That is absolutely correct, as Hans Havermann implied in his original message. But the Easter algorithm used by most Western churches, which I summarized earlier as given by Meeus, is a *formal* algorithm which makes no claim to astronomical accuracy. Note its finicky insistence on the * Ecclesiastic* full moon, which even in the present era can deviate from the astronomical full moon by as much as two days, and in the course of ensuing centuries will doubtless diverge even further. Even the Gregorian calendar's basic assumption (that 400 solar years equals 146,097 solar days) is wrong enough to be inaccurate after about 3 millennia, and that's without considering the secular slowing of the Earth's rotation. On Sat, Apr 11, 2009 at 5:47 PM, Dave Dyer <ddyer@real-me.net> wrote:

You're not taking into account "leap seconds", which are necessary to keep midnight at midnight, and the gradual recession of the moon, which is gradually changing the length of the lunar month.

For example, over the quoted span of 5 million years, leap seconds at the rate of one a year would result in 57 fewer days between now and then.

.. exposing the futility of calculations about dynamic systems over these time scales. You're really doing the equivalent of arguing how many angels can dance on the head of a pin.

_______________________________________________ math-fun mailing list math-fun@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun