FOURTH YEAR OF THE WAR ^REVOLT OF MITYLEXE. 235 rect their blows, and all remained at their posts, except a reserve of three hundred men, kept constantly in readiness for special emergencies, who marched out and patrolled the outside of the ditch to intercept any fugitives from within. At the same time, fire-signals were raised to warn their allies at Thebes, but here again the Platseans in the town had foreseen and prepared fire-signals on their part, which they hoisted forthwith, in order to deprive this telegraphic communication of all special mean- ing. 1 Meanwhile, the escaping Platoeans, masters of the two adjoin- ing towers, on the top of which some of them mounted, while others held the doorway through, so as to repel with spears and darts all approach of the blockaders, prosecuted their flight without interruption over the space between, shoving down the battlements in order to make it more level and plant a greater number of ladders. In this manner they all successively got over and crossed the outer ditch ; every man, immediately after crossing, standing ready on the outer bank, with bow and javelin, to repel assailants and maintain safe passages for his comrades in the rear. At length, when all had descended, there remained the last and greatest difficulty, the escape of those who occu- pied the two towers and kept the intermediate portion of wall free : yet even this was accomplished successfully and without loss. The outer ditch was, however, found embarrassing, so 1 Thucyd. iii, 22. typvuroi re tjpovro sf ruf 0^/3af TroTiSfiioi, etc. It would seem by this statement that the blockaders must have been often in the habit of transmitting intelligence to Thebes by means of fire-signals ; each particular combination of lights having more or less of a special meaning. The Platajans had observed this, and foresaw that the same means would be used on the night of the outbreak, to bring assistance from Thebes forth- with. If they had not observed it before, they could not have prepared for the moment when the new signal would be hoisted, so as to confound ita meaning OTTW^ uaafyy -<L ai)p.ia ij. . . . Compare iii, 80. I agree with the general opinion stated in Dr. Arnold's note respecting these fire-signals, and even think that it might have been sustained more strongly. "Non enim (observes Cicero, in the fifth oration against Verres, c. 36), sicut erat nnper consuetudo, prajdonum adventure, significabat ignis & spcculd tiiblatus aut tumulo : sed flamma ex ipso incendio navium et calamitatenj
acceptam et periculum reliquum nuntiabat.
