348 



G. Bar us — Resistance of Stressed Glass. 



cated by the inside arrows ; but these currents are considerably 

 reenforced by the action of the element sodium amalgam /hot 

 glass / mercury, as shown by the arrows crossing the tubes a 

 and b. The electromotive force of this element is easily found 

 by reversing the action of E. In an actual experiment I meas- 

 ured KaHg/hot glass / Hg= 1*4 volts, a datum somewhat 

 affected by polarization and depending for its value on the 

 strength of the amalgam and the purity of the mercury. 



Besides this large electromotive force there is another of 

 smaller value, due to the fact that the tubes a and b with 

 appurtenances, represent two elements switched against each 

 other. The currents are indicated by the outside arrows in the 

 diagram, and they are necessarily so circumstanced as not to 

 flow through the galvanometer G differentially. Their occur- 

 rence is therefore a serious and annoying disturbance, such that 

 measurements at 360° can not, without unreasonable painstak- 

 ing, be made with the same accuracy as measurements at 100°. 

 I measured the electromotive force in question, as about *2 

 volt; but it is necessarily variable even as to sign, containing 

 as it does the polarization inconstancy of both elements. 



Figure 3. — Apparatus for 360°. 



Since the resistance of glass near 360° is enormously low 

 relative to its value at ordinary temperatures (in some practical 

 cases the apparatus showed less than 10U0 ohms), the extran- 

 eous electromotive force jB'can be withdrawn altogether. The 

 present measurements of the electrical effect of traction are 

 therefore made with the NaHg / hot glass / Hg element in the 

 apparatus, figure 3. Notation being as above 2p 1 = *53 em , 

 2 ( 2 =4O cm . The small resistance at the boiling point is not 

 available ; owing to the formation of bubbles at the surface of 

 contact between mercury and glass the resistance is too variable 

 even for approximate measurement. Hence I observed at a 

 lower temperature, encountering somewhat larger resistances R. 



