Flyingmac,
The altimeter doesn't contain mercury. Some altimeters do have dual Kollsman scales or windows (milibars and inches of mercury). These don't mean that the altimeter contains mercury, but are measurements that reference mercury. Specifically, they reference the readings that would be obtained from a mercury barometer, or a barometer calibrated to read in "inches of mercury."
Typical measurements for pressure readings include both inches of water and inches of mercury. The standard measurement in inches has been set using mercury, for the simple reason that it's more dense that water and produces easier numbers with which to work. (a column of mercury at sea level is only 30 inches high...about two and a half feet, or a little less than a meter. A column of water at sea level is about 35 feet high, or nearly six meters).
Standard air pressure at sea level will be adequate to support a column of mercury in a glass tube to a height of 29.92 inches. That is, a glass tube, open at one end and placed in a bowl of mercury, will have the mercury pushed up the glass tube by air pressure on the mercury in the bowl, to a height of 29.92 inches. Above the mercury is a vacuum, sealed by the endof the tube. The mercury barometer compares existing or ambient atmospheric pressure acting on the bowl of mercury against the vacuum in the tube, and determines the pressure by measuring the height to which the mercury is pushed in the tube. Mercury is dense, heavy, and doesn't get pushed as high as water...which is why we reference mercury rather than something else like water. This harks back to an old fashioned mercury barometer. These days most barometers don't use mercury, but use electronics to determine pressure. They are still calibrated and read to the standard of mercury, however.
If you fly an airplane with a manifold pressure gauge, it also uses inches of mercury as the reference (which is why, on a standard day, the manifold pressure gauge at rest on the ramp at sea level will read 29.92 inches of mercury). There is no mercury found in the manifold pressure gauge, and none is found in the altimeter.
Have you heard the one about the dog and the pilot.
The job of designers is to take the pilot out of the equation for exactly this reason.
Forgetting to adjust the altimeter, forgetting to arm the pressurisation, forgetting that pilots are human, forgetting why people wish a lot, are all good reasons and some why I supect we will end up with a more autonomous mechanism for altimeteres - I quite like IO540s suggestion.
The is actually for a dog, a cat, a duck, and a brick. A fully equipped cockpit. If IMC and disoriented, and unable to navigate or find up,one opens the cockpit window and tosses out the brick. Follow it down to find the ground. If one doesn't wish to go down, toss up the cat. The cat always lands on it's feet, and so long as it's declawed, where it lands is down. If it's not declawed, where it lands may be up. The duck loves wet weather, and when in doubt and lost in the clouds, toss out the duck and follow it. The dog is there to bite the pilot's hand if he touches anything.
Why not simply make the airplane a UAV...unmanned aerial vehicle? Take man out of the equation completely? Put autothrottles in Cessnas, and create Category III landing capability in the Cherokee? Why have the pilot at all?
We have simple mechanisms in airplanes because they tend to be most reliable. We learn the foibles of instrumentation because all instruments lie to us. Learning the errors is part of learning to fly instruments. Perhaps you have never been IMC and had electronic displays go dark; I have, and I understand perfectly why the revision for a full EFIS cockpit is to go to manual "steam gauges," or instruments. When all the gee-whiz gadgets go south, then one reverts to the stuff that works.
As we continue to see progression into magenta lines, GPS for navigation in all counts, and increasing sophistication in the cockpit, we continue to see a degradation in the skills of the pilot body at large. Increasingly I note that without a magenta line to follow on a display, some pilots couldn't find their backside with both hands, and without automation, some are completely lost. The availability of GPS in the cockpits of even the most simple light airplanes today with quasi-FMS capability does offer some wonderful capabilities, but also heralds laziness in learning and practice.
A company for whom I worked some years ago fired a chief pilot. He captained a medical flight into a remote high desert airport in mountainous terrain. The approach was nearly always flown at night, because night equated to instrument conditions. His altimeters worked perfectly, but he was quite convinced the field elevation was a thousand feet lower than it actually was. He set the alerter to field elevation, expecting to have it ring out a thousand feet above the field.
His charts gave the correct information, and he was flying an instrument approach. He had flown there before, but none the less, for some unknown reason, set field elevation a thousand feet low. A medic who had made the flight many times before caught the error. Once on the ground the medical crew refused to fly with the man, and within 24 hours he was released from employment.
Pilots make errors but GPS wouldn't have changed his view any more than a barometric altimeter. The airplane was quite capable with GPS on board, FMS, electronic displays, etc, yet he still made the error. Until one elects to go completely automated or unmanned, one will continue to have errors. I can tell you, however, that even in unmanned situations, the result isn't as peachy as you might think. Highly sophisticated unmanned assets are often not where they are supposed to be...and in fact are often off their altitude by a considerable value.
We put pilots in airplanes and train for a good reason, just as we continue to use barometric altimetery vs. GPS altitude, for good reason.