227 
1912-13.] Torsional Oscillations of Metallic AVires. 
to the rotating oscillator. Six recorders of a seven-record chronograph 
can he placed one in each of the six circuits furnished by the copper 
strips, while the seventh is connected with a clock. Thus it is possible 
to register the instants at which the oscillator attains successive angular 
positions at intervals of 2°. 
In the actual experiments it was found that registers at intervals of 
10° were sufficient, so the first slit only in each of the sets was filled 
with mercury. Further, three alternate sets were joined together 
metallically by means of a wire connecting the corresponding binding 
screws, and the remaining three were connected each to one of three 
registers of the chronograph. The connecting wires, including those to 
the clock and the accumulators, are seen in fig. 1. The figure also shows 
the chronograph * and the small motor which drives the paper strip. The 
strip passes between two brass plates (the front one seen dark against 
the paper) both of which are pierced by seven circular holes, the holes 
in one plate facing those in the other. The records are made upon the 
paper as it passes these holes. The drums on or off which the paper 
is rolled have their axes vertical, and no two holes are in one horizontal 
line. Thus the records are kept distinct. Each record is made by a 
needle point attached to a shaft which carries on its other end the 
armature of a small electro - magnet. AVhen the armature is attracted 
by the magnet, in spite of the control of a spring, the needle point is 
driven into the paper and makes a small sharply defined hole. The 
needle point is attached to the shaft by a joint so that it can turn 
horizontally in the direction in which the paper is drawn. This prevents 
tearing of the paper. Finally, the cessation of the magnetising current 
allows the withdrawal of the needle by the controlling spring, and the 
needle point, being also under the control of a spring, reassumes the direc- 
tion of the shaft to which it is attached. Four of the horizontal shafts 
can be seen against the dark background of the brass plate in fig. 1, and 
the armatures attached to them are seen nearer the edge of the clock. The 
switch, with two rows of seven binding screws, is visible, against the side of 
the clock, at the right-hand edge of the wooden base of the chronograph. 
Fig. 2 gives a diagrammatic plan of the connections. It is evident 
that chronographs 3, 5, and 7 should give, apart from instrumental 
irregularities, identical records. But these records are established by 
contacts made at slits whose angular positions differ by 120°. Hence 
errors due to slight want of symmetry of construction, or to slight 
* The cost of the chronograph and motor was defrayed by part of a grant given by 
the Trustees of the Carnegie Fund. 
