With the Romans it was still worse. Nothing shows how much better they understood fighting than astronomy, than the way they managed their year. Under Romulus it was said to have consisted of only ten months. It is not easy to see how this could be adjusted on any mode of computation, and yet the numerical names, some of which have come down to our own calendar, would seem to present some proof of it. The last month in the year is yet called _December_, or the _Tenth_. In the days of Numa it consisted of twelve lunar months, with a system of intercalation something like that of the Greeks. The two added months were January and February, which, in numerical order would have been Undecember, and Duodecember, or the Eleventh and Twelfth. The year, however, by the clumsiness of these methods, and by the whole matter being left in the hands of the Pontifices who seem to have had little science, and still less honesty, became turned so completely topsy-turvy, that instead of being put at the end, these two new months were finally arranged at the beginning. The first was called January from the great (some say the greatest) Latin deity, Janus, whose original name was Djanus or Di-annus, _The God of the Year_ (similar to the Greek Kronos or Time), and who was most expressively represented with two faces, one ever looking back upon the past, and the other forward to the coming period. In the hands of the Pontifices the Roman year had again been getting more and more out of order, until, in the days of Julius Caesar, the first of January had retrograded nearly to the autumnal equinox. This very useful despot determined to take the matter in his own hands, and make a thorough reform; but, as a preliminary, was obliged to have an extraordinary year of 445 days, which was called the _year of confusion_. Before this, there had been, too, a continual neglect of the fraction of a day, although its existence seems to have been known at a much earlier period. Caesar arranged the months as they now stand, and made provision for the fraction by ordering a day to be added to February every fourth year. This seemed to answer every purpose, until, after the lapse of more than fourteen centuries, it was found that the seasons began to disagree with the almanac, and the religious festivals to fall somewhat out of place. The error was estimated to amount to eleven days; the correction of which was assumed by the Roman Pontifex, but with the aid of a science far more accurate than had been possessed by the Pontifices of the older time. The modes now adopted, for preserving accuracy in future, are known to most well-informed readers, so that we shall not dwell upon them farther than to say, that they consist generally in such omissions of the leap year, from time to time, as will correct the very small excess by which a quarter of a day exceeds the actual fraction of the tropical year.