Page:Decline and Fall of the Roman Empire (1827) Vol 2.djvu/398

 380 THE DECLINE AND FALL CHAP, same time constructed, and moved forwards on wheels, till the soldiers, who were provided with every species of missile weapons, could engage almost on level ground with the troops who defended the rampart. Every mode of resistance which art could suggest, or courage could execute, was employed in the defence of Amida ; and the works of Sapor were more than once destroyed by the fire of the Romans. But the resources of a besieged city may be exhausted. The Persians repaired their losses, and pushed their ap- proaches ; a large breach was made by the battering- ram, and the strength of the garrison, wasted by tlie sword and by disease, yielded to the fury of the assault. The soldiers, the citizens, their wives, their children, all who had not time to escape through the opposite gate, were involved by the conquerors in a promiscuous massacre. Of Singaia, But the ruin of Amida was the safety of the Roman A*^b 360. pi'ovinces. As soon as the first transports of victory had subsided, Sapor was at leisure to reflect, that to chastise a disobedient city, he had lost the flower of his troops, and the most favourable season for con- quest". Thirty thousand of his veterans had fallen under the walls of Amida, during the continuance of a siege which lasted seventy-three days ; and the disap- pointed monarch returned to his capital with affected ti'iumpli and secret mortification. It is more than pro- bable, that the inconstancy of his barbarian allies was tempted to relinquish a war in which they had en- countered such unexpected difficulties ; and that the " Ammianus has marked the chronology of this year by three signs, which do not perfectly coincide with each other, or with the series of the history. 1. The corn was ripe when Sapor invaded Mesopotamia ; " cum jam stipula flavenle turgerent; a circumstance which, in the latitude of Aleppo, would naturally refer us to the month of April or May. See Harmer's Observations on Scripture, vol. i. p. 41 ; Shaw's Travels, p. 335. edit. 4to. 2. The pro- gress of Sapor was checked by the overflowing of the Euphrates, which ge- nerally happens in July and vVugust. Plin. Hist. Nat. v. 21; Viaggi di Pietro della Valle, tom. i. p. 696. 3. When Sapor had taken Amida, after a siege of seventy-three days, the autumn was far advanced. " Autumno praecipiti, haedorumque improbo sidere exorto." To reconcile these apparent contradictions, we must allow for some delay in the Persian king, some in- accuracy in the historian, and some disorder in the seasons.