350 0. Barus — Resistance of Stressed Glass. 



Again the traction effect in case of electrolytic conduction, 

 being a decrement of resistance, is of the opposite sign of the 

 traction effect in cases of metallic conduction* (increment of 

 resistance). The former is also of decidedly greater magnitude. 

 If, therefore, conduction in metals is essentially the same phe- 

 nomenon]- as in electrolytes, then the soft metallic state must 

 be singularly well adapted to promote molecular reconstruction. 

 This fine adaptation of structure is destroyed by strains of 

 dilatation, by heat, by alloying,:]; etc. In the data given, the 

 electrical traction-coefficient, as well as the electrical tempera- 

 ture-coefficient (resistance), are similar in sign and in relative 

 magnitude, both in metals and in electrolytes. They are posi- 

 tive in metals and small, negative in electrolytes and large. 

 This is additional evidence in favor of a volume effect dis- 

 cussed at some length elsewhere. 



12. The chief result of the present paper is the emphasis 

 thrown on the fact that, independently of the passage of cur- 

 rent, such a solid as glass must be conceived as undergoing 

 spontaneous molecular reconstruction at all temperatures. For 

 if the reconstruction in question were superinduced by the 

 electric field, then the current passing would vary at a power 

 higher than the first of electromotive force ; whereas it may 

 be taken for granted that currents of the intensity of those 

 discussed above, pass through glass in accordance with Ohm's 

 law.§ Recently J. J. Thomson! among many results of his 

 development of the Lagrangian function, investigated an ex- 

 pression for the number of times, n, the electric field is dis- 

 charged at any point, in case of conduction through either 

 metals or electrolytes. If X be the specific conductivity, K the 

 specific inductive capacity of the medium, then n=2rrftX/ JS~; 

 where ft is a coefficient the value of which is less than unity 

 and depends on the relative time of destruction and existence 

 of the electric field. Accepting provisional values for ft and 

 K, Thomson computes a table of values for the superior limit 

 of n, in cases both of metals and of electrolytes. From this 

 table it appears that n for mercury for instance, is less than 

 8X10 16 . Similar values for the limit of n in case of glass at 

 the above temperatures of observation, 100°, 200°, 360°, may 

 be deduced. In round number the specific resistances of glass 

 at the temperatures stated were, in ohms, 10 9 , 10 5 , and 5Xl0 3 , 



* Mousson: Neue Schweizer, Zeitschr., xiv, p. 33, 1855; H. Tomlinson : Proc. 

 Eoy. Soc, xxv, p. 451, 1876; id., xxvi, p. 401, 1877. 



f J. J. Thomson : Applications of Dynamics to Physics and Chemistry, p. 296 ; 

 Macmillan, 1888. 



i This Journal, xxxvi, p. 427, 1888. 



§ Fitzgerald and Trouton (Rep. Br. Assoc, 1886, p. 312) show that electrolytes 

 obey Ohm's law accurately. 



|| J. J. Thomson: 1. c, p. 299. 



