302 



comiiensator as shown in Fig I. It consists of a fine 

 platinum wire A, wliicli is partially immersed in mer- 

 cury B. When the bulb is heated the mercury in the 

 capillary tube expands and short circuits the platinum 

 loop, thus diminishing the resistance of the circuit. 

 This balances a change in e. m. f., due to a rise of 

 temperature of the cold junction. 



All the contacts of the different parts of the cir- 

 Fig. 1. cult should be carefully made, and wherever possible 



this should be done by soldering. The hot junction of 

 the wires used in the couple should be fused together. For easily fusible 

 metals, such as copper, this can be done in the Bunsen flame, but for 

 platinum oxygen is reciuired. Platinum may also be fused in the electric 

 arc. At the cold junction the lead wires should be soldered to the thermo- 

 element wires. The wires composing the couple, which are subjected to 

 high temperatures, should be insulated throughout their entire length by 

 glass tubes or pipe stems. Asbestos thread may also be used for tempera- 

 tures below 1300° C. Small fire clay tubes pierced by two holes may also 

 be procured and are very convenient. For industrial work the couple 

 should be inclosed by an iron or porcelain tube. The former should not 

 l)e us-ed for temperatures over 800° C. 



Radiation Pyrametry. — From the fact that the intensity of light emit- 

 ted from a body increases very rapidly with rise of temperature the op- 

 tical method is well adapted to the measurement of high temperatures. 

 For example, the luminous intensity of the red part of the light emitted 

 by a body of 1500° C. is 130 times the intensity of 1000° C. and at 2000° C. 

 it is more than 2100 times as great as at 1000° C. It thus appears that a 

 comparatively rough measurement of the luminous intensity of an incan- 

 descent body would give a pretfj- accurate measurement of its temperature. 

 This conclusion, however, is modified by the fact that different bodies at 

 the same temperature emit very different amounts of radiant energy. Th«? 

 radiating power of a body depends not only upon the temperature but also 

 the composition and nature of the surface. In order that the radiation and 

 optical methods can be used for comparison of temperatures it is necessary 

 that the effect of differences of surfaces be eliminated. This can be done 

 by rediicing the radiation from all surfaces to the radiation that would 

 occur from some ideal surface arbitrarily taken as a standard of com- 

 parison. 



