C. Barus — Resistance of Stressed Glass. 



349 



Even under favorable conditions these data are only quali- 

 tatively satisfactory. They are important, however, because 

 they indicate that at 300°, the diminution of resistance due to 

 traction is not larger in numeric value than at 100° ; and since 

 this would be the case if the decrements 8R observed were 

 due to elastic change of dimensions, I have here in hand addi- 

 tional evidence against this assumption. 



Table III. — Resistance of stretched glass at 360°. 



Apparatus. 



P 



Maximum 

 oscillation. 



cm. 



R 



6R 



10 3 x6R/R 





kg. 



ohms. 



ohms. 





I 



6 



•50 



13000 



—30 



-2 





10 



•50 



. - - - 



— 30 



-2 





15 



•70 





— 50 



-4 





19 



•70 







-50 



-4 



III 



6 



1-00 



17000 



-15 



-1 





10 



1-40 



- . _ 



-22 



— 1 





15 



2-00 







— 31 



-2 



The present experiments are attended with much annoyance. 

 As the load increases, the tube is apt to break in such a way 

 as to spill the hot mercury ; and with all reasonable care 

 several tubes are usually sacrificed before a full series of obser- 

 vations can be obtained. 



11. The above paragraphs summarized, prove that a solid 

 electrolyte like glass is a better conductor of electricity (i. e. 

 manifests smaller specific resistance), when in a state of strain 

 (traction, torsion), than when free from strain. Inasmuch as 

 the necessary concomitant of conduction in this case is molecu- 

 lar decomposition* and recombination, stress of the given kind 

 must promote such decomposition. The rate at which molec- 

 ular reconstruction occurs per unit of volume increases nearly 

 proportionally to the intensity of stress ; and it may in case of 

 traction carried as far as the limit of rupture of glass amount 

 to an increment of one per cent. In case of torsion the effect 

 is not much larger than about 1/10 of this ; and the increased 

 break-up due to torsion is therefore studied with greater diffi- 

 culty. The influence of temperature in changing the value of 

 the electrolytic effect of stress is not marked. So far as ob- 

 served the same pull per unit section does not increase the 

 conductivity of glass more at 350° than at 100°, if indeed it 

 increases it as much. 



* It is best to avoid the term dissociation here. The term molecular recon- 

 struction is used in preference. 



