OF DENISON UNIVERSITY. I5 



The distribution of energy in the internodal spaces was next in- 

 vestigated. Bridges were placed at two nodal positions external to the 

 box and the Leyden jars moved from one end of the intervening half 

 wave to the other at intervals of lo cm. In each position four readings 

 of the galvanometer were made and the mean values taken as ordinates 

 for a curve whose abscissae were bridge positions, gave directly the dis- 

 tribution of electrical energy in the wires. For positions between suc- 

 cessive nodes a very smooth and regular curve was obtained, which 

 differs but little from a sinusoid. This is shown as curve N, Plate II. 



When a bridge was placed at the box end, to force a node there, 

 filling the box with distilled water left the wave system in the air-sur- 

 rounded portion of the wires unchanged, but only slight and unsatisfac- 

 tory traces of a wave system could be detected within or beyond the 

 box, whether by Leyden jars placed at the end, or by those placed on 

 the wires within the liquid. E. Cohn i had detected and measured the 

 wave-length of resonance waves of this sort in water, and deduced 

 therefrom the specific inductive capacity of water for long electrical 

 waves. Of course the introduction of liquid, by the change in capa- 

 city, might be expected to destroy the resonance within and beyond 

 the box, but it was hoped that by careful adjustment of a variable 

 capacity at the end, the capacity of this part of the system might be 

 increased exactly to some simple multiple of its former value and re- 

 sonance thus restored. 



These hopes proved delusive. The capacity at the ends was grad- 

 ually raised by different means through wide limits, but such changes 

 seemed to make no difference whatever in the resonance system within 

 the liquid. This attempt was therefore abandoned, and improvement 

 sought along the following lines : (a) securing great galvanometer sen- 

 sitiveness, (b) purity in the liijuids used, (c) using vessels of such mate- 

 rials that the li(]uids used would have no action upon them. Zinc had 

 been used in earlier experiments. Now glass and glass coated with 

 pure silver, alone were used, (d) Taking readings at very short inter- 

 vals along the wire. 



By the use of these various precautions well defined maxima were 

 obtained, both with water and alcohol. (See Curve Q, Plate II). The 

 maxima were less sharp and far weaker however in the part of the 

 wires surrounded by water than in that surrounded by air. Since the 



' E, Cohn, Wied. Annalen, Vol. XLV, p. 370 (1892.) 



