effects of the eruption. Some or all of this damage and dislocation could have been due to causes (earthquakes, shock-waves, tsunamis, tephra-fall) connected with the eruption of Thera.

1. Knossos. The palace and the city round it were severely damaged by a great earthquake in Late Minoan I A. This horizon of destruction is one of the few at Bronze Age Knossos not accompanied by traces of fire.
2. Vathipetro. The important country house excavated by Marinatos here was finally destroyed by fire it seems in Late Minoan I B. But the preliminary reports of the excavations suggest that the large storeroom with a couple of pillars may have been filled with debris after damage to the house (by earthquake?) in Late Minoan I A. The clay vases illustrated trom this room appear to be Late Minoan I A rather than Late Minoan I B in date (Marinatos 1949, 107 fig. 8; 1950, 248 f. figs. 10 -11; 1952, 600 ff. figs. 10 - 16).
3. Katsamba. A horizon with pottery assignable to Late Minoan I A was recognised here by Alexiou (1955) who suggested that the destruction responsible for it might have been connected with the eruption of Thera c. 1500 B. C.
4. Mallia. There is no evidence for a destruction involving the whole site in Late Minoan I A. But a deposit in the eastern part of House Epsilon south of the palace is assignable to that period (Pelon 1970). This is phase III A, represented by couche 6 (which is couche 2 in coupes 3 -5). The relative thickness of the deposit and the number of complete vases recovered from it suggest that it reflects quite extensive damage to this part of the building in Late Minoan I A. The deposit is shown in the published sections as consisting of "terre et charbon", but it is not altogether clear that the "charbon" resulted from a destruction by fire. The latest pottery from the "Maison de la Plage" in Area Theta (Effenterre 1976) might go down into Late Minoan I A, and a building like this by the sea could have been seriously damaged by a tsunami if this stretch of coast was afflicted by one at the time of the eruption of Thera.
5. Mochlos. Seager (1909, 293 cf. 290 f.) noted that part of the large and pretentious House D which was destroyed by fire in Late Minoan I B had been "cut into rooms belonging to the earlier part of Late Minoan I". But this does not necessarily imply damage to the town in Late Minoan I A.
6. Pseira. Seager (1910) speaks of rebuildings in the little town here in the Middle Minoan III and Late Minoan I periods. He also refers to "extensive rubbish heaps found among the Late Minoan I houses" with "a kind of ware which immediately precedes the typical Late Minoan I style of pottery" (Seager 1910, 20). He called this ware Middle Minoan III to "distinguish it from the later Late Minoan I wares of which it is the forerunner". But he notes: "This style of pottery may be said to occupy an intermediate stage between the true Middle Minoan III ware of Knossos and the typical Late Minoan I fabrics, and cannot truly be said to belong to either class" (Seager 1910, 21). It is just possible that the "rubbish heaps" with this pottery noted by Seager reflect some major catastrophe which overwhelmed the town in Late Minoan I A.
7. Gournia. The excavator, Harriet Boyd, had the impression that many of the houses in the town had gone out of use at a distinctly earlier period in Late Minoan I than the final destruction by fire assignable to Late Minoan I B. Some of the published vases appear to support this view, notably those from the great cache of ritual vessels found in C 58 in House Cm on the eastern slope of the hill below the palace (Boyd Hawes 1908, 40 pl. vii nos. 25 - 32, 34 - 41). One of these vases has indeed been described (Popham 1967, 339 pl. 80, d) as belonging to the Late Minoan I B period, but this seems difficult to accept in view of the consitently Late Minoan I A character of the decoration on it and on the other vessels of the group, which was one of the largest and most important of any period recovered at Gournia. Harriet Boyd herself was certainly convinced that the eastern region of the town where this cache of ritual vases was found "had an older tinge than the West Slope or the crest of the Acropolis" with the palace (Boyd Hawes 1908, 39).
8. Palaikastro. There is some evidence for dislocation and subsequent rebuilding in the town here during Late Minoan I A. In Block Chi, for instance, below the site of the later Greek temple, floor deposits with complete vases assignable to Late Minoan I A were uncovered in a gap between two later houses (Dawkins 1904 - 05, 272 ff., Bosanquet 1923, 27 f.). The house in which these vases were buried does not appear to have been destroyed by fire, - but it was never rebuilt.
9. Zakro. The reports of the recent excavations suggest that there was damage of some kind in the town here in Late Minoan I A.

There is thus a good deal of evidence for destruction or dislocation of one kind or another at sites in the northern part of Crete during Late Minoan I A.

At Knossos at any rate damage in Late Minoan I A was clearly due to an earthquake of quite exceptional severity. Such an earthquake could have been connected in some way with the eruption of Thera, but this is not certain. The exact nature of the damage, and in some cases the date of it, at other sites are less clear. No positive evidence has yet been recognised linking any of this damage with the eruption in an indisputable manner.

Evidence for material damage caused by the eruption outside Thera is therefore at present at any rate somewhat ambiguous. The great earthquake which afflicted Knossos in Late Minoan I A mayor may not have been connected with it. As to shock-waves and tsunamis, it appears to be an open question whether, in spite of the scale of the eruption, any were generated sufficiently powerful to cause serious damage in surrounding areas. There was undoubtedly a heavy fall of tephra, especially in the neighbouring islands and in the central and eastern parts of Crete; but it is not yet known how deep the tephra lay, or how lasting the damage done by it was.

One thing, however, can be inferred with a high degree of certainty. The impression made by the eruption, its spiritual and psychological impact, upon the population of surrounding areas must have been tremendous. Even the modest eruption of the nineteenth century was spectacular enough to attract comment from an English journalist engaged in fighting on the side of the Cretans in revolt against the Turks in 1867. He describes the smoke by day and the glow of fires on the horizon by night as seen from the coast of Crete west of Fodele (Skinner 1868, 146f.). The impact of the immense Bronze Age eruption on the Cretans of that time can hardly be exaggerated. It would not be surprising therefore to find instances of cult in Crete reflecting the eruption. One highly probable and two possible examples of such cult are indeed to hand.

1 - Zakro Votive Pits.

These two pits on the edge of the Bronze Age town were first noted by Halbherr in 1892 and were cleared by Hogarth (1900 - 01). Hogarth recovered vast quantities of pottery from the pits and observed that there was no sign of any stratification in them. It seems agreed that virtually all the pottery from the pits was assignable to one period, Late Minoan I A, and the pottery appears to have been remarkably homogeneous: the bulk of it consisting of cups of the Vapheio shape with an overall dark wash and decoration in white, and bowls with designs in dark on light, notably floral patterns and tortoise shell ripple; very little plain domestic ware was recovered. Hogarth (1900 - 01, 127f.) deduced that the contents of the pits were votive offerings cleared from a shrine. Dawkins (1903, 260) suggested that "The shrine connected with the pits must have been the sacred place of the town". The homogeneity of their contents, however, is against the idea that the pits were filled with offerings which had accumulated over a period of years in a shrine. It rather suggests that they were filled upon the occasion of some unique act of public or official devotion. What could be more calculated to inspire such an exceptional demonstration of piety in a relatively primitive people than the rain of tephra and the obliterated light of day, to name only two of the effects attendant upon the great eruption in this part of Crete?

2 - Palaikastro Block Pi.

Here immediately across the street from Block Chi, and like it below the later Greek temple area, debris from a Bronze Age sanctuary assignable to the period of the Zakro pits, that is to Late Minoan I A, was identified in the excavations of 1902 - 06 (Dawkins 1904 - 05, 287 f.; Bosanquet 1939 - 40, 67.; Rutkowski 1972, 327.; Hood 1977, 167 f. pl. C fig. 1). This may have been the site of a public sanctuary, if not of the chief one in the town, throughout the Bronze Age. As Dawkins (1904 - 05, 287) noted: "Viewed in connexion with the building of the" later Greek "Temple on almost exactly the same spot, these remains are significant for the traditional sanctity of the place, showing it deeply rooted in the pre-Hellenic past". But there seems to have been something of an upsurge of cult activity here in Late Minoan I A. This did not escape the notice of the excavators of 1902 - 06, and they were aware of the comparison with the votive pits at Zakro. Commenting on the fact that nothing later than Late Minoan I A as recovered from this sanctuary debris or from the Zakro pits Bosanquet (1923, 21) remarked: "This may be mere coincidence, or may point to some simultaneous purification or rebuilding, in which accumulated offerings at Zakro were removed to grottoes" (i.e. the two pits), "at Palaikastro buried beneath a higher floor level", referring to houses destroyed in Late Minoan I B of which scanty traces were noted in the area.

The pottery from Block Pi at Palaikastro seems to be assignable to Late Minoan I A, but the excavators had the impression that it represented a steady accumulation over a period of years. They kept it separated in four arbitrary layers during excavation and thought that the material from the lowest of these corresponded with that from the pits at Zakro. But 50 % of the pottery from this lowest layer was decorated with designs in light on dark in the Kamares tradition, while Hogarth had remarked that four fifths of the pottery from Pit 1 at Zakro was referrable "to the best Knossian Mycenaean period", that is to say, it was decorated in dark on light, leaving only one fifth or 20 % with light on dark Kamares style decoration. This equates better with the percentage of 26 % and 5 % of light on dark ware reported from the two upper levels of the Palaikastro deposit. At all events this deposit in the Bronze Age sanctuary in Block Pi appears to have come to an end in Late Minoan I A, which taken in conjunction with other evidence from the town may reflect some kind of serious dislocation of life there during that period.

3 - Nirou Khani.

The votive deposit discovered by N. Platon in 1946 during the course of repairs to the large country house or villa here is more suggestive of a specific connection with the eruption of Thera (Platon 1947, 636; 1954, 449 f.). The deposit was in a little walled enclosure like a box under what appears to have been the threshold of a doorway which had once connected the two rooms (7, 7a) set apart as a shrine. The door itself appears to have been blocked at the end of Late Minoan I A. The deposit consisted of "hundreds" of little conical clay cups of the type characteristic of both phases of Late Minoan I, and most of them held a lump of pumice. Platon (1947, 636) immediately suggested a connection between this cult and the eruption of Thera c. 1500 B. C.

CONCLUSIONS

The eruption of Thera appears to have begun and ended within the Cretan period of Late Minoan I A. There is no real certainty as yet as to the character and extent of damage in Crete and other neighbouring areas due to causes (earthquakes, shock-waves, tsunamis, tephra-fall) connected with the eruption. Such damage as there was may have been repaired or obliterated relatively soon afterwards by the efforts of the surviving inhabitants. Houses and settlements may have been abandoned for a time in areas where the tephra lay deep, but no definite evidence for this has yet been recognised. The spiritual and psychological impact of the eruption, however, must have been immense throughout the Aegean area and is likely to have left traces in religious cult. Such traces have indeed been noted in Crete.

FUTURE WORK

1. It would be interesting to know for certain if the pumice in the votive deposit at Nirou Khani was from the Bronze Age eruption of Thera (cf. Rapp & Cooke 1973, Rapp, Cooke and Henrickson 1973).
2. A systematic programme of borings from the sea-bed might give more definite evidence for the depth of the tephra-fall on islands near Thera and in different parts of Crete (cf. Ninkovich & Heezen 1965).
3. Further excavation attended by careful observation in levels of the Late Minoan I A period, especially in islands near Thera and in the eastern parts of Crete, may throw new light on the problem of damage and dislocation caused by the eruption.

---------------------------------------------