With a permit granted to the American School of Classical Studies by the Ministry of Culture and the Second Ephoreia of Byzantine Antiquities, the Ohio State University Excavations at Isthmia conducted a topographic survey of the medieval Kastro of Dokos in 1996. In 1997 we continued that survey and carried out a brief exploration of the walls of the presumed early Byzantine basilica located in a saddle to the south of the Kastro. It has already been suggested (Lakvnikaiq Spoydaiq 21  97-118) that this church may have housed the relics of the Spanish martyrs Valerius, Vincent, and Eulalia, shipwrecked somewhere on their journey from Spain to Constantinople in the seventh century.
Exploration in 1997 was carried out in a very short season, 23 to 27 July, limited primarily to the cleaning of four small areas in the vicinity of the small church of St. John the Theologian, where walls of a predecessor might be encountered. Living conditions and supply issues on this barren island hampered the project and limited the length of time we could work on the site. In addition, exploration revealed that bedrock in this area is extremely close to the surface while vegetable gardens in the churchyard also made cleaning somewhat difficult.
Trench 97-1, in the area of the presumed apse, revealed only bedrock and washed fill without any trace of a built structure. A local informant told us that a pit had been dug there in the 1940's for mixing of mud for construction of a nearby building. Exploration in this trench was thus was a salutary lesson in precaution and demonstrated the need for cleaning before full announcement of the site was made.
Exploration in trench 97-2 was likewise somewhat disappointing. This trench was located near the presumed east-west wall which had been tentatively identified as the north wall of the earlier basilica. The stones that had been previously seen above the surface were cleaned and at least one other block was found that lay on the same orientation, but unfortunately the blocks were simply set in sterile soil, and no stratigraphy was found associated with them. Thus, we were able to suggest that this did indeed represent an earlier wall, but that any stratigraphy associated with its use had been destroyed as a result of time and subsequent activity in the area.
Trench 97-3 was laid out at a point symmetrical to that of 97-2 in the hope of identifying what we thought might be the south wall of an earlier structure. Fortunately in this area some stratigraphy was preserved intact between the surface disturbance and the bedrock. Thus, in this area were many pieces of broken rooftile and small stones, along with pottery of the 7th century, apparently undisturbed from where it had fallen in late antiquity. This debris was notably concentrated in the northern part of the trench, so it was decided to extend the exploration to the north in a series of 1x1 m. squares. As we moved northward the debris increased in quantity, although exploration showed that it rested on a level of hard-packed sterile soil, and not on any paved surface. Finally, at a point roughly on the line of the south wall of the present church of St. John we encountered a fill of stones that made up an east-west wall which had collapsed, almost certainly with its tiled roof, to the north and the south. This must have been the wall of the earlier basilica on the site. To the north, as to the south, of the wall there was no paved flooring still in situ, but a single slab of marble paving, with some plaster on one side, is a probable indication that such a floor had existed but had been torn out, undoubtedly to reuse its materials for another purpose. In addition, among the debris just to the south of the wall was a marble acanthus leaf, undoubtedly part of the architectural decoration of the building once located on the site. The acanthus leaf is made of relatively fine white marble, heavily drilled with considerable care. On the basis of technique, this architectural decoration can be assigned to late antiquity, perhaps to the fifth or sixth century.
Trench 97-4 was a small (1x1 m.) exploration located just to the east of the southeast corner of the modern church. Again, the fill was very shallow and the surface disturbed. Nonetheless, running more or less northeast-southwest across the cutting are the remains of a powerful mortar-poured foundation. The extent of the exploration precludes certainty, but it is possible that this foundation was for an apse at the eastern termination of the wall discovered in Trench 97-3.
Because of the shallowness of the fill and modern disturbances no secure date could be established for this earlier building. The marble acanthus decoration, however, together with the marble pavement slab, suggest that the building whose remains we explored was of monumental character and presumably to be assigned to the fifth or sixth century. Ceramics found among the debris were exclusively of the 7th century, and we may tentatively at least date the collapse of the building to that time. Again, certainty is impossible at this point, but it is likely that the remains found in the 1997 season are those of a church which was a predecessor of the contemporary chapel. Whether this was dedicated to Spanish martyrs is still an open-if tantalizing-possibility.