Tephrochronological Evidence for the Time of the Bronze Age Eruption of Thera

TEPHROCHRONOLOGICAL EVIDENCE

Since the first Thera congress in 1969 the authors have visited Greece on two occasions to collect soil samples to be examined for traces of volcanic glass particles from the Bronze Age eruption of Santorini (Thera).

On the first visit, more than 80 samples were collected, during a ten-day intensive search on Crete, from road cuts, natural outcrops, and archeological sites at Amnisos, Eremopolis, Gournia, Hagia Pelagia, Kato Zakro, Knossos, Mallia, Nirou Khani, and Vathypetro. In 1974 the search was extended to Phylakopi on Melos, where a total of 37 samples were collected from sections in different trenches. Three other suites of samples were very kindly collected for us and forwarded by Dr. Joseph Shaw (3 samples from Kato Zakro), Mr. L.H. Sackett (10 samples from the Unexplored Mansion, Knossos), and Prof. A.C. Renfrew (39 more samples from Phylakopi). All these were examined under the petrographic microscope in search of traces of volcanic glass particles having a refractive index in the range of tephra from the Bronze Age eruption of Thera (n = 1. 509 ± 0.002).

Although no accumulations were found in soil profiles on Crete that might represent the original ash layer, altered or unaltered, enough traces of Thera tephra were found to prove beyond doubt that at least the eastern end of Crete had at one time been blanketed by ash fallout from the eruption, thus confirming the findings of Ninkovich & Heezen (1965) in deep-sea cores. At the same time, the presence of these particles in the soils cannot throw any light on the question of just when that ash fell.

Evidence pertaining to the time of the fall can only be obtained from archeologically different levels in Minoan ruins, preferably from as many sites as possible. While it is not at all surprising, in view of the topography and climate of Crete, that no remnants of the original ash layer were found in situ, in the nature of things it is to be expected that some particles would have remained lodged in crevices in some dwellings occupied at the time of the fall. Their chances of survival would decrease the longer the buildings continued to be occupied. At the same time, the possibility of the introduction of particles from the outside would continue to exist for as long as the buildings were in use. Particles of Thera tephra, some of which are still present in Cretan soils, could also be tracked into houses built at any time later, right up to the present, or could be exhumed if stones from structures contemporaneous with the eruption were "quarried" to build new houses. Thus the lowest level in which particles are consistently found should be the level occupied at the time of the eruption. The occasional finding of a few particles in levels antedating the eruption would also have to be allowed for, since a certain amount of contamination of unsealed older levels cannot be ruled out. Other things being equal, however, a sufficiently large statistical sample should show greatest concentrations in levels dating from the time of the ashfall, decreasing more or less steadily upward through the section, and only slight sporadic findings or none at all in levels predating the eruption. 

Thus attention must be focussed on the Late Minoan IA level or its stratigraphic equivalent. The consistent presence of tephra in samples from that level, in deposits sealed before Late Minoan IB, would argue for an LM IA date for the eruption, whereas its consistent absence in that level would suggest an LM IB date.

So far, the evidence seems to point toward an LM IA date, which is consistent with the lack of definite LM IB artifacts on Thera itself.

Our most significant samples were collected at Kato Zakro, where we were able to sample the LM IA level in several rooms, one sample being from a freshly dug deposit believed to have been sealed before LM IB. Not only did all five of the samples from this level contain tephra particles, but in four out of the five it was noticeably more abundant than in LM IB levels from the same site. These results have been published (Vitaliano & Vitaliano 1974).

The results from Phylakopi, though not as clear-cut as those from Kato Zakro, are fully in accord with a pre-LM IB date. Barring a trace of tephra in one sample from a level belonging to the Middle Cycladic period, which has to be due to contamination, tephra particles were found in levels dating from the later part of the Late Bronze I sequence to Late Helladic III. The greatest concentrations (if that word can be used for amounts measured in a few grains per thousand after removal of soluble constituents) were found in and above an LB I floor sequence in trench IIA. The first occurrences antedate the end of the LB I constructional phase, indicating that the "Third City" survived long enough after the Tephra fall for one or more floor renewals. According to Prof. A.C. Renfrew (1977) the evidence is consistent with the eruption's having occurred at a time when LM IA imports were reaching the site and before LM IB imports reached it; however, as of the 1976 excavation season, it cannot be firmly documented that the tephra fall occurred before the arrival of any LM IB imports.

The findings at Kato Zakro were highly suggestive but, by themselves, not fully conclusive. But now that Thera tephra has also turned up in the only other pre-LM IB levels examined so far, those at Phylakopi, the evidence becomes more compelling. Add to that the fact that Rapp et al (1973) had previously reported the finding of water-borne Thera pumice in the dig at Nichoria on the Messenian Gulf, in a context (Late Helladic IIA) which they correlated with LM IA of Crete, the case for an LM IA date for the Bronze Age eruption looks even better. We hope to be able to continue the search in the relevant levels at other sites, with the help of archeologists actively digging in Minoan sites on Crete and elsewhere.

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