248 PROCEEDINGS OF SECTION A. 
appeared to be reversible, but upon trying repeatedly the deflec- 
tion became less regular, so that the deflection in one direction 
was greater than that in the other. On leaving the apparatus 
for an hour or two without disturbance and then repeating the 
experiment a deflection was obtained which would not reverse-on 
reversing the electrification of the quadrants. These experiments 
with the needles showed that if there were any contact effect 
constant in character the imperfections of the needle method were 
such that it could not be satisfactorily established in this way. 
The experiments with the quadrants were much more satisfac- 
tory and consistent. for the first two sets of quadrants ordinary 
roll sulphur was used. The first set was made in a dish 11 cm. 
in diameter, the second was 11:5 em. in diameter. 
The zine dish containing the quadrants was first of all placed 
upon the brass quadrants of the instrument and the aluminium 
needle hung immediately over them. The needle was charged, as 
in the ordinary use of the instrument, through the sulphuric acid 
of the condenser. 
For the first few trials no very consistent results were obtained 
owing, as it afterwards appeared, to insufficient diselectrification. 
Using first a potential of 50 volts a small deflection towards the 
insoluble quadrants was noticed when the charge on the needle 
was positive and towards the soluble when the charge was 
negative, but the deflection towards the insoluble was always 
much smaller than that towards the soluble. On rotating the 
quadrants through an angle of 90° the deflection was reversed. 
The zine dish was insulated from the brass quadrants on mica 
plates, and the potential used raised to 100 volts. In this case 
the deflection could not be reversed by changing the sign of the 
charge. The deflection was always towards the soluble quadrants, 
being larger in one direction than the other. These results 
showed that some disturbing influence was at. work, so the quad- 
rants were taken off and thoroughly diselectrified. 
Upon replacing them the electrophorus was used to charge the 
needle. The deflection was then much larger: the spot could 
easily be sent completely off the scale. The displacement was 
now invariably towards the soluble quadrants, the charge being 
positive. The quadrants were rotated repeatedly through 90° 
each time, and every time the deflection was reversed. The 
reversal showed that the deflection could not be due to induction: 
or other effects between the needle, case, and brass quadrants, as 
the only part moved was the sulphur dish. The deflection must 
have been due to something in the sulphur. The case was: some- 
times insulated and sometimes joined to earth. The quadrants 
also were sometimes insulated on mica, but no alteration of the 
effect was noticed ; the needle was always attracted towards the 
