According to this, we consider that the tempest took place after four days, counted from the day of landing; that the full moon fell on the following night; and lastly, that this period coincided not with the highest _tide_, but with the highest _tides_ of the ocean. Thus we believe that it would be sufficient for ascertaining the exact day of landing, to take the sixth day which preceded the full moon of the month of August, 699; now this phenomenon, according to astronomical tables, happened on the 31st, towards three o’clock in the morning. On the eve, that is, on the 30th, the tempest had occurred; four full days had passed since the landing; this takes us back to the 25th. Cæsar then landed on the 25th of August. Mr. Airy, it is true, has interpreted the text altogether differently from our explanation: he believes that the expression _post diem quartum_ may be taken in Latin for the third day; on another hand, he doubts if Cæsar had in his army almanacks by which he could know the exact day of the full moon; lastly, as the highest tide takes place a day and a half after the full moon, he affirms that Cæsar, placing these two phenomena at the same moment, must have been mistaken, either in the day of the full moon, or in that of the highest tide; and he concludes from this that the landing may have taken place on the second, third, or fourth day before the full moon. Our reasoning has another basis. Let us first state that at that time the science of astronomy permitted people to know certain epochs of the moon, since, more than a hundred years before, during the war against Perseus, a tribune of the army of Paulus Æmilius announced on the previous day to his soldiers an eclipse of the moon, in order to counteract the effect of their superstitious fears.[383] Let us remark also, that Cæsar, who subsequently reformed the calendar, was well informed in the astronomical knowledge of his time, already carried to a very high point of advance by Hipparchus, and that he took especial interest in it, since he discovered, by means of water-clocks, that the nights were shorter in Britain than in Italy. Everything, then, authorises us in the belief that Cæsar, when he embarked for an unknown country, where he might have to make night marches, must have taken precautions for knowing the course of the moon, and furnished himself with calendars. But we have put the question independently of these considerations, by seeking among the days which preceded the full moon of the end of August, 699, which was the one in which the shifting of the currents of which Cæsar speaks could have been produced at the hour indicated in the “Commentaries.”