366 C. Barus — Expression in Thermo- electrics. 



Art. XL. — An Elementary Expression in Thermo-electrics / 

 by Carl Barus. 



1. The thermo-electric equation due respectively to Aveua- 

 rius* and to Tait,f does not reproduce the observations satis- 

 factorily when long- temperature ranges 01000° C.) are dealt 

 with.:}: "Within an interval of a few hundred degrees, how- 

 ever, the equation is usually in good accord with the experi- 

 mental results. It is desirable, therefore, to endeavor to find 

 a more general relation, from which the Tait equation may be 

 derived as an approximation. This is my object in the present 

 paper. I have tried to throw some light on the subject with- 

 out explicit reference either to the Peltier or to the Thomson 

 effect, and in a deductively experimental way, as follows : 



2. In an investigation of the relation between the thermo- 

 electric power and the specific resistance of steel varying with 

 hardness, Dr. Strouhal and I found a linear relation§ to obtain. 

 Some 90 states of hardness were examined, all tempered with 

 scrupulous care. I have since endeavored to test this relation 

 further, by making as many different platinum alloys as I 

 could (54 in all), but found || that marked changes of specific 

 resistance (10-65 microhms) were as common, as marked 

 changes of thermoelectric power were rare, and no law was 

 apparent. Nevertheless I am unwilling to concede that the 

 results for steel in their bearing on the subject in hand are 

 devoid of suggestion ; at least I believe that the relation which 

 holds for temper will also hold for temperature because in the 

 latter case the structure of the metal is to a less extent inter- 

 fered with. 



3. When long ranges of temperature are considered the re- 

 lation of metallic resistance^ to temperature is not such that 

 its nature can be closely enough inferred, experimentally. 

 Electrolytic resistance, however, presents a promising case. 

 The thermal variation of the resistance of an electrolyte, as I 

 found both in the case of aqueous solutions (zinc sulphate) 



*Avenarius: Pogg. Ann., cxix, p. 406, 1863. 



fTait: Trans. R. S. Edinbgh., xxvii, p. 125, 1872-73. 



% Prof. Tait does not claim that the formula will do so. 



§ Wied. Ann., xi, p. 969 et seq., 1880. It will be remembered that the resist- 

 ance of hard steel is 3 to 4 times as large as that of soft steel. Cf. Phil. Mag., 

 (5), viii, p. 341, 1879. 



|j Barus: this Journal, xxxvi, p. 427, 1888; Bull. U. S. G. S., No. 54, pp. 143, 

 146, 1889. 



Tf The striking results of Dewar and Fleming (Phil. Mag., xxxvi, p. 271, 1893), 

 from which an absence of metallic resistance at the absolute zero of temperature 

 may be inferred, are rather of the nature of an initial tangent to the curve above 

 in question. . 



