_N_ in the same plate is a restoration of the building with the four quadrangles, only lately excavated. It is the largest building yet discovered at the Hieron, being nearly 90 yards square. Each of the four quadrangles is surrounded by a number of rooms. In all there were between seventy and eighty of these apartments, each of which opened into its own quadrangle, so far as I could judge. A colonnade ran round the interior of each quadrangle. Query, what is it?—a gymnasium, a palæstra, a college for the priests, or a great hostel? I confess the last-named seems the most probable. When one considers the large number of the sick who came to the Hieron, it is obvious that extensive accommodation must have been provided for them somewhere. The two chambers of the abaton could not have held more than 120 beds, supposing these to have been placed in two rows; or if we suppose the almost dark lower story of the western end to have been a dormitory also, 180 would then have been the greatest possible accommodation. This, if the extreme number to be entertained, scarcely accords with the accounts given by ancient writers of the multitudes who came for healing to the sanctuary. It appears likely, therefore, that this and other undetermined buildings were hostels for the accommodation of those whose ailments were slight or who were convalescent. [Illustration: PLATE XIII—North-Eastern Colonnade] The remains of this curious structure are shown as seen from a distance in Plate XVII below. _O_ in Plate I is a Roman building. Cavvadias thinks that α is the temple of Asklepios and the Egyptian Apollo. _P_ is a building also of the Roman Period, and evidently contained baths. There are traces of a hypocaust. The remains of hot air or hot water pipes are abundant, and certain curious apse-like recesses in the walls, containing a seat and terminating below in a bath or deep basin, were evidently a form of sitzbath. When we remember that the French have lately discovered at Delphi no less than three extensive bathing establishments, adjacent to the walls of the precinct on the east, west, and south sides respectively, it is not surprising that we should find at least two such buildings at Epidauros. A part of this Roman bath-house is seen in the distance in Plate XIII. [Illustration: PLATE XIV—Figure of Aphrodite] _Q_ in Plate I is a quadrangular building between the Temple of Artemis and the South Portal. Round three if not four of its sides were rooms, as in the case of the great four-quadrangle building; many remains of columns are seen. Its purpose is doubtful. Probably it is the Colonnade of Kotys which Pausanias mentions. (It is to be hoped that no shrine of the dissolute Thracian goddess of that name existed here.)