What Beloch’s arguments do emphasize is the fact that the situations during Peisistratus’ two periods of exile were in some ways very similar. The sameness of the two situations may in fact be the reason why so little has been remembered about the earlier. It raises the question whether the tyrant mined and coined during his first exile. There is no certainty that he did either, but the probability is that he did both. As regards Thrace we know that Miltiades, probably with Peisistratus’ permission and approval[276], had settled in the Gallipoli peninsula soon after the tyranny was first established at Athens[277]. Thrace is the one region that we can be sure that the tyrant must have considered as a possible place of exile. [Sidenote: is connected by Babelon with the Athena-head coins of Athens] As regards the coinage it has been suggested on the high authority of Babelon[278], that the famous series with the owl on one side, and the head of Athena on the other (fig. 6), which remained for centuries the coin types of the city, was actually started to commemorate the help that the tyrant claimed to have received from his patron goddess at the time of his first restoration. [Illustration: Fig. 6. Coin of Athens with Athena and owl.] The evidence is not conclusive. The arguments for and against this date are based on a few literary references that are too vague to be of much use, on points of style and technique from which it is notoriously dangerous to draw conclusions, on a comparison of the coin and pottery statistics from Naukratis which it is no less dangerous to use as evidence, on a hoard found in 1886 among the pre-Persian remains on the Athenian Acropolis which, as far as the circumstances of the find are concerned, may have been lost or deposited there long before the catastrophe, and only establish a _terminus ante quem_ that nobody would think of disputing, and on certain alliance coins (Athens-Lampsacus, Athens-Sparta?, Athens and the Thracian Chersonese)[279]. These last look more promising at first sight, but only the Athens-Lampsacus coins can be dated with any certainty, and they, unfortunately, are very small, and may have been struck under difficulties, so that it is not easy to be sure of their chronological position in the Athenian series. We are driven back therefore on to the impressions of experts, most of whom agree with Babelon that the owl-Athena series cannot begin either much before or much after 550 B.C.[280]. That is to say that this double type was certainly in vogue when the tyrant secured his second restoration by means of his Thracian silver[281], and “rooted his tyranny” in revenues derived “partly from the river Strymon, partly from home.” [Sidenote: nicknamed (probably just about this time) girl, virgin, Pallas.]