1896 - 97 .] 
Chairman's Opening Address. 
199 
might come off the horizontal arm of the second lever. In the 
first case, when the pressure was too great, the curves recorded on 
the paper would he less in height than they ought to be, and, if 
the point of the lever came off the horizontal arm of the second 
lever, or touched it with a series of knocks, either no curves at all 
were obtained or the curves were not transcripts of the marks on 
the cylinder. 
I endeavoured to meet this difficulty by turning the screw q, 
on plate, which raises or depresses the arm, bearing, in the usual 
phonographic arrangement, the reproducing diaphragm, or, as in 
my instrument, the aluminium lever. Still it was difficult to do 
this accurately. At last the difficulty was got over by an auto- 
matic device. On the top of the screw for raising or elevating the 
aluminium lever was fixed a round shallow drum, r, 80 m. in 
diameter, and covered, on its outer surface, with sand-paper so as 
to be rough. Then the cord passing from the electro-motor to the 
phonograph, for driving slowly the mandrill bearing the wax 
cylinder, c IV , was brought round a pulley, s, and then transversely 
in front of the side of the roughened drum, r. The cord then 
passed round another pulley, t, attached to the keeper of an electro- 
magnet, marked v in the plate, and then back to another pulley, 
passing transversely to the roughened drum and parallel to the 
first cord. It then passed round a second pulley, marked v on 
plate, then round the pulley on the phonograph, and back to the 
electro-motor. When the electro-motor worked, the cord moved 
slowly in front of the roughened drum and one portion moved in 
one direction while the other moved in the reverse direction. If 
the lower cord, t 1 , touched the roughened drum, it would be moved, 
carrying the screw with it, in the direction say of the hands of 
a watch, but if the upper cord, t n , touched the side of the roughened 
drum, then the direction of the rotation of the screw would be in 
the opposite direction. It will be evident that the one movement 
would lower the aluminium lever while the other would raise it. 
In the next place, to make the apparatus work automatically, a 
slender platinum wire, w, was attached to the aluminium lever, about 
15 mm. from its point. This was caused to dip into a little trough, 
x, containing mercury resting on a horizontal metal arm passing out 
from the support bearing the aluminium lever, but immediately 
