136 Prof. W. E. Ayrton on the Electrical 



represents the lowest conductivity I was able to measure with 

 certainty directly with the galvanometer. It may here be 

 mentioned that such high resistances could be measured with 

 the galvanometer, since one Daniell's cell, through a resistance 

 oi' 600 megohms, gave 130 scale-divisions deflection on a scale 

 about lJr metre distant. The various curves obtained for con- 

 ductivity between 115° and 65° C. agree so closely that they 

 may all be represented by the one carve L M. The curves for 

 the conductivity between 80° and 40° C. are all quite regular, 

 but not all of exactly the same slope, the difference depending 

 on the highest temperature to which the wax was heated before 

 cooling on the particular day of experimenting, this being 

 sometimes about 130° C. and at other times about 90° 0. All 

 the curves, however, agree so closely that they are all contained 

 between the two limiting curves N P and Q R shown in fig. 3. 

 In no curve was there the slightest appearance of a rise of con- 

 ductivity at the melting-point, which would probably have 

 been obtained had the copper plates approached one another 

 an appreciable distance on the wax solidifying. 



We may therefore conclude that in the previous experiments 

 the rise in the capacity at melting indicates a true increase in 

 the specific inductive capacity coincident with an increase in 

 the index of refraction for light. 



As regards apparent increase of resistance by electrification, 

 which, as mentioned, I observed during repeated experiments, 

 extending over some months, with the wax condenser buried 

 underground, the general conclusions arrived at were that 

 not only did the conductivity usually increase by electrification, 

 but that it steadily increased day by day — a result indicating 

 that the wax was deteriorating, probably from damp penetrating 

 through the joints of the wooden box in spite of leather having 

 been inserted between the different parts of the wood before 

 they were screwed together. This conclusion appeared to be 

 correct, since, on digging up the condenser and keeping it 

 near a fire for many days, it regained its original high 

 resistance. The damp must therefore not only have entered 

 through the joints in the wood, but through some small cracks 

 that were observed in the mass of wax when the condenser 

 was opened ; and it was probably due to this damp that the 

 peculiar effects of polarization were observed similar to those 

 noticed by the Comte du Moncel when testing stones, and by 

 Mr. T. Warren in certain insulating oils. 



I now made a number of experiments with lead chloride as 

 a dielectric — a substance to which my attention was especially 

 drawn by some remarks of M. Buff in Ann. Chem. Pharm. ex. 

 p. 258 (1859), in which he says that this substance conducts 



