200 On the Condvcting-powe)- of Insulators at high Temperatures* 



ing the two places where platinum and glass are in contact, 

 must be carefully avoided ; for this reason the sand-bath was 

 always employed here ; the polarizations produced by the actions 

 of galvanic currents are so strong that I could never observe any 

 disturbance caused by thermo-electric currents. 



By means of a commutator I investigated the relation which 

 exists between the polarization, the temperature, and the inten- 

 sity of the current. Two glass bars were heated in the same 

 sand-bath to 250*^ C, then through the one a current from a 

 single element, through the other a current from two elements 

 of a zinc and platinum battery was passed, until the polarizations 

 each attained a maximum ; both bars were afterwards made to 

 produce opposite ciFects upon a galvanometer with which they 

 were connected by means of a commutator. The polarization 

 produced by two elements was strongest, and similarly it con- 

 tinued to increase according as four, six, or ten elements were 

 employed. The magnitude of the polarization does not depend 

 upon the acting force, but upon the actual intensity; and as 

 this is still small in this case, it explains why, with so great an 

 electromotive force, an increase of polarization always occurred. 



Further, the same current was passed successively through a 

 glass bar which was heated to 130°, and through a second heated 

 to a cherry-red heat. By turning the commutator, it was found 

 that the platinum wires which were heated to 130° were most 

 strongly polarized. Accordingly, heated glass before it fuses 

 behaves in every respect like other electrolytes; so that Bec- 

 querel's* pyro-electric battery differs from common batteries 

 in no other respect, except that in the former the electrolytes 

 have to be prepared by heat, whereas in the latter they are always 

 ready at hand. We cannot estimate what advantage may be 

 thus gained, inasnmch as the same intensity of current corre- 

 sponds, in both kinds of batteries, to the same destruction of 

 metal f. 



* Comptes Rendus, vol. xxxviii. p. 905. 



t Since com})leting, and for the most part writing down, the ahove ex- 

 periments, I have received a memoir from Buff, which has yet to appear in 

 the Ann.der Chem. und Pharm. (vol. xc. p. 25/), wherein the conducting 

 power of glass is also examined. [See p. 12 of the present volume of this 

 Journal.] Although the experiments have been conducted quite differently, 

 our results agree perfectly, e. g. in reference to the commencement of con- 

 ducting power, to the polarization, and to its dependence upon the inten- 

 sity of the current and the temperature. Whilst Buff arrived at the last 

 by single measurements, wherein he had to assume the equality of the pola- 

 rizations which, according to his own statement, had not quite constant 

 values when the cuiTcnts had different intensities, I discovered the same 

 dependence by means of the commutator ; nevertheless, the; differences 

 between the polarizations at different temperatures are so great, according 

 to Buff's measurements, that we willingly overlook the small deviations 



