e purposes. Lunar months were observed in the regulation of
temples, and lunar years, &c., have been suspected. To find uniformity
in any department in Egyptian practice would be exceptional. By the
decree of Canopus, Ptolemy III. Euergetes introduced through the
assembly of priests an extra day every fourth year, but this reform had
no acceptation until it was reimposed by Augustus with the Julian
calendar. Whether any earlier attempt was made to adjust the civil to
the solar or Sothic year in order to restore the festivals to their
proper places in the seasons temporarily or otherwise, is a question of
great importance for chronology, but at present it remains unanswered.
Probably neither the Sothic nor any other era was employed by the
ancient Egyptians, who dated solely by regnal years (see below). An
inscription of Rameses II. at Tanis is dated in the 400th year of the
reign of the god Seth of Ombos, probably with reference to some
religious ordinance during the rule of the Seth-worshipping Hyksos;
Rameses II. may well have celebrated its quater-centenary, but it is
wrong to argue from this piece of evidence alone that an era of Seth was
ever observed.

From the Middle Kingdom onward to the Roman period, the dates upon
Egyptian documents are given in regnal years. On the oldest monuments
the years in a reign were not numbered consecutively but were named
after events; thus in the Ist Dynasty we find "the year of smiting the
Antiu-people," in the beginning of the IIIrd Dynasty "the year of
fighting and smiting the people of Lower Egypt." But under the IInd
Dynasty there was a census of property for taxation every two years, and
the custom, continuing (with some irregularities) for a long time,
offered a uniform mode of marking years, whether current or past. Thus
such dates are met with as "the year of the third time of numbering" of
a particular king, the next being designated as "the year after the
third time of numbering." Under the Vth Dynasty this method was so much
the rule that the words "of numbering" were commonly omitted. It would
seem that in the course of the next dynasty the census became annual
instead of biennial, so that the "times" agreed with the actual years of
reign; thenceforward their consecutive designation as "first time,"
"second time," for "first year," "second year," was as simple as it well
could be, and lasted unchanged to the fall of paganism. The question
arises from what point these r