[22] The Parian marble mentions a mission of Croesus to Delphi in the year 556. The date of the year for the capture of Sardis is destroyed, and cannot be even approximately restored, as the nearest dates are either mutilated or destroyed. The dates in Eusebius are derived from Apollodorus, who in turn draws from Eratosthenes. Eusebius puts the testing of the oracles in Olymp. 57,3 = 550 B.C., the march of Cyrus against Croesus in Ol. 57,4 = 549 B.C., the capture of Croesus in Ol. 58,3 = 546 B.C. Jerome represents Croesus as beginning the war in Ol. 57,3 = 550 B.C. and puts his capture in Ol. 58,1 = 548 B.C. According to the statement of Syncellus (1,455, ed. Bonn.), Croesus was defeated in the 14th year of Cyrus, which would give 547 B.C., if with Eusebius, who allows Cyrus to reign 31 years, we put his accession in 560 B.C. (V. p. 381 _n._). The interval of three years which Eusebius (549-546) and two years which Jerome (550-548) places between the beginning of the war and the capture of Croesus, appears to be due to the three years for which, according to Herodotus, Apollo delayed the overthrow of Croesus; the presents came to Delphi three years before the fall (Herod. 1, 91). According to Herodotus the campaign occupies only one summer and autumn. The temple of Delphi was burned down in 548 B.C. (Ol. 58,1; Pausan. 10, 5, 13), and as Herodotus represents the temple as intact at the time when Croesus sent to Delphi after his fall, this must have taken place before 548 B.C., and therefore Croesus must have been conquered by Cyrus in 549 B.C. If the justification of the oracle in Herodotus is merely an invention of the priests, yet in things so well known the existing circumstances could not be left out of sight. It is certain that if the presents of Croesus had been injured by the burning of the temple before his fall, this evil omen would not have been left out of sight by the legend, or by Herodotus, who himself saw and mentions the lion of Croesus at Delphi which had been injured by the fire (1, 50). I have therefore no scruple in putting the fall of the Lydian kingdom in the year 549 B.C. Though the reign given by Herodotus for Croesus, fourteen years and fourteen days, may have arisen out of the fourteen Lydian boys who wished to be sacrificed with their king (p. 12), yet Eusebius, Jerome, and Syncellus put the reign of Croesus at 15 years. It may therefore be regarded as an established fact that his reign ended in the fifteenth year. According to Herodotus (1, 64, 65), it might seem as though he were of opinion that Croesus sought allies in Hellas at the time when Pisistratus was tyrant for the third time over Athens. But this would be an error due to Herodotus' habit of anticipation. We can only be concerned with the second tyranny of Pisistratus, which belongs to the years 550 and 549 B.C. Against the argument here used--that the priests could not leave out of sight the actual circumstances in things