water. Looking at the position of this lake relatively to the foundations on the shore, I am inclined to think that it must in ancient times have been a harbour. Indeed, I am assured by M. Ducci, the Russian vice-consul here, that he remembers to have heard from old inhabitants of Rhodes a tradition that a canal formerly connected this lake with the sea. If we suppose that another canal anciently communicated between this lake and Port Mandraki, ships would have been able to pass in and out without having to weather the sandy point. Such an hypothesis woidd give a more definite meaning to the rhetorical statement of Aristides (see ante, p. 148), that the harbours of Rhodes were arranged as if for the express piurpose of receiving the ships of Ionia, as well as those of Caria, Cyprus, and Egypt. It may be observed that the row of windmills on the N.W. shore stands on a ridge running parallel with the edge of the sea. It is not improbable that this ridge marks the line of the wall of the ancient city, in which case the foundations uncovered by me may be those of a square tower. The margin of shore at the foot of this ridge has probably been thrown up, and the sandy spit prolonged by deposit from the sea since the time of the ancients.
It will be seen by comparing the plan of Rhodes, Plate 4, with the view, Plate 5, that Port Mandraki is separated from the great harbour by a narrow isthmus at the N.E. angle of the fortress.
Within this angle is a level area, covered with rich vegetable soil, and occupied by gardens.
