28S Rood — Experiments on High Electrical Resistance. 



weather, the hygrometer having a range from 12 to 30 per 

 cent of moisture. At the end of this paper certain determi- 

 nations that were made in damp weather are added. 



Preliminary expe7'iments. — The case of a sealed glass tube 

 l m 30 c in length, will illustrate partially what has been said 

 about the behavior of electricity moving over a bad conductor. 

 When the tube was made to establish connection between a 

 Leyden jar charged to a potential of 12,000 volts and the 

 electrometer, the electricity advanced slowly over its entire 

 length, the electrometer not being at all affected for some 

 time. In about five minutes it gave regular strokes ; the tube 

 had assumed the stationary condition and had nearly the same 

 potential over its whole length, except near the electrometer, 

 where for 3 or 4 cm it was lower. For half an hour it was 

 allowed to deliver charges to the electrometer with much regu- 

 larity, the rate being observed with a stop-watch. The elec- 

 trometer was then connected with the middle of the tube ; at 

 first the rate of discharge was greatly increased, as the poten- 

 tial was higher here than at the drained end, and the electricity 

 was flowing into the electrometer from both sides, but in 

 about five minutes the flow had fallen nearly to the old rate 

 and become quite regular. An air space of a centimeter was 

 now made between the tube and the electrometer ; the latter 

 continued to give discharges, but more slowly; the potential of 

 the middle part that had been somewhat drained was rising 

 and acted inductively on the electrometer, and this process 

 continued for some minutes, till the original level had been 

 established. The electrometer was then again connected with 

 the farther end of the tube and matters allowed to get into a 

 stationary condition. The strokes having become regular, the 

 middle of the tube was connected with the ground by wet 

 paper and wire. The electrometer strokes however continued 

 for a considerable time, but with diminishing frequency, till 

 the supply of electricity at the farther end of the tube had 

 exhausted itself. In this and in all similar experiments it is to 

 be understood that the two ends of the conducting body were 

 provided with metallic armatures, tinfoil fastened with thread. 

 A great many experiments were made on conductors of this 

 class without any armatures at all ; the tinfoil however facili- 

 tates matters, and makes the flow more regular. 



All this illustrates the case of a glass tube with rather good 

 insulating properties, and I pass on now to more ordinary 

 samples. A sealed tube of hard Bohemian glass, l m 30° in 

 length, was washed in water, dried in sunshine and connected 

 with jar and electrometer. In a second or two the electro- 

 meter began its discharges, the rate being about three seconds 

 per stroke. The entire tube was then heated quite hot with a 



