I was reading Jewett's "The World Of Physics," and came across something
on page 178. He states that a mercury thermometer reads accurately when
just the bulb is immersed, while an alcohol thermometer only reads
accurately when the ENTIRE thermometer is in the liquid. I was intrigued
by this idea. His explanation is that mercury is a metal and a good
conductor, and the column reaches a uniform temperature quickly; this does
not happen with an alcohol thermometer, which is why the entire thing must
be immersed, or else the column expansion will not be consistent with the
calibration of the thermometer.
This makes some sense to me. However, thermometer bulbs are made of very
thin glass, so I thought that immersing just the bulb will always do the
job -- no matter what type of thermometer is used. The thick glass on most
of the thermometer is a poor conductor anyway. Many schools are getting
rid of their mercury thermometers for safety reasons. Is it really true
that the widely-used alcohol thermometers should be fully immersed for
accurate readings?