444 On the Electrical Insulating -power of various kinds of Glass. 



in communication with the inner coating of a Leyden jar 

 charged as in the preceding experiments, and touched the lower 

 coatiDg with a conductor which communicated with the outer 

 coating of the Leyden jar. I could then receive sparks from 

 the coating of the plate at intervals of less than a second, until 

 the Leyden jar was nearly discharged. When the communica- 

 tion between the plate and the jar was kept up constantly, the 

 latter had entirely discharged itself in twenty minutes. The re- 

 sult with the plate No. 2, under exactly the same circumstances, 

 was widely different. When the electricity that first arose in 

 the lower coating by induction had been got rid of, I had to 

 wait several minutes before a scarcely perceptible spark could be 

 obtained, and with continuous communication between the two 

 coated surfaces of the plate and those of the Leyden jar, the 

 latter, after an interval of an hour, was still found to be strongly 

 charged. 



It has been remarked that glass, when newly blown, has a 

 conducting outer crust, and only loses its conducting-property 

 by wear and tear after being used for some time. In order to 

 assure myself that the strong conducting-property of the plate 

 No. 5 did not arise from some such cause, I repeated the above- 

 mentioned experiment, only with the difference that the outer 

 coating of the jar communicated with a narrow rim of tinfoil 

 round the edge of the whole plate. It is clear that if the con- 

 ducting-power of the plate depended on its surface and took 

 place along that and under the varnish, the discharge of the 

 Leyden flask ought now to take place about twice as rapidly 

 as before; if, on the contrary, it depended principally on 

 the conducting-power of the glass itself, the discharge ought 

 to take place far more slowly than in the former experiment. 

 The latter was found to be the case ; for the Leyden jar was 

 still strongly charged after an interval of three quarters of an 

 hour. 



If we now go back to the analyses in order to obtain from 

 them some indication of what caused the difference of conduct- 

 ing-power in the substance itself of the different glasses, it is 

 evident that the phenomenon cannot be satisfactorily explained 

 either by the amount of silica or of lime contained, whether we 

 take account of their percentage or of their equivalent propor- 

 tions. But there is an ingredient the quantity of which in the 

 glass continually increases in the measure in which its electrical 

 insulating-power decreases : that ingredient is soda ; and I con- 

 sider that I may combine the results of this research in the fol- 

 lowing statements : — 



(1) The non-conducting property of the glass does indeed in 

 some measure depend upon the more or less non-conducting 



