TRACING GRAMOPHONE AND PHONOGRAPH RECORDS. 29 



a rotation once in five hours to be perfectly safe with the most sensitive 

 of levers. Vibrations of the lever itself do not occur unless the apparatus 

 is jarred or the tracing point is so badly adjusted that it catches on the 

 paper. The period of the lever is so short that its vibration appears in 

 the former case as a white dot and in the latter as a cross Une. The vibra- 

 tions of the lever itself have no influence whatever on the movements 

 for the curve traced ; they can even be used to widen the tracing line by 

 putting on the table some automatic jarring apparatus. 



The tracing point must move without catching; this is tested by rap- 

 ping on the near drum when it is recording a large side deviation. If 

 it catches, it will move by jerks. 



The tracing point must he in the bottom of the groove; this is tested 

 by rapping very gently on the gramophone disc. If the point is not at 

 the bottom it will fall to it, and a sudden jerk of the lever will be seen. 

 The trouble can be removed by weighting the lever. 



The gramophone disc must turn evenly. The movement is best 

 studied by mounting a piece of mirror over the center by a bit of wax, 

 and reflecting a beam of light on the wall. With good lubrication and 

 a non-stretching belt the beam will move evenly. Otherwise it will move 

 with jerks that become less as the ideal conditions are approached. Raps 

 on the table produce sudden jerks; these indicate how much the belt 

 stretches. The evenness of rotation of the screw for side movement can 

 be studied in a similar way. 



It is necessary to keep a careful watch over these points at all times, 

 because any irregularity of motion appears in the record with the same 

 magnification as the speech curve. The belts should be very flexible, 

 but non-stretching; certain threads and strings answer the requirements, 

 but they can only be selected by careful tests. I have obtained the best 

 results by using fine iron wire. 



The hard rubber ring or "contact disc," referred to above (p. 24), 

 serves two purposes. The ten sets of metal contact pins — the first con- 

 taining one pin, the next two pins, etc. — pass over a copper brush in circuit 

 with a battery and a magnetic time-marker, which writes on the strip of 

 paper. As the rotator turns, the sets of pins pass in succession over the 

 brush and make momentary contacts, whereby the time-marker registers 

 successively 1, 2, 3, etc., checks. In this way it is possible to know the 

 position of the disc at each point of the tracing. This convenient arrange- 

 ment is not absolutely necessary, except for a few turns in order to estabhsh 

 the relation between the rate of rotation of the disc and the length of the 

 curve, that is, the "time equation" (see below); it is useful, however, in 



