rural life was so closely dependent upon the administration and its accounts. We may compare the fact that the lunisolar calendar of Greece was introduced as an ecclesiastical calendar, and succeeded in establishing itself as the civil calendar owing to the close connexion between the religious and the political life; but the old reckoning from the phases of the stars persisted alongside of it. In the same way we must suppose that in Egypt alongside of the numerical calendar the old method of reckoning by the concrete appearance of the moon originally persisted, but since by this time it had lost its practical importance it vanished without leaving any other traces than the length of the arithmetical month (as a round number) and the name ‘month’. On the other hand it must have been intended to give to the year the length of the solar year: the five extra days were accordingly introduced outside the series of months. Hence the same word _wepet ronpet_ means both the first day of the civil shifting year and also the day of the actual morning rising of Sirius; hence too the three four-month divisions of the shifting year are called after the seasons. The first of these, the time of inundation, began exactly with the morning rising of Sirius when the Nile began perceptibly to rise. Here the Egyptians went wrong because they did not realise that the year does not consist of exactly 365 days, but contains an additional fraction of a day. The consequence was that the Egyptian year got out of place in relation to the solar year, but so slowly that no inconvenience was caused in practical life: the linguistic difficulty, that _wepet ronpet_ acquired two different meanings and that e. g. the season called the time of inundation might fall in the actual seed-time or harvest, the conservative minds of the Egyptians enabled them to tolerate. A contributing factor was the practical convenience of the calendar. The dislocation must however very soon have been recognised, since the actual morning rising of Sirius, so far as we know, was always celebrated, i. e. it was a movable feast in relation to the calendar. The error is included in the well-known formula of the Sothic period (1461 Egyptian = 1460 Julian years).