608 PROF. J. J. THOMSON AND MR. G. F. C. SEARLE ON THE RATIO OF THE 
The supports for the contact wires of the commutator and for the brushes were two 
pillars screwing into the base of the framework. One of these is shown in the figure, 
it is shown broken, and the two parts separated in order to show both parts of the 
commutator. Fig. 6 shows the arrangement as it would look in an end view if the 
Fig. 6. 
stroboscopic disc were absent. The wires for making contact with the commutator 
are clamped under the two binding screws and are made with a curl Q so as to be 
very elastic and to ensure contact with the commutator. Before the curl was put 
into the wires it was found that the jarring of the apparatus gradually caused the 
wires (though made of hard drawn brass) to relax their pressure on the commutator 
and to make uncertain contacts. With the curled wires no trouble was found in this 
respect, and the pressure could be made much less, saving both friction and wear, 
'fhe binding screws clamping the wires are fixed to a piece of ebonite which can slide 
on the pillar and be fixed in any position by a screw. The pillar also carries a support 
ibr the brush. The Inrush itself is made of very fine hard dravrn brass wire, soldered 
into a brass piece of a suitable shape. The way in which the brush acted depended 
a good deal upon the regularity and straightness of its wires. We found it best not 
to have a thick bunch of wire, but a tliin layer oidy a few wires thick. The brass 
piece into which the brush wires were soldered, fits on the end of an ebonite rod 
passing through the brush holder and capable of being fixed in its proper position 
by means of a screw. 
The whole arrangement was clamped down to a thick iron slab, resting on a strong 
table. In this way the vibrations, which would otherw'ise have been set up, were 
avoided. 
The tuning fork, by means of which the speed was observed and regulated, was 
placed on a separate table. At first the two were on the same table, but it was 
found that at the speed at which the commutator made one revolution for each 
