1884.] 



On the Electrical Qualities of Glass. 



491 



coils from the same side, and the current is usually sent through the 

 coils, so that one pair cause their horseshoe to move outwards, and the 

 other pair their horseshoe to move inwards, thus turning the needle 

 system round the suspension fibre. This system of needles when 

 rightly adjusted is practically astatic in a magnetic field of uniform 

 intensity. A magnet moved in azimuth by a tangent screw, and 

 vertically by sliding along the vertical supporting rod, is used to 

 give a difference of intensity to the magnetic field at the upper and 

 lower ends of the needles, which are placed with similar poles 

 turned in dissimilar directions. 



The instrument is fitted with a distributing plate, the construction 

 of which is fully described in the paper. By means of this plate the 

 coils can be joined in series or in multiple arc, or in any possible 

 combination of these modes of arrangement, so as to give all possible 

 variations of resistance and sensibility, or can be arranged for use 

 as a differential galvanometer when required. 



The resistance of the coils when joined in series is slightly over 

 30,000 ohms, and it can easily be arranged to have a sensibility such 

 as to give, when placed in series with a resistance of 10 11 ohms in 

 the circuit of a single Daniell's cell, a deflection of one division on a 

 scale placed at a distance of a metre from the mirror. The 

 instrument can easily be made still more sensitive, but when this is 

 done the long period of the needle system renders it difficult to take 

 readings quickly. This disadvantage may, however, be in great 

 measure obviated by using a vane wholly immersed in liquid, or 

 contained in a nearly closed air-vessel of proper size. 



With a considerably less degree of sensibility we have found the 

 instrument very convenient for the measurements described below. 

 These were conducted in the following manner. The flask was 

 heated by the sand-bath to a temperature above 100° C, and readings 

 were then taken of the current sent through the glass when the 

 battery was applied by means of the key. That the current 

 measured by the galvanometer might not be affected by leakage 

 through the table and other supports of the apparatus, the wire 

 connected to the inside mercury coating of the flask was carried 

 through air only direct to the galvanometer terminal, and tests were 

 made as to whether the whole current shown by the galvanometer 

 actually passed through the glass by withdrawing the connecting wire 

 from the mercury inside the flask and bringing it into contact with 

 the neck of the flask outside by twisting it round the glass. Such 

 tests always showed that no leakage current existed, and that the 

 current measured was really that passing through the substance of 

 the glass from one mercury coating to the other. After a reading of 

 the galvanometer in one direction had been obtained and recorded, 

 with the temperature of the glass when the reading was taken, the 



