Dielectric Constants by means of Hertz Oscillations. 181 



Time of Charge 

 in seconds. 



Dielectric Constants of 



Plate glass 

 (0-8797 

 centim.). 



"Solinglas" 

 (0-4338 

 centim.). 



Ebonite 

 (0-7164 

 centim.). 



Petroleum 

 (1-9272 

 centim.). 



Water 

 (1-9342 

 centim.). 



5xl0 _1 



5xl0~ 4 



3X10" 8 



4-67 

 5-34 

 7-31 



464 

 5-09 

 6-50 



2-64 

 281 

 3-01 



Not measured. 

 2-35 oo 

 242 oo 



The dielectric constant, if we reckon it simply from the 

 capacity as in this paper and as J. J. Thomson has done, is 

 for quick oscillations not only not smaller, but, on the contrary, 

 much larger. In all the substances investigated we have found 

 the same phenomenon, and the differences found are too great 

 for us to assume that they arise from secondary sources 

 of error. I am therefore of opinion, contrary to J. J. Thom- 

 son, that in general we shall not be able to arrive at the true 

 dielectric constant even by the aid of Hertz oscillations. 



For the time of oscillation, neglecting certain terms which 

 need not here be considered, we have 



T 2 =LS(\ 2 + 7r 2 ), 



where L is the self-induction, S the capacity, and X the 

 coefficient of damping. This damping (X) increases as the 

 oscillations become more rapid, and it is probable that in our 

 Hertz oscillation the lag produced in tho swing by damping- 

 is the chief cause of the increase which we have found in the 

 dielectric constant. The phenomena taking place in the 

 dielectric with such rapid oscillations are, however, in any case 

 of too complicated a nature for me to advance this theory as 

 an explanation of the increase of the dielectric constant with 

 rapidity of oscillation *. 



* That the phenomena are not so simple, and that in many cases, 

 among which J. J. Thomson's experiments may also be counted, a decrease 

 of the dielectric constant occurs with Hertz oscillations, is shown by the 

 following experiments : — 



The condenser-plates are separated by about 28 centim., and a trough 



of water 27 centim. long is placed between them. The capacity of this 



condenser, measured by my method w r ith Euhmkorff oscillations, is just 



the same whether the vessel contains distilled water, or water to which 



Phil Mag. S. 5. Vol. 31. No. 190. Mar. 1891. P 



