248 
DR. J. H. POYNTING AND MR. P. L. DRAY IN' SEARCH OF A 
observations were continued usually for 70 or 80 periods. Adding up the columns 
and dividing’ l:)y the number of periods, any forced oscillation would be indicated by 
a periodicity in the quotients. The periodicities found were too irregular to be taken 
as evidence of the existence of a couple. 
Description of the A'pparatus. 
Tlie (juartz spheres were placed in a cellar at Mason College, Birmingham, below 
the room in which the observing telescope and rotating apparatus were fixed. 
The smaller sphere, 0'9 centim. diameter and weighing 1’004 grams, was held in 
an aluminium wire cage, and was suspended by a long, fine quartz fibre in a brass 
case from a torsion-head at the top of the case. 
A light plane mirror was fixed to the cage, and opposite this mirror was a glass 
window in the case ; in front of the window wms a plane mirror at 45°, by means of 
which the light from the scale was reflected into the case and back again to the 
telescope, as shown in fig. 2. 
The case wms surrounded by a double-sided wooden box, lined within and without 
with tin-foil, and with cotton-wool between its inner and outer walls. The box was 
supported on indiarubber blocks to lessen tremors. 
The larger sphere, 6'6 centims. diameter and weighing 399’9 grams, was held at 
the lower end of a vertical brass tube which terminated in a very carefully turned 
shallow brass bell, in which the sphere was held by tapes. The tube passed upwards 
through the top of the wooden casing without contact, a kind of air stuffing-box 
indicated in the figure serving to prevent currents through the hole. The tube came 
into the room above, and was there connected with a train of wheels, driven by an 
electromotor, the rotation of the motor being geared down from 1000 to 1. The 
observing telescope was fixed to a heavy stone slab resting on indiarubber blocks, 
standing on a brick-j-tillar, which was built on the brick arches forming the cellar- 
roof. A diagonal scale (of half-millimetre graduations, divided into tenths by the 
diagonal ruling) was clamped to the telescope-tube and illuminated by an incandes¬ 
cent lamp, aided by a concave mirror. A tenth of a division could be read with 
certainty, and as the distance from scale to mirror was 358 centims., the position of 
the suspended sphere could be determined within a little more than one second of 
arc. 
The steady rotation of the larger sphere was maintained by a regulator, for which 
we are indebted to Mr. B. IT. Holtsman. It consisted of two parts:— (l) the 
governor proper, which automatically maintained approximate steadiness, and (2) a 
fine hand-acljustment, by which the motion could be accelerated or retarded when it 
got “ out of time.” 
(Jne lead to the mofor went through tv/o mercury-cups, and the circuit was 
completed by a fork of platinum-wire dipping into the cups. This wire was fastened 
