540 



Mr. G. U. Yule on the Passage of Electric 



Fig. 5. 



The figures given by StschegtiaefF agree so nicely with 

 Maxwell's law that one regrets being compelled to treat them 

 as negligible ; and the low value found for alcohol by Bouty is 

 also in a minority and requires explanation. It should be 

 noted that it was not determined directly, but deduced from 

 the constants found for mixtures of small proportions of 

 alcohol with other liquids. 



My figures agree roughly with the high values given in 

 the table, but are somewhat low. This is very likely to be 

 ascribed to the fact that a certain portion of the wave always 

 flows through the dielectric outside the jar instead of going 

 through the liquid. This portion interferes with the first 

 in such a way as to shift the maximum of the curve forwards ; 

 i. e. the whole arrangement behaves as if the wave had flowed 

 through a layer of air and water mixed instead of through 

 pure water. 



The indefinite error arising from this cause might have 

 been avoided in one way, namely, by making a certain length 

 A B (fig. 5) of the one wire into a tube sur- 

 rounding the other, and using this tube A B C D 

 also for holding the liquids. The whole of the 

 induction must then lie between the inside wire 

 and the tube, and, consequently, the whole of 

 the radiation must pass through the liquid. 



This was in fact the arrangement that I 

 originally wished to use. It possesses, however, 

 several serious disadvantages which finally led 

 to its rejection in favour of the simple wires and 

 glass jar. For one thing, such a condenser as 

 the tube forms reflects of itself a considerable 

 portion of the incident energy. Again, the 

 variations in the position of the electrolyte sur- 

 face, with reference to the top of the jar, would 

 introduce fresh interference phenomena, as wave- 

 trains are reflected equally well from the latter 

 and the former (as in the apparatus used by 

 Mr. Barton*). The large surface of metal in contact with 

 the liquid would cause additional inconvenience, as it would 

 rapidly render distilled water or very dilute electrolytes 

 impure. 



In addition to the liquids I have mentioned, salt and 

 soda crystals were used with a view to try and determine 

 their dielectric constants. No definite results were, however, 



* Loc. cit. 



