TROWBRIDGE. — ELECTRICAL DISCHARGES, 441 



the surface of the water inside the tubes which vaporize the water and 

 thus lead to an explosion. The layer of air conducts more readily than the 

 water. The same phenomenon can be shown by interposing a conductor 

 made of plumbago and infusorial earth of about ten thousand ohms re- 

 sistance between the terminals of the apparatus. A spark passes over 

 the surface of such a conductor through the air, if the length of such a 

 conductor does not exceed ten or twelve inches. The breaking down 

 resistance of atmospheric air for such high voltages between terminals 

 this distance apart is therefore less than ten thousand ohms. The same 

 phenomenon is observed in the case of paraffine oil, which exhibits between 

 pointed terminals a resistance of only a few thousand ohms for distances 

 of six to eight inches. The actual resistance after breaking down fills 

 to two or three ohms. This T have shown in a previous paper.* To still 

 further study this breaking down of air insulation I carefully measured 

 by Kohlrausch's Bridge method a liquid resistance consisting of diluted 

 sulphate of copper of about one thousand ohms. The terminals in the tube 

 were disks of copper about one square centimeter in area. It was found 

 that the spark preferred to jump through five centimeters of air to pass- 

 ing through the sulphate of copper. The entire range of my experiments 

 precludes to my mind the idea that this effect is due entirely to polariza- 

 tion of the electrodes. The air evidently breaks down with increasing 

 readiness when the electromotive force is increased beyond a certain 

 limit. This is true also of liquid dielectrics like oil. One of the most 

 striking experiments in this connection can be performed by coating 

 a board with a thin layer of plumbago, which is polished upon the 

 surface in such a manner as to make a resistance of about one 

 thousand ohms between broad terminal bands of copper. "When a dis- 

 charge under a difference of jwtential of one million volts passes between 

 the terminal bands which rest on the coated surface, the entire sur- 

 face is luminous. If an orange is interposed between terminals of the 

 apparatus which gives over two million volts, it glows in the dark like 

 a golden Chinese lantern, and the discharge passes beneath the rind and 

 over the liquid portion of the conductor. These experiments make it 

 clear to my mind that the disruptive effects of lightning in rending trees 

 and shattering structures is due in large measure to the expansion of con- 

 fined air or to the sudden formation of steam produced by the electric 

 spark preferring to break down the air to overcoming the resistance of the 

 poor solid conductor. 



* These Proceedings, XXXII. 253. 



