286 Rood — Experiments on High Electrical Resistance. 



harden on the walls, and to a certain extent obstruct the flow 

 of the fluid by contracting their cross-section. This illustra- 

 tion corresponds to a polarizing action which often sets in, and 

 gradually diminishes the flow of the electricity, through or 

 over a substance of very high resistance, and the greater the 

 resistance the larger will be this counter action. What has 

 been said so far applies to the cases of glass, mica, paraffin 

 paper, silk and most artificial insulating substances. 



If we now imagine our tank to be quite narrow, offering 

 thus great resistance to the passage of the pitchy fluid, then at 

 the end where the fluid is introduced, the height or potential 

 will equal that of the supplying source, but as the distance 

 from this end increases, the potential falls and finally becomes 

 zero. This 1 case corresponds to the surface conduction of elec- 

 tricity by black sealing wax, jade and gutta percha. They all 

 furnish to the electrometer only small quantities of electricity, 

 which can only be drawn from the neighborhood of the reser- 

 voir, and here the polarization is very marked. 



Finally, imagine the walls of the tank to be still farther con- 

 tracted and that the pitch-like fluid refuses to enter them in 

 perceptible quantities, and we have the cases of ebonite, ordi- 

 nary rosin and exfoliated mica (in very dry weather). 



Evidently, if two tanks of the same length and cross-section 

 deliver equal amounts of the pitchy fluid in equal times, then 

 their conducting powers should be set as equal, and in general, 

 if there be no loss in transit, the amounts delivered per minute, 

 by tanks differing from each other, will be a measure of their 

 relative conducting powers. Finally, if the tanks almost 

 refuse to deliver up any of the pitchy fluid, still, if their 

 potentials or the heights of the fluid at various points can be 

 measured, in a manner analogous to that employed in the 

 study of the conduction of heat by solid bodies that have 

 reached the stationary condition, then it becomes possible to 

 determine their relative conducting powers, and by comparison 

 with somewhat better conductors, an estimate of their resist- 

 ance in standard units can be formed. 



Apparatus. — I modified a single leaf electroscope, so that it 

 became a serviceable electrometer for this kind of work. Its 

 plan may be seen in the diagram. A brass rod AR was sup- 

 ported on a bar of ebonite, which was placed on two rods of 

 the same substance fastened on a base of ebonite. The rods 

 and the bar were painted with melted rosin. The rod AR 

 carried an aluminum leaf, bluntly pointed below, and attached 

 above to the point R by a gold leaf hinge. Aluminum leaf 

 was used, it having been found that under the action of the 

 minute but very numerous electrical discharges, the gold leaf 

 slowly wasted away at its corners, which it always presented to 



