month, upon the mountains of Ararat."

The five months during which the waters prevailed upon the earth were,
therefore, reckoned as of thirty days each. If all the new moons, or
even that of the seventh month, had been actually observed, this event
would have been ascribed to the nineteenth day of the month, since 150
days is five months and two days; but in the absence of such
observations a sort of "dead reckoning" was applied, which would of
course be corrected directly the return of clear weather gave an
opportunity for observing the new moon once again.

A similar practice was followed at a much later date in Babylon, where
astronomy is supposed to have been highly developed from remote
antiquity. Thus an inscription recently published by Dr. L. W. King
records that--

"On the 26th day of the month Sivan, in the seventh year, the
day was turned into night, and fire in the midst of heaven."

This has been identified by Mr. P. H. Cowell, F.R.S., Chief Assistant at
the Royal Observatory, Greenwich, as the eclipse of the sun that was
total at Babylon on July 31, B.C. 1063. The Babylonians, when bad
weather obliged them to resort to dead reckoning, were, therefore, still
reckoning the month as precisely thirty days so late as the times of
Samuel and Saul, and in this particular instance were two, if not three,
days out in their count. Had the new moon of Sivan been observed, or
correctly calculated, the eclipse must have been reckoned as falling on
the 28th or 29th day of the month.

The Athenians in the days of Solon, five hundred years later than this,
adopted months alternately twenty-nine and thirty days in length, which
gives a result very nearly correct.

The Jews after the Dispersion adopted the system of thus alternating the
lengths of their months, and with some slight modifications it holds
good to the present day. As will be shown in the following chapter, the
ordinary years are of twelve months, but seven years in every nineteen
are "embolismic," having an extra month. The names employed are those
learned during the Babylonian captivity, and the year begins with the
month Tishri, corresponding to September-October of our calendar. The
lengths of most of the months are fixed as given in the following
table, but any adjustment necessary can be effected either by adding one
day to Heshvan, which has usually twenty-nine days, or taking away one
day from Kislev, which has usually thirty--