Hello everyone,
This is a small point but could be significant. If you have a mercury
maximum thermometer then read the maximum when the temperature is as high as
possible. The mercury thread above the nick will contract as the thermometer
cools and the reading will be lower than that obtained by reading when the
temperature is still high.
I noticed this today and have done so on a few other occasions. My max
was 32.8°C, read when the temp was still about 30. I didn't shake it down, for
some reason. I looked in the screen about midnight with the temperature 20 and
read the max again, and it was 32.6°C, 0.2°C lower. I suspect (and hope) that
the higher figure is the correct one, if the thermometer itself is OK.
The error is proportional to the temperature difference between the max
and the ambient at time of reading, multiplied by the length of the mercury
thread above the nick, at maximum, expressed in thermometer degrees. In my
case the figures are 13° and 64° respectively, the nick being at equivalent to
-31°C. Multiply by the coefficient of expansion of mercury, 0.000182 per deg
and you get 0.15°C as the difference, which agrees near enough with my 0.2 deg
error.
The number of times this will make a sensible difference is probably
pretty small, but this was one of them.

Best wishes, Tudor Hughes, Warlingham, Surrey.

"TudorHgh" wrote in message
...
Hello everyone,
This is a small point but could be significant. If you have a
mercury
maximum thermometer then read the maximum when the temperature is as
high as
possible. The mercury thread above the nick will contract as the
thermometer
cools and the reading will be lower than that obtained by reading when
the
temperature is still high. snip

.... I can confirm the same effect seen here (Bracknell-Tawfield), and
the difference seems to be a maximum of 0.2degC; yesterday it was
0.1degC. It's not always possible to monitor the hour-by-hour readings
though (and not desirable to keep opening the screen), so there must
have been other occasions when lower maxima have been recorded than
actually occurred (or were indicated). A good debating point whether
we should note the reading 'as is' at 18Z (or 09Z next day), or the
highest seen. The Observer's Handbook doesn't appear to give guidance
on this point.

.... What do others think? Should we note a reading that has been
'enhanced' by the marked expansion of the mercury column, rather than
just the expansion of the mercury reservoir in the bulb?

--
Martin Rowley: data via:
http://homepages.ntlworld.com/booty....r/metindex.htm
Bracknell (Wooden Hill/Great Hollands), Berkshire.
NGR: SU 854 667
Lat: 51DEG23MIN30SEC(N); Long: 00DEG46MIN28SEC(W)
Height(amsl): 80 metres (262 feet)