212 Dr. L. J. Blake on the Production of 
a quadrant-electrometer constructed by R. Voss of Berlin, 
according to the suggestions of Von Helmholtz. The charge 
of the quadrant pairs of this electrometer was maintained by 
means of two Zamboni's piles of about 1500 pairs each. The 
aluminium needle of the electrometer, suspended by a fine 
silver thread, had the form of two nearly right-angled sectors 
of a circle whose apices were joined with each other. In 
order to make the instrument more sensitive, the distance 
between the two horizontal surfaces of the quadrants between 
which the needle swung was less than in former instruments 
of a similar kind. The mirror attached to the needle swung 
beneath the quadrants. The upper end of the silver thread 
was fastened in a torsion head, and served also for uniting any 
source of electricity with the needle. The metal frame of the 
instrument, which supported the Zamboni piles, was connected 
with the earth. One Daniell (Cu— CuS0 4 — ZnS0 4 — Zn), 
one pole of which was connected with the electrometer-needle, 
and the other with the earth, gave at the distance of 3 millim. 
between the millimetre-scale and the mirror a deflection from 
67 to 71^ millim. After six months, without disturbing in any 
way the charge of the quadrants maintained by the Zamboni 
piles, the deflection for 1 Daniell under similar arrangement 
of experiment was from 64^ to 72 millim. 
The condenser, the electrometer, and all other apparatus 
mentioned hereafter were enclosed in metal cases connected 
with the earth. All the insulated connecting wires were 
enclosed within brass tubes similarly connected with the 
earth to neutralize the effect of outside electricity. The brass 
axis, which supported at one end the collector plate B of the 
condenser, was prolonged at the other end, and about 25 
centim. from B was bent into the form of a ring. In this 
ring rested a porcelain dish which contained in a sand-bath 
a second smaller dish of the same material. By this arrange- 
ment the insulation in the following experiments could be 
affected only through the support for the axis of B. The 
insulating power of this support, upon which evidently the 
reliability of the experiment depends, was measured both be- 
fore and after each experiment. When B was charged by 
4 Daniells, and the needle of the electrometer which was in 
connexion with A, distant i millim. from B, gave a deflection 
of 140 millim. through influence from B, then the loss in 
electricity upon B, during 10 minutes, varied from 6^ to 11 
millim., i. e. from 4^ to 7 per cent. 
Further, there was a peculiar electrical action through this 
insulating support of the collector plate B. B, unelectrified 
