THE RESONANT RECORDER 13 



advantages of an intermittent over a continuous contact. 

 What has been said of the writer vibrating ten times in 

 a second holds good equally in those cases where the 

 vibration-frequency is much higher. 



One great difficulty we encounter in carrying out 

 this idea hes in giving the recording-point an impulse 

 exactly perpendicular to the direction of its recording 

 movement. If the recording- writer be made of fine steel- 

 wire, and if we place behind it a small electro-magnet — 

 the pole consisting of a rectangular piece of soft iron, at 

 right angles to the wire — then, by sending a momentary 

 strong current through this electro-magnet, a pull will 

 be exerted on the wire which will make its recording-tip 

 strike for an instant against the glass recording-surface. 

 As the steel wire has to be made extremely fine, in order 

 to reduce to a minimum the inertia of the recorder, the 

 resulting pull exerted on it is very slight, unless an exces- 

 sively strong current be sent round the electro-magnet. 

 Again, unless the intermittent closures of the electric circuit 

 be properly timed, the writing-index may be subjected to 

 attraction in the course of its journey, now to the recording- 

 surface and again away from it. In the latter of these 

 cases its vibration, on which the intermittent contact 

 depends, is totally destroyed. 



But the most serious difficulty of all is that introduced 

 by the edge of the attracting electro-magnet. It is known 

 that the magnetic intensity of a pole is strongest at its edges. 

 Should the writing-index by chance be placed exactly 

 symmetrically, as regards the right and left edges of the 

 pole, then the two lateral pulls, being equal, will neutralise 

 each other, and the index will vibrate to and fro perpendi- 

 cularly to the recording-surface. But should it be placed, 

 however slightly, nearer to one edge than to the other, then 

 one of the two pulls will be in excess, and the index will 

 be drawn to one side, thus producing a disturbance in the 

 record not due to the excitatory pull of the leaf. Even 

 if, at the beginning, the index had been placed in a 



