A thermometer will indicate short-term changes when used below 100°C, and it will recover an error of 0.01 to 0.02 degrees in several days. Admittedly, these are small amounts of error. Greater amounts of temporary error can be exhibited when a thermometer is used for relatively high temperature readings and then is immediately used for relatively low temperature readings. Thus, it may be advisable to separate thermometers not only by their temperature ranges, but also by the temperature ranges for which they are used. This practice is especially important if your work requires accurate thermometric readings.

Long-term changes are generally caused when a thermometer is maintained at high temperatures for extended periods of time (> several hundred hours).16 Such conditions can cause errors as great as 12°C. It is also possible to cause permanent changes from repeated cycling at lower temperatures between -30°C and room temperature.

It is uncommon to observe an alteration due to a temperature-induced change in the readings along selective parts of the thermometer scale. Typically they are more of a dropping of the entire scale reading. Thus, periodically checking the zero point with an ice bath, or checking the water boiling point with a steam bath, can ensure the accuracy of your thermometer. It is not necessary to recalibrate the entire stem of the thermometer.

2.5.7 Thermometer Calibration

Steam bath calibration for the liquid-steam point of H2O is not easy for many labs to use because of its expense, and the equipment is difficult to use. In addition, the user needs to account for atmospheric pressure variations and the effects of local variations of gravity on the barometer. Fortunately, the ice-point calibration is easy to set up and use.

Any thermometer with 0.0°C on either its main or auxiliary scale can be calibrated with an ice-bath calibration apparatus that can be assembled with the following material:

1.A 1- to 1.5-liter Dewar flask, cleaned and rinsed with distilled water

2.Crushed ice made from distilled water

3.A (clean) siphon to remove excess melted ice

Fill the Dewar approximately one-third full with distilled water. Place the end of the siphon tube to this level, and then fill the Dewar to the top with crushed ice.

The thermometer should be inserted so that the thermometer is immersed to the 0.0°C level (if calibrating a complete-immersion thermometer) or inserted to the immersion line (if calibrating a partial-immersion thermometer). The thermometer is likely to require support to maintain its proper position. Let the entire apparatus sit for 15 to 30 minutes, to reach equilibrium. Periodically add more ice, as needed, and remove any excess water with the siphon. If the ice is kept clean and

columns. If both thermometers have breaks in their liquid columns, it is almost impossible for them to agree across a range of temperatures. Detection of liquid breaks is virtually guaranteed. Unfortunately, this technique will not identify which thermometer is at fault.

There are two techniques for rejoining separated liquid ends: heating and cooling. The option chosen will depend on the location of the break, as well as on the existence and location of contraction or expansion chambers.

Generally it is better (and safer) to cool the liquid in thermometer than it is to heat it. First, try to cool the thermometer with a (table) salt-and-ice slush bath.* This method should bring the liquid into the contraction chamber or bulb. Once the liquid is in the chamber or bulb, it should rejoin, leaving the air bubble on top. If there is not a clean separation of the air bubble, it may be necessary to softly tap the end of the thermometer. This tapping should be done on a soft surface such as a rubber mat, stopper, or even a pad of paper. Alternatively, you may try swinging the thermometer in an arc (such as a nurse does before placing it in your mouth)." Once joined, the liquid in the thermometer can slowly be reheated.

If the scale or design of the thermometer is such that these methods will not work, try touching the bulb end to some dry ice. Because dry ice is 38°C colder than the freezing point of mercury, you must pay attention and try to prevent the mercury from freezing. As a precaution, when reheating, warm from the top (of the bulb) down to help prevent a solid mercury ice plug from blocking the path of the expanding mercury into the capillary. With no path to expand into, the expanding liquid may otherwise cause the bulb to explode.

An alternate technique to rejoin broken liquid columns is to expand the liquid into the contraction or expansion chamber by heat. Be careful to avoid filling the expansion chamber more than two-thirds full, as extra pressure may cause the top of the thermometer to burst. Never use an open flame to intentionally heat any part of a thermometer as the temperature from such a source is too great and generally uncontrollable.

If the thermometer has no expansion or contraction chamber, heating the bulb should not be attempted because there is no place into which the liquid may expand to remove bubbles. Occasionally it may be possible to heat the upper region of the liquid along the stem (do not use a direct flame; use a hot air gun or steam). Observe carefully, and look for the separated portion to break into small balls on the walls of the capillary. These balls may be rejoined to the rest of the liquid by slowly warming the bulb of the thermometer so that the microdroplets

'This assumes that the location of 0°C is sufficiently close to either the bulb or the contraction chamber; otherwise a different slush bath may be required.

fMedical thermometers have a small constriction just above the bulb that serves as a gate valve. The force of the mercury expanding is great enough to force the mercury past the constriction. However, the force of contraction is not great enough to draw the mercury back into the bulb. The centrifugal force caused by shaking the thermometer is great enough to draw the mercury back into the bulb.