43'S APPENDIX. 



TABLE 58O. 

 TEMPERATURE MEASUREMENTS. 



The ideal standard temperature scale (Kelvin's thermodynamic scale, see introduction, p. xxxiv) is in- 

 dependent of the properties of ,nv substance, and would be indicated by a gas thermometer using a perfect 

 gaT?he scale indlaLl by any actual .as can be corrected if the departure of that gas from a perfect gas 

 be known (see Table 206, p. 195,- also Buckingham, Bull. Bur. Standards, 3, 237). The thermodynamic 

 omoto of the constant-pressure scale at any temperature is very nearly proportional to the constant pres- 

 STeT^h thegas is kept and that for the const ant- volume scale is approximately proportional to the 

 fdUal pressure at the ice-point. The gas thermometer has been earned up to the melting point of palla- 

 dium 1822 K (1549 C) (Day and Sosman, Am. J. Sc., 29, p. 93. iQio). 



A proposed international agreement divides the temperature scale into three intervals. The first inter- 

 val --40 to 450 C uses the platinum resistance thermometer calibrated at the melting point of ice, o C, 

 at saturated steam, 100 C, and sulphur vapor, 444-6 C, all under standard atmospheric pressure. Points 

 on the temperature scale are interpolated by the Callendar formula?: 



where t is the temperature, R, the resistance, Pt, the platinum temperature, and 5, a constant. 



Temperatures in the second interval are measured by a standard platinum-platinum-rhodium couple cal- 

 ibrated say at the freezing points of zinc, 419-4 C, cadmium, 320.9 C, antimony, 630 C, and copper free 

 from oxide, 1083 C. These points furnish constants for the formula, e =a + bt + ct 2 (see Sosman, Am. J. Sc., 

 30, p. I, 1910). 



For the region above 1100 C most experimenters base their results upon certain radiation laws. 

 laws all apply to a black body and the temperature of a non-black body cannot be determined directly with- 

 out correction for its emissive power. For standard points the melting- points of gold, 1336 K and palla- 

 dium 1822 K, are convenient, 



Above 1336 K the optical pyrometer is generally used with a calibration based upon Wien's equation 



By comparing the brightness of a black body at two temperatures and applying this equation, the following 

 formula results: 



"- 



where R is the ratio of the brightnesses, \, the wave-length used, TI and Ti, the two temperatures, and C2 

 = 14.250 n deg. Thus if R is measured and one temperature known, the other can be calculated. 



A table of the standard fixed points is given in Table 207, p. 195. With these determined there comes the 

 difficulty of maintaining this temperature scale both from the standpoint of the standardizing laboratory 

 and the man using the temperature scale in the practical field. In the region of the platinum-resistance 

 thermometer and the thermocouple, standards of either can be obtained from the standardizing laboratories 

 and used in checking up the secondary instruments. It is not very difficult to actually check up a resist- 

 ance thermometer at any one of the standard points in the region 40 C to +450 C. It is a little more 

 difficult to check the thermocouple in the region 450 C to 1100 C. Most of the standard fixed points in 

 this region are given by melting points of metals that must be melted so as to avoid oxidation. This re- 

 quires a neutral atmosphere, or that the saTnple be covered with some flux that will protect it. 



Both the gold and the palladium, used to calibrate the scale above 1300 K, can be successfully melted in 

 a platinum wound black -body furnace. The whole operation can be carried out in the open air, requiring 

 neither a vacuum nor neutral atmosphere within the furnacs. But because of the trouble necessitated by a 

 black-body comparison, much time can be saved if a tungsten lamp with filament of suitable size is stand- 

 ardized so as to have the same brightness for a particular part of the filament, when observed with the 

 optical pyrometer, as the standard black-body furnace for one or more definite temperatures. With such 

 lamps properly calibrated, any one may maintain his own temperature scale for years, if the calibration 

 does not extend higher than that of the palladium point and the standard lamp is not accidentally heated to 

 a higher temperature. 



(See 1919 Report of Standards Committee on Pyrometry, Forsythe, J. Opt. Soc. of America, 4, p. 205, 

 i<)2o; The Measurement of High Temperatures, Burgess, Le Chatelier, 1912, The Disappearing Filament 

 Type of Optical Pyrometer, Forsythe, Tr. Faraday Soc., 1919.) 

 SMITHSONIAN TABLES. 



