Archaeological Observations at Akrotiri Relating to the Volcanic Destruction

There was at least a small village on this site in the Late Neolithic, as attested by the scattered pottery sherds of that period. The development of this village into an Early Cycladic settlement is confirmed by both the moveable finds - pottery, marble vases and figurines etc. - and the architectural remains and structures. There is similar evidence for the smooth transition to - and development of - the Middle Cycladic town, even though it is still difficult to determine its extent. This urban center must have achieved a not inconsiderable economic acme towards the end of the MC period, when a major destructive earthquake seems to have occurred. Enormous quantities of debris from the ruins were arranged in the streets and squares or contained within special structures. Although many of the MC buildings were rebuilt, others were demolished completely, considerably altering the town's layout. In those buildings which were rebuilt the ground floor was transformed into a semi-basement, due to the debris which, as mentioned above, had been deposited in the streets and is up to two meters deep in places. Within this debris - which has yielded MC and early LC I pottery - the drainage conduits of the LM IA city were laid. This city, which emerged from the ruins of the prosperous urban center of MC Akrotiri, was at its zenith when it was destroyed by the volcanic eruption.

In the course of two decades of excavation at the site the sequence of phases of this catastrophic event may be outlined as follows:

1. Minor earth tremors warned the inhabitants who, taking their precious possessions with them, vacated the city. Thus the absence of victims and precious vessels and objects from the excavated buildings is explained.
2. The slight tremors were followed by more severe earthquakes, causing serious damage to the buildings, the ruins of which are frequently found beneath the layers of volcanic material, indicating that they were destroyed prior to the eruption.
3. It seems that a period of quiescence followed the severe earthquakes, during which the inhabitants, or at least special work teams, were sufficiently reassured to return to the city and set about demolishing the rickety buildings, opening up pathways through the ruins and perhaps commencing repairs. Substantial evidence and indications of such activities have been found.
4. It seems that the eruption of the volcano commenced during this phase of 'rebuilding' activity. In several places works were unfinished while tools have been found hastily abandoned on top of piles of rubble, not underneath them. The eruption must have begun with some warning signs (e.g. emission of gases, visible smoke etc.) since once again those working in the ruined city were able to abandon it, without victims. That the time interval between the commencement of clearing operations and the onset of the eruption was brief is borne out by the following: a) there are no traces of erosion on the exposed standing walls of the ruins and b) no accumulation of earth in the narrow spaces between the buildings has been observed, despite the fact that they are located on easily eroded mounds of rubble. The succession of the phases and paroxysms of the eruption is graphically recorded in the stratigraphy of volcanic deposits.
5. A thin layer (2 - 3 cm thick) of fine pumice covers the entire island, constituting a terminus ante quem for all human activity. At no point in the excavation has violation of this thin layer been observed. It is difficult to estimate the time lapse between the creation of this layer and the subsequent phase of the eruption. There are, however, indications that this was long enough for a considerable quantity of fragments of basaltic stones, erosion products (?), to accumulate within the cone of the crater. These fragments are found in the lower layers of pumice, created by the following paroxysms.
6. The lowest layer (10 - 15 cm thick) comprises of pea-sized pellets of pumice and a large quantity of fragments, in a ratio of approximately 50/50
7. The immediately overlying layer (15 - 20 cm thick) is of pumice of slightly larger diameter and a significantly lesser proportion of basalt fragments.
8. The proportion of basalt fragments is even further reduced in the succeeding layer (25 - 30 cm thick), though the size of the pumice pieces is about the same as in the underlying layer.
9. The topmost layer of pumice (50 cm thick and more) is of much larger pieces and is entirely free of basalt fragments. We are of the opinion that the gradual decrease in quantity of basalt fragments in the upper layers of pumice indicates that these already existed in the cone of the crater when the paroxysms began. With each new paroxysm the quantity of fragments diminished until, finally, only pumice was ejected. The position of these fragments at the bottom of each layer also indicated the mechanism by which they were deposited; they were expelled together with the pumice, but being heavier material fell to the ground first. The absence of any traces of erosion shows that these layers were created one after the other with a minimal intervening time lapse, and are presumably the products of successive paroxysms of the same eruptive phase.
10. In the uppermost layer of pumice, materials from - and even entire sections of - fallen walls are found. This means that during the course of the eruption there were still some walls standing, which collapsed as a result of seismic disturbances and/or shock waves. Many of these walls have collapsed outwards from the building, regardless of orientation, and in some instances this was perhaps due to the accumulation of volcanic material inside the buildings, after the caving in of the roof.
11. Immediately above the layer of pumice, which is over one meter thick in many places at the site, a thick layer of tephra (pozzuolana) was laid, the horizontal bedding of which attests its violent horizontal ejection, due to the base surge mechanism. Within the excavation this layer is as much as 5 - 6 meters thick.
12. More or less simultaneously with the ejection of the pozzuolana, enormous boulders each weighting several tons, were detached from the crater wall and propelled over considerable distances. Some of them hit houses at Akrotiri, leaving vivid traces of their impact, from which their trajectory can be calculated. Such boulders, scattered in the surrounding vineyards, damaged the House of the Ladies and Xeste 4.

From even a cursory study of the present configuration of the island, it is apparent that the great mass of volcanic material fell towards the east, creating the areas of plain from the north-west to the south-east of the island, since it correspondingly filled up the sea. This eastward dispersal of Theran tephra is being constantly confirmed by discoveries in the east Mediterranean seabed, on Rhodes, Kos, and in the Smyrna region. Taking into account that westerly winds usually blow in the Aegean in early summer, it could be maintained that the eruption took place at this time of year. And this view gains further credence from the fact that the pithoi, inside which cereals and pulses were usually stored, were found almost empty. Perhaps the Theran farmers had not begun to harvest and store their crops before the eruption took place.

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