250 PROCEEDINGS OF SECTION A. 
The electrophorus was used to charge the needle. Upon a 
charge being given, there was a deflection towards the soluble quad- 
rants, as there had been before. On increasing the charge, the 
deflection increased to the extent of about 123 divisions on the 
scale. Still increasing the charge, the deflection began to 
decrease, till at last the spot of light stood again at zero. 
Still increasing the charge the deflection reached 130 divi- 
sions on the other side of zero. On allowing the charge to 
gradually dissipate, a reverse series of movements was gone 
through ; the deflection towards the insoluble quadrants disap- 
peared and changed back to the original direction. When it 
reached eighty-five divisions in this direction it again altered the 
direction of movement and gradually settled down to zero. 
The needle was then rotated through 90°, everything else 
remaining as before. On charging (positively) there was a large 
deflection as before towards the soluble quadrants. When the 
charge was increased, the deflection diminished again, and then a 
deflection towards the insoluble quadrants took place, which was 
increased to about 20° by sufficiently increasing the charge. On 
allowing the charge to leak away, the effects were gone through 
backwards just as before. Rotating the needle several times the 
same results were obtained. 
This reversal effect upon increasing the positive charge on the 
needle could be accounted for by induction in the insoluble 
quadrants. Insoluble sulphur has a higher specific inductive 
capacity, and certainly a much higher electric conductivity than 
the soluble modification. Surface charge, due to difference of 
specific inductive capacity of sulphur from that of air, would, 
therefore, be greater on the insoluble quadrants. Induction 
charges, such as those formed in the ordinary way upon con- 
ductors, would also be much more rapidly formed on the insoluble 
sulphur ; they would also leak away more quickly. 
Starting, then, with a negative charge on the soluble quadrants, 
and a positive on the insoluble, due to the contact of the two 
modifications, the deflection on giving the needle a positive charge 
would be towards the soluble sulphur as usual. On increasing 
the charge, the induced charge would begin to have effect. The 
positive charge on the needle would induce a negative charge on 
the upper surface of the insoluble quadrants, which would tend 
to reverse the deflection, and when this induced charge became 
sufficiently great would actually cause a reversal. On allowing 
the charge to leak away, the induction charge would disappear, 
and so the deflection would return to its original direction, but 
would not attain its first maximum owing to the diminishing 
charge on the needle. 
To test whether this was the true explanation, it was con- 
sidered that, if the needle were charged negatively, no such 
