on our left. Here were foundations of houses and walls, and an ancient well. A little further on we passed on our left a hill called Kemisalla, where are also some Hellenic walls. From this spot the road led straight to Marmaroulia, the spot already noticed when we branched off in going from Siana to Agios Phokas. The N.E. end of Akramytis seen from this road presents a curious study for geologists; some of the strata being vertical, others twisted and curled round like volutes.
Having heard of some other ancient remains in the neighbourhood of Siana, I determined to explore them. At about an hour's distance to the N., near the foot of Mount Atabyron, is a place called Agros, where is an angle remaining from a peribolos of Hellenic masonry. One wall running N.E. and S.W. is continuous for 29 feet, and may be traced in the same direction for 73 feet, where the corresponding angle has been carried away by mountain torrents. The other walls of this enclosure may also be traced at intervals. The height of the wall at the angle is 12 feet; the blocks are about 3 feet 9 inches by 1 foot 10 inches. There are traces of an inner wall running parallel mih the outer enclosure.
From Agros, turning nearly east, we rode for half an hour along the side of Atabyron, having Akramytis on our right, till we came to a ruined church called Hepta Amartias. In the walls are squared Hellenic blocks, and a sepulchral cippus with festoons. Thence we proceeded to a hill called Castello, about an hour south of Hepta Amartias, where
