24 8 Table 226 



OPTICAL PYROMETRY 



(The following discussion is abbreviated from Dushman, Rev. Mod. Phys., 2, 387, 1930.) 



Data on various substances are now available by which accurate determina- 

 tions may be made of the temperature of an emitting surface. For a compre- 

 hensive review see Lax and Pirani, Handb. Phys., 19, 1-45 ; 21, 190-272, Julius 

 Springer, Berlin, 1929. Data on total radiation from various bodies have been 

 summarized by Coblentz (I.C.T. 5, 238-245, 1929). 



Undoubtedly the most accurate method consists in determining the brilliancy 

 (candles/cm*). The temperature coefficient (dB/B)/(dT/T) where B = 

 brightness in international candles/cm 2 , for W varies from 22.75 at iooo K. 

 to 8.45 at 3000 K. (Jones, Langmuir, Gen. Elec. Rev., 30, 310, 354, 408, 

 1927). With the exception of electron emission and the rate of evaporation, 

 the candlepower shows the greatest temperature variation. 



While the values of B as a function of T are available for several substances, 

 a fair approximation is possible to the true T of a substance for which the 

 average luminous emissivity is unknown by the following methods. (Average 

 luminous emissivity means the ratio of the total normal brightness to that of a 

 black-body at the same T. For W this value is from 0.464, r=iooo° K. to 

 0.440, 7 = 3000° K., Forsythe, Worthing, Astrophys. Journ., 61, 126, 1925). 



With a photometer determine the temperature T c at which substance emits 

 light of the same color as a black-body — a value higher than the true tempera- 

 ture. The following table is due to Worthing-Forsythe : 



The brightness may be compared with a standard lamp for a given wave 

 length, usually 0.665^. The temperature found is called the brightness tempera- 

 ture, T s , lower than the true temperature, increasingly so with decrease in e x , 

 the emissivity for this wave length, and increasing with the temperature. Thus 

 iorzv, e x for A = o.665/x varies from 0.456, T— 1000 ° K. to 0.415, 7 = 3000° K. 

 and the observed values T s are given in the preceding table. From these 

 determinations of T c and T s fairly approximate values of T may be deduced. 



It is possible to determine the actual value of e x by methods described by 

 both Langmuir, Worthing and Forsythe, then calculate T, the true tempera- 

 ture from optical pyrometer measurements of T. See Forsythe, Journ. Opt. 

 Soc. Amer., 16, 307, 1928; Journ. Amer. Ceram. Soc, 12, 780, 1929; Foote, 

 Fairchild, Symposium on Pyrometer, Amer. Inst. Mining and Metallurg. 

 Eng-> 33 8 > 3 2 4, 1920; also loc. cit, p. 291, 367, 285. 



At temperatures below 1100-1200° K. optical methods become impractical, 

 and radiation in watts radiated per unit area or resistance must be used. Data 

 on energy radiated as a function of the temperature have been summarized by 

 Lax and Pirani, Handb. Phys., 21, 236-240, for W, Mo, Ta, Pr, Os, Au, Ni, 

 Fe, C, Ag, Cu, and Zr. 



Smithsonian Tables 



