I discovered an eclipse cycle

[Helios]

I discovered an eclipse cycle. It will be called the Wells cycle ( Garrick Wells ). It is simply a cycle of 51649 days. I believe that if you see an eclipse, then 51649 days later, another will likely be seen from your same location. I've have indeed found an example of this while probing the NASA Phases-of-the-Moon web site. The series lasted seventeen centuries, with just one fail. Should someone offer some explanation, I'd love to hear it.

[Borek]

Compare with:

http://en.wikipedia.org/wiki/Saros_cycle

I guess these are somehow related.

[fluidistic]

Should someone offer some explanation, I'd love to hear it.
Maybe http://en.wikipedia.org/wiki/Poincar%C3%A9%27s_recurrence_theorem.

[negitron]

I found a discussion of thirteen-month calandars in which the figure in the OP falls naturally out of the conversation:

The total number of days is 2,014,311 = 3 * (13^2) * 29 * 137.
So not only is the number (39) of 1749-month cycles is divisible by 13,
but the number of days (51649) in each 1749-month cycle.
Each 1749-month cycle
= 107 yerms
= 1749 months
= 3979 trecena
= 32 drifts over 29-day cycle
= 51649 days
= 2 yerm eras

http://www.nabble.com/Re:-5515-Year-Luni-Solar-Cycle-p20917812.html