EXCAVATIONS AT OOURNTA, ORETR. 567 



mass of bronze, evidently reduced by heal, lay in a dooiway of llie 

 palace, and may have formed a part of the Iriinniiiius of the door. 

 As a rule the house walls are not sufficiently hiirh for windows (o have 

 been preserved, but three openini>:s in walls on the east slope were 

 certainly intended to admit li<>ht and air. Floors were made of 

 beaten earth, " terrazza ' (a cement of pebbles covered with a layer 

 of plaster), stone slabs, or i)avino- stones like those in the roads. As 

 for roofs, the evidence seems all in favor of the flat terrace forms 

 conmion to-day in the East, Pieces of plaster still bearin*; impres- 

 sions of reeds show Avhat the ceiling must have been. In a o;r()nnd- 

 iloor room of the palace a large tree trunk was found fallen and 

 burned, completely charred through, but retaining its original shape; 

 this supported either the flooring of the ui)per story or the roof. The 

 central hall of the palace was choked with such timbers. 



In plan the houses are simple, conforming to the lay of the land 

 rather than to a fixed form. When similarity of plan can be detected, 

 as in certain houses on the east slope, the arrangement is modern 

 rather than classical and is in agreement with the mosaic pictures of 

 Minoan houses found in the palace of Knossos in IDOi. As in the 

 mosaic, so at Gournia we see the houses built flush with the streets 

 and usually provided with a good stone threshold; crossing this we 

 enter a ])aved antechamber with doors leading to the gi'ound-floor 

 rooms and steps mounting to the second story: cellar ste[)s may de- 

 scend directly from the antechamber or from an inner room. Certain 

 cellar rooms are finished in plaster and provided with doors; others 

 were entered, if at all, by ladders from above and can have served 

 only for storerooms; still others were mere substructures. Several 

 houses on the east slo])e have open courts which seem to have been 

 generally omitted in the private dwellings on the top of the hill for 

 lack of space. We knoAv that there were second stories, because five 

 stone staircases are well preserved and the former existence of wooden 

 steps at many other ]X)ints is clearly indicated. Moi'eover, nuiny ob- 

 jects, and these usually the best, were found in the earth at varying 

 heights above the floor level, and except where theiv was proof that 

 these had stood on a wooden shelf, since rotted away or burned, they 

 must have fallen from an upper story. 



No satisfactory exjilanation has yet been given for a stone object 

 which is very often found just within the street door. It looks like a 

 large mortar, and either stands ui)on the paved floor or is sunk 

 beneath it to the rim. It would make an awkward basin, for there 

 is no way of removing water except by dipping; on the other hand, 

 its position, invariably close to the outer door, makes us think that it 

 must have served some other purpose than the one of pounding and 

 grinding Avhicli its form suggests, or at least that some special sig- 

 nificance Avas attached to its use. No i)estle has yet been found with 



