In article ,
Col writes:

"John Hall" wrote in message
...
In article ,
Col writes:
I recall asking a similar question some time back and was told that
the early readings were derived from a tiny handful of sites (hence
their accuracy to only 1 degree)

I think that you're thinking of the claimed accuracy of the earliest CET
monthly averages to 1F, which equates to roughly 0.5C.

In the very earliest years (to 1670) the figures are given in whole
degrees only
so surely that is an accuracy to 1 degree?
From 1671 to 1698 it is half a degree and thereafter 0.1C.

You're probably right. My memory was that the whole series - including
the extension back in time as far as 1659 - was originally produced by
Manley using degrees Fahrenheit, but I may well have been wrong in that
and or the precision may have subsequently been revised.
--
John Hall
"One half of the world cannot understand
the pleasures of the other."
From "Emma" by Jane Austen (1775-1817)