In preparing the temperature scale of a pyrometer for practical use, the instrument is subjected successively to a number of the temperatures indicated in the table, and in this manner several fixed points are established on its scale. The space between these points is then suitably subdivided to represent intermediate temperatures.

Table of Fixed Points.

Substance.Physical Condition.Deg.Deg.
Cent. Fahr.
Water (ice)At Melting Point0 32
Water” Boiling ”100 212
Aniline”   ”   ”184 363
Naphthalene”   ”   ”218 424
Tin” Melting ”232 449
Lead”   ”   ”327 620
Zinc”   ”   ”419 786
Sulphur” Boiling ”445 833
Antimony” Melting ”631 1167
Aluminium”   ”   ”657 1214
Common Salt”   ”   ”800 1472
Silver (in air)”   ”   ”955 1751
Silver (free from oxygen)”   ”   ”962 1763
Gold”   ”   ”1064 1947
Copper (in air)”   ”   ”1064 1947
Copper (Graphite covered) ”   ”   ”1084 1983
Iron (pure)”   ”   ”1520 2768
Palladium”   ”   ”1549 2820
Platinum”   ”   ” 1755 3190

It is necessary to point out that the figures given in the table refer only to pure substances, and that relatively small quantities of impurities may give rise to serious errors. The methods by which the physical condition to which the temperatures refer may be realised in practice will be described in the succeeding chapter.

National Physical Laboratory Scale.—Exact agreement with regard to fixed points has not yet been arrived at in different countries, and an effort to co-ordinate the work of the National Physical Laboratory, the United States Bureau of Standards, and the Reichsanstalt, with a view to the formation of an international scale, was interrupted by the war. In 1916 the National Physical Laboratory adopted a set of fixed points on the Centigrade thermodynamic scale, in conformity with which all British pyrometers have since been standardised. It will be seen that the figures differ very slightly from those given in the previous table, which represent the average results of separate determinations in different countries.

National Physical Laboratory Scale (1916)

Substance.Physical Condition.Deg.Deg.
Cent. Fahr.
Water (ice) At Melting Point0 32
Water ” Boiling  ”  (760 mm.) 100 212
Naphthalene ”   ”   ”   ”217·9 424
Benzophenone ”   ”   ”   ”305·9 582
Zinc At Melting Point419·4 787
Antimony ”   ”   ”630 1166
Common Salt ”   ”   ”801 1474
Silver (in reducing atmosphere) ”   ”   ”961 1761
Gold ”   ”   ”1063 1945
Copper (in reducing atmosphere) ”   ”   ” 1083 1982

For higher temperatures the melting points of nickel (1452° C.) and palladium (1549° C.) are employed, but the accuracy in these cases is not so certain as with the substances named in the table. A useful point, intermediate between copper and nickel, has been established by E. Griffiths, and is obtained by heating nickel with an excess of graphite, when a well-defined eutectic is formed which freezes at 1330° C., or 2426° F.

Temperatures above the Present Limit of the Gas Thermometer. —As it is not yet possible to compare an instrument directly with the gas thermometer above 1550° C., all higher temperatures must be arrived at by a process of extrapolation. By careful observation of a physical change at temperatures up to the limit of 1550° C., the law governing such change may be discovered; and assuming the law to hold indefinitely, higher temperatures may be deduced by calculation. An amount of uncertainty always attaches to this procedure, and in the past some ludicrous figures have been given as the result of indefinite extrapolation. Wedgwood, for example, by assuming the uniform contraction of clay, gave 12001° C., or 21637° F., as the melting point of wrought iron, whereas the correct figure is 1520° C., according to the gas scale. Even in recent times, the extrapolation of the law connecting the temperature of a thermal junction with the electromotive force developed, obtained by comparison with the gas scale up to 1100° C., led Harker to the conclusion that the melting point of platinum was 1710° C., a figure 45 degrees lower than that now accepted. The laws governing the radiation of energy at different temperatures, however, appear to be capable of mathematical proof from thermodynamic principles, and temperatures derived from these laws are in reality expressed on the absolute or thermodynamic scale. Extrapolation of these laws, when used to deduce temperatures by means of radiation pyrometers, appears to be justified; but it is still desirable to extend the gas scale as far as possible to check such instruments. Assuming the radiation laws to hold, it is possible to determine the highest temperatures procurable, such as that of the electric arc, with a reasonable degree of certainty.