748 PROFESSOR FLEEMING JENKIN AND J. A. EWING ON THE 
as to dip into box qg, holding ink, whilst the lower end was turned so as to point 
towards the periphery of a large wheel, which it approached but did not touch. 
A long band of paper an inch wide was coiled round the circumference 
of W, and as this revolved past the siphon o it received the tracing in the form 
of a succession of fine spots of ink, which were deposited by the siphon as 
the result of a continuous electrification, which came as a discharge from the 
rod 7 to the plate « The ink box g and therod7 were supported and insulated 
by vulcanite brackets, and the rod was connected by a wire to a small 
* mousemill ” or inductive electrical machine. This method of registering the 
movements of a pointer was invented by Sir Witi1aAmM THomson, and is 
used in his telegraphic recorder. It has been used by us before in experi- 
ments on friction, and a full account of it will be found in the paper describ- 
ing these experiments (“ Phil. Trans.,” vol. clxvii. p. 509). It has the immense 
advantage of doing away with all friction between the recording pencil and 
the paper on which the tracing is drawn. 
The directive force needed to bring the siphon o quickly back to the vertical 
position after having been displaced was given by the spring w, which, although 
shown as a spiral in the drawing, actually consisted of a delicate straight thread 
of india-rubber with one end fastened to the siphon at the same point as the 
fibre y, while the other end was secured to the fixed support v. This spring 
was stretched sufficiently to make the system of levers very ‘‘ dead-beat ;” that 
is to say, their recovery after displacement was rapid and unaccompanied by 
any gradually dying away oscillation. All the fixed supports formed part of a 
very strong framework which projected from the iron table on which the 
phonograph stood, so that any shaking or movement of that table, such as 
might have been caused by a footfall in its neighbourhood, had no effect in 
altering the relative position of any of the parts. Glass was chosen for the 
connecting thread 4, after many other substances had been rejected as unsuit- 
able owing to their change of length under the varying stress caused by the 
greater or less extension of the spring 2. 
With this arrangement each displacement of the pointer of the phonograph 
through any very small distance produced a proportional displacement of the 
end of the siphon through about four hundred times that distance. 
Hence when the phonograph containing the record of a spoken sound was 
placed in position as shown in the figure, and the pointer adjusted so as to 
press gently into the embossed furrow in the tinfoil, if the cylinder of the 
instrument were slowly turned the end of the siphon 0 would copy on a greatly 
enlarged scale the movements which had originally been made by the pointer 
and disc when under the influence of the voice. And if at the same time the 
electrical machine were set in action, and the wheel W turned so as to draw 
the paper ribbon past the siphon, an enlarged copy of the wave-forms of the 

