to keep this position of peculiar
prestige. Hence it was that Sothis came to be associated with Isis, one
of the most important divinities of Egypt, and that the day in which
Sothis was first visible in the morning sky marked the beginning of
the new year; that day coinciding, as already noted, with the summer
solstice and with the beginning of the Nile flow.

But now for the difficulties introduced by that unreckoned quarter of
a day. Obviously with a calendar of 365 days only, at the end of four
years, the calendar year, or vague year, as the Egyptians came to call
it, had gained by one full day upon the actual solar year--that is to
say, the heliacal rising of Sothis, the dog-star, would not occur on
new year's day of the faulty calendar, but a day later. And with each
succeeding period of four years the day of heliacal rising, which marked
the true beginning of the year--and which still, of course, coincided
with the inundation--would have fallen another day behind the calendar.
In the course of 120 years an entire month would be lost; and in 480
years so great would become the shifting that the seasons would be
altogether misplaced; the actual time of inundations corresponding with
what the calendar registered as the seed-time, and the actual seed-time
in turn corresponding with the harvest-time of the calendar.

At first thought this seems very awkward and confusing, but in all
probability the effects were by no means so much so in actual practice.
We need go no farther than to our own experience to know that the names
of seasons, as of months and days, come to have in the minds of most of
us a purely conventional significance. Few of us stop to give a thought
to the meaning of the words January, February, etc., except as they
connote certain climatic conditions. If, then, our own calendar were
so defective that in the course of 120 years the month of February had
shifted back to occupy the position of the original January, the change
would have been so gradual, covering the period of two life-times or
of four or five average generations, that it might well escape general
observation.

Each succeeding generation of Egyptians, then, may not improbably have
associated the names of the seasons with the contemporary climatic
conditions, troubling themselves little with the thought that in an
earlier age the climatic conditions for each period of the calendar were
quite different. We cannot well suppose, however, tha