Gen. xviii. 14, “At the time appointed I will return”; and
Gen. xxi. 2, “At the set time of which God had spoken.”
Gen. i. 14 is therefore, “They (the sun, moon, and stars) shall be for signs (things to come) and for cycles (appointed times).”
Here, then, we have a distinct declaration from God, that the heavens contain not only a Revelation concerning things to come in the “Signs,” but also concerning appointed times in the wondrous movements of the sun, and moon, and stars.
The motions of the sun and moon are so arranged that at the end of a given interval of time they return into almost precisely the same position, with regard to each other and to the earth, as they held at the beginning of that interval. “Almost precisely,” but not quite precisely. There will be a slight outstanding difference, which will gradually increase in successive intervals, and finally destroy the possibility of the combination recurring, or else lead to combinations of a different character.
Thus the daily difference between the movement of the sun and of the stars leads the sun back very nearly to conjunction with the same star as it was twelve months earlier, and gives us the cycle of the year. The slight difference in the sun's position relative to the stars at the end of the year, finally leads the sun back to the same star at the same time of the year, viz., at the spring equinox, and gives us the great precessional cycle of 25,800 years.
So, too, with eclipses. Since the circumstances of any given eclipse are reproduced almost exactly 18 years and 11 days later, this period is called an Eclipse Cycle, to which the ancient astronomers gave [pg 179] the name of Saros;[77] and eclipses separated from each other by an exact cycle, and, therefore, corresponding closely in their conditions, are spoken of as being one and the same eclipse. Each Saros contains, on the average, about 70 ± eclipses. Of these, on the average, 42 ± are solar and 28 ± are lunar. Since the Saros is 11 days (or, more correctly, 10.96 days) longer than 18 years, the successive recurrences of each eclipse fall 11 days later in the year each time, and in 33 Sari will have travelled on through the year and come round very nearly to the original date.
But as the Saros does not reproduce the conditions of an eclipse with absolute exactness, and as the difference increases with every successive return, a time comes when the return of the Saros fails to bring about an eclipse at all. If the eclipse be a solar one before this takes place, a new eclipse begins to form a month later in the year than the old one, and becomes the first eclipse of a new series.
This is the history of one such eclipse: On May 15 (Julian), 850 a.d., there was a (new) eclipse of the sun, and it occurred as a partial eclipse. On August 20 (Julian), 1012 a.d., this new eclipse became total. From that time it has been an annular eclipse, the latitude of the central shadow gradually shifting southward from the north, until on December 17 (Julian), 1210, it had reached N. Lat. 24°. It turned northward again after 1210, until March 14 (Julian), 1355, when it fell in N. Lat. 43°. Then it turned south, [pg 180] and has moved steadily in that direction, until on March 18 (Greg.), 1950, its last appearance as an annular eclipse will take place. On May 22 (Greg.), 2058, it will fall so far from the node that a new eclipse will follow it on June 21. It will make three more appearances as an ever-diminishing partial eclipse, and be last seen on June 24 (Greg.), 2112. Its total life-history, therefore, will have been 1,262 years and 36 days, and will have occupied 70 Sari.
In the above life-history of an eclipse[78] there is not the slightest difficulty as to its identification. The Saros shows no break, and no interruption; nor does the character of the eclipse suffer any abrupt change. The district over which it is visible moves in a slow and orderly fashion from occurrence to occurrence over the earth's surface.