A calculation based on the difference in latitude between Rome and the Straits shows that they are only about 9° south of west, and that the sun, when setting in this direction is about 6° south of the equator: this is the case 15 days after the equinox, which in the beginning of the fourteenth century fell on September 14. Therefore the date at which the sun was seen from Rome setting in the direction between Sardinia and Corsica was the 29th of September, and its position in the zodiac was then the middle of the sign Libra. (Owing to precession, it would be in the _constellation_ Virgo, but we have seen that Dante ignores precession in these references.) But the moon ought now, on this second night in Purgatory, to have reached at least the last degrees of Scorpio, since on the previous night she rose when the first part was already above the horizon. Dante has therefore made a flagrant mistake with his moon, running her back into Libra again, and falsifying all his previous descriptions—_if_ he knew as much about geography as we do now. But it is only necessary to glance at the map of Henry of Mainz to see that such a supposition would be grotesque. The portolan maps did indeed show the relative positions of Rome, Sardinia, and Corsica, with remarkable accuracy;[474] but as we have already said, it is not in the least likely that Dante had ever seen one of these sailing charts. It was on the classics, and especially on Orosius, that he relied for his geography,[475] and the suggestion of Dr. Moore seems most plausible, that he was here thinking of the Spaniard’s description of Sardinia:—“Habet _ab oriente et borea_ Tyrrenicum mare quod spectat ad portum urbis Romae.”[476] This implies that Sardinia lay south-west of Rome, in a direction such as the sun might have had when setting in November, in Scorpio or Sagittarius. ------------------------------------------------------------------------ [474] See for instance the “Carte Pisane,“ and the Central Mediterranean map of Vesconte, dating from about 1300 and 1311 respectively, in Beazley’s _Dawn of Geography_, iii. Latitudes and longitudes are not given, but from certain centres lines radiate to all points of the compass, like great spiders’ webs. [475] _De Mon._ II. iii. 87-90. etc. [476] “It is bounded _on the east and north_ by the Tyrrhenian sea, which lies towards the port of Rome,” Moore, _Studies in Dante_ iii. p. 72. ------------------------------------------------------------------------ Or possibly, when the poet was in Rome in 1301, he may have been watching a sunset in the second week in November, and may have been told, among other pieces of misinformation which are so readily given to visitors to great cities, that he was looking towards the straits which separate Sardinia from Corsica. We can imagine his replying, or thinking: “Then the sun sets behind them when he has nearly finished his course in Scorpio, and is about to yoke his steeds under another star!”[477]