es the degree of accuracy of high temperature measurements.

TABLE 7

ACCURACY OF HIGH TEMPERATURE MEASUREMENTS[7]

+------------------------+------------------------+
| Centigrade | Fahrenheit |
+-------------+----------+-------------+----------+
| | Accuracy | | Accuracy |
| Temperature | Plus or | Temperature | Plus or |
| Range | Minus | Range | Minus |
| | Degrees | | Degrees |
+-------------+----------+-------------+----------+
| 200- 500 | 0.5 | 392- 932 | 0.9 |
| 500- 800 | 2 | 932-1472 | 3.6 |
| 800-1100 | 3 | 1472-2012 | 5.4 |
| 1100-1600 | 15 | 2012-2912 | 27 |
| 1600-2000 | 25 | 2912-3632 | 45 |
+-------------+----------+-------------+----------+

Mercurial Pyrometers--At atmospheric pressure mercury boils at 676
degrees Fahrenheit and even at lower temperatures the mercury in
thermometers will be distilled and will collect in the upper part of the
stem. Therefore, for temperatures much above 400 degrees Fahrenheit,
some inert gas, such as nitrogen or carbon dioxide, must be forced under
pressure into the upper part of the thermometer stem. The pressure at
600 degrees Fahrenheit is about 15 pounds, or slightly above that of the
atmosphere, at 850 degrees about 70 pounds, and at 1000 degrees about
300 pounds.

Flue-gas temperatures are nearly always taken with mercurial
thermometers as they are the most accurate and are easy to read and
manipulate. Care must be taken that the bulb of the instrument projects
into the path of the moving gases in order that the temperature may
truly represent the flue gas temperature. No readings should be
considered until the thermometer has been in place long enough to heat
it up to the full temperature of the gases.

Expansion Pyrometers--Brass expands about 50 per cent more than iron and
in both brass and iron the expansion is nearly proportional to the
increase in temperature. This phenomenon is utilized in expansion
pyrometers by enclosing a brass rod in an iron pipe, one end of the rod
being rigidly attached to a cap at the end of the pipe, while the other
is connected by a multiplying gear to a pointer moving around a
graduated dial. The whole length of the expansion piece must be at a
uniform temperature before a correct reading can be obtained. This fact,
to