[tz] Earth's day lengthens by two milliseconds a century

[Paul Eggert]

On 12/08/2016 03:45 PM, Steve Allen wrote:
> Except that Morrison and Stephenson and Hohenkerk only show the past.
> NASA polynomials attempt to predict the future, but (as seen in the
> data of this new publication) to their peril.

Yes, the Stephenson et al. paper is newer and presumably more 
authoritative and 'Theory' cites it first. I assume Espenak will update 
the other page eventually, and in the meantime it's an entertaining set 
of estimates.

The 'Theory' file cites these two sources only to support the statement 
that we don't know historical solar time to more than about one-hour 
accuracy, a statement that I hope is vague enough to pass muster. If we 
wanted to get picky, we'd have to determine what's "historical": back to 
3200 BC? or back to the earliest reliable astronomical records? if the 
former, we don't have even one-hour accuracy; if the latter, it depends 
on one's definition of "reliable". Still, the statement is in the right 
ballpark, which is all that's needed there.