Theran Ash in Minoan Crete: New Excavations on Mochlos

This work involves the cleaning of remains excavated by Richard Seager in 1908, but also includes for the first time since 1908 extensive new excavation in areas outside those previously excavated. In the course of these excavations, during the month of July, an important discovery was made bearing on the subject of Thera and the destruction of Minoan Crete.

In an area along the south coast of the island, which Seager designated as Block C of the Minoan settlement, a large quantity of volcanic ash was found in two rooms of a house that Seager had already excavated in part (Fig. 1, 2). The house is located directly on the coast and opens to the east in what must have been the eastern side of Block C. Seager believed that the Late Minoan levels in this block had been cleared away by the later Roman occupation and described the house in his 1909 report with this belief in mind: 'The easternmost rooms of this block were not destroyed but actually rebuilt by Roman settlers, who took part of the old house and made a new façade of small ashlar blocks on the eastern side, which was the front of this later house.' (Seager 1909, 284.) The 1989 excavations provide a much different picture of events.

Four trenches were opened in this area (E3 900, E4 9800, 9900, 10000), and walls belonging to Seager's 'Roman' house were soon found together with a large amount of roof tiles. These walls belong to a large rectangular house that extends much farther to the north where trenches were opened in areas undug by Seager. Two amphorae found on the floors of these rooms date the building to the Early Byzantine period, probably the 6th and 7th centuries AD, and the reoccupation here belongs not to the Roman period as Seager suggested but to a later time. To the south, in Trenches 900, 9900 and 10000, walls of this building were found to rest directly on top of earlier Minoan walls. The Byzantine building was divided into two rooms here, only the north-east corner of one room and the north-west corner of the other still preserved. The western room was provided with a low bench on its eastern side, and both rooms were paved with floors of green schist which also rest in part on Minoan walls. 65 to 70 cm below these floors the floor of the Minoan house was found. It is not true therefore that the Byzantine settlers rebuilt the earlier Minoan house; they built on top of it and probably destroyed the upper part of its walls and reused stones for building materials, but they did not penetrate to the Late Minoan floor level.

The Minoan house below also displays typically Minoan architectural features that neither Romans nor early Byzantines could ever have produced. The façade of the house is indeed built of 'small ashlar blocks', just as Seager reported, but these blocks are cut sandstone blocks that come from the Minoan quarry across from the island on Crete itself (Soles 1983). They are cut flat only in front and on top and bottom, then laid in place on top of a low levelling course or euthynteria against a packing of earth and rubble. It is characteristic Minoan construction that was not duplicated after the Bronze Age. The threshold which provided the main entrance to the house is a large flat slab of green limestone that has identical parallels in the threshold stones of the houses at Gournia. Immediately to the right as one enters the house, a U-shaped staircase is located that also has many good Minoan parallels. The bottom five steps of the staircase in the eastern cage were constructed in ashlar sandstone blocks, which have collapsed backwards; while the steps in the western return of the staircase were wood. The early Byzantine wall runs down the centre of this staircase and was probably responsible for displacing a number of the stone steps, one of which was found reused in a Byzantine level to the east of the house. To the west of the staircase there appears to have been a large paved room with a round column near its centre, the stone base for which was found still in situ. And to the west of this room lay a long narrow room and a wall that preserves the western façade of the house. The southern section of the house is destroyed by the modern beach line. The northern section was partly exposed in Trench 900 where a cooking hole was found in the bedrock floor of the house.

Seager excavated the entire western part of the house (Trench 9800) down to bedrock. He also excavated the entrance, the threshold of which was actually exposed at ground level, and the entry alcove just inside the entrance at the base of the staircase, but here only to a lower Minoan floor level. Sufficient pottery was found, however, in Trench 900 in association with the cooking hole and in the area between the column base and the western return of the staircase, where Seager did not excavate, to date the house to the LM IB phase (Fig. 3, 4, 5). This pottery includes fragments of Marine style including one, on the left of Fig. 5, with two tentacles of an argonaut, which is probably imported from Knossos, although the rest appear to have been locally made. It also includes many fragments of a spouted jug, found scattered in trenches 900 and 9900 (Fig. 3, 4), which belongs to what P. Betancourt calls the LM IB standard tradition (Betancourt 1985). Good parallels for its decoration are found on some IB vases from Palaikastro (Bosanquet and Dawkins 1923, 58, 59; Fig. 45, 46) and Pseira (Betancourt 1985, Pl. 18G).

Beneath the floor of the LM IB house a large quantity of volcanic ash was found in the western cage of the staircase and in the area between the column base and the western wall of the staircase (Fig. 6, 7, 8). This ash averaged 5 cm in depth; it rested on a stone pavement and in pockets between stones collected to a depth of 10 cm. It extends beneath the westernmost wall of the staircase, and the ash exposed in the excavation covered an area of c. 7.5 m². The actual area of preserved ash must be greater since it also extends beneath the eastern cage of the staircase. Originally, it probably also extended over the pavement in the alcove in front of the staircase where Seager must have removed it without commenting on it. It is unclear how far to the west it might have extended into Trench 9800. It was not found in the street east of the house (in Trench 10000), where it must have been washed away by rain shortly after deposition, nor was it found to the north in Trench 900 where the LM IB floor rests on bedrock and earlier floors must have been swept clean at the time of the IB rebuilding.

The ash has not been analysed yet although several bags were kept as samples which will be sent for analysis. It is clearly volcanic tephra, however, since it contains tiny particles of pumice and tiny glass crystals that sparkle in the sun, and contains no carbon; two pieces of pumice the size of golf balls were also found with the deposit, one with signs of burning. (Since the delivery of this paper a sample of the deposit has been identified by C. Vitaliano as Theran tephra and 'the product of the Minoan eruption.')

Beneath the ash another floor level was uncovered in these same areas; it lay c. 20 cm below the LM IB floor and extended out into the alcove in front of the staircase. This whole floor appears to have been paved with stone slabs and the ash rested directly on top of this pavement. At one point, midway between the column base and the westernmost wall of the staircase, a large quantity of pottery was uncovered in and around a built stone pit that had been sunk beneath the pavement. The pottery includes several cups, two of which are complete, as well as a jug and a ewer decorated with running spirals in a characteristic LM IA style with added red and white paint (Fig. 9, 10, 11).

In sum, the excavation of this house has uncovered an extensive layer of volcanic ash sealed between LM IA and LM IB floors in a house that was substantially rebuilt after the deposition of ash occurred. The plan of the earlier LM IA house on the site is unclear although it must have included the westernmost walls, the bases of which lie a good 40 cm below those of the easternmost LM IB walls, and probably included an open area to the east which permitted the ash to fall onto the paving slabs noted above. The house appears to have been damaged at the time of the ash fallout since the wall running east-west (north of the column base), which also belonged to it, lies somewhat askew and was in fact destroyed at its east end. The house was rebuilt then with a new east façade, with its threshold and ashlar masonry, with a new staircase resting directly on top of the ash, and a new column base which was sunk in a hole dug through the ash to set it firmly in place. This rebuilding must have occurred immediately, within the year, or the autumn and spring rains would have washed the ash away before it could be sealed beneath the new floor.

CONCLUSIONS

Until now, no real stratum of ash has been reported in Crete in an LM I context, and this has led some people to doubt that there was any (Barber 1987). The best evidence on Crete itself for a Late Minoan ash fall has been the ash reported by G. Cadogan from soil samples retrieved from destruction levels of Myrtos, Pyrgos (Cadogan and Harrison 1978), and by C.J. and D.B. Vitaliano from soil samples taken at Zakros (C.J. and D.B. Vitaliano 1974, 19). In each case, however, minuscule quantities of ash are involved and it is impossible to estimate the volume of ash fall. It has also not been possible to date the ash fall clearly to LM IA or IB. At Pyrgos Cadogan reports that it could have blown from Thera either in LM IA at the time of Thera's abandonment or in LM IB if the eruption occurred then. And at Zakros the precise contexts are still vague and those reported to be LM IA may well not be sealed contexts (Warren 1980). More recent evidence has been provided by Betancourt and Davaras from Pseira where small pieces of pumice have been reported in an LM IA context (Betancourt and Davaras 1988, 218). They appear to have been water worn, however, and are thought to have been washed ashore and carried up to the site. While they point to an LM IA eruption in Thera contemporary with LM IA in Crete, they might have survived from a much earlier eruption of the volcano, and they do not provide evidence for a general ash fall over the island.

The excavation of the Mochlos house demonstrates for the first time that there was an extensive ash fall over the island, as the research of Ninkovitch and Heezen indicated that there should be (Ninkovitch and Heezen 1965), and that it occurred while LM IA pottery was being produced. At the same time, it demonstrates that large quantities of ash from this fallout would have been preserved only under two conditions. The ash had to fall into an open area which then had to be covered up immediately to prevent the ash from being swept or washed away. If the ash fell on the roof of a house or in the street of a settlement or came through the windows of a house that continued to be occupied, it would not have been preserved except in trace amounts that would escape the archaeologist's eye. Since this is true, the precise date of the ash fall should be placed at or near the end of the LM IA phase. Some pottery which might be classed stylistically as IA is found alongside the newer styles above the ash in House 1, and the IB standard tradition is, as Betancourt has recently noted, very much a continuation of the earlier style. Nevertheless, House 1 dates to the IB phase, and the eruption of the volcano seems to mark the dividing point between LM IA and IB.

In addition, three important conclusions may be drawn about the eruption of the Santorini volcano and its effects on Crete.

1. Our picture of how the volcano on Thera actually erupted needs revision. Attempts to explain the apparent absence of ash in Crete by positing a lateral ejection of ash and pumice rather than a vertical ejection up into the air need to be revised (Pichler and Friedrich 1980; Barber 1987). The huge quantity of ash in Block C, House 1 did not wash ashore; it was wind-borne. The widespread fallout of ash to the south-east of Thera reported by Ninkovitch and Heezen was also wind-borne, and the estimates of Watkins and Thorarinsson of a 5 to 11 cm fallout of ash from Knossos to Zakros must be correct (Watkins et al. 1978; Thorarinsson 1978).
2. The eruption of the volcano on Thera occurred towards the end of the LM IA phase and was not responsible for the destruction of Minoan civilization. It is no longer possible to argue, for example, that there was a gap of many years between the original earthquake and abandonment of Thera and the final eruption of the volcano and that during this gap LM IB pottery was produced in Crete. Nor may it be argued that IB pottery was a rarity, manufactured only at the palaces, and therefore late to appear in outlying areas such as Thera. Recent scientific discoveries have indicated that the eruption of Thera occurred well before 1500 (see Chen 1989), and the archaeological evidence in Crete now confirms an early dating for the volcano's eruption.
3. The LM IA eruption of Thera did indeed cause some destruction in Crete, as Hood has suggested (1978), perhaps the result of associated earthquakes, but it was followed by a period of extensive rebuilding which occurred immediately and which often made use of architectural features once associated with 'palatial' dwellings. At Mochlos this included not only the immediate rebuilding of Block C, House 1, with a fancy new ashlar staircase, but also the rebuilding of the large house in Block D, as Seager noted (1909, 277), where a pier and door partition and a lightwell were introduced into the house plan. The sandstone quarry that provided the ashlar for these houses, and for the LM IB House of the Pillar Partition at Pseira (Betancourt and Davaras 1988, 211-213), also provided the ashlar for the palace at Gournia and the remodelling of this palace probably also occurred now and testifies to the continuing vitality of the region. The eruption of the Santorini volcano may have had an adverse psychological impact on Minoan Crete and may have weakened its hold on the Cyclades, but it marked a period of renewal and continuing prosperity in Crete itself.

The duration of this period, usually allotted no more than 40 to 50 years, is problematic. It ended sometime between 1500 and 1450 BC, but it began earlier than previously thought, perhaps as early as the late 17th century if the scientific evidence is correct. In this case, the LM IB phase would have extended over a much longer period of time and the brilliant achievements of the period, when Minoan civilization reached its zenith, would not have been quite so ephemeral.

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

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