24 



Mr. Carl Barus on the 



Table X. — Breakdown of Resistance in the case of divers 

 Glass arrangements. Specific resistance at 215°, 19 x 10 6 . 



Glass in 



Sperm-oil. 



Gasolene. 



Petroleum. 



Thin mineral 

 machine-oil. 



Thick mineral 

 machine-oil. 



100 

 215 

 310 



500 

 1 





1400 

 1 

 •2 



300 

 I 

 •2 



1 



•05 



Deductions. 



20. It appears -from Table I. et seq., that the pressure- 

 coefficient does not decrease with the size of the pressure inter- 

 val &p. The tendency is rather in favour of an increase, but this 

 is obscure. Hence the results obtained are properties of 

 either the oil or of the glass, and not incidental disturbances. 

 Moreover, I am warranted in taking the means of the divers 

 pressure-coefficients found at any temperature as the probable 

 value at that temperature. 



21. Remarks on the discrimination of temporary and per- 

 manent effects have been made, §§11, 13. 



22. In order to obtain a clear insight into the data of 

 Tables I. to VIII., it is best to construct pressure-coefficient 

 as a function of temperature. It would be more in keeping 

 with the present work to coordinate this coefficient with re- 

 sistance ; but the latter data are not sufficiently accurate. In 

 fig. 5 the number of the series is marked on the curve. 



23. Turning attention first to the behaviour of the solid 

 insulator, it appears that the electrical resistance of glass in- 

 creases in very marked degree with the pressure which is 

 brought to bear on it. In my earlier paper *, in which I 

 subjected glass to tensile stress (presumably dilatational), I 

 found that the resistance had pronouncedly decreased. I also 

 found that the results at 360° were not as evident as at 100°. 

 Hence the present and the earlier observations, though 

 attacking the question from opposite points of view, are in 

 accord. I am thus justified in announcing that the chemical 

 equilibrium of a solid changes with each change of strain (§ 30). 



24. The tables further show that the pressure-coefficient of 

 resistance decreases with rise of temperature ; or that the 

 effect of pressure decreases algebraically as the number of free 

 ions available for the transfer of charges becomes greater. 

 In an earlier paper t I showed that the resistance of a liquid 



* American Journal, xxxvii. p. 349 (1889). 

 t American Journal, xl. p. 219 (1890). 



